Positive Interactions and Interdependence in Plant Communities
Positive Interactions and Interdependence in Plant Com...
124 downloads
943 Views
4MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Positive Interactions and Interdependence in Plant Communities
Positive Interactions and Interdependence in Plant Communities by
Ragan M. Callaway University of Montana, Missoula, MT, U.S.A.
A C.I.P. Catalogue record for this book is available from the Library of Congress.
ISBN 978-1-4020-6223-0 (HB) ISBN 978-1-4020-6224-7 (e-book)
Published by Springer, P.O. Box 17, 3300 AA Dordrecht, The Netherlands. www.springer.com
Cover figures redrawn by Wendy Ridenour
Printed on acid-free paper
All Rights Reserved © 2007 Springer No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work.
This book is dedicated to Bruce Mahall
ACKNOWLEDGEMENTS
I got a lot of help writing this book. Most recently, the editing by Giles Thelen was priceless and he takes full responsibility for all typos. I also have had tremendous help from many others, including the Alpine Pals – Lohen Cavieres, Chris Lortie, Richard Michalet, Paco Pugnaire, Zaza Kikvidze, and Alfonso Valiente-Banuet. The Pals have been relentlessly supportive. An inordinate amount of the ideas, much of the information and a lot of the editing in this book comes from them. Chris’ inputs on meta-analysis are the only reason I could write about it here. I was also facilitated in the whole process by Mark Bertness, the king of positive interactions who years ago invited me to join him as an author on an idea paper, and now I get far too much credit for the idea. It has been Mark that has dragged facilitation into modern community ecology. Fernando Maestre and Katja Tielbörger helped me, and still are, to think though new angles on the stress gradient hypothesis; and I am particularly grateful to Fernando for reworking some of his data and providing me with a figure. I have also received a great deal of assistance from many of my students; Erik Aschehoug, Beth Newingham, Marnie Rout, Kerry Metlen, Kurt Reinhart, Dean Pearson, Wendy Ridenour, Dayna Baumeister, and Leigh Greenwood either read and edited parts of the book or contributed lively discussion. Wendy also drew several of the figures and the cover of this book. Over the years I have benefited tremendously from talking over ideas about facilitation, indirect interactions, soil microbes and succession with Lars Walker, John Klironomos, Joe Connell, Jonathan Levine, Joe McAuliffe and Steve Pennings. These times are truly appreciated. Many, many people have responded positively to my pestering for copies of the figures or the data to remake the figure, and I have tried very hard to appropriately acknowledge all of these folks; their contributions are deeply appreciated. Finally, I owe quite a debt to my PhD advisor, Bruce Mahall, for years of mentorship and friendship.
vii
TABLE OF CONTENTS
Chapter 1 Introduction ........................................................................................1 Chapter 2 Direct Mechanisms for Facilitation ..................................................15 2.1 Water Relations: Hydraulic Lift ................................................18 2.2 Water Relations: Canopy Interception .......................................22 2.3 Shade…......................................................................................23 2.4 Water Relations: Soil Moisture ..................................................45 2.5 Nutrients ....................................................................................56 2.5.1 Climate and Variation in Canopy Effects on Soil Fertility.................................................................59 2.5.2 Experimental Approaches to Canopy Effects on Soil Fertility.................................................................63 2.5.3 Conditionality and Canopy Effects on Soil Fertility ........66 2.5.4 Canopy Facilitation of Soil Fertility in Mesic Habitats ....67 2.5.5 Canopy Facilitation of Available Soil Nutrients...............67 2.5.6 Litter Type, Mixtures, and Canopy Facilitation................69 2.5.7 Experimental Studies of Facilitation by Canopies on Soil .............................................................................71 2.5.8 Nitrogen Fixation and Nutrient Facilitation......................73 2.5.9 Canopy Interception of Nutrients .....................................77 2.5.10 Interactions between Nutrients and Shade......................78 2.5.11 Nutrients and Conspecific Facilitation............................82 2.5.12 Nutrients and the Chemical Signature of Neighbors ......83 2.6 Wind ...........................................................................................87 2.7 Soil Oxygenation........................................................................94 2.8 Substrate.....................................................................................99 2.9 Disturbance ..............................................................................104 2.10 Population Size and Positive Density-Dependence................108 2.11 Seed Shadows.........................................................................110
ix
x
Table of Contents
2.12 Communication ......................................................................113 2.13 Conclusion..............................................................................116 Chapter 3 Indirect Mechanisms for Facilitation..............................................117 3.1 Herbivore-Mediated Facilitation ..............................................118 3.1.1 Shared Defenses..............................................................120 3.1.2 Associational Resistance.................................................129 3.2 Other Herbivore-Mediated Positive Effects .............................137 3.3 Reproductive Feedback, Pollinators, and Population Size.......140 3.4 Dispersers.................................................................................150 3.5 Mycorrhizae .............................................................................154 3.6 Plant-Soil Microbe Feedbacks .................................................160 3.7 Positive Indirect Interactions Among Competing Plants .........164 3.8 Conclusion ...............................................................................176 Chapter 4 Interaction Between Competition and Facilitation .........................179 4.1 Competition, Facilitation and Abiotic Stress ...........................192 4.1.1 Stress Gradients and the Importance of Productivity......195 4.1.2 Stress Gradients and Meta-Analysis ...............................196 4.2 Spatial Scales, Time Scales and the Balance of Facilitation and Competition on Stress Gradients.......................................219 4.3 Facilitation and Stress: Importance Versus Intensity ...............237 4.4 Facilitation and Life History Stage ..........................................240 4.5 Competitive Advantages Provided by Benefactors..................248 4.6 Indirect Effects and the Balance of Competition and Facilitation.........................................................................250 4.7 Pollution and Shifts in Facilitation and Competition ...............253 4.8 Conclusion ...............................................................................253 Chapter 5 Species-Specific Positive Interactions............................................255 5.1 Are Beneficiary Species Non-Randomly Associated with Potential Benefactors?..............................................................256 5.2 What Mechanisms Cause Species-Specific Facilitation?.........267 5.2.1 Species-Specific Direct Effects ......................................267 5.2.1.1 Shade..................................................................267 5.2.1.2 Nutrients.............................................................272 5.2.1.3 Oxygen ...............................................................277 5.2.1.4 Hydraulic Lift.....................................................277 5.2.1.5 Disturbance ........................................................279
Table of Contents
xi
5.2.2 Species-Specific Indirect Effects ....................................280 5.2.2.1 Consumers..........................................................280 5.2.2.2 Pollinators and Dispersers..................................281 5.2.2.3 Soil Microbes .....................................................283 5.2.3 Species-Specificity Due to Similar Positive Effects, but Different Negative Effects .......................................286 5.2.4 Species-Specific Interactions and Life Histories ............288 5.2.5 Communication...............................................................290 5.2.6 Implications for Community Ecology ............................293 Chapter 6 Positive Interactions and Community Organization.......................295 6.1 Positive Interactions and the Expansion of Niche Space .........296 6.2 Positive Interactions and the Role of Diversity in Community Function ...............................................................296 6.3 Positive Interactions and Spatial Scale.....................................305 6.4 Positive Interactions and Stability in Plant Communities ........309 6.5 Facilitation and Productivity ....................................................312 6.6 Positive Interactions and Exotic Invasion ................................315 6.7 Facilitation and Conservation...................................................319 6.8 Facilitation and Evolution in Plant Communities ....................319 6.9 Replacing the Notion of Individualistic Communities with the “Integrated Community” ............................................327 6.10 Conclusions ............................................................................333 References.......................................................................................................335 Index................................................................................................................413
CHAPTER 1 INTRODUCTION
Over the last 50 years there has been profound growth in the discipline of ecology. To name just a few advances, ecologists have demonstrated, quantified and explained global changes in temperature, developed elegant mathematical models for competitive interactions, constructed wonderfully complicated food webs, integrated soil biota into aboveground processes, and experimentally explored the intricacy of indirect interactions among many species in communities. Some things, however, have not changed very much. This book was written in part to address a surprisingly static idea; the individualistic conceptual paradigm for plant communities. This is the perspective that plant communities are solely the product of population phenomena, and therefore are assemblages of individual species merely because they share adaptations to particular abiotic conditions (Gleason 1926). This leads to the conclusion that plant communities are simply a handy typological construct. In this book I argue that plant communities are not simply suites of species that happen to be dispersed to and adapted to the same biotic conditions at a given place. I argue that many if not most plant communities have fascinating interdependent characteristics, with some species creating conditions that are crucial for the occurrence and abundance of other species. Most ecologists are not card-carrying members of either individualistic or interdependent guilds, but our perception of plant community organization, and the way we conduct research, is affected by a historical dichotomy with lingering and powerful heuristic impacts; the dichotomy of the individualistic versus organismal nature of plant communities (Clements 1916, Gleason 1926). For example, the legacy of Gleason’s triumph lives on in almost all ecology textbooks, neutral theory (Hubbell 2001, Whitfield 2002) and assembly rules (see Lortie et al. 2004). The fundamental thesis of this book is that the current individualistic model is inadequate in the light of the last 20 years of empirical research on facilitation and indirect interactions. This is because these interactions demonstrate that plant communities frequently contain plant species that would not be present at all, or that would be present at much lower abundances, if it where not for the presence of other plant species (Callaway 1995, 1997). The individualistic view of plant communities has led to very successful research on the importance of the abiotic environment and 1
2
Introduction
competition as factors structuring plant communities. Negative interactions predation, competition for resources and allelopathy – have been central to the study of ecology and evolution. However, it has become clear that organisms can greatly enhance the performance of their neighbors as well as modify the environment in ways that benefit other species. Positive interactions among plants, or facilitation, occur when the presence of one plant enhances the growth, survival, or reproduction of a neighbor. Much like the way the term “competition” is used in the literature, the term “facilitation” is also used in a loose manner, and facilitation may occur in concert with negative, positive, or neutral reciprocal responses from neighbors. Facilitation does not have to be mutualism, an interaction where both participants gain (+,+), but some experiments have shown mutualistic bi-directional facilitation. In some cases facilitation may occur as commensalism (+,0) in which one species benefits from another, but does not affect it in return. In a review of interaction types Schoener (1980) noted that “[documented] examples of commensalism are relatively rare”; and commensalism among plants remains relatively unstudied. However, this rarity may be an artifact of scientific disinterest rather than ecological frequency. Futuyma (1979) suggested that commensalism may be so common “that we often do not notice it”. Commensalism may be common, but the empirical research explored in the following chapters indicates that facilitation probably occurs most often as a positive effect of one species on another with a reciprocal competitive effect from the species receiving the benefits on its benefactor (+,-). It is important to grasp the broad semantic usage of the term facilitation. As one colleague has put it, “I hate the word facilitation because why would a plant make it easier for a competitor to grow next to it”? My colleague’s problem is that of perceived intent, as if plants were trying to befriend their neighbors. The word “facilitation” means nothing like this. The nuances of language are complex and facilitation, like most words, suffers from subtleties in its gestalt. The term facilitation describes a process and not purpose. The fact that seedlings of saguaro cacti occur almost exclusively under shrubs and trees in the Sonoran Desert is almost certainly a by-product of the changes the shrubs and trees create in the environment simply because the shrubs and trees exist, not because they are altruistic. By analogy, in another (+,-) interaction gazelles try hard not to be eaten, but they have a strong positive effect on lions and cheetahs anyway. Like the beneficial effect of gazelles on lions, positive interactions among plants are produced simply by benefactors with characteristic effects on the abiotic and biotic environment that other organisms can utilize. Just like the suite of different mechanisms we group into the term “competition”, it is not necessary to identify the two-way interaction signs or the precise mechanisms behind a particular facilitative interaction to show that facilitation occurs. Furthermore, casual semantics and unknown mechanisms do
Chapter 1
3
not diminish the fundamental conceptual importance of facilitation for community theory. For over 50 years most plant ecologists have accepted the notion that the distribution of plant species, and their organization into groups or communities, is determined individualistically, that is by the adaptation of each species in a “community” to a particular abiotic environment, highly stochastic dispersal events, competition among these similarly adapted species, and the disruption of adaptive and competitive distributions by consumers. Definitions of the “individualistic” paradigm of plant community organization can be controversial (see Nicolson and McIntosh 2002), but have emphasized “the fluctuating and fortuitous immigration of plants and an equally fluctuating and variable environment” (Gleason 1926). Moore (1990, see Nicolson and McIntosh 2002, Chapter 6.9) re-phrased the individualistic concept as “vegetation as an assembly of individual plants belonging to different species distributed according to its own physiological requirements as constrained by competitive interactions.” Even a loose definition of facilitation suggests something fundamentally different than this. If the presence of one species can increase another species’ fitness, or the probability that another species will occur in the same place, plant communities cannot be individualistic. In the last 20 years, hundreds of peer-reviewed papers have been published on the positive effects of plants on each other. These papers implicitly challenge the adequacy of a strict definition of the theory of individualistic plant communities (Gleason 1926), one of the most basic and widely accepted conceptual models in ecology, as a foundation for understanding how groups of plant species are organized. The implications of rethinking plant individualism go beyond academic quarreling; if plant communities are even just a little less individualistic than we have thought, the conservation implications are profound (see Byers et al. 2006, Padilla and Pugnaire 2006). Interdependence in plant communities means that the loss of some plant species will have important negative effects on others. Most general conceptual models of community structure are either explicitly or implicitly based on competition, and this perspective has a historical legacy that is intertwined with individualistic theory. After the trouncing given to holistic community concepts in the 1950’s by John Curtis (1959) and Robert Whittaker (1951, 1953, 1956), espousal of ideas with a hint of Clements’ (1916) organismal mysticism was likely to bring disapproval from one’s peers. As a graduate student in the 80’s I was encouraged by some to avoid the word ‘community’ and instead refer to ‘assemblages’. ‘Assemblage’ is a perfectly good word, and stigma for supporting ultra-holistic Clementsian views was certainly warranted. Furthermore, the proscription on holistic theory fostered the successful emphasis on plant competition over the last few decades, an interaction that has no holistic implications (Connell 1983, Schoener 1983,
4
Introduction
Fowler, 1986, Aarrsen and Epp 1990, Goldberg and Barton 1992). But the reaction against holism created an environment that was not conducive to exploring facilitation. The emphasis on competitive effects and downplay of facilitative effects has also been exaggerated by the study of plant interactions in the greenhouse. While the isolation of some mechanisms is possible only under controlled conditions and must be conducted in pots and in greenhouses, studies of plants interacting in greenhouses almost always show competition. This may be because cramming several plants into a restricted area reduces their niche options, or because conditions in greenhouses tend to be so benign that neighbors can have no real effects on the harsh conditions that exist in the real world. For example, if there are plenty of nutrients then soil amelioration is inconsequential, if there is no wind the effect of neighborhoods as buffers against wind cannot be important, if there are no herbivores there can be no shared resistance, or if ambient humidity is high then the effects of neighbors on moisture around leaves is minimal. Without the normal stress of real life, studies are far less likely to demonstrate facilitation (see Chapter 4). Much like research on Paramecium in aquaria oversimplified theoretical perspectives on interactions among organisms in general; research on plants in greenhouses and pots has overemphasized competition. To my knowledge, the first experiment on facilitation was published in 1914 by G.A. Pearson. Pearson noticed that conifer species appeared to regenerate better after fires in clones of Populus tremuloides (quaking aspen) than in the open and that “herbaceous growth is invariably more luxuriant under the aspen than in the openings.” He then planted seedlings of Pseudotsuga menziesii (Douglas-fir) under aspens and in openings and found greater survival under aspens. Recognizing the possibility that site effects may have differed (aspens may simply have been growing in sites with generally superior abiotic characteristics) he measured wind speeds and evaporation rates and hypothesized that amelioration of these effects and those of shade benefited Douglas-fir regeneration independently of site effects. Until the late 1980’s, data such as Pearson’s (including a large number of other experimental results) were rarely interpreted as conceptually important in any general way - with the exception of a few ecologists focusing on facilitative interactions driven by herbivores. In 1976, Peter Attsat and Dennis O’Dowd of the University of California at Irvine published a review in Science titled “Plant defense guilds” with the leader titled “Many plants are functionally interdependent with respect to their herbivores” (my italics). They went on to argue that the probability that a plant will suffer from herbivory depends on the chemistry, morphology, distribution, and abundance of neighboring plants. Such indirect forms of facilitation, such as described by Atsatt and O’Dowd are treated in detail in Chapter 3 of this book. However, Atsatt and O’Dowd’s strikingly
Chapter 1
5
non-individualistic perspective had no impact on plant community theory. Other early papers, such as published by J.D. Ovington as early as 1955 on the speciesspecific (see Chapter 5) effects of trees on understory composition and productivity also did not stimulate any general interest in the conceptual ramifications of facilitation. To my knowledge, the first broadly conceptual appreciation of positive interactions emerged in two foundational, but under-cited, publications with strong theoretical stances. The first, “Positive Feedback in Natural Systems” by Don DeAngelis et al. in 1986, explored the general role of positive effects in ecosystems, and the second, “Plants Helping Plants” by Hunter and Aarssen (1988), explicitly argued for facilitation as an important and common process in plant communities. A third under-appreciated paper was published in Europe with the title of “Positive Interaktionen Zwischen Pflanzenarten” by Gignon and Ryser (1986). The most powerful effect on the resurgence of interest on facilitation, however, came from a series of experimental studies conducted by Mark Bertness and colleagues at Brown University (Bertness 1988, 1991, Bertness and Shumway, Bertness and Hacker 1994). Since the late 1980’s, a large number of reviews and commentaries have refined the theoretical role of direct and indirect positive interactions in natural plant communities, and organized the body of evidence that has accrued supporting positive interactions as important and general phenomenon affecting plant distributions, productivity, diversity, and reproduction (Wilson and Agnew 1992, Bertness and Callaway 1994, Callaway 1995, Callaway and Walker 1997, Callaway 1997, Callaway 1998a, Dodds 1997, Brooker and Callaghan 1998, Bertness 1998, Stachowicz 2001, Bruno et al. 2003). These reviews have coincided with an explosion in empirical research on facilitation (Figure 1.1).
6
Introduction
1100
# of hits for “facilitation”
1000 900
40
800 700
30
600 500 20
400 300
10
200
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
0
1990
100 0
# of hits for “competition” and journal totals
50
Figure 1.1. Web of science search hits on the terms: [(‘positive interactions’ or facilitation) and plant], shown as bars, [(‘negative interactions’ or competition) and plant] in filled points and total publications of American Naturalist, Ecology, Journal of Ecology, Oecologia and Oikos, shown as open points, from 1990 to 2000. Reprinted from Dormann and Brooker (2002) with permission from Acta Oecologia.
As noted in many of the recent reviews, experimental studies of facilitation and competition rarely provide unbiased neutral estimates of the relative importance of these interactions in communities. This is an important problem for those attempting to understand the fundamental role of interactions in community organization. In an effort to solve this problem, Walter Dodds (1997) constructed a general neutral model based on a number of field studies in which seven or more species were manipulated. As predicted from earlier simulation models (Dodds and Henebry 1996), he found that positive interactions among species were as likely as negative ones in communities as long as relatively large numbers of species and connections were considered. If fewer species were considered in a single interaction matrix, the probability of finding either positive or negative interactions decreased. Dodd’s models are intriguing, and in several other empirical studies involving large numbers of species, generally designed to examine competition, the results have indicated some positive interactions. However, the proportion of positive interactions demonstrated in empirical studies has usually been lower than that predicted by Dodd (Wilson and Tilman 1995, Twolan-Strutt and Keddy 1996, Thomas et al.
Chapter 1
7
1999). A review by Goldberg and Barton (1992) indicated that neighbors promoted the survival or growth of individuals in approximately 10% of experiments. Peter Ryser (1993) found that two of six species studied in a Swiss grassland required shelter by neighboring plants. In a field comparison of spatial patterns and growth correlations Freeman and Emlen (1995) found more competitive than facilitative effects, but for some traits there were large numbers of apparent positive effects of species on each other. Hoffman (1996) found that eight of 12 tree and shrub species in cerrado savanna vegetation of Brazil responded favorably to canopy cover, whereas only one species experienced lower establishment under canopies. Out of a total of 35 species in a Chilean desert community, Gutiérrez et al. (1993) found that five appeared to be facilitated by shrub canopies and five appeared to be inhibited. However, it is unclear how species were chosen for analysis or experimentation in many of these experiments. However, there are many recent studies in which high proportions of species in communities participate in positive interactions (e.g. Choler et al. 2001, Callaway et al. 2002). Furthermore, the distributional positions of particular experimental species on environmental gradients appear to be crucial for predicting the proportions of species involved in facilitation (Choler et al. 2001). The bottom line that we can gain from studies that incorporate multiple species appears to be that competitive interactions are usually more common than facilitative interactions, but facilitative interactions are not rare, and can be common even in communities composed of species with similar morphologies. Studies of spatial association are not as powerful as experimental evidence, but they also provide important insight into the relative importance of positive interactions versus negative interactions in plant communities. In some cases the consistency of spatial relationships among species can be impressive. Consider the relationship between Ziziphus lotus and Asparagus albus illustrated by Reyes Tirado and Francisco Pugnaire (2003) in oceanic dunes in southern Spain (Figure 1.2). Not only was the latter species virtually always found inside Ziziphus patches, transplanted Asparagus seedlings had higher survival rates in patches than in the open and produced more flowers, fruits, and showed a higher mass of seeds in patches than when isolated. This facilitative effect seemed to be due to nutrient enrichment in the patches.
8
Introduction Ziziphus lotus
Asparagus albus
South-North distance (m)
150
100
50
0 0
50
100
150
East-West distance (m)
Figure 1.2. Distribution map of Ziziphus lotus and Asparagus albus in a sand dune-strip on the Almeria coast of Spain. Shrub symbols are represented in three size classes (<1 m, 1–4 m, and >4 m) for Z. lotus and two (<0.2 m and >0.2 m) for A. albus. Clear triangles represent Asparagus plants with other shrub species. Reprinted from Tirado and Pugnaire (2003) with permission from Oecologia.
In a correlative study of desert perennials in Namaqualand, South Africa, Eccles et al. (1999) argued that spatial patterns for 23% of species pairs in shrubby ‘short strandveld’ suggested positive interactions, whereas only 6% appeared to be driven by competition. In ‘medium strandveld’ positive interactions appeared to determine spatial associations for 38% of species pairs and negative interactions only 13%. Based on spatial patterns, similar proportions of species also appear to be facilitated or nursed in Sonoran desert systems (McAuliffe 1988). These proportions are only based on correlations, but experiments conducted in alpine plant communities to examine the connection between spatial pattern and the interactions among species found that negative correlative spatial associations between species rarely pointed to competitive interactions (as determined though removal experiments) but positive spatial patterns often signaled facilitation (Choler 2001). Many of the facilitative mechanisms discussed in Chapters 2 and 3 would not be manifest as discrete spatial associations. Therefore empirical comparisons and a clear understanding of mechanism indicate that spatial patterns may be conservative in their estimation of the relative importance of facilitation. Other evidence for the relative importance of facilitative interactions comes from comparisons of the performance of plants in mixtures to performance in monocultures. Darwin (1858) mused about the potential for species mixtures to be more productive than monocultures, and since then many
Chapter 1
9
studies have been conducted to compare the biomass productivity of mixtures of different field crops to the productivity of the same species grown at the same densities alone (see Chapter 6.2). Legume-grass mixtures commonly “overyield’ (produce more than either species grown alone) due to the nitrogen-fixing properties of legumes, but studies of mixtures of nonleguminous species are much less conclusive. In 1974, Trenbath summarized studies on the productivity of 344 different mixtures and found that the means of 60.2% of the mixtures were above the mean yields of the monocultures, a proportion that was highly significantly different (P<0.001). Eighty three of the mixtures were more productive than the most productive monoculture, whereas only 45 mixtures yielded less than the least productive monoculture. Twenty years after Trenbath’s analysis, Peter Joliffe re-examined comparative studies on the total productivity of species mixtures to monocultures and arrived at similar conclusions (Jolliffe 1997). In 38 of 54 published experiments with two-species mixtures the mixtures were significantly more productive, and significantly lower in 8 studies. On average, mixtures were 12 to 13% more productive than pure stands, depending on the criteria used for inclusion of studies in the analysis. Overyielding is not necessarily produced by facilitation, and can be a by-product of niche partitioning, different temporal patterns of growth and development, and nutritional complementation. However, the more frequent occurrence of overyielding than underyielding in natural and man-made communities suggests profitable mechanistic research directions in community theory. The relationship of species mixtures to productivity is highly relevant to the current interest in species diversity and ecosystem function (Tilman et al. 1996, Hector et al. 2002, see Chapter 6.2). For example, Symstad et al. (1998) conducted pot experiments in which 4, 8, or 12 species from 4 different functional groups were combined into assemblages with either all species present or assemblages with random deletions of one species. The effect of losing a single species from the assemblage generally reduced the total biomass of the community, but the effects were highly species-specific. Decreasing diversity by one species had either negative, positive, or neutral effects depending on the species – the “idiosyncratic” hypothesis proposed by Naeem et al. (1994). However, Symstad et al.’s results were not completely idiosyncratic as only nitrogen-fixing legumes elicited positive effects on community biomass. The potential importance of facilitation in the diversity-ecosystem function relationship was more clearly demonstrated in a study by Mulder et al. (2001) of bryophyte communities exposed to short-term drought. They found that productivity increased significantly with the species richness of the community. Mulder and her colleagues argued that an increase in positive interactions in drought conditions, and not in more mesic conditions, among plants drove the relationship between diversity and productivity. A similar argument was made by Caldeira et al. (2001) in a semi-arid climate in Spain.
10
Introduction
To my knowledge there have been no other studies explicitly investigating the potential for facilitation to account for the effects of species diversity on community function, but understanding positive interactions has great potential to shed light on the biodiversity-community function debate. Also, whether idiosyncratic or not, the positive effects of many species on both ecosystem and community attributes suggest a level of interdependence in plant communities that challenges the individualistic status quo. Understanding the degree to which plant communities are individualistic or interdependent is not just an academic problem. These concepts have strong implications for conservation theory and application. For example, the view that plant species are fully individualistic and “interchangeable” in communities has been used to advocate active human involvement in “shaping and synthesizing new ecosystems, even in the ‘natural’ environment.” (emphasis added, Johnson and Mayeux 1992). If maintaining functional plant communities is simply a matter of finding a suite of species that can form a stable individualistic competitive hierarchy, then Johnson and Mayeux’ ideas may not be so farfetched. However, if interactions among plants are more complex and interdependent, as suggested by research on facilitation and the indirect effects of herbivores and mycorrhizae (Callaway et al. 1999, Marler et al. 1999), networks of direct and indirect interactions within the plant community (Miller 1994, Pennings and Callaway 1996, Takahashi 1997; Callaway and Pennings 1998, Levine 1999), and novel interactions among exotic invasive plants (Callaway and Aschehoug 2000, Callaway and Ridenour 2004) such shaping and synthesizing will lead to unforeseen and disastrous results. Interestingly, conservationists often assume a high degree of interdependence in communities when they argue for the preservation of natural systems and biological diversity (Freedman 1989 Erlich 1990, Erlich and Wilson 1991, Miller 1993, Noss 1994). The Ecological Society of America, in an assessment of the use of science in achieving the goals of the Endangered Species Act (Carroll et al. 1996) recommended consideration of the following priorities: “does the species play an especially important role in the ecosystem in which it lives? Do other species depend on it for their survival? Will its loss substantially alter the functioning of the ecosystem?” If applied to plants, these priorities assume interdependence in communities. Mechanistically, negative and positive interactions can be quite different. Negative direct interactions among plants appear to depend mainly on the effects of plants on common limited resources and the responses of plants to these same resources (Goldberg 1990, Miller and Travis 1996) and the biochemical effects of neighbors on each other (Williams 1990, Mahall and Callaway 1992, Inderjit and Del Moral 1997, Wardle et al. 1998, Bais et al. 2003, Vivanco 2004). Direct positive interactions incorporate a wider
Chapter 1
11
range of different mechanisms than direct negative interactions (see Chapter 2), but like competition facilitation may occur through resource effects, one species increasing nutrient, water, or light availability to another, or through chemical effects. Facilitation may be also driven by non-resource processes. Most commonly, species that are tolerant to non-resource stress such as cold, heat, wind, salinity, and disturbance buffer these factors on other species. Other non-resource facilitative processes are indirect. In an intraspecific example, Fischer and Matthies (1998) observed that individual Gentianella germanica plants, a rare species restricted to central Europe, produced more seeds per plant when they occurred in large populations than in small populations (40 to 5000 flowering individuals), and that population growth rates of large populations were higher than those of small populations. They also conducted a common garden experiment in which seeds from plants from the different populations were germinated in a greenhouse and transplanted into a common garden. They found that seed number and survival rates were significantly correlated with the size of the source population. Other experiments demonstrated that seed bank size and seed production were important to maintaining population size for G. germanica (Fischer and Matthies 1998). At the scale of individuals it is no surprise that an outcrossing plant needs another of its own kind nearby, but Fischer and Matthies’ results suggest that an interesting positive density-dependent mechanism operates at larger scales. In the field, pollinator limitation appears to decrease individual fitness, and common garden experiments indicate that individuals in larger populations benefit from amelioration of pollinator limitation and maintain higher fitness levels. Others have shown positive correlations between population size and plant fecundity and these correlations have been attributed to pollen limitation (Jennerston 1988, Petanidou et al. 1993, Lamont et al. 1993, Widen 1993) and genetic deterioration (Menges, 1991, Heschel and Paige 1995, Menges and Dolan 1998). Facilitation affects plant community structure and diversity in very different ways than competition. Inherently, competitive interactions limit coexistence among species, and therefore competition-based theory focuses on how species avoid competitive exclusion (Lotka 1932, Gause 1934, Hardin 1960, Hutchinson 1961). Coexistence in a world dominated by competition has been attributed to 1) “niche partitioning” (Parrish and Bazzaz 1976, Cody 1986), 2) variation in the physical environment and subsequent subtle differences in competitive advantages, 3) disturbance that continuously provides patches of competition-free microhabitat and alters competitive hierarchies (McNaughton 1985), 4) heterogeneity in the ratios of limiting resources that alter competitive hierarchies (Tilman 1976, 1985, 1988), 5) the development of local and species-specific resource depletion zones that, under certain conditions, do not strongly affect the resources available to neighbors
12
Introduction
(Huston and DeAngelis 1994, Grace 1995), and 6) spatial structures that suggest niche partitioning (Van der Maarel et al. 1995). In contrast to the suite of theories that attempt to explain species coexistence despite competition for the few resources that are shared by all plants, positive interactions suggest that some interactions among plants expand niches (Chapter 6.1) and directly promote coexistence and community diversity. Positive interactions do not increase community diversity in a haphazard manner. The ways in which plants modify their environments create conditions in which the beneficiaries are likely to be functionally different than their benefactors. Therefore we have legumes facilitating non-nitrogen-fixing grasses, trees facilitating shade tolerant grasses, and woody perennial shrubs and trees facilitating stem-succulent columnar cacti. This fundamentally inherent process of plant-driven environmental modification creates a situation in which something functionally different than the benefactor can thrive is a very important aspect of positive interactions in general. An excellent example of this process occurs in savannas of southern Africa. Acacia nilotica, a tree with very small drought-deciduous leaves, is the predominant species colonizing open grassland (Smith and Goodman 1987). Acacia cannot recruit under conspecifics, but many other broad-leaved evergreen shrubs and trees can. This broad-leaved evergreen functional group apparently would not occur in this environment without the positive effect of Acacias. Acacias are maintained as a dominant species in the system by large-scale disturbance by elephants. Where elephants have been eliminated many areas once dominated by Acacia are now dominated by thickets of the evergreen shrub Euclea divinorum. Positive interactions do not only increase species diversity, they also increase functional diversity. Environmental modification by plants is generally assumed to facilitate the growth or reproduction of other species, or even the replacement of themselves by other species. However, in a review of positive-feedback switches in plant communities, Wilson and Agnew (1992) make a convincing case for processes in which plants, or communities of plants, can also modify their environment in ways that favor themselves. They argue that positive-feedback switches may produce stable mosaics in originally uniform habitat, sharp boundaries between plant communities on environmental gradients, and either accelerate or retard succession. These positive-feedback switches may be particularly important in exotic invasions. In this book, I focus on several fundamental questions about positive interactions in plant communities: Are positive interactions widespread among different biomes and climates? What kinds of mechanisms drive direct and indirect positive interactions? How do positive and competitive interactions function together? Are positive interactions species-specific? How do positive interactions affect community dynamics? What do positive interactions mean for
Chapter 1
13
community theory? By organizing the literature and concepts around these questions I hope to place positive interactions on solid theoretical footing in plant ecology and support a new conceptual paradigm for the nature of plant communities.
CHAPTER 2 DIRECT MECHANISMS FOR FACILITATION
This chapter and the next are reviews of the empirical research that provides evidence for positive interactions among plants. These chapters are organized by mechanism and I emphasize both experimental and correlative studies. The latter approach is crucial for connecting the processes demonstrated in fine-scale experiments to community-scale organization (see Kikvidze et al. 2005), but correlative studies have a hard time distinguishing between biological effects, shared physical microhabitat requirements, or the tendency of large perennials to act as foci for seed deposition. Positive spatial correlations among plant species that have been explored experimentally are generally supported in terms of facilitation, but not always (Moen 1993, Meiners and Gorchov 1998; Choler et al. 2001). Field experiments are the strongest evidence for the existence and importance of interspecific facilitation in plant communities and the mechanisms behind the phenomenon. The mistaken notion that positive interactions are not well demonstrated with field experiments may be largely responsible for perceptions of facilitation as an interesting, but not fundamental organizing process in plant communities. Positive interactions can be direct, simply the effect of one species on another, or positive interactions can be indirect, requiring an intermediate species in order to occur (Strauss 1991, Wooton 1994, Callaway and Pennings 2000). Indirect facilitation, mediated by parasitic plants, fungi, animals, microbes, and other plants within the same trophic level is discussed in Chapter 3. Although direct and indirect mechanisms can be difficult to separate operationally, the purpose of this chapter is to focus on the direct mechanisms that drive positive interactions. In following chapters I address how these mechanisms may interact with each other, establish community structure, and affect community productivity, diversity, and composition. The most common experiment performed to investigate the positive effects of plants on each other is no different that those typically done to test for competition, with the exception that facilitation has been rarely studied in the greenhouse or experimental gardens. Typically, one or more species is removed from a pair or group of species and growth, survival, reproduction, or some metric is acquired for the remaining target plant. Other approaches include manipulating the canopies or root systems of suspected benefactors separately, but without removing the entire plant. Although removal 15
16
Direct Mechanisms for Facilitation
experiments may not confirm the specific active mechanism (e.g. shade vs. canopy throughfall) they provide a good way to distinguish between biotic and microsite effects – a distinction that is difficult, if not impossible to accomplish without experiments. But experimental approaches are not without problems. If removal is not thorough, the experiment may create conditions in which the remaining “beneficiary” species is subjected to greater stress than would be experienced in habitats without the benefactor at all. For example, if removal eliminates positive effects such as shade, but does not substantially reduce root competition (e.g. if regrowth is abundant) the remaining target plant may do much worse than if were just exposed to the full impact of the abiotic environment alone. Therefore, removal would result in overly poor performance of target plants and the experiment would overestimate the importance of facilitation. Spatial associations and experiments dovetail when trying to understand immediate effects versus net effects of interactions over the lifespans of the interacting species. For example, removal and other manipulative experiments may provide insight into processes that may last several years, but without evidence from the longterm spatial patterns that integrate interactions over long time periods, the fundamental ecological importance of the interactions is hard to determine. An excellent example of the benefit of examining the relationships among patterns and processes plant communities was provided by Zaal Kikvidze and colleagues who found correlative links among temperature, precipitation, productivity, experimentally documented plant interactions, spatial pattern, and community richness in alpine communities around the world (Kikvidze et al. 2005). The suggested that the relationship between positive interspecific spatial patterns and increased community richness was due to niche construction by facilitators, which allows for the coexistence of more species than would be possible if niches were not built by some of the species in the community. Long-term effects of facilitative mechanisms that do not disappear with the removal of putative benefactors such as higher soil nutrients or decreased soil density also complicate interpretation of removal experiments. Facilitation is indicated if plants perform significantly worse after the removal of a neighbor. However, if residual facilitative effects make plants perform significantly better after the removal of a neighbor, or if they drive a neutral response, interpretations may be inaccurate. Interpreting improved performance after neighbor removal as competition may be even more problematic. The enhanced performance of the target plant may have more to do with the high-quality conditions left behind by the removed plant than by the elimination of its competitive effects. Considering the strong effects that plants have on the soils they grow in, overestimating the importance of competition is probably common. For example, if soil modification by species A (e.g. increased soil nutrients) produces a strong
Chapter 2
17
positive effect on species B, and competition for water produces a weak competitive effect of A on B, removal of A may result in higher performance by species B due to the residual high nutrient conditions, and be misinterpreted as strong competition. Interpretive problems are magnified by the fact that plants often interact with their neighbors via a number of different, but simultaneous, positive and negative mechanisms (Bertness and Callaway 1994, Callaway and Walker 1997, Chapter 4). Jack Greenlee and I discovered how important spatial associations can be to corroborate experiments in a study of Lesquerella carinata, a rare mustard in Montana (Greenlee and Callaway 1996). Spatially, Lesquerella was highly associated with bunchgrass tussocks, but experiments conducted in a wet year found no evidence for facilitation by bunchgrasses. Instead we found strong evidence for competition. The spatial associations convinced us to conduct another experiment the following year, which turned out to be exceptionally dry. In this year shade from the bunchgrasses had substantial facilitative effects on Lesquerella. The experimental results suggested that immediate effects may be either positive or negative, but the spatial results suggested that net effects are positive. Understanding these important processes would not have been possible without integrated spatial correlations and experiments. Greenlee’s and my results could not have been demonstrated in the greenhouse, and this is probably true for most studies that have demonstrated facilitation. If two plants are grown together in a pot with adequate water, nutrients, and light, they are quite likely to compete with each other. This is because most facilitative effects occur because a benefactor ameliorates some harsh aspect of the environment, often while simultaneously competing with their beneficiary (Chapter 4). If there is nothing to ameliorate, all that is left is competition. There are facilitative mechanisms that may become apparent only in greenhouse studies, some microbially mediated effects for example, but for the most part greenhouses are bad places to study facilitation. Other experimental approaches include separation of abiotic microsite effects and biotic facilitation with combinations of removal experiments, controls for treatment effects, and nurse plant “mimic” experiments. The latter are experimental manipulations in which nurse plant characteristics are simulated by constructing structural mimics that provide comparable levels of shade or protection from herbivores, but not long-term substrate effects. Comparison of the performance of beneficiary plants with and without benefactor mimics can provide good evidence for the importance of nurse plants and the mechanisms by which they may aid their neighbors. The understanding gained by documenting the mechanisms that drive facilitation, or any biological interaction for that matter, is not trivial. If mechanisms for facilitation are few and biologically simple, then facilitation is
18
Direct Mechanisms for Facilitation
less likely to be important in a wide variety of ecosystems and climatic conditions and inanimate objects such as rocks, stumps, or shade cloth may elicit the same effects as living plants (see Chapter 5). As we shall see, however, facilitative mechanisms are complex. Furthermore, if the specific facilitative mechanisms described below change substantially in intensity and importance along environmental gradients (see Chapter 4), the potential for highly variable hierarchies of advantages and disadvantages for different species in a community will be even greater. Species diversity and coexistence in plant communities may depend on such variation in competitive and facilitative hierarchies.
2.1. WATER RELATIONS: HYDRAULIC LIFT One of the first published reports of positive spatial associations among plant species was written by Phillips (1909), who found that seedlings of Pinus monophylla (pinyon pine) were found often under Artemisia tridentata (Great Basin sagebrush) and rarely in the open. This “nurse plant” spatial pattern was later described by others for P. monophylla and the closely related species P. edulis (Drivas and Everett 1988, Everett et al. 1986, Welden et al. 1990, Callaway et al. 1996, Sthultz et al. 2006). In woodlands of New Mexico Martens et al. (1997) found that young Pinus edulis and young Juniperus monosperma were highly associated with adults of different shrub species. Almost 90 years after Phillips’ observations I conducted removal and transplant experiments with colleagues at the University of Illinois and University of Nevada, Reno (Callaway et al. 1996) to study the nature of the relationship between these species. Confirming the claims of others, we found that A. tridentata shrubs significantly improved the survival rates of P. monophylla seedlings in comparison to open inter-shrub spaces and plots where A. tridentata had been removed. Shrubs provided indirect facilitation by reducing herbivory, but also directly reduced mortality due to ameliorating desiccation and heat stress. Shrubs may simply shade P. monophylla seedlings during the hottest and driest times of the year, but other, more complex facilitative mechanisms appear to be involved. Not long after Phillips’ observations, Magistad and Breazeale (1929) hypothesized that deep-rooted plant species might extract water from far below the surface and lose a portion of this water into dry soils at the surface. Five
Chapter 2
19
Figure 2.1. Deuterium-labeling experiment designed to show if water absorbed by the deep roots of Artemisia tridentata (left) would appear in the stem water of neighboring Agropyron desertorum grasses. Intact deep roots of A. tridentata were immersed in vials of D2O, and DHO content of stem water in the grasses was measured. The dashed line shows the presumed pathway of deuterium. Reprinted from Caldwell et al. (1990) with permission from the Israel Journal of Botany.
decades later Harold Mooney observed that surface soils under Prosopis tamarugo in the Atacama Desert were relatively moist, despite the almost total lack of precipitation, and hypothesized that the moisture might come from the roots of the Prosopis itself (Mooney et al. 1980). Since then, the redistribution of soil water through root systems has been shown for a large number of species in a wide variety of conditions (Caldwell et al. 1998) including relatively shallow rooted species (Wan et al. 1993). The phenomenon, christened “hydraulic lift” was first clearly documented for Artemisia tridentata when Jim Richards and Martin Caldwell recorded substantial diurnal cycles in shallow soils under Artemisia tridentata and hypothesized that the shrub was transporting water from deep, moist soils to dry surface soils during the night (Richards and Caldwell 1987; also see Chapter 5.2). In a later experiment, light provided throughout the
20
Direct Mechanisms for Facilitation
night significantly reduced the diurnal cycles, presumably because open stomata maintained a steep gradient of water potential to the atmosphere (Williams et al. 1993). Hydraulic lift is a passive process, in which nocturnal stomatal closure establishes a water potential gradient running from high-water potential deep soils, along the gradually decreasing gradient within the plant, and then to the low water potential endpoint in the dry surface soils. When stomata open, the low water potential endpoint is re-established in the atmosphere. By caching water in shallow soils during the night Artemisia plants can increase their total daily transpiration rates. Facilitation occurs when cached water at the surface becomes available to other species. Cached water may play a role in the nursing of Pinus monophylla, but movement of water through Artemisia via hydraulic lift to P. monophylla has not been demonstrated. However, Caldwell and Richards (1989) and Caldwell (1990) demonstrated that deuterated water absorbed by the deep roots of Artemisia appeared 11 hours after application in stems of neighboring Agropyron desertorum tussocks (Figure 2.1). However, the amount of water transferred from Artemisia to Agropyron was small, suggesting that any positive effects caused by this facilitative mechanism were probably not strong. Thus this form of facilitation may not be important relative to the intense competition that has been shown to occur between these species (Caldwell 1990). In another effort to understand hydraulic lift as a facilitative mechanism, Todd Dawson (1993, also see Brooks et al. 2002, Chapter 5) used stable isotopes to investigate the magnitude of the water lifted by Acer saccharum (sugar maple) in northeastern forests of the United States and the effects of the hydraulically lifted water on understory plants. He quantified hydraulically lifted water in the xylem of all understory plant species examined, and found that the proportional use of hydraulically lifted water by understory species ranged from 3 to 60%. Within a species, individual plants that that used large proportions of hydraulically lifted water had more favorable water potentials, conductances and growth than those that did not. For 12 understory species that varied in morphology from herbs to woody perennials, being close to hydraulically-lifting A. saccharum (0.5 m versus 5.0 m) resulted in 2-4 times higher rates of conductance and much higher water potentials. Comparison of the amount of hydraulically lifted water in understory plant xylem and the amount calculated for soil water budgets suggest that understory plants may preferentially take up lifted water due to its higher matric potential, or that the roots of understory plants are spatially associated with the roots of A. saccharum (Emerman and Dawson 1996). The positive effects of A. saccharum on the species that grow underneath them also depended on the size of the A. saccharum. Only trees larger than 10-cm dbh used large amounts of ground water, a prerequisite for hydraulic lift (Dawson 1996). Trees that were smaller than 10-cm in diameter
Chapter 2
21
used surface soil water exclusively and did not lift. Mature A. saccharum may hydraulically lift 100 L (25% of their entire daily water budget) (Emerman and Dawson 1996), giving them the potential to have large-scale effects on ecosystem processes (Dawson 1996) and function as “ecosystem engineers” (Jones et al. 1994, 1997). This potential for large-scale ecosystem engineering through hydraulic lift has recently been emphasized in a study of evergreen tropical trees in the Amazon rainforest. In a fascinating scaling up of hydraulic lifting by individual tropical trees (see Oliviera et al. 2005), Lee et al. (2005) used an atmospheric general circulation model to estimate that large numbers of hydraulically lifting trees could have strong effects on climate in the Amazon region. A widely described positive effect of Quercus douglasii canopies on understory grass productivity has been attributed primarily to the way that canopy throughfall and litterfall increase soil nutrients near the trees (Holland and Morton 1980, Holland 1980, Callaway et al. 1991, see Chapter 4). However, Ishikawa and Bledsoe (2000) observed gradual increases in soil water potential at night and rapid decreases during the day in soils under Q. douglasii trees. These diurnal fluctuations in water potential are indicative of hydraulic lift. Hydraulic lift is discernable only in relatively dry soils; when soils are wet their high water potential, relative to that in shallow tree roots, does not allow water to passively move into the soil. For example, Dawson’s findings were reported from an unusually dry year, and Ishikawa and Bledsoe found that diurnal hydraulic lift patterns developed a month earlier in a dry year than in a wet year. But it is unlikely that hydraulic lift plays a role in the general facilitative effect of Q. douglasii trees because diurnal patterns in soil moisture do not develop until later in the summer, and by this time the annual grasses that are facilitated by trees are dead. Alternatively, water released by Q. douglasii roots could delay the rate of soil water depletion and increase the growing season for the annual understory species or affect other soil processes that ultimately benefit understory grasses. Hydraulic lift has now been reported in the literature for at least 27 species (unofficially over 59, personal communication, T. Dawson) and the process occurs in a diverse number of biomes including shrub steppe, savannas, temperate forests, and tropical forests (Caldwell et al. 1998). The ecosystem effects of hydraulic lift may be large (Horton and Hart 1998, Lee et al. 2005), and clearly hydraulic lift may have broad importance as a facilitative mechanism. Although the facilitative effects of hydraulic lift have been emphasized primarily in the context of perennial trees and shrubs benefiting herbaceous understory beneficiaries, this mechanism may also be important for the survival and growth of seedlings of perennials growing under nurse plants, such as the pinyon pine seedlings that are found so often under the canopies of Artemisia.
22
Direct Mechanisms for Facilitation
A form of hydraulic lift, or “water transfer”, has also been described between Zea mays (corn) and Medicago sativa (alfalfa). Corak et al. (1987) grew these species in an experimental apparatus designed to examine the transfer of water from Medicago to Zea. A single Medicago plant was grown so that its roots connected with a Zea plant through two tubes separated by a 5 cm air gap, bridged only by the Medicago roots. The Zea roots were unable to cross the air gap between the tubes. When both the top tube and the bottom tubes were watered frequently, water potentials of Zea remained ≈-0.20 MPa for the duration of the 50 day experiment. When only the bottom tube was watered, the water potentials of Zea in the top tube decreased to ≈-4.0 MPa after 50 days. However, when Medicago was present, and its roots bridged the gap between the tubes, water potentials of Zea were above -2.0 MPa. Furthermore, high levels of labeled tritium supplied only to Medicago roots were detected in Zea tissue. These results indicate that water transferred from moist soil to dry soil by Medicago facilitated the survival of Zea in otherwise lethal drought conditions. Although hydraulic lift can have strong facilitative effects in many different systems, it should be noted that the very trees that lift can also have strong competitive effects on understory species for water, to the point that the competition overwhelms facilitation (Ludwig et al 2004, see Chapter 4).
2.2. WATER RELATIONS: CANOPY INTERCEPTION Many studies have shown that water input or soil moisture is higher around tree canopies or in forest stands where moisture from the air is intercepted and condensed (Vogelmann et al. 1968; Azevedo and Morgan 1974; Ingwersen 1985; Schemenauer et al. 1988; Huntley et al. 1997, Rigg et al. 2002). Furthermore, when canopy trees are removed the water input from fog drip and stream flow declines (Ingwersen 1985). This circumstantial evidence suggests that canopies can intercept and condense water from air may facilitate neighbors and create more mesic habitats. In an elegant study of Sequoia sempervirens (coastal redwood), perhaps the world’s greatest fog collector, Todd Dawson combined extensive sampling of fog and rainwater input with isotope analyses to quantify the amount of water collected by Sequoia trees and the acquiring of Sequoia-collected fog water by its smaller neighbors. Dawson found that forested sites received more total water inputs than nonforested sites and on average 35% of the total forest water inputs was due to fog drip from the Sequoia trees. In nonforested sites fog accounted for only 17% of total input. Isotopic analyses indicated that the average Sequoia obtained 19% of its annual water input from fog,
Chapter 2
23
with smaller trees receiving almost 40% of their annual budget from fog. Understory plants benefited from the water intercepted by Sequoia trees, with shallow rooted herbs acquiring up to 100% of their moisture from fog inputs in dry years. While smaller species appear to collect a little fog themselves, the presence of Sequoia, a far better fog collector, provided understory species with much greater water inputs. Dawson noted that “loss of the canopy tree S. sempervirens is not only a loss of biomass and the nutrients contained within it, but will lead to a loss of the diverse canopy ‘community’ … as well as the organic-rich forest soils to post-disturbance erosion. Tree loss will also convert a once moist, cool, closed ecosystem into a more drought prone, warmer, open ecosystem. Plants and animals which depend upon the moisture input from fog drip or other microclimatic benefits caused by the presence of fog will experience more frequent water stress when S. sempervirens is removed. In addition, both S. sempervirens seedlings and understory plant species which require moist and cool conditions to regenerate could suffer or disappear if inputs of fog decline.” In New Caledonia, Araucaria laubenfelsii, a species that is morphologically similar to S. sempervirens, collects large amounts of water from fog, even on days with no recorded rainfall, and deposits this water beneath the tree canopy (Rigg et al. 2002). They found that A. laubenfelsii facilitated succession to rainforest by reducing stress experienced by late seral species and acting as “nuclei for forest species invasion of the maquis”. Once mature A. laubenfelsii establish nuclei, rain forest develops by expansion from these patches and their coalescence.
2.3. SHADE The benefits of shade include maintenance of plant tissues below lethal or near-lethal temperatures, decreasing respiration costs, lowering transpirational demands by decreasing the vapor pressure difference between leaves and air, reduction of ultraviolet irradiation, and increased soil moisture due to lower evaporative demand. Most plants suffer substantial physiological damage at temperatures between 50 and 60oC because at these temperatures enzymes, cell membranes and thylakoid membranes begin to degrade (Larcher 1995). However, mitochondrial respiration rates increase exponentially with temperature, so temperatures much lower than 50oC can have negative effects on the carbon balance of plants. Shade can also reduce the vapor pressure difference between plant tissues and the air surrounding them by increasing ambient humidity (Geiger, 1965, Keeley and Johnson 1977, Larcher 1983), and by decreasing the temperature difference between plant tissues and the air.
24
Direct Mechanisms for Facilitation
Furthermore, for plants that have limited access to water and cannot conduct gasexchange during hot periods, such as succulents or seedlings without fully developed root systems, stomatal closure and limited CO2 uptake may allow light to damage photosystems. Also, for many species of woody plants, and in particular conifers, soil temperatures greater than 45oC can cause seedling mortality by destroying the cambial tissue and phloem (Hartley 1918, Toumey and Neethling 1924, Levitt 1972). In hot climates, subcanopy temperatures are not only cooler than ambient temperatures, but they are also more stable. For example, soil temperatures at 5 cm and 10 cm under an Acacia tortilis tree in the winter ranged from 23-28oC and 23-26oC, respectively. At the same time temperatures at these depths in the open ranged from 25-36oC and 24-34oC. It is important to note; however, that canopy-induced decreases in temperature or other microclimatic conditions do not always appear to drive facilitative effects even though they may be correlated with positive associations among species (Callaway et al. 1991, Gass and Barnes 1998). Many species reach their maximum photosynthetic rates (light saturation) at photosynthetically active radiation (PAR) levels far below the general natural maximum (≈2000 μmol m-2 s-1). These species may benefit from the effects of shade from taller neighbors on leaf temperature and transpirational loss without any cost of decreased carbon gain. For example, Arnica cordifolia is a perennial herb that is commonly found in the understory of conifer forests in the northern Rocky Mountains. At three sites near Missoula, Montana I found the average percent cover of Arnica in stands dominated by Pseudotsuga menzeisii and Pinus ponderosa was 9.2 ± 3.0 versus 0.2 ± 0.7 in nearby open meadow. This pattern may be explained simply by the direct positive effects of moderate shade cast by the conifer canopies on the carbon/water balance of Arnica. Interestingly, Donald Young and Bill Smith found that a 30% decrease in light on the forest floor during cloudy days in the Medicine Bow Mountains of Wyoming resulted in a 37% increase in carbon gain for Arnica and an 84% reduction in transpiration (Young 1983). They found that the photosynthetic rates of Arnica remained near saturation even on very cloudy days. In other words Arnica gained from the lower transpiration rates associated with decreased light levels without an accompanying cost of lower photosynthesis. The potential positive effects of Artemisia on Pinus monophylla seedlings via hydraulic lift described above may be negligible compared to the effect of shade from Artemisia canopies. For example, in field experiments, artificial shrub mimics had strong positive effects on P. monophylla (Callaway et al. 1996). Since shrub mimics do not have roots facilitation via hydraulic lift is not possible. Shade may interact with water availability as a positive effect, but shade from overstory species can also have facilitative effects that are completely independent of effects on water
Chapter 2
25
relations. However, the effects of shade on understory plants are complex, and despite the win-win situation described above for Arnica the positive effects of shade on water relations and temperature may be counteracted by the negative effects of shade on light availability for many understory species. Interactions between the effects of shade on light as a resource, and temperature and transpiration rates as a stress, are unavoidable and must taken into account in any study of shade as a competitive or facilitative mechanism. Such interactions are likely to create the “seed-seedling” conflicts described in Chapter 4 (see Schupp 1995). Milena Holmgren and colleagues (1997) took these potential tradeoffs into account by modeling the relative importance of shade from a nurse plant as a facilitative effect or a competitive effect depending on the synergy between light and moisture in a plant’s environment. Assuming that plant growth is a relatively simple product of light and moisture, and that soil moisture is affected by canopies to a greater degree in xeric environments than in mesic environments, they argued that plant growth will increase as light increases. However, in xeric environments increasing light is correlated with rapidly decreasing soil moisture and plant growth decreases at higher light intensities. This sets the stage for shade-derived facilitation in water-limited environments, and decreased likelihood for such facilitation in mesic environments (Figure 2.2, also see Chapter 4). For example, in mesic successions after glacial retreat in southeastern Alaska, Alnus sinuata facilitates the later successional species, Picea sitchensis, via soil amelioration, but its shade effects are inhibitory (Chapin et al. 1994). In mesic forests in New Zealand, Bellingham et al. (2001) found that shade effects were consistently negative for tree seedlings, but litter and mineral soils from late seres were facilitative for two late successional species. In an experiment designed to test Holmgren and colleagues’ hypotheses about shade, soil moisture, and plant interactions, Davis et al. (1999) conducted a field experiment in which seedlings of Quercus macrocarpa (bur oak) and Q. elipsoidialis (northern pin oak) were grown in different levels of shade, nitrogen, water, and neighbors. They found that photosynthetic rates of Quercus seedlings increased with grass neighbors in dry conditions, but tended to decline in the shade in wet conditions, supporting the general model of Holmgren et al. (1997).
Direct Mechanisms for Facilitation
Growth
26
MESIC HABITAT
XERIC HABITAT
Canopy
Light Open
Figure 2.2. Intersection of the hypothesized correlation between microsite light and moisture with seedling growth as a function of those factors. This shows the potential for decreasing growth with increasing light in xeric conditions, but not in mesic conditions. Redrawn from Holmgren et al. (1997) with permission from Ecology.
Foreshadowing Holmgren et al.’s argument by over 40 years, Shirley (1945) conducted an experiment investigating the interactive effects of hardwood roots and canopies on the survival and growth of conifer seedlings and saplings in the forests of Minnesota. He found that under dry conditions, moderate shade improved conifer survival, although the overall effect of hardwoods on conifers under most conditions was negative. These experiments indicated that overstory facilitative effects occurred, but that interference was usually more important. More recently, Thomas et al. (1999) conducted similar overstory thinning experiments, but in mesic Pseudotsuga menziesii forests in the northwest US. They found that 85% of the species in their plots were released from competition and only four of the 54 most common species showed evidence for facilitation. In more arid systems, as might be predicted from Holmgren et al.’s (1997) model, the positive effects of the shade may be more common. In another test of her theory Holmgren (2000) examined responses of Liriodendron tulipifera, a common tree in mesic communities of the temperate
Chapter 2
27
deciduous forest of the eastern USA, along gradients of PAR and water availability in greenhouse experiments. She found no evidence that increasing PAR compensated for the negative effects of drought on photosynthesis and growth; instead as predicted growth rates decreased with increasing PAR under dry conditions. Specifically, estimates of daily carbon gain with plentiful water increased in high light (27% of ambient) to 686±84 μmol g-1d-1 from 161±86 μmol g-1d-1 in low light, but in dry treatments carbon gain decreased in high light to -12±12 μmol g-1d-1 from 198±34 μmol g-1d-1 in low light. Growth decreases in high light and dry soil were associated with strong decreases in sapflow. Interestingly, Liriodendron is not particularly shade tolerant, nor is it unusually adapted to drought, even in comparison to other tree species with which it occurs. Holmgren made the case that if plants do not show a tradeoff between shade and drought tolerance in general, but do experience amelioration of drought in the shade, facilitative effects of shade should be common where drought limits the establishment of plants. Drought limits the establishment of plants in deserts, and facilitation in deserts appears to have created unusually strong patterns of spatial associations and many reports of facilitation (Flores and Jurado 2003). Deserts provide striking examples of shade as a facilitative mechanism in plant communities, and strong positive spatial associations in arid environments have attracted ecologists for decades (see Agnew 1997, Robinson 2004). As early as 1910, Forrest Shreve noticed that young saguaro cacti were usually found beneath shrubs or trees (Shreve 1910, 1917). In 1929, Compton reported strong positive associations among several species of shrubs in the Karoo Desert of South Africa and argued that stress-tolerant species provided shelter for less tolerant species. In more recent studies in the Karoo of southern Africa, over 60% of pair-wise spatial comparisons suggested positive interactions (Eccles et al. 1999). In the Sonoran Desert, Shreve (1931) observed that saguaro cacti (Carnegia gigantea) seedlings were common under shrubs and trees, but not in open spaces. This particular association between saguaro seedlings and other desert perennials has been studied intensively and coined the “nurse plant syndrome” in some of the earliest detailed studies of positive interactions in plant communities (Niering et al. 1963, Turner et al. 1966, 1969, Steenberg and Lowe 1969, 1977). Saguaro seedlings are sheltered by many different species of perennial plants, but predominantly by Cercidium microphyllum (paloverde tree). Turner and colleagues (Turner et al. 1966, 1969) investigated mechanisms behind the associations between young saguaro cacti and nurse trees and shrubs by measuring spatial associations and experimentally transplanting 5-cm and 15-cm tall saguaros in factorial treatments of shade and caging for protection from herbivores. They found that predation on seedlings was very high, but that the survival of seedlings that were transplanted in the shade of nurse plants was much higher than survival in the open.
28
Direct Mechanisms for Facilitation
All non-shaded saguaro seedlings died, in comparison to a 65% mortality rate of shaded seedlings. In these experiments canopy shade appeared to interact with soil albedo created by litter from different plant species (see species-specific effects in Chapter 5); for example more saguaro seedlings died on darker soils from under Prosopis than on light soils from under Cercidium (Turner et al. 1966). The strong effect of Cercidium nurse trees on saguaros and the strong eventual competitive effect of saguaros on their nurse trees have been hypothesized to drive some aspects of community dynamics in the Sonoran desert (Vandermeer 1980, McAuliffe 1988). Vandermeer suggested that higher mortality rates of Cercidium with saguaros under their canopies combined with the positive effect of Cercidium on saguaro populations could establish dramatic fluctuations in population numbers of the cactus (Steenberg and Lowe 1977) determined only by the consequences of interacting populations, Alfonso Valiente-Banuet and Exequiel Ezcurra (1991) experimentally compared the relative importance of protection from predators and shade from neighbors for Neobuxbaumia tetetzo, a columnar cactus in the Viscaino Desert and the Gran Desierto de Altar in Mexico that is commonly found under Mimosa luisana. They found that cages improved survival, but that long-term survival only occurred in shade treatments. In the Sonoran Desert, three species of cacti, Lophocereus schottii, Mammillaria thornberi and Peniocereus striatus, are highly associated with the dominant shrub Olneya testota (Humberto, et al. 1996). The spatial relationships between these species and Olneya are much stronger than with other potential nurse plant species. These differences were attributed to the observation that Olneya reduced subcanopy temperatures by 15 oC, which was 3-5oC more than any other species tested. Furthermore, the spatial relationships between cacti species and Olneya were stronger on southfacing exposures than on north-facing exposures, which also emphasize the importance of shade in more stressful conditions (Humberto et al. 1996). Using matrix models based on field data, Godinez-Alvarez et al. (1998) showed that survivorship was the most important life-history parameter for the finite rate of increase in Neobuxbaumia populations, underscoring the substantial importance of facilitative interactions on the long-term population dynamics of this species. Large columnar cacti have been the subject of many studies of facilitation, probably because their association with other species is so obvious. Desert cacti have exceptionally high tolerances to temperature, but columnar cacti may be particularly dependent on benefactors because large succulents have exceptionally low surface to volume ratios and cannot dissipate heat as efficiently. Ferocactus covillei and F. wislizenii, two species of barrel cacti from the Sonoran Desert, maintain intact cell membranes and cell viability at 69oC (Smith et al. 1984), but temperatures near the ground in deserts can exceed 75oC. For succulent plants, and especially young plants near the ground surface, the
Chapter 2
29
direct effects of shade on maintaining sublethal temperatures and decreasing respiration costs are probably the most important; succulents cannot efficiently cool themselves by transpiring because of the scarcity of water. The commonly described dependence of cacti species and other succulents on nurse plants (Yeaton 1978, Vandermeer 1980, McAuliffe 1984a, Yeaton and RomeroManzanares 1986, Parker 1988, 1999, 2000, Franco and Nobel 1989, Parker 1989, Yeaton and Elser 1990, McAuliffe 1991, Valiente-Banuet 1991, ValienteBanuet and Escurra 1991, Valiente-Banuet et al. 1991, Arriaga et al. 1993, Silvertown and Wilson 1994, Suzan et al. 1994, Tewksbury and Petrovich 1994, Humberto et al. 1996, Carrillo-Garcia et al. 2000a, Reyes-Olivas et al. 2002, Flores and Jurado 2003) may be due to inefficient heat dissipation of seedlings with low surface to volume ratios and exposure of these seedlings to exceptionally hot temperatures on the desert floor. For example, young barrel cacti (Ferrocactus acanthodes) experience 11oC decreases in maximum stem surface temperatures in the shade of nurse plants (Nobel 1980b). Franco and Nobel (1989) found that 89% of all saguaro seedlings occurred under Ambrosia deltoides or Cercidium microphyllum shrubs in Organ Pipe National Monument in the Sonoran Desert, but only 29% of Ferocactus acanthodes (barrel cactus) seedlings were found under nurse plants at a second site in the Mojave Desert. Saguaro seedlings can survive apical temperatures up to 66oC, whereas Ferocactus seedlings can tolerate temperatures up to 71oC (Nobel 1984a,b). Ambrosia deltoides and C. microphyllum shrubs reduced surface soil temperature maxima from ≈60oC to ≈47oC and ≈42oC, respectively; whereas another nurse, Hilaria rigida, decreased maximum temperatures from >70oC to ≈55oC. These kinds of species-specific positive relationships are discussed further in Chapter 5. Smaller pad-forming cacti also need nurses. Cody (1993) measured spatial distributions of three different species of Opuntia cacti at three different sites in the Mojave Desert and found that the smallest (and perhaps the youngest) Opuntias were highly associated with various species of nurse plants (also see Yeaton 1978). Spatial patterns suggested that mature Opuntia may in turn serve as a nurse for the seedlings of some of the shrub species. Archer et al. (1988) described similar relationships between Opuntia and Prosopis in southern Texas. Acacia species have generally been noted in the literature for their positive effects on subcanopy fertility because they are nitrogen fixers (see below, Radevanoki and Wickens 1967, Dancette and Poulain 1969, Singh and Lal 1969, Van Auken et al. 1985, Facelli and Brock 2000, Pandey et al. 2000). However, shade from Acacia canopies may also enhance the fitness of understory plants and structure savanna plant communities. Seedlings of the very small-leaved, Acacia nilotica in savannas of southern Africa virtually never occur under conspecific adults, perhaps because small leaves have small boundary
30
Direct Mechanisms for Facilitation
layers and are very good at dissipating heat. In contrast, almost all seedlings of broad-leaved species occur in A. nilotica understories (Smith and Goodman 1987). For example, 100% of the sampled seedlings of the evergreen Euclea divinorum occurred underneath the canopies of A. nilotica. Shade may be crucial to leaf temperature balance of larger and darker leaves in hot arid climates, especially before the young seedlings have established extensive root systems. As for Euclea, the effect of shade on temperature is also important for many other non-succulent species. For example, Shumway (2000) found that soil temperatures beneath Myrica pennsylvanica were almost 10oC cooler than in the open. This difference was associated with higher growth and reproduction and more favorable water relations of understory herbs. Pseudotsuga menziesii seeds remain viable and germinate at much higher rates under shrub canopies in the Argentine steppe where they have been introduced (Caccia and Ballaré 1998). High viability and germination is directly related to cooler temperatures in the shade of other plants. It is important to note that even strong spatial correlations between canopy shade and the presence of particular species can be misleading. For example, Tewksbury et al. (1998) analyzed the spatial distributions of Capsicum annuum, wild chilies, with respect to different nurse species and found that Capsicum had much stronger associations with some shrubs than others. Celtis reticulata (desert hackberry) was by far the most common nurse species in proportion to its abundance, and experimental transplants also showed a preference for Celtis. They also found that the canopies of Celtis attenuated sunlight approximately twice as much as any other species. Even though the exceptionally shady canopies of Celtis appeared to be particularly facilitative for Capsicum, the high degree of specificity between these species was driven by indirect interactions with bird dispersers due to similarities in the color of their fruits (see Chapter 3). So far, I have emphasized the effects of shade on the temperature of understory plants, but the effects of shade are more complex than this. Therefore, some ecologists have attempted to compare the relative importance of shade effects on temperature to that of other factors affecting plant growth. In arid environments, the difference in soil moisture between the subcanopy and the open may be as strong as differences in temperature (Shreve 1931, Abd El Rahman and Batanouny 1965a,b). However, to my knowledge the importance of the direct effects of shade on temperature in cacti-nurse plant relationships have rarely been compared to the effects of shade on soil moisture in experiments. In one of the best comparisons, irrigation of unshaded saguaro seedlings had minor effects on survivorship relative to the effects of shade, suggesting that lower temperatures in the shade were more important than improved water relations (Turner et al. 1966).
Chapter 2
31 2
χ = 18.03 P<0.0004 observed expected
16 12 8 4 0
N
16 12
16
*
12
8
Shrub
4
8 4
0
W
0
16
E
12 N
8 4 0
S
Figure 2.3. Observed and expected azimuths under shrubs where seedlings of Austrocedrus chilensis were located in southern Argentina. Reprinted from Kitzberger et al. (2001) with permission from Ecology.
Strong spatial associations among seedlings of Austrocedrus chilensis and shrubs in the shrub-steppe of Patagonia (Kitzberger et al. 2001) suggest nurse relationships like those between Artemisia tridentata and Pinus monophylla in shrub-steppe in the northern hemisphere (Callaway et al. 1996). As for Pinus monophylla, survival of planted Austrocedrus seedlings was much higher under shrubs than in the open, and naturally occurring tree seedlings were disproportionally more common on the pole-facing cooler sides (south) and east sides of shrubs than the hotter equator-facing sides (north) (Figure 2.3). To separate the direct effects of shade versus the effects of shade on soil moisture, Kitzberger et al. conducted an experiment in which they planted seedlings under shrubs and in the open and then watered some replicates in each environment. The addition of water greatly increased survival in the open, but only shaded seedlings survived through the second summer (Figure 2.4). Corresponding with general relationships between Austrocedrus and nurse plants, and supported by experimental tests on how shade and climate interact, Kitzberger et al. (2001) and Villalba and Veblen (1997) also found that the population age structure of Austrocedrus (studied using dendrochronology) closely tracked decadal-scale climatic variability. Years with greater than average seedling establishment were associated with the cold phase of the El Niño-Southern Oscillation (La Niña) which provides more moisture during the summer. During the wetter La Niña period Austrocedrus appears to be able to establish to some degree even in open habitats. The summers are much warmer and dryer during the El Niño phase and recruitment of Austrocedrus does not occur; either with or without nurse shrubs.
32
Direct Mechanisms for Facilitation 100 Aa Ab
10
Cd
1
Bc
Shade, watered Shade, nonwatered Interspace, watered Interspace, nonwatered
F M A M J J A S O N D J F M Summer 1996
Fall
Winter
Spring
Summer 1997
Figure 2.4. Survivorship curves of cohorts of Austrocedrus chilensis in southern Argentina under different treatments from February 1996 to March 1997. Solid symbols represent shade treatments whereas open symbols represent open interspaces. Triangles represent watered treatments and circles non-watered treatments. Reprinted from Kitzberger et al. (2001) with permission from Ecology.
However, during years that are intermediate in the effects of La Niña and El Niño, which are dry but not excessively so, establishment of Austrocedrus appears to require nurse shrubs. Although shade may be crucial for establishment, such annual variation in long term recruitment suggests that soil moisture may be more important than indicated in short term experiments. Other conifer species also benefit from the shade of nurse shrubs. For example, in the semiarid conditions of southern Spain, Jorge Castro and colleagues found that survival of experimentally planted seedlings of Pinus sylvestris and P. nigra was almost two times higher when grown under Salvia lavandulifoia shrubs than in the open, and nurse shrubs did not reduce growth (Castro et al. 2002). Facilitative shade effects are not restricted to the amelioration of high temperatures. Nurse plant canopies protect young saguaros from freezing temperatures in the winter (Steenberg and Lowe 1977, Nobel 1980a, Humberto et al. 1996). Many cacti species from the southwestern U.S. cannot survive temperatures lower than –6 to –10oC (Nobel 1982), in part because their succulent tissues are susceptible to internal ice formation. Nobel (1980a) measured apical stem temperature of saguaro cacti in four different sites in
Chapter 2
33
southern Arizona. He found that nurse canopies (primarily Cercidium microphyllum) increased apical stem temperatures on smaller individuals by as much as 16oC during the winter. The buffering of cold temperatures decreased with the height of cacti, apparently because taller plants were less affected by the temperatures at the soil surface. Freezing in open microsites may severely limit the recruitment of saguaros (Steenberg and Lowe 1976), and facilitation by shrubs and small trees during early life stages may extend the northern distribution this cactus (Nobel 1980a,b). Buckley et al. (1998) found similar, but larger-scale positive affects of Quercus rubra and Pinus resinosa canopies on tree seedlings in the northern USA. They planted Q. rubra seedlings in forests dominated by either Q. rubra or P. resinosa trees and in treatments where 0, 25%, 75%, or 100% of the overstory trees were left intact. When stands were either clear-cut (no trees left) or thinned to 25%, Q. rubra seedlings damaged by frost during a late spring freeze increased from 0-20% to over 90%. In 1970, Frank Ronco and colleagues at the Rocky Mountain Experiment Station observed that open-grown Picea englemannii seedlings at high elevations were chlorotic whereas seedlings in the shade remained “a normal green”. Furthermore, needles shaded by experimental shingles remained green until their tips grew out from under the shingle, after which chlorosis developed. Ronco then conducted an experiment in which he compared photosynthetic rates of potted seedlings growing in different artificial shade treatments at 3,000 m elevation. He found that P. englemannii seedlings that had been grown in full shade where able to photosynthesize at 2-3 times greater rates, across a range of light intensities, than seedlings that had been grown without shade. This difference was not associated with lower water relations, but was attributed to “solarization, a phenomenon in which photosynthesis is inhibited by high light intensities”. This light induced “photoinhibition” has been studied extensively over the 35 years since Ronco’s work and we know now that the negative effects of high light interact in very complex ways with other factors. Prevention of photoinhibition by shade from neighbors may contribute to facilitation in montane environments. Several studies have shown that recruitment of Abies lasiocarpa is higher near other trees or microhabitats with grass cover (Habeck 1969, Rebertus et al. 1991, Callaway 1998b, Miller and Halpern 1998, Germino et al. 2002) and in some cases these patterns of recruitment appear to create dense tree islands surrounded by herbaceous vegetation. Recruitment of P. englemannii also appears to be facilitated by the mature canopies, but generally to a lower degree (Rebertus et al. 1991). However, in some environments facilitation of P. englemannii appears to be crucial. Noble and Alexander (1977) experimented with seedling recruitment of P. englemanni in the central Rocky Mountains of Colorado and concluded
34
Direct Mechanisms for Facilitation
that successful natural regeneration in openings on south aspects “is not possible”, and shade is “absolutely” essential for survival. However, in this case, desiccation appeared to be the primary source of mortality. Annual precipitation at their site was quite low for subalpine forests (58 cm yr-1), possibly explaining P. englemannii’s need for shade at the south-facing site, but not at sites facing other directions. Shade effects on photoinhibition are also important for Picea glauca, which like P. englemannii, is also often found underneath tree canopies when it is young. A common associate of young P. glauca is Populus tremuloides, a winter deciduous species. In P. tremuloides stands, leaf-off in the spring and autumn allows high light intensities to reach evergreen seedlings and saplings, such as P. glauca, in the understory. Man and Lieffers (1998) transplanted P. glauca seedlings in P. tremuloides understories and in surrounding open grassland-shrubland and measured their photosynthetic responses. They found that photosynthetic parameters were more depressed in open-grown seedlings compared to seedlings in the understory in the spring and autumn when frosts were common at night. However, net photosynthetic rates in the summer were similar even though understory seedlings were growing with two to four times less sunlight. Additionally, shoot and needle growth of open-grown seedlings was significantly less than of seedlings growing in the shade. Freezing air makes the high-light inhibition of the light-saturated rate of photosynthesis more likely. Therefore, the positive effects of canopies on the growth and physiology of P. glauca may not have been simply due to temperature, but to synergistic effects of temperature and light. In high light, low temperatures can amplify photoinhibition; a light-dependent depression of photosynthetic rate that occurs when leaves absorb more light than can be used (Krause 1994). For P. glauca, decreased summer photosynthetic rates were not due to lower stomatal conductance (Man and Lieffers 1998) but instead to the respiratory cost of repair to chloroplast, chlorophyll, enzyme, and electron transport systems that were damaged during the spring. Other studies have shown that the amelioration of cold temperatures can improve photosynthesis by Picea species and other conifers (Germino and Smith 1999). DeLucia and Smith (1987) found that decreases in photosynthetic rate of adult P. englemannii were correlated with the minimum air temperature of the previous night during the summer. Hellmers et al. (1970) found that minimum nighttime temperatures had stronger negative effects on the growth and survival of P. englemannii in greenhouse experiments than low daytime temperatures. Nunez and Bowman (1986) found that the minimum temperature tolerated by target plants increased as a function of stand density for Eucalyptus delegatensis at high elevations, a process that may have a positive effect on understory plants by reducing photoinhibition. At tree line in Australia seedlings of Eucalyptus
Chapter 2
35
pauciflora (snowgum) establish in higher densities and are more vigorous around adult trees and on the southern (more sheltered) sides of trees (Ball et al. 1991). Similar patterns have been observed for Nothofagus solandri timberlines in New Zealand (Ball 1994). Egerton et al. (2000) experimented with vertical shade screens to reduce excess light without reducing minimum air temperatures, and found that seedlings of E. pauciflora at high elevations were less photoinhibited, maintained higher photosynthetic rates, lost less leaf area, and maintained higher leaf area ratios when they were protected from excess light in the winter. These physiological differences were consistent with greater growth rates of sheltered seedlings over the winter. In another experiment with E. pauciflora, seedlings experienced higher levels of photoinhibition in forest clearings than those under canopies, but photoinhibition was not related to significant differences in seedling growth (Blennow et al. 1998). These studies suggest that cold temperature amelioration by shade from nurse plant canopies at tree line may have important facilitative effects beyond simply preventing neighbors from freezing to death. As described above, young cacti and conifer seedlings experience increases in temperature minima under whole-stand and solitary tree canopies because solar radiation is trapped during the night and re-radiated as long-wave irradiation. Leaf temperatures at night are strongly influenced by exposure to the sky, especially on calm nights with low convective heat exchange (Jordan and Smith 1994, 1995). Leaves exposed to the open night sky experience low levels of long-wave irradiation and leaf temperatures may drop to levels as much as 7oC lower than air temperatures beneath canopies (Jordan and Smith 1994). For example, conifer seedlings near timberline and beneath 50% canopy cover receive approximately 50 W m-2 more energy on a clear night than seedlings in the open. At night, these processes resulted in an approximately 4oC increase in mean needle temperature for seedlings with cover. Such temperature amelioration may be crucial at timberline, where many plant species are growing on the edge of existence, and are probably highly affected by “modulatory forces, including the presence of certain taxa” (Korner 1998). Tree establishment and growth at treeline is highly dependent on temperature, either for overall carbon balance, or sufficient production of new cells, or the development and differentiation of functional tissues (Korner 1998). Seedlings may be facilitated by canopies that simultaneously reduce exposure to low temperatures and high light, as emphasized above. However, the lowest temperatures typically occur at night when there is no light, and the highest light levels occur when temperatures are relatively warm. Exposed sites (without canopies for example) experience exceptionally low temperatures at night that are followed by exceptionally high light during the following day. This temporal decoupling of low temperatures and high light in natural
36
Direct Mechanisms for Facilitation
environments has raised some doubt about the ecological significance of canopy shade effects on photoinhibition. How the diurnal sequence of nighttime cold and daytime high light affect conifer seedling physiology was examined in a series of elegantly designed field experiments by Matt Germino and Bill Smith (1999, 2000) and Johnson et al. (2004). They found that Abies lasiocarpa and Picea englemanii seedlings were most common in habitats with ≈40-80% of the overhead area open to the sky, and demonstrated that seedlings in these habitats benefited from the overstory canopies because of ameliorated nighttime temperature and lower light levels the following day (also see Callaway 1998b). Light-saturated photosynthetic rates of newly germinated, potted, Abies lasiocarpa seedlings increased seven-fold in response to increased long-wave irradiation at night (warming), approximately five-fold in response to shading during the day, and eight-fold by a combination of night warming and shading during the day. Reductions in photosynthetic rate were associated with ≈50% increases in “slowly reversible” or “irreversible” long-term photoinhibition. Picea englemanni was much less affected by warming and shading treatments than A. lasiocarpa. These results were supported by measurements on natural seedlings, and demonstrated that typical diurnal patterns of low nighttime temperatures, and high light the following day, severely inhibited the photosynthetic ability of A. lasiocarpa seedlings. The susceptibility of A. lasiocarpa to cold nights coupled with sunny days, and the relative tolerance of P. englemanni to these conditions further indicates how important the positive effects of shade can be in subalpine communities. Many studies have demonstrated the importance of overstory canopies for ameliorating photoinhibition in understory recruits. But many of these understory recruits eventually reach the sunny overstory. Why these species appear to suffer from photoinhibition when they are seedlings, but function normally in full sunlight when adults is not clear. However, the eventual development of high root:shoot ratios by adults may play a large part. Photoinhibition may be minimized by increasing carbon sink capacities of older roots (Arp 1991), and as trees mature their roots may become much larger sinks for fixed carbon. Shunting excess fixed carbon out of the chloroplasts may allow the photosynthetic apparatus of mature trees to continue functioning in full sunlight. Shade from neighbors clearly facilitates many species simply due to the buffering of temperature extremes and through direct alteration of the light regime. However, as noted above for cactus-shrub relationships, shade also reduces soil water loss and plant transpiration. In other words, canopies may “engineer” (Jones et al. 1997) more mesic environments. Mesic environments may be more important to non-succulent plant species than succulents in a broad geographic context (Shreve 1931, Abd El Rahman and Batanoury 1965a,b, Holmgren et al. 1997), but shading appears to be crucial for many different plant
Chapter 2
37
morphologies at the local scale and particularly in the early stages of development and establishment. In northern Europe, severe water stress limits seedling establishment on coastal dunes, and shade from mature plants improves seedling water relations and survival (De Jong and Klinkhamer 1988). They found that seedling survival for Cynoglossum officinale and Cirsium vulgare was highly correlated with tree and shrub cover (Figure 2.5) and moisture in the upper 10 cm of the soil. Experiments in which either water or fertilizer were added to seedlings at different sites found that water was by far the most important factor determining survival of Cynoglossum seedlings. Furthermore, the effect of watering differed substantially depending on the particular habitat. Water had very strong effects in xeric habitats, increasing survival from 0 to 50% for Cirsium. However, in the habitats where thicket canopies shaded the soil and maintained high levels of surface moisture, watering had a much smaller effect. 12
1982
10 8
Established seedlings per flowering plant
6 4 2 0 12 10
Open Scrub Thickets
1983
8 6 4 2 0 12
1984 10 8 6 4 2 0
May 1
November 1
Figure 2.5. Seedling establishment of Cynoglossum officinale in three vegetation types. Redrawn from De Jong and Klinkhammer (1988) with permission from the Journal of Ecology.
38
Direct Mechanisms for Facilitation
In the exceptionally xeric upper zones of California salt marshes, spring ephemeral species are facilitated by the sub-shrub Arthrocnemum subterminale (Callaway 1994). Shade from Arthrocnemum enhances soil moisture, and the ephemerals are highly sensitive to annual differences in the amount and timing of rainfall and the experimental addition of water. Interestingly, similar processes appear to be caused by Arthrocnemum macrostachyum in marshes in Spain (Rubio-Casal et al. 2001, also see Espinar et al. 2002). In other coastal communities in North America Shumway (2000) found that the transpiration rates of two dune herbs, Solidago and Ammophila, were 15 to 20% lower under the canopies of Myrica than in the open. Decreased water losses through transpiration were associated with significantly lower midday water potentials, growth rates, and reproduction of the herbs in the open. Shumway found no differences in soil moisture under versus outside the canopies, suggesting that either beneficiaries used all of the extra soil water or canopy facilitation functioned through direct effects on beneficiary water relations and not through changes in soil water status. Facilitative effects of shade from shrubs are important for the recruitment of oaks in central California savannas and woodlands (Callaway et al. 1991, Callaway and D’Antonio 1991). In field experiments, unshaded cages reduced mortality rates of Quercus douglasii due to predation, but eventually all unshaded seedlings died over a two year period. In contrast 35% of shaded and caged seedlings survived for one year (Callaway et al. 1991). For seedlings without shade, most mortality was due to desiccation. Nearer the coast, Quercus agrifolia seedlings are much more common under shrubs than in the open grassland (Callaway and D’Antonio 1991). We conducted experiments by planting oak seedlings under shrubs and found that 31% of coast live oak seedlings that emerged after being experimentally planted under shrubs survived for more than two years, but no seedlings that emerged in the open survived. Again, mortality was caused by desiccation. However, these effects of shade are conditional, as shrubs and Q. agrifolia seedlings are much more associated on drier south-facing aspects than on north-facing aspects (Callaway and Davis 1998). Once facilitated by shade from nurse shrubs, shade from mature oaks facilitates other species (Owens et al. 1995). For example, Quercus agrifolia seedlings that survive to adulthood provide positive benefits to understory grasses by shading them. Parker and Muller (1982) conducted reciprocal transplant experiments with Bromus diandrus and Pholistoma auritum, two herbaceous species that are much more abundant under oak canopies than in the open grassland. They found that these species were more tolerant to shade than Avena fatua, a species that is generally excluded from habitat beneath Quercus agrifolia. In laboratory experiments B. diandrus had higher relative growth rates
Chapter 2
39
in low light than Avena fatua (Mahall and Parker 1981). These differences in growth rates were not produced by changes in photosynthesis, which was higher per unit leaf area for Avena in both shade and sun treatments. Instead, Bromus maintained higher whole-plant assimilation rates by allocating proportionally more biomass to leaves. In experiments in which whole soil blocks were transplanted between the understories of Quercus agrifolia, adjacent open grassland, and artificial shade treatments, Maranon and Bartolome (1993) found that reduced radiation under the canopies, not soil source, limited the distribution of the species that naturally dominated in the open. Competition from grass species that were dominant in the open inhibited understory species from establishing in the open grassland. Shade from canopies has dramatic effects on understory community composition and productivity in deserts (Patten 1978, Schmida and Whittaker 1981, Holzapfel and Mahall 1999). Shrubs also establish heterogeneous microgradients of radiation and soil temperature within canopies that appear to provide different niches for different species (Moro et al. 1997). Moro and colleagues quantified the positioning of several herbaceous species under the canopy of Retama sphaerocarpa, a leguminous shrub in semi-arid Spain, in response to several gradients including irradiance and temperature. Radiation at the soil level in a central position under the canopy was 60% of that outside, and temperature differences between the shrub understory and open microhabitats exceeded 7°C, and different understory species were associated with particular understory microclimates. It is not particularly surprising that shade from large woody trees and shrubs can facilitate understory herbs and seedlings. However, other studies indicate that morphologically similar herbaceous species can have important facilitative effects on each other via shade (Eckstein 2005). In northern Switzerland, shade from grasses prevents desiccation and frost heave on limestone soils, increasing seedling survival of other species by up to 10 times that of open patches (Ryser 1993). In a study of six species he found that two, Arabis hirsuta and Primula veris, had very low establishment in gaps, and much higher success in microsites with vegetation. In abandoned vineyards of southern France, the biennial herb, Picris hieracioides is an important early to midsuccessional colonizer. Sans et al. (1998, also see Sans et al. 2002) transplanted rosettes of Picris into plots where herbaceous vegetation had been killed with herbicide, and then monitored survival and growth over a seven month growing season. They found that Picris seedlings survived at much higher rates when the herbaceous vegetation matrix was present, and that the facilitative effect was stronger in fields with denser vegetation (Figure 2.6). The mechanism for the positive effect of herbaceous neighbors on Picris was not clear; however, most Picris mortality was correlated with desiccation and extreme temperatures.
40
Direct Mechanisms for Facilitation
Survival (%)
100
neighbors & resources neighbors
80
a
60 no neighbors & resources
40
20
a
b
no neighbors
b 0 0
5
10
15
20
25
30
Weeks after transplanting Figure 2.6. Survivorship curves of Picris hieracioides in the presence of neighbors (solid lines), absence of neighbors (dashed lines) and with or without the addition of resources. Survivorship curves within each field with the same letter are not significantly different based on log-rank tests. Redrawn from Sans et al. (1998) with permission from Oikos.
The facilitative effects of shade within herbaceous communities such as those demonstrated by Ryser (1993) and Sans et al. (1998, 2002) appear to be produced, in part, by effects of the general vegetation matrix on microclimate. With this perspective, Zaal Kikvidze (1996, also see Kikvidze et al. 2006) measured strong bimodality in insolation due to variable and dynamic cloud cover in subalpine meadow communities in the Caucasus Mountains. He conducted removal experiments in different densities of vegetation, which was produced by different grazing intensities. In the most productive and dense plant communities (low grazing) he found that leaf temperature increased for plants without neighbors in sunny weather by an average of 3oC (-0.8 to 5.5oC depending on the target species). In heavily grazed communities there was no effect of removing neighbors on leaf temperature. Kikvidze then measured photosynthetic rates of Trifolium ambiguum either in intact swards of vegetation or in clearings throughout the day as clouds passed over and the sky cleared. He found that plants with intact neighbors increased their photosynthetic rates by 50% in sunny conditions, but when clouds appeared plants with neighbors decreased photosynthetic rates by 26%. As described above, many sun-loving species light saturate well below full sunlight, so decreasing PAR from ≈2000
Chapter 2
41
μmol m-2sec-1 to ≈1200 μmol m-2sec-1 probably did not light-limit T. ambiguum. However, if leaves are heated much above ambient air temperatures, leaf-to-air vapor pressure differences increase dramatically and may cause stomata to close in order to avoid a decrease in water-use-efficiency. Whole sward boundary layer effects may also promote stable and higher humidity near the plants, increasing water-use-efficiency and stimulating higher photosynthetic rates. For example, Antheleme et al. (in press) found that the highly drought-tolerant tussock-grass, Panicum turgidum, strongly diminished the vapor pressure deficit (VPD) within its canopies, and facilitated other species. Similarly, Caldeira et al. (2001) found that found target species in experimental plots with higher plant species richness and cover had carbon isotope ratios indicative of higher water-use-efficiencies. As found by Kikvidze, this was probably due to the overall effect of enhanced community boundary layer, retention of more humid air around the leaves of plants, and greater shading within the community. Similar facilitative processes have been demonstrated in other herbaceous communities. At the Cedar Creek Experimental Station in Minnesota, Wilson and Tilman (1995) found that Chenopodium album grew faster in the presence of other herbaceous neighbors than when neighbors were removed. But of eight species tested, only C. album showed this positive response to neighbors. Hillier (1990) found facilitative effects among herbaceous species in communities on dry calcareous slopes in Great Britain. Jack Greenlee and I (1996) found that the rare herbaceous mustard, Lesquerella carinata, was highly associated with bunchgrasses in dry montane grassland in western Montana. In experiments that included bunchgrass removal and the construction of artificial shade, we found that shade facilitated Lesquerella in an unusually dry year and at a dry site. David Wardle et al. (1999) conducted a large-scale experiment in perennial grasslands in New Zealand, continuously removing subsets of functional groups or species over a three-year period and monitoring the responses of other species. Most responses indicated competition; however, Trifolium repens responded to the removal of Lolium perenne in a much more complicated fashion. During the winter Trifolium was strongly stimulated by Lolium removal, indicating competition, but in summer when conditions were dryer, removal of Lolium strongly inhibited Trifolium, indicating facilitation. Other experiments have shown that Lolium tends to be less competitive with Trifolium in drier environments (Thomas 1984). Much like plants in temperate or semiarid grasslands of similar stature facilitate each other through community-scale shading, temperature reduction, and increasing ambient humidity, such community-scale shading processes may function in arctic and alpine communities to maintain temperatures above that of the ambient air. The shoot morphology of many alpine and arctic species allows
42
Direct Mechanisms for Facilitation
them to trap heat and thus modify their environment. Compact cushions and rosettes hug the ground and possess dense schlerophyllous leaves which increase temperatures at the leaf surface. For example, Korner (1999) found that temperatures within cushion plants may be 20oC higher than the ambient air. In alpine vegetation in Patagonian Chile, Arroyo et al. (2003) found that at 700 m in elevation the cushion plant, Azorella monantha, increased temperatures at the leaf surface approximately 5oC above ambient air temperature and at 900 m the leaf-air temperature difference was 8oC. They also found that temperatures of the leaves of the cushion plant, Diapensia lapponica, can be 3-5oC above ambient on cloudy days and 12-15oC higher on sunny days. Other species can benefit from the warmth of these cushions. It is difficult to separate the effects of temperature from the effects of other harsh environmental characteristics at high elevations such as wind (see below), but many studies have shown that large proportions of species are spatially associated with cushions (see below, Whitehead 1951, Griggs 1956, Bonde 1968, Blundon 1983, Sohlberg and Bliss 1984, Alliende and Hoffman 1985, Kikvidze 1993, 1996, Kikvidze and Nakhutsrishvili 1998, Anderson and Bliss 1998, Pokatzhevkaya 1998, Nunez 1999, Arroyo et al. 2003, Cleavitt 2004, Caveries et al. 2006), and several have shown strong facilitative effects experimentally (Choler et al. 2001, Callaway et al. 2002, Klanderud and Totland 2005). Experiments conducted by Maranon and Bartolome (1993, (described above) suggest an alternative interpretation for some experiments in which shade has been assumed to play a predominately direct role. Shade from a canopy dominant may have a positive effect on understory species, or it may suppress superior competitors and indirectly facilitate other species by providing them relatively competition-free habitat (Chapter 3 addresses such indirect facilitative effects). Seedling establishment of oaks (Quercus velutina and Q. alba) is inhibited by dense herbaceous vegetation (Harrison and Werner 1984) and Werner and Harbeck (1982) found that the highest densities of oak seedlings occurred in patches of low vegetation and with the canopies of staghorn sumac shrubs. Similar effects have been described for junipers, which provide openings for oak seedlings by suppressing herbaceous vegetation (Bard 1952). Petranka and McPherson (1979) found that drought-intolerant tree species were not able to invade tall grass prairie vegetation without the prior invasion of the shrub Rhus copallina, and that the colonization of prairie by more drought-tolerant tree species was enhanced by the shrub (Figure 2.7). Invasion by Rhus increased levels of soil moisture and nutrients, but apparently interfered with the growth of prairie grasses by allelopathic suppression and decreasing available light.
Chapter 2
43
Tree seedling density (#/m 2)
A
Inside Rhus shrubs In open near Rhus shrubs
0.4
0.3
0.2
0.1
0.0
0
0
0
B
0.4
0.3
0.2
0.1 0 U am lmus eri can a J un i p vir gin erus ia n a
0 Ce spe ltis c ie s
Q ma uerc rila us nd ic a
0 Qu e ste rcus lla t a
0.0
Figure 2.7. Densities of tree seedlings within and near clones of Rhus copallina at the prairie-forest ecotone in Oklahoma. A) Near the forest border. B) Far into the prairie. Redrawn from Petranka and McPherson (1979) with permission from Ecology.
Shade from plants may elicit positive responses from neighboring species even in exceptionally wet environments. Spartina alterniflora (cordgrass) is very closely associated with algal species, including the brown alga, Ascophyllum nodosum, in marsh plant communities on much of the eastern coastline of North America (Chock and Mathieson 1976, Gerard 1999, Chapman and Chapman 1999). On Long Island, New York, removal of Spartina caused Ascophyllum to grow slower and Ascophyllum without Spartina neighbors experienced greater dehydration (Gerard 1999). Farther north on the coast of Nova Scotia, however, removal of Spartina in a similar experiment produced the opposite results. Removal of Spartina did not have significant effects on the growth of Ascophyllum, but reduced dehydration (Chapman and Chapman 1999).
44
Direct Mechanisms for Facilitation
They proposed that Spartina was important for the stabilization of substrate throughout its range, hence potentially explaining the consistently strong association between Spartina and Ascophyllum, but that the facilitating effect of substrate stabilization operated at a far larger scale than that at which they conducted their experiment. Tree canopies may also have very large scale effects on climate, and apparently increase regional temperatures. Gordon Bonan (1999) used maps of predicted natural vegetation and true vegetation maps to develop climate simulations for the midwestern United States. He found that the removal of trees may have contributed to temperature decreases of as much as 2.5oC in the region. Such effects may not be readily discernable in experiments conducted by most ecologists, but large-scale environmental modifications, or “ecosystem engineering” (Jones et al. 1997) should not be dismissed as potential facilitative processes. Subsumed within the large-scale effects of shade in forests, very small scale facilitative shade effects may be caused by species-specific patterns in light transmission through canopies (Canham et al. 1994, Seiwa 1998) and phenological niches caused by seasonality in light availability (Mahall and Bormann 1978). For example, Canham et al. (1994) measured the light transmission through canopies of nine deciduous and evergreen tree species in forests of southern New England. They found that light extinction, as light passed through canopies, differed among species with shade-tolerant species casting the deepest shade (1% of full sunlight, ≈20 μmol m-2 sec-1) and early successional species casting the least shade (6%, ≈120 μmol m-2 sec-1). Furthermore, sunflecks contributed little of the transmitted light under the shadiest species, but 40-50% of the transmitted light under the less shady species. Species-specific variation in subcanopy light regimes correlated with models of predicted sapling mortality under the different canopies. Shade intolerant species in general transmit more light to their understory than shade tolerant species (Messier et al. 1998), because the latter have a more efficient leaf display (Bond et al. 1999). Species-specificity in canopy effects is not only determined by canopy structure. Seiwa (1998) found that the seasonal growth of Acer mono seedlings in temperate deciduous forests in Japan was greatly affected by the timing of leaf fall and flush of different overstory species. Growth rates of experimentally planted seedlings were higher under trees that flushed leaves later in the spring and dropped them earlier in the fall (Ostrya japonica, Ulmus davidiana, and Juglans ailanthifolia), than under trees with early flushes and late drops (Quercus mongolica). The strongest differences in species-specific understory community composition have been observed when tree species differ morphologically in
Chapter 2
45
ways that affect shade, such as evergreen versus deciduous, but many studies have described specific canopy-understory species associations among species with similar morphologies (Veblen et al. 1979, Hicks 1980, Sydes and Grime 1981, Beatty 1984, Boettcher and Kalisz 1990, Dzwonko and Loster 1997, Pabst and Spies 1997). These patterns are the basis for predictive transition matrix models of forest dynamics (Horn 1975, Woods and Whittaker 1981). Frelich et al. (1993) described “neighborhood effects” in temperate forests in which several overstory species appeared to have strong effects on the composition of the understory community. Models based on these data indicate that these positive neighborhood effects may also augment small differences in the environment and result in large mono-specific patches (Frelich et al. 1998). In models including the possibility of such neighborhood effects, groups of species were initiated in a random pattern and then allowed to interact in positive and negative ways. These modeled species formed different and distinct communities.
2.4. WATER RELATIONS: SOIL MOISTURE In the section above, I focused on how shade may directly benefit plants by reducing leaf temperatures, reducing transpiration, and limiting light damage, and introduced the potential for how shade creates a more mesic environment. I also discussed how hydraulic lift can contribute to the effects of overstory species on soil moisture. In this section I go into more detail on the effects of shade on soil moisture and discuss other mechanisms by which plants can improve soil moisture conditions and facilitate their neighbors. Tree canopies often intercept precipitation, keeping it from reaching the understory, but often these same understories have wetter soils than soils of surrounding open areas (McLeod and Murphy 1977, Belsky et al. 1989, Ko and Reich 1993). The proportion of precipitation reaching the understory is affected by the intensity and duration of rainfall events. For example, in rainfall events in a semi-arid savanna in Kenya that were less than 2 mm, no throughfall reached the understory (Belsky et al. 1989). But in events that exceeded 20 mm average throughfall equaled ambient rainfall. Similar patterns have been observed in California oak woodlands (Callaway and Nadkarni 1991) and deserts (Robinson 2004). Different tree species can have widely variable effects on the amount and distribution of precipitation reaching the ground based on tree size, bark characteristics, and leaf area (Haworth and McPherson 1995). Despite the interception of precipitation by oak and Acacia canopies, and increased productivity of the herbaceaous understory, subcanopy soils in both California oak woodland and Kenyan savanna are often wetter than in the open,
46
Direct Mechanisms for Facilitation
corresponding with many other studies showing that canopies can create mesic habitats. In savannas (dehesas) in southwestern Spain, the canopies of Quercus rotundifolia and Quercus suber cause significant delays in soil water loss relative to soils in the open (Joffre and Rambal 1988). They measured water movement through the upper 150 cm of soil and found that increased soil moisture due to shade, and the effects of the oaks on physical soil characteristics, correlated with large differences in species composition under trees versus in the open. Similar effects occur in Acacia-dominated savannas. Ovalle and Avendaño (1987) found that herbaceous species under Acacia craven were more productive than in open areas and attributed this to improved soil moisture conditions. In Kenyan savannas herbaceous communities under Acacia tortilis and Adansonia digitata (baobob) tree canopies are strikingly different in comparison in to the surrounding matrix of open grassland (Belsky et al. 1989, Belsky 1994). Others have reported similar strong canopy effects on species distributions in South African savannas (Kennard and Walker 1973). Improved water relations may play a role in these patterns as leaf water potential of grasses under Acacia tortilis is significantly higher than grasses in the open (Maranga 1984). Anderson et al. (2001) found that woody seedling densities and the survival of transplanted Prosopis glandulosa seedlings were higher under Quercus fusiformis in a Texas savanna. However, many Q. fusiformis individuals in the populations had been infected by the oak wilt, which increased their general facilitative effects. Soil moisture under symptomatic trees was significantly higher than under healthy trees, presumably because leaf drop in the infected canopies reduced transpiration. Drought is one of the primary causes of mortality for seedlings of Olea europea (wild olive) in southern Spain (Rey and Alcántara 2002). They planted seeds of Olea in a variety of habitats, and found that survival was negatively correlated with the light intensity of the environment and was much higher under several species of shrubs where shade retained soil moisture than in the open. They also followed seedling germination and survival of a natural cohort and found that germination and emergence of seedlings was higher on open ground, but only seedlings that were under shrubs survived the two year duration of their experiment. Breshears and colleagues (1998) measured micrometerological dynamics under Pinus edulis and Juniperus monosperma canopies in New Mexico in detail, providing a relatively complete picture of the long-term effects of these canopies on understory moisture. Subcanopy temperatures were as much as 10oC cooler in the summer, but subcanopy soils were warmer during the coldest parts of the day in the winter. They used these measurements, soil volumetric water content, and measurements of the rates of soil drying from excised samples
Chapter 2
47
to calculate evaporation from subcanopy and inter-canopy spaces. During the growing season drying rates in the inter-canopy spaces were as much as 2% per day higher than drying rates under the canopies of the trees. Furthermore, these rates were amplified at low soil water contents. These results show that the commonly reported facilitative effects of canopies on understory moisture are temporally dynamic, perhaps more so than in open habitats, and this may increase microhabitat heterogeneity for understory species Soil moisture under canopies is also affected by the way soils are modified by long-term deposition of litter (Eckstein and Donath 2005). Soil bulk density (soil mass per unit volume) decreases as litter from canopies accumulates and degrades into soil organic matter, which creates a more favorable rooting substrate and soils that can hold more water. Root channeling may also increase the rate of water infiltration and storage. Joffre and Rambal (1988, 1993) found that subcanopy soils beneath two evergreen oaks, Q. suber and Q. rotundifolia, had significantly lower dry bulk density and lower moisture release curves than soils from the surrounding open grassland. These differences were correlated with much higher subcanopy soil water storage capacities, higher seasonal soil water contents, and shifts in species composition. Georgiadis (1989) found that the soil bulk density under shrubs in West African savannas was 22% lower than in the open, and the soil compaction index was less than half of that of open soils. Low densities increased water infiltration rates of subcanopy soils over three times more than that of open soils. In southern African savannas soils under trees are deeper and contain higher percentages of organic carbon, a pattern that appears to affect the mosaic of herbaceous communities (Carter and O’Connor 1991). I found that soil bulk density in the upper 10 cm under Quercus douglasii in Californian savannas and woodlands ranged from ≈0.7 to 0.8 g/cm3, whereas in the open density was >1.0 g/cm3 (Callaway et al. 1991), but differences in soil moisture were not clear. Tiedemann and Klemmedson (1986) found that soil bulk density was lower under Prosopis juliflora trees than in the open spaces between the trees. The effects of trees on subcanopy soil characteristics are difficult to separate from the effects of subcanopy soil characteristics on trees. For example, it is possible that Prosopis juliflora trees (or any other tree species) are more likely to establish and grow in sites with lower bulk densities than sites with high bulk densities. Because physical soil characteristics change very slowly, there are few options for experimentally separating cause from these kinds of spatial correlations. However, there are three approaches to this dilemma. First, if trees actually are causing the soil characteristic, then such changes should be very consistent - in my opinion much more consistent than subcanopy vegetation which is affected by many other factors. Second, if trees are causing the change there should be a relationship between the age of the tree and the magnitude of
48
Direct Mechanisms for Facilitation
the effect. For example, Baumeister and Callaway (2006) found that the age of Pinus flexilis (limber pine) trees on the eastern front of the northern Rocky Mountains was highly correlated with the depth of the A horizon beneath them, potentially explaining how limber pine might affect understory plant communities. Finally, it is reasonable to expect tree canopies to increase litter and the organic content of soil, and trees may certainly reduce the density of soils, but trees are much less likely to alter soil texture several centimeters below the soil surface. If soil texture is similar, but organic content is not, it is reasonable to assume that site differences are biotically determined. In a few cases, herbs (which are usually the beneficiaries of shade from woody perennials) have positive effects on the water relations of the shrubs that shelter them. Understory herbs cannot shade overstory perennials, but they can shade soil and affect soil moisture. Holzapfel and Mahall (1999) conducted experiments using thatch additions and removal of annuals and found that understory annuals had positive effects on the predawn water potentials of associated Ambrosia dumosa shrubs. In southern Spain, one of the driest regions of Europe, Retama sphaerocarpa shrubs have higher water potentials when the understory herb Marrubium vulgare occupies the understory than when the understory is bare ground (Pugnaire 1996a), suggesting that a dense understory helps to retain rainfall. In greenhouse experiments, Espigares et al. (2004) found that herbs found in the understory of Retama exerted strong competitive effects on Retama seedlings and claimed that the positive effect of understory herbs occurred only after a shift away from competitive effects with age. This may be true, but greenhouses are very poor places to study facilitation because watering, fertilization, and protection from wind and herbivores ameliorate stress in the greenhouse and make facilitation less likely. Greenhouse experiments almost always demonstrate competition. Although understory effects on overstory shrubs and trees have not been investigated frequently, research like that of Holzapfel and Mahall (1999) and Pugnaire et al. (1996a) has the potential to provide fresh insight into the generality of positive interactions in communities. Facilitative effects on moisture are stronger in arid conditions, but relatively drought tolerant species can modify the environment in ways that facilitate the water relations of less tolerant species even in very mesic habitats. Eriophorum vaginatum, and other Eriophorum species (cottongrass), readily germinate and grow on open, bare surfaces of peatlands and are common species on boreal and temperate peatlands around the world. In peatlands that have been cut for fuel, Eriophorum appears to facilitate the colonization and survival of some the world’s best “ecosystem engineers”, the Sphagnum species that gradually form peat through the accumulation of their dead and living biomass (see below). Even though peatlands have high water tables and water-saturated
Chapter 2
49
soils just below the surface, the uppermost layers tend to be dry when plant cover is low because of the low albedo of dark peat and desiccating winds (Salonen 1987, 1991). In the Jura Mountains of eastern France the majority of Sphagnum colonization occurs under the leaves of Eriophorum vaginatum tussocks (Grosvernier et al. 1995), and in Canadian peatlands E. angustifolium appears to facilitate revegetation of other species (Ferland and Rochefort 1997). Tuittila et al. (2000) studied the spatial associations of several species with E. vaginatum tussocks in a Finnish peatland 20 years after fuel cutting and abandonment. They found that almost all of the other species showed a positive association with the early-colonizing E. vaginatum. Furthermore, the smaller the E. vaginatum tussock, the closer other plants occurred to it. Not only did other species occur more frequently near the tussocks, they occurred more frequently on the northern, shadier, aspects of the tussocks. These spatial associations were assumed to be related to the more humid microclimate under the tussocks. Previous examples show that some plant species modify physical habitat in ways that create mesic conditions. In some cases, however, a beneficiary host plant may provide mesic environments with its own body. Epiphytic plants live in conditions where moisture is highly variable. On Sapelo Island, Georgia, colleagues and I conducted field experiments to study mechanisms behind the preference of two vascular epiphyte species, Tillandsia usneoides and Polypodium polypodioides for particular host species (Callaway et al. 2002). The abundances of these epiphytes on Celtis laevigata (hackberry), Quercus virginiana (live oak) and Juniperus virginiana were over twice that of any of the other host species, and this correlated with the much greater water-holding capacity of the bark of these species. After 24 hours of experimental drying, 42-66% of the initial water remained for Celtis, Q. virginiana, and Juniperus, compared to 0-14% for Pinus elliotii, P. taeda, Magnolia grandifolia, and Ilex opaca which were all very poor epiphyte hosts (Figure 2.8). The saturated capacities of the three best hosts for epiphyte species was more than twice that of the “bad hosts”. Because atmospheric epiphytes lack access to consistent water supply from the soil, they are susceptible to water stress. Bark with a high water-holding capacity may improve the performance of air plants such as Tillandsia by increasing humidity near the tree and decreasing leaf-to-air vapor pressure differences. Low leaf-to-air vapor pressure differences promote high rates of gas exchange in Tillandsia usneiodes and other species in the genus (Lange and Medina 1979, Martin and Siedow 1981). Interestingly, Garth (1964) found that atmospheric humidity was the single climatic factor which best fit the northern limit of the distribution of Tillandsia usneoides. Unlike Tillandsia, Polypodium is able to grow roots into the bark of host trees; therefore, bark with high water-holding capacity may provide a consistent water supply. Polypodium can regain full
Direct Mechanisms for Facilitation
a
2
saturated (r =0.76) 2 after 24 hours drying (r =0.77)
800
b
600
b b
400
b
b
b
n
b
b
b no
200
Ac er
P. tae da M ag no lia
P. e
lli oti i
0
LOW
o
o o
o
o
Rank order of Tilandsia preference
Ce lti s
o
igr a Li qu id am ba r Ju nip er us Q. vir gin ian a
o
Q. n
o
Ile x
2
Water held in bark (mg/cm )
50
HIGH
Figure 2.8. Water-holding capacity at saturation and after 24 h for branch bark of different tree host species for Tillandsia usneoides. Host species are displayed in order of field ranking for epiphyte density with the poorest hosts on the left. Bars indicate means + 1 standard error. Shared letters within each variable indicate no significant difference (post-ANOVA Tukey test). Inset figure shows regressions for water-holding capacity and duration and field abundance rank for each host species. One-way ANOVA, Fcapacity=14.23, df=9,49, P<0.001; F24 hrs=14.48, df=9,49, P<0.001. Reprinted from Callaway and Pennings (2002) with permission from Oecologia.
photosynthetic capacity after complete dehydration (Stuart 1968), but Polypodium is susceptible to photoinhibition as leaves begin to dry, and even partial drought decreases carbon gain (Muslin and Homann 1992). The spatial distributions of other Polypodium species in tree canopies in a Mexican cloud forest correlate with their ability to tolerate drought (Hietz and Hietz-Seifert 1995, Hietz and Briones 1998). The positive effects of canopy trees on subcanopy soil moisture are affected by the age or size of the overstory tree. Kellman and Kading (1992) found that densities of young Pinus strobus and P. resinosa on sandy soil near Lake Huron were about six times more abundant under the canopies of Quercus rubra than in the open. Pine seedlings were much more common under old oaks than young oaks. Experiments with transplanted pine seedlings resulted in high emergence and survival under medium and large oak trees (Figure 2.9), but young oaks did not have this effect. Summer high temperatures at ground level
Chapter 2
51
Pinus resinosa
'large' oaks
'medium' sized oaks
5 4 3 2 1
0 25
0
0 0
0 0 0
5 4 3 2 1
1 2 3 4 5
0 0 0
0 0 0
5 4 3 2 1
0
1 2 3 4 5
'large' oaks
'medium' sized oaks 20
Pinus strobus
Number of seedlings
6
15
10
5
0
0 0
0 0
0 0 0
5 4 3 2 1
1 2 3 4 5
0
0 0 0
5 4 3 2 1
0
0 0 0
1 2 3 4 5
Quadrats at proportional distances from tree trunks Figure 2.9. Emergence and survival of Pinus resinosa (red pine) and P. strobus (white pine) seedlings planted beneath and outside of the canopies of oaks of different sizes. n=100 seeds for each location. Shaded boxed and dotted lines depict the position of the oak canopies. Redrawn from Kellman and Kading (1992) with permission from the Journal of Vegetation Science.
ranged from ≈32-54oC (depending on position under the canopy) under small oaks, but only from ≈22-45oC under large canopies. Seedling emergence and survival were highly correlated with subcanopy soil moisture, which was much higher under the larger tree canopies. Not only trees improve local water relations. After a stand-replacing fire in a matorral community in southern Argentina, Raffaele and Veblen (1998) removed half of the canopies of two different shrub species. They measured soil moisture and the density and growth of other species resprouting under intact canopies and where canopies had been removed. They found that soil moisture was 2-3 times higher, and the number of resprouting species was 2-10 times
52
Direct Mechanisms for Facilitation
higher, when canopies were left intact. Total resprout density was 6-7 times higher under canopies. Påhlsson (1974) compared microclimatic conditions in tussock grassland on south- and north-facing slopes in Sweden. He found that relative humidity and temperature within tussocks of Onomis repens was similar to that on the north-facing slope. Onomis tussocks were associated with higher soil moisture than the surrounding open matrix and some herbaceous species common on the north-facing slopes could only be found on the south-facing slopes when associated with Onomis. Fonteyn and Mahall (1981) removed interspecific and intraspecific neighbors from around target Larrea tridentata shrubs in the Mojave Desert and found that Ambrosia dumosa competed strongly with Larrea for water. But in the Chihuahuan Desert of Mexico the xylem pressure potential of Larrea shrubs was more than 1.0 MPa higher (less water stress) in some months when interspecific neighbors were left intact than when they were removed (Briones et al. 1998). Algae may also facilitate vascular plants in dune environments by improving soil moisture. In one of the rare examples of a greenhouse experiment demonstrating facilitation (Vazquez et al. 1998) found that the presence of filamentous algae increased the germination rates of four herbaceaous species common in tropical dune slacks to approximately 2-3 times that occurring on bare sand by maintaining substrate humidity. Most studies of positive effects of plants on each other have focused on interspecific relationships. However, despite the general assumption that intraspecific interactions should be exceptionally competitive because different individuals require exactly the same resources and occupy the same rooting and canopy spaces, there are a number of examples of facilitation for water relations among conspecifics. Ptelea trifoliata (hop tree) is a shrub that colonizes mid- and fore-dunes at the edge of Lake Michigan. Ptelea reproduces entirely by seed so germination and seed survival in the xeric, sandy environment of the dunes are limiting stages in this plant’s life history. Ptelea canopies reduce precipitation reaching the ground by 40%, but maximum soil temperatures are reduced by 1520oC, light intensity is reduced by an average of 85%, and evaporation under shrubs is only 42% of that in the open (McLeod and Murphy 1977). These changes in microclimate resulted in a two-fold increase in soil moisture under shrubs compared to the open. Increased surface moisture under mature shrubs was associated with 24.5% survival of seedlings in the understory microhabitat during the 6 month growing season, whereas all seedlings in the open died within 2 months of the beginning of the experiment (Figure 2.10).
Chapter 2
53
Survival (%)
100 80 Sheltered microhabitat 60 40 20 0
Open microhabitat 19 25 5
May
26 5
June
19 28
July
25
Aug
11
Sept
13
Oct
11
Nov
Figure 2.10. Survivorship curves for Ptela trifoliata seedlings on sand dunes near Lake Michigan, USA. Redrawn from McLeod and Murphy (1977) with permission from The American Midland Naturalist.
Even in alpine environments adult plants may facilitate conspecific seedlings by improving moisture conditions (Wied and Galen 1998). They found that the litter from the decaying flowering and fruiting structures of mature adults sheltered the soil below them. This created microsites in which soil moisture was higher and evaporation rates were lower. Experimentally planted seedlings had higher water potentials in sites that were protected by litter from adults than in sites where litter had been removed. A little lower in elevation, subalpine conifers improve soil moisture conditions near them by altering the flow of wind near the ground surface which has direct effects on other plants in the lee of their canopies, and other effects by increasing snow deposition. Holtmeier and Broll (1992) found that snow accumulated inside and leeward of tree islands at high elevations in Colorado, increasing soil moisture. Similar processes appear to create snowdrifts and more moisture on the leeward sides of mature trees and develop “ribbon forests” in subalpine regions of the Rocky Mountains (Billings 1969). Billings stated that seedling establishment is inhibited in these drifts but the higher moisture left after the drifts melt enhances the growth of seedlings that do manage to establish. Alpine “tree islands” at Niwot Ridge in Colorado create large snowdrifts in their lee which create mesic microenvironments that are beneficial to other species (Marr 1977). For example, Marr found that Ribes montigenum only occurred within or in the lee of Picea engelmanii and Abies lasiocarpa islands at Niwot Ridge. These tree islands appear to move across the landscape over decades, driven across the tundra by wind-induced desiccation and mortality on the
54
Direct Mechanisms for Facilitation
windward sides of the island and growth on the leeward side where stems and buds are protected from the wind by layered branches (Marr 1977). Pontecorvo and Bokhari (1975) also asserted that Juniperus excelsa facilitated similar islands via the same mechanisms. At timberline, tree islands and krummholtz may function as focal points for snow and moisture accumulation. However, Brooke et al. (1970) found that snow next to the trunks of trees at lower elevations in subalpine meadows in coastal British Columbia melted faster than snow in the open. In this environment earlier snowmelt appeared to facilitate the growth of shrubs and other trees in these “snow craters”, eventually creating tree islands (also see Wipf et al. 2006). This phenomenon may be important in establishing interspecific aggregation such has been described for Abies lasiocarpa (subalpine fir) around Pinus albicaulis (whitebark pine) in the northern Rocky Mountains of Montana (Habeck 1969, Callaway 1998b), conifers in the central Cascades of Oregon (Miller and Halpern 1998), and Picea engelmannii (Englemann spruce) and Abies lasiocarpa around Pinus flexilis (limber pine) in the Colorado Front Range (Rebertus et al. 1991). Interestingly, other species may facilitate seedlings of P. albicaulis, a common benefactor when an adult, by improving moisture conditions (Maher et al. 2005). Abies lasiocarpa recruitment may also be facilitated by moisture provided by shrub species. Weisberg and Baker (Weisberg 1995) reported a much higher spatial relationship between A. lasiocarpa seedlings and Vaccinium species in the Rocky Mountains of Colorado than would be predicted by the proportional cover of Vaccinium. Other researchers have noted this relationship, but generally at larger scales (Stahelin 1943, Franklin et al. 1971). As for virtually all positive spatial associations, the putative benefactor, Vaccinium, may simply be an indicator of some aspect of site quality for A. lasiocarpa, or Vaccinium shrubs may actually have a direct positive effect through amelioration of dry soils. Much like the ‘ghost of competition past’, facilitative effects may not always be obvious because the effects often disappear as the interactions among species mature (see Figure 2.15), and because multi-mechanistic interactions among benefactors and beneficiaries may shift in relative strength over time (see Chapter 4). For example, tussock grasses are facilitated by shrub species in arid Patagonian shrub-steppe, apparently due to seed accumulation and shadeincreased soil moisture under the shrub during the growing season (Aguiar et al. 1992, Aguiar and Sala 1994). However, once grasses have established under shrubs, soil moisture decreases dramatically and the survival of grass seedlings becomes lower. As the original beneficiaries grow they reach the point where they can use all shrub-enhanced water and more, obscuring the facilitative effect of their benefactors. When mature grasses were removed from beneath shrubs,
Chapter 2
55
soil water potential increased from –1.5 to –0.3 MPa during the spring, and correspondingly, grass seedling survival was enhanced as the facilitative effect of the shrub was unmasked. Aguiar and colleagues argued that all of these interactions together established dynamic interactions and community development. When the shrubs are young and small, facilitation through shade is the most important interaction, resulting in a dense ring of tussock grasses associated with shrubs. When shrubs increase in age and size, competition from the shrub and from the previously facilitated grasses overshadows the benefit of shade. After the death of the shrubs grass rings fragment, and then establish patches of grass-dominated communities. Shrub patches also affect the distribution of sexes for the dioecious grass species, Poa ligularis, but genderbiased patterns depended on the quality of the shrub patch (Bertiller et al. 2000). When patches were either short and dense or tall and sparse, conditions in which abundant populations of P. ligularis did not develop, males and females coexisted. However, when shrub patches were large and tall enough to facilitate large numbers of P. ligularis, females dominated the patches, and males occupied the edges of the patch and the bare soil outside of the patch where no females occurred. Shade may also affect the water relations of plants by altering soil salinity. Saline soils stress plants by decreasing the osmotic potential of water in the soil and toxic effects of salt ions on plant tissues. Shade provided by salt tolerant species in salt marshes reduces evaporation from subcanopy soils and consequently maintains lower soil salinities, and therefore higher soil water potentials, than in adjacent soils exposed to direct insolation (Bertness and Hacker 1994, Callaway 1994). In a New England salt marsh, Steve Brewer et al. (1997) manually removed Juncus gerardi at three different elevations and measured soil responses. They found that soil salinity increased significantly at all elevations in the marsh when Juncus was removed, but the difference was greater at low elevations. Other experimental manipulations of soil salinity in the upper zones of a New England salt marsh have shown that Distichlis spicata, a salt tolerant colonizer of saline bare patches created by disturbance, facilitates the growth of Juncus gerardi when soil salinities are high (Bertness 1991, Bertness and Shumway 1993). When soil salinities were experimentally reduced by watering, growth of Juncus was not improved by Distichlis. This process depended to some degree on the size of the disturbed patch; in large patches salinities were very high and salt-tolerant species played facilitative roles. In smaller patches, however, soil salinities were not as extreme and interactions among early colonizers and latecomers were competitive (Shumway 1994). As discussed above (Brewer et al. 1997), once Juncus was established its turf morphology shaded the soil. This maintained much lower soil salinities than experimentally cleared patches, decreasing evaporation and therefore salt
56
Direct Mechanisms for Facilitation
accumulation (Hacker and Bertness 1995, 1999). At higher elevations in the marsh Bertness and Hacker (1994) and Bertness and Yeh (1994) found that shade provided by the shrub Iva frutescens also decreased soil salinity and facilitated the growth and survival of Juncus gerardi.
2.5. NUTRIENTS The development of vegetation is a fundamental driver of soil development, and it is clear that shrubs and trees alter the characteristics of the soil microhabitat in ways that favor some species over others. As stated by Ovington in 1955, “Trees are the dominants of the community, and by virtue of their size and longevity they are able to determine to some extent the site conditions under which the associated plants and animals live. Under different tree stands distinct woodland environments exist. Whilst the trees cannot alter primary site factors such as bedrock or topography, they may modify some secondary factors. Nutrients are removed from the soil and are returned in part as litter fall so that the tree’s influence those soil processes which affect the physical and chemical condition of the soil. A closed tree canopy forms an effective barrier and the climate beneath the canopy is vastly different than that above. The floristic composition and luxuriance and ground vegetation in woodlands are largely dependent on secondary site conditions such as these…”
Whilst trees may tilt the competitive playing field towards certain understory species by creating a ‘vastly different’ climate, Ovington also noted that deeply rooted perennials take up nutrients that are unavailable to more shallowly rooted understory plants and deposit them on the soil surface via litterfall and throughfall. Understory plants may eventually acquire these nutrients after the litterfall from the overstory decays. The importance of this process for species that are adapted to life in the understory may be accentuated because shadegrown or shade-adapted plants typically allocate proportionally less biomass to roots and have lower root:shoot ratios than sun-grown plants (Bellingham et al. 2001). Root symbionts such as nitrogen-fixing bacteria or mycorrhizae may also contribute to the concentration of nitrogen and phosphorus around understory plants. Organic matter deposited through litterfall provides energy to these microbes and also stimulate decomposition and mineralization rates in the understory (facilitative interactions mediated by soil microbial communities are discussed in Chapter 3). Other mechanisms by which plants may enhance
Chapter 2
57
nutrient availability for other plants include trapping of airborne particles by foliage and eventual deposition at the base of the plant, and the indirect effects of nitrogen fixation. The effects of tree canopies on nutrient availability may be confounded by indirect interactions with animals because trees provide perches for birds and shady habitat for ungulates. Animals that are attracted to trees often spend disproportionally large amounts of time near them and defecate there more often than in open environments, transporting nutrients from the surrounding open, treeless areas to understory habitats (McNaughton 1983). As noted in another study of forest canopy effects on communities and ecosystems by Madgwick and Ovington (1959), nutrients in throughfall and litterfall greatly exceed those deposited by precipitation and dry deposition. Those nutrients that are retranslocated to a lesser degree such as calcium and potassium can be especially abundant under canopies (Callaway and Nadkarni 1991, Callaway et al. 1991). Leaf litter is generally the primary source of nutrients from canopies, but non-leaf litter may account for a relatively large amount of nutrient deposition; under Quercus douglasii canopies in central California approximately 15-35% of nutrient deposition is in twigs, bark, flowers, and acorns. For some tree species, throughfall (precipitation coming directly through the canopy) nutrient deposition can equal litter as a nutrient source. Hart and Parent (1972) found that concentrations of sodium, calcium, magnesium, potassium, phosphorus, and nitrate were 3-16 times higher under the canopies of Pseudotsuga menziesii and Juniperus scopulorum than in the open and thought that canopy enrichment of precipitation was a major source of ecosystem resources. Stemflow may be an even richer supply of resources to the understory. Voight (1960) compared throughfall to stemflow under Pinus resinosa, Tsuga canadensis, and Fagus grandifolia in the northeastern U.S. and found that nitrogen was over two times higher in stemflow than in throughfall. Other nutrients showed similar patterns. By altering nutrient cycles, canopy species provide habitat in which other plant species thrive. Throughfall can be chemically complex, and therefore is not simply a source of nutrients, but can also affect basic soil characteristics. In what may be the most dramatic quantified example of a species’ effect on its soil, Amiotti et al. (2000) found that the influence of Pinus radiata, an invasive in the study area in Argentina, “triggered changes in the evolutionary trends of the soils of such magnitude as to be reflected at the highest taxonomic level in soil taxonomy”. Clear evidence of acid hydrolysis of primary silicates, decreasing values of pH, CA, and exchangeable Mg, and increasing values of exchangeable H and Al under the trees resulted in the soil next to the trunks of trees being classified as umbric rather than the mollic soils of the grassland matrix. High levels of nutrients in throughfall and litterfall, relative to the amounts deposited by precipitation in the open, are usually reflected as higher
Direct Mechanisms for Facilitation
Nutrient concentration (mg/g)
58 10
a
"positive" trees "negative" trees OPEN
8 6
b
a a
4 c
b
2 0
c
a a
a a a
a b c NITROGEN PHOSPHORUS POTASSIUM CALCIUM
MAGNESIUM
Figure 2.11. Total nitrogen and phosphorus and exchangeable potassium, calcium, and magnesium under Quercus douglasii and in the open in woodland, and savanna soils. Shared letters within sites and nutrients designate means that were not significantly different. “Positive” and “negative” trees refer to overall effects on understory productivity and are discussed in Chapter 4. Reprinted from Callaway et al. (1991) with permission from Ecology.
nutrient concentrations in understory soils than in open soils. But understoryopen soil differences tend to be less dramatic than throughfall and litterfall versus open deposition differences. For example, the total annual deposition of nitrogen and phosphorus in the throughfall and litterfall of Quercus douglasii was 5-10 times that of precipitation in the open grassland between the trees, yet total soil phosphorus and nitrogen were only 15-60%, and 2-3 times greater, respectively, in the subcanopy soils (Figure 2.11; Callaway and Nadkarni 1991, Callaway et al. 1991). Most documented cases of soil enrichment by perennial canopies are from semi-arid environments where there are clear boundaries between canopy and shrub patches (see review by Scholes and Archer 1997). In these cases the differences in physical soil characteristics and soil nutrients between subcanopy and open microhabitats are often substantial and litter deposition consistently develops higher organic matter contents in subcanopy soils than open soils (Vetaas 1992). For example, Belsky et al. (1989) measured levels of total phosphorus, calcium, and potassium roughly two times higher under Acacia trees than in the open in African savannas. High nutrient concentrations under Acacia canopies were correlated with high herbaceous productivity. Soil nitrogen concentrations were also higher under Sericocomoosis pallida shrubs in other Kenyan savannas (Georgiadis 1989). As for Acacias, understory productivity
Chapter 2
59
was much higher under S. pallida canopies and a common understory grass grew larger in subcanopy soil than in open soil. In addition to these examples, nutrient enrichment, relative to open areas, has been found under trees and shrubs in African savannas (Radevanoki and Wickens 1967, Dancette and Poulain 1969, Weltzin and Coughenhour 1990, Belsky 1994), Asian savannas (Singh and Lal 1969, Tupas and Sajise 1977), evergreen and deciduous mediterranean-climate savannas and woodlands (Jackson et al. 1990), deserts (Went 1942, GarciaMoya and McKell 1970, Charley and West 1975, Halvorson and Patten 1975, Hazlett and Hoffman 1975, Patten 1978, Whittaker et al. 1979, Schmida and Whittaker 1981, Mott and McComb 1974, Aggarwal et al. 1976), grasslands (Tiedemann and Klemmedson 1973, 1977, Barth and Klemmedson 1978, Yavitt and Smith 1988), alder shrublands (Goldman 1961), Patagonian shrublands (Rostagno et al. 1991), dune systems (Yarranton and Morrison 1974), and open Pinus contorta (lodgepole pine) stands (Zincke 1962). These positive effects are often associated with enhanced understory productivity or shifts in community composition, but by no means is facilitation the only interactive outcome produced by tree canopies. The competitive effects of overstory trees may be as strong as or stronger than their facilitative effects (Scholes and Archer 1997). Bastow Wilson (1989) demonstrated a unique positive interaction between the grasses Deschampsia flexuosa and Festuca ovina involving nitrogen. Deschampsia naturally grows in nitrogen-poor habitats and nitrogen-rich conditions in greenhouse actually reduce its growth. However, when grown with Festuca in high nitrogen conditions Deschampsia biomass increased relative to when it was grown alone in high nitrogen. Wilson presumed that this was due to the uptake of nitrogen by Festuca which established nitrogen levels more suitable for Deschampsia. Whether or not this would occur at nitrogen levels found in the field is unknown.
2.5.1. Climate and variation in canopy effects on soil fertility The positive effects of canopies on soil fertility in such a broad range of ecosystems supports the generality of this type of facilitation, but the relative influence of tree canopies on soil fertility varies with climate. Scholes (1990) developed a conceptual model for the general interactive effects of soil fertility and annual rainfall on herbaceous productivity in South African savannas (Figure 2.12). This effect of soil fertility on rainfall-productivity relationships has important implications for predicting the relative intensity of tree canopies on soil nutrients and understanding the balance of competitive and facilitative mechanisms in communities in general (see Chapter 4). Scholes argued that the effect of rainfall on herbaceous productivity depends on soil fertility, with fertile
Direct Mechanisms for Facilitation -1
Herbaceous productivity (kg ha )
60 6000
5000
4000
3000 A
2000
B 1000
0 0
100
200
300
400
500
600
700
800
900
1000
-1
Annual rainfall (mm yr ) Figure 2.12. Relationship between above-ground annual herbaceous productivity and annual rainfall. Lines with large dashes represent upper and lower limits of the general relationship between rainfall and productivity in semiarid regions (Rutherford 1980). The line labeled “A” represents the relationship found on a fertile soil soils, “B” on infertile soils. Note the more rapidly increasing herbaceous productivity on fertile soils. Redrawn from Scholes (1990) with permission from Journal of Biogeography.
sites hypothesized to increase rapidly in productivity and infertile sites to increase slowly in productivity with increasing rainfall. This relationship between moisture and soil fertility suggests that positive canopy effects on soil nutrients should have stronger effects on understory productivity in mesic climates; at least until canopy density limits light (Holmgren et al. 1997), than in xeric climates where moisture may constrain the potential benefits of soil nutrients. There are not many studies that have addressed this issue directly, but several have quantified production or density of the plants under canopies versus in the open in different years or sites with different amounts of rainfall. A second conceptual model was suggested by Garner and Steinberger (1989) who proposing that “fertile islands” (see Cross and Schlesinger 1999, Stock 1999) are prominent around woody perennials in deserts because “concentrating biological mechanisms” such as water transport and biological deposition rates are more important in arid climates than “dispersive physical mechanisms” such as soil water diffusion, leaching, atmospheric washout, and gaseous diffusion, which predominate in mesic environments (Figure 2.13).
Transport rate, distance/time
Chapter 2
61 dispersion mechanisms dominate physical transport concentration mechanisms dominate biological transport
Relative moisture Figure 2.13. Relationship between relative moisture availability and the transport rate of nutrients in soils. Redrawn from Garner and Steinberger (1989) with permission from the Journal of Arid Environments.
McClaran and Bartolome (1989) quantified productivity under canopies of Quercus douglasii versus in the open in different years and sites with different amounts of rainfall. They found that the effect of canopies on understory productivity (relative to open grassland) varied by about 50% in five sites over two years, and along a rainfall gradient ranging from 40 to 90 cm yr-1. Unlike results reported by others (Holland and Morton 1980, Callaway et al. 1991, Ratliff et al. 1991, Connor and Willoughby 1997), who showed that Q. douglasii increased subcanopy productivity, Q. douglasii canopy effects were primarily neutral at more xeric sites and negative at mesic sites in both years. However, reanalyzing McClaran and Bartolome’s results indicates that productivity in the fertile soil under the canopies increased ≈45% in an aboveaverage rainfall year; whereas, in the open productivity increased only ≈25%. Although the comparative fertility of the different sites is not known, these patterns and what is known about the fertilizing effect of Q. douglasii on soils support Scholes’ general hypothesis. A reanalysis of my data (Callaway et al. 1991) and unpublished data (R.M. Callaway) also supports Scholes’ hypothesis. Understory productivity and total soil nitrogen (0-10 cm) in the subcanopy soils was quantified for each of 12 “positive” (trees that showed a general facilitative effect) Quercus douglasii trees in central California. Productivity was measured
62
Direct Mechanisms for Facilitation
Percent difference in producitivity (1987/1988)
30
25
20
15
10
5 6
7
8
9
10
11
12
13
Total soil nitrogen (mg/g soil) Figure 2.14. Relationship between total soil nitrogen beneath Quercus douglasii trees in a central California savanna and the difference in the proportional effect of Q. douglasii on understory productivity in a wet year (1987) and a dry year (1988). R2 = 0.61, P<0.05. Percent difference = ([(subcanopy-open)/open in 1987] – [(subcanopy-open)/open in 1988]) * 100.
in a wet year (1987) and a dry year (1988), and soil nitrogen was measured in 1987. Understory productivity for each tree in each year was calibrated with the productivity in the surrounding open savanna or patches of grassland within woodland. Understory productivity was approximately 2.5 to 5.0 times greater than open productivity under these trees. Quercus douglasii trees with higher soil nitrogen showed the highest productivity response to increased rainfall (Figure 2.14). For example, the facilitative effect of trees with soil nitrogen contents of 67 mg/g soil only increased ≈10% in the wetter year, whereas trees with 11-12 mg/g soil increased their facilitative effect by 20-30%. Research by Tielborger and Kadmon (1997) shows an opposite relationship between canopy effect and annual precipitation. Their work in the Negev desert showed that the mean density of annual plants under various shrubs was higher than in the open in wet years, but lower in dry years (see Chapter 4). As in McClaran and Bartolome’s study, soil fertility was not measured, but since many other researchers have found that desert shrubs create “resource islands” the subcanopy was probably a much more nutrient-rich microhabitat than the intershrub spaces.
Chapter 2
63
Conflicts between studies that support Scholes’ hypothesis and those that do not may be due to synergistic interactions between soil conditions and the facilitative effects of nutrients that can produce conditional results. Virginia and Jarrell (1983) found that nitrogen accumulation under Prosopis glandulosa canopies was always higher than in open grassland, but the improvement was much greater where Prosopis was utilizing ground water and where surface soils had high clay contents. Virginia (1986) also found that the mount of nutrient accumulation under Prosopis glandulosa and Dalea spinosa trees was positively correlated with estimated water availability at particular sites.
2.5.2. Experimental approaches to canopy effects on soil fertility There have been few large-scale removal experiments conducted to test litterand throughfall- enrichment of soil nutrients versus other mechanisms. However, in one of these Kay (1987) removed all Q. douglasii from areas by either cutting, or cutting and applying herbicide. Prior to these treatments, herbaceous productivity under oak canopies tended to be less than productivity in open grassland (as for the ‘negative’ trees of Callaway et al. 1991). However, after the trees were killed, herbaceous productivity did not simply increase to equal that of open grassland, instead ex-understory productivity exceeded that of open grassland for 12 of the next 14 years (Figure 2.15).
Productivity (10's of pounds per acre)
350 NS
300 NS
250 200
NS NS NS
NS
150 NS 100 50 0
live trees open grassland trees cut & herbicide 1967
1972
1977
1982
Date Figure 2.15. Annual productivity of open grassland, under live Quercus douglasii, and where Q. douglasii had been cut and herbicided. All points not labeled with “NS” showed significant differences between the cut-herbicide treatment and the live trees. Drawn from data presented in Kay (1987).
64
Direct Mechanisms for Facilitation
Only after 15 years did the residual effect of the trees on herbaceous productivity disappear. While not specifically demonstrating nutrients as the mechanism, these results indicate that Q. douglasii trees modified soils in ways that lasted for years after the removal of the tree itself, and that these effects eventually went away. Thirteen years after cutting Prosopis juliflora trees, the subcanopy-open difference in soil organic carbon decreased from 3.0x to 1.6x and total nitrogen from 2.8x to 1.9x (Tiedemann and Klemmedson 1986), also illustrating the impermanent nature of plant-driven soil enrichment. Experiments with soils and bioassays from this same system provide a more ecologically relevant picture of long-term changes in soil enhancement. Klemmedson and Tiedemann (1986) grew two grass species, Digitaria californica (Arizona cottontop) and Hordeum vulgare (barley) in soils collected from under Prosopis canopies and in the open soil matrix, and also where Prosopis canopies had been removed 13 years prior to the study. Where Prosopis was intact the growth of Digitaria and Hordeum was 3.2x and 1.6x higher in subcanopy soil than in soil from the open spaces between the trees. In sites where Prosopis had been removed the effect of subcanopy soil on growth of Digitaria was much lower. The effects of Prosopis on understory fertility and plant growth existed 13 years after tree removal, but the effect was decreasing with time. In an innovative approach to exploring causal relationships between canopies and soil fertility, Kleb and Wilson (1997) moved understory soils from aspen forests to prairie habitats, and prairie soils to understory forest habitats, and then measured changes in soil characteristics. They found that the nitrogen-poor soils from the aspen stands increased in soil nitrogen concentration when they were moved to the prairie and that prairie soils decreased in nitrogen concentration when moved to aspen stands, indicating a strong biotic effect rather than simple microsite differences. Other experiments demonstrated similar effects of vegetation on soil moisture. When they transplanted cores from aspen forests to prairie, soil moisture in the cores was similar to that in prairie soils. Further evidence for trees and shrubs causing resource islands (versus preferentially establishing in them) comes from measurements of correlations between the age of the tree forming the canopy and soil characteristics beneath the tree. For example, Josè Facelli and Daniel Brock (2000) at the University of Adelaide in Australia found no differences in any soil characteristic between samples from under very young Acacia papyrocarpa trees and samples from open areas. However, organic matter, total and available phosphorus, total nitrogen, and soil salinity increased dramatically with the age of Acacia papyrocarpa until maturity, and then declined as the tree canopy senesced (see Figure 4.36, Chapter 4). Total nitrogen in the soil under trees declined precipitously after the death of trees. Organic matter and total and available phosphorus remained higher than that in the open grassland matrix soil for
Chapter 2
65
approximately 50 years after the death of the trees. Documentation of these tree age-soil relationships is one of the few studies confirming that the resource islands associated with the trees were actually caused by the trees, rather than the tree preferentially establishing in high resource patches. Interestingly, there were no differences in sodium, potassium, calcium, and magnesium, nutrients that typically accumulate through rainfall washing through the canopy. Only nutrients associated with organic matter deposition changed in availability with the influence of the tree, indicating that the accumulation of organic debris was probably the most important facilitative mechanism. The resource patches that developed under A. papyrocarpa facilitated many understory species, and some species were almost entirely restricted to A. papyrocarpa understories. Furthermore, Enchylaena tomentosa, an herbaceous species found almost exclusively under the canopies of A. papyrocarpa, established and grew significantly larger in soil collected under tree canopies than in soil from the open matrix. For some resources the degradation of the resource islands started with tree senescence rather than death, suggesting that the effects of tree canopies on soil chemistry may be more complex than simply the deposition of nutrients on the ground via throughfall and litterfall. Facelli and Brock hypothesized that the fragmented canopies of older senescing plants transmitted more sunlight, increasing soil temperatures and faster decomposition of organic matter as well as reducing deposition rates. Other experiments and field studies have shown that when vegetation develops on previously denuded or bare substrate, soil nutrients increase (see Bellingham et al. 2001), and others have shown that nutrient concentrations and organic carbon content increases with tree age or size (Barth 1980, BernhardReversat 1982). Also, nutrient accumulation over time in the presence of perennial woody plants has been demonstrated in microcosms by Silber and Raviv (1996). They measured increases in water-soluble phosphorus, calcium, and magnesium in volcanic tuff when roses were grown in the pots for two years. The effect of vegetation on nutrient availability and cycling is probably one of the most general and broad facilitative effects in natural systems. The modest experimental evidence for canopies causing “resource islands” calls for more research, but the correlations among canopies and increased soil fertile is so strong that other interpretations are unlikely. However, it is reasonable to expect that in particular systems trees or shrubs preferentially select unusually fertile sites or that animals defecate and urinate more under the canopies and eat more in the open.
66
Direct Mechanisms for Facilitation
2.5.3. Conditionality and canopy effects on soil fertility The relative importance of deep-rooted facilitators, through the deposition of litterfall and throughfall from their canopies, depends on the inherent soil fertility at a particular site. At fertile sites the effects of litterfall and throughfall are unlikely to make large differences, but at infertile sites the effect of canopies can be strong. For example, soils under junipers on infertile sand dune substrate showed 5-12 times higher concentrations of manganese, potassium, and organic carbon than open dune soils (Yarranton and Morrison 1974); whereas I found that in the relatively fertile soils of California oak woodlands canopy-open differences in nutrients typically range from <1.5 to 3 times (Callaway et al. 1991). Few studies have addressed the issue of canopy effects on soil nutrients over gradients of fertility (such as commonly addressed for other stress gradients, see Chapter 4), but in a study of arctic tundra, not a biome renowned for fertility, on Devon Island, Gold and Bliss (1995) reported that comparative differences in nutrient concentrations between fertile “resource islands” caused by cryptogams and the open spaces between them varied with larger scale environmental conditions. Jim Reynolds et al. (1999) addressed the issue of temporal conditionality of shrub effects on soil fertility by conducting a 3-year study in New Mexico using rainfall exclusion shelters to manipulate the effects of drought on the nutrient dynamics of resources islands associated with Larrea tridentata and Prosopis glandulosa. They found that concentrations of nitrate increased under shrubs exposed to fall drought, relative to that of controls, creating steeper gradients in resource availability between resource islands and open substrate. As mentioned in other places in this chapter, a number of other researchers have found that nitrogen-fixing Prosopis species have pronounced positive effects on soil fertility (Tiedemann and Klemmedson 1973, 1977, 1986, Virginia and Jarrell 1983). Parsons et al. (1994) measured the responses of Empetrum hermaphroditum, Vaccinium myrtillus, C. uliginosum, and V. vitis- idaea to increased temperature, nutrients, and water in the field. They found that all species responded significantly to warmer temperatures. Additionally they found that temperature and fertilization treatments together elicited strong and positive synergistic effects on plant growth. Their results suggest that some facilitative mechanisms may be poorly expressed if other limits to plant growth are not ameliorated. In the study of Parsons and colleagues, for example, the benefits of fertilization were not realized until low temperatures were ameliorated.
Chapter 2
67
2.5.4. Canopy facilitation of soil fertility in mesic habitats Canopy effects on soil nutrients in mesic forests are more difficult to observe and quantify than in savannas. However, many studies have reported strong correlations between particular canopy species and the soil nutrient characteristics beneath them (Lodhi 1977, Beatty 1984, Boettcher and Kalisz 1990, Dzwonko and Loster 1997, Finzi et al. 1998a,b, Lovett and Rueth 1999). Others have reported correlations among canopy and understory species, but as for many studies in savannas these studies do not separate canopy effects from the confounding effects of the physical environment (Lippmaa 1939, Whittaker 1951, 1956, Bratton 1975, McCune and Antos 1981, Austin 1985, Rheinhardt 1992, Fensham and Butler 2004). For example, Maguire and Forman (1983) sampled canopy-overstory species associations in an old-growth forest and argued that the distribution of understory herbaceous communities was affected by canopy foliage of the different species. The potential for different tree canopies to affect soil nutrients to affect understory communities is illustrated by research conducted by Diekman and Falkengren-Grerup (1998). They measured soil nitrogen mineralization rates for ammonium, nitrate, and total-N in incubation experiments for 661 plots in deciduous forests in southern Sweden. From these measurements they calculated a “functional N index” for understory species and found that community composition and species distributions were highly correlated with the index. It was not clear how specific overstory species correlate with patterns of the functional N index, but Diekman and Falkengren-Grerup’s results suggested that the canopy-specific patterns in nutrient deposition and accumulation described by others (Boettcher and Kalisz 1990, Finzi et al. 1998a,b, Lovett and Rueth 1999) affect the composition and productivity of understory communities. Different sampling approaches may exacerbate the problem of detecting facilitative canopy effects in complex mesic forests. For example, if plots are established randomly, as in classic community studies, fine-scale correlations among canopy and understory species may be obscured by variation both in the canopy layer and in the understory layer. If plots are scaled to the size of individual canopies the probability of detecting differences in species composition may be higher.
2.5.5. Canopy facilitation of available soil nutrients The majority of studies of spatial associations between perennial canopies and soil nutrient concentrations have focused on total nutrients, and available nutrients show similar, but more complex patterns. Parker and Muller (1982)
68
Direct Mechanisms for Facilitation
measured far higher levels of nitrate in soils under Quercus agrifolia than in open grassland soils (Figure 2.16), but similar concentrations of ammonium. These patterns in nitrate concentration correlated with plant distributions, with Bromus diandrus and Pholistima auritum dominating the understory, and Avena fatua dominating the open grassland. Others have shown that fertilized microsites favor Bromus diandrus over Avena (Jones 1963, Hull and Muller 1976), but Parker and Muller (1982) found that all three species showed similar increases in growth when grown in soils from under Quercus, and other experiments suggested that variation in light was more important for herbaceous spatial patterns than nitrogen. I found that ammonium and nitrogen were significantly higher at the beginning of the rainy season in central California under Quercus douglasii than in open grassland around the trees (Callaway et al. 1991). Later in the rainy season, after herbaceous species had matured, these patterns were less pronounced (Figure 2.17). Soil under trees with high densities of fine oak roots in the upper 50 cm (“negative trees”, see Chapter 4) showed sharp decreases in nitrate in March, when fine oak roots undergo a flush of growth. 80
Grassland soil Oak subcanopy soil
Nitrate (ppm)
60
40
20
0 October December January February March
May
Figure 2.16. Nitrate in soil beneath Quercus agrifolia canopies and in the open grassland around the trees. Error bars show 95% CI. Reprinted from Parker and Muller (1982) with permission from the American Midland Naturalist.
Chapter 2
69
30
30
Savanna
Savanna
25
25 "Positive" trees "Negative" trees Open grassland
20
20
15 10 5
a
c
b b
Woodland 25 a a a
15
10 b
a
b
b
5
b
10 c
b
5
March 1988
Woodland
a
25 20
a
15
b
10 b b
November January 1987 1988
b
0 30
b
0
a
15 b
0 30
20
a a
Nitrate (ug/g soil)
Ammonium (ug/g soil)
a
a
a
b b
a a b
5 0
November January 1987 1988
March 1988
Figure 2.17. Ammonium and nitrate in soils in woodland and savanna habitats under “positive” and “negative” Quercus douglasii trees (see Chapter 4). Error bars show 2 SE, and different letters within a date denote significant differences (Tukey, P<0.01). Redrawn from Callaway et al. (1991) with permission from Ecology.
2.5.6. Litter type, mixtures, and canopy facilitation The type of litter a tree produces can cause widely varying effects. In general, persistent tree litter, that which does not decompose rapidly, tends to reduce the growth of understory species (Sydes and Grime 1981). In field experiments using litter from four deciduous canopy species from British forests, Acer psuedoplatanus, Fagus sylvatica, Fraxinus excelsior, and Sorbus aucuparia, Barret (1931) found that the grasses Poa trivialis and Holcus mollis were strongly inhibited by persistent litter from certain species, and high amounts of litter per ground surface area. In contrast, Endymion non-scriptus and Viola
70
Direct Mechanisms for Facilitation
riviniana were unaffected by litter source or quantity, and a fifth species, Galeobdolon luteum, slightly benefited from persistent litter. Field distributions of these species fit these results with Poa occurring only in relatively litter-free sites and Endymion and Galeobdolon occupying sites with heavy litter. The litter of Fagus, which was the most persistent of all species, had stronger effects than those of plastic litter treatments, which of course was even more long-lasting. This result suggests that leaf chemistry had effects in addition to leaf persistence. Monk and Gabrielson (1985) experimented with litter from the floor of different deciduous species in South Carolina in the field and found that litter generally improved the overall productivity in understory plots, but the effects of litter on individual species ranged from facilitation to interference. They attributed these effects primarily to leaf chemistry. Facelli and Pickett (1991) found that white oak (Quercus alba) leaf litter indirectly facilitated the growth of seedlings of Ailanthus altissima by suppressing herbs that competed with Ailanthus. Litter may also provide “safe sites” for seed germination and recruitment (Enright and Lamont 1989, see Chapter 3). The effects of leaf litter from a species can be altered by the way that litter decomposes in mixtures of litter from other species. Nutrient release from leaves of species that decay rapidly can stimulate the decomposition of more recalcitrant leaves (Seastedt 1984) resulting in enhanced nutrient release, increased litter decay rates, and increased respiration rates (Carlyle and Malcolm 1986, Chapman et al. 1988, Ineson and McTiernan 1992). In contrast, some mixtures result in decreased respiration rates, suggesting that they decay more slowly. Briones and Ineson (1996) compared the decomposition rates, respiration, nitrogen release, and acidity of leachates of Eucalyptus globulus leaf litter alone, to rates when combined with Quercus petraea, Fraxinus excelsior, or Betula pendula litter. They found that mixing Eucalyptus leaf litter with Quercus litter resulted in substantially enhanced total respiration from the litter and weaker enhancements with Fraxinus and Betula litter. Quercus and Betula litter greatly reduced ammonium release from Eucalyptus litter, Fraxinus litter enhanced ammonium release from Eucalyptus litter, and surprisingly, adding Eucalyptus litter almost doubled the ammonium release from Fraxinus litter (Figure 2.18). Although Eucalyptus is not naturally associated with the other three species, such complex effects of different litter types on each other, and the associated microbes, have the potential to maintain high community productivity, synchronize nutrient release and plant uptake (Briones and Ineson 1996), and generate unusual positive interactions.
Chapter 2
71
NH4 release rate (µg N g-1)
400
Eucalyptus litter Expected values of mixture Fraxinus liter Eucalyptus, Fraxinus mixture
300
200
100
0 1
3
5
7
9
11
13
Weeks Figure 2.18. Ammonium release (μNH4-N g-1) over two weeks from microcosms containing ‘pure’ litter and (a) eucalyptus litter mixed with litter from oak, ash or birch. Pure eucalyptus litter is shown as a solid circle, the pure litter from the focal tree species in each panel (e.g. oak, ash, birch) is represented with an open circle, the expected additive value is shown as an open square, and the observed value for the mixture is shown as a open diamond. Error bars are 1 SE. Reprinted from Briones and Ineson (1996) with permission from Soil Biology and Biogeochemistry.
The effects of litter from the same species can range from suppression to facilitation. I found that fresh leaves of blue oak strongly suppressed the growth of an understory grass species, whereas blue oak litter that dropped naturally at the end of the growing season facilitated grass growth (Callaway et al. 1991). Sydes and Grime (1981) found that Fagus sylvatica litter collected at different times or applied in different amounts elicited contrasting and highly variable results.
2.5.7. Experimental studies of facilitation by canopies on soil Other studies on the positive effects of canopies via nutrients have used soil collected from under different species in comparisons of plant performance. Turner et al. (1966) found that saguaro seedlings in field garden tests survived best on soil collected from under Cercidium microphyllum than on soils from under either Prosopis juliflora or Olneya tesota, but the potential effect of soil
72
Direct Mechanisms for Facilitation
nutrients was confounded by soil albedo and temperature. In another Sonoran desert site, soils under C. microphyllum were almost 3 times higher in total nitrogen than soils collected from the open matrix (Franco and Nobel 1989). As noted above, surface soils under the canopies of Quercus douglasii are much more nutrient rich than surrounding open grassland (Holland 1980, Callaway et al. 1991, Dahlgren 1997) and in greenhouse experiments understory soils improved the growth of Bromus diandrus, an understory species, when compared to open grassland soils. When Q. douglasii litter was added to open grassland soil the growth of B. diandrus also improved (Callaway 1991). Walker et al. (1986) used greenhouse experiments to demonstrate the facilitative effects of nitrogenrich soil from under Alnus tenuifolia on Salix alaxensis and Populus balsamifera seedlings and found that these effects were due to litter-mediated increases in growth or nutrient uptake. Prosopis cineraria (khejri) is a dominant tree on alluvial and sandy plains of the Thar Desert near Johdpur, India. Farmers keep Prosopis on their fields because their crops grow better under the trees than in the open fields. Aggarwal et al. (1993) found that soil nitrogen, phosphorus, and potassium were higher under Prosopis canopies than in open fields, and that the biomass of Penesetum typhoides was three times higher when grown in Prosopis soil than in open soil. Adding nutrients reduced these differences, but Penesetum grown in Prosopis soils was always at least two times larger than when grown in open soils regardless of the nutrient additions, suggesting a possible microbial effect. Soil nutrients are also higher under canopies of Eucalyptus xanthoclada and E. drepanophyllla in northeastern Australia (Jackson and Ash 2001). These canopies were also associated with higher pasture productivity. Native Australian grasses produced 42% more biomass when grown in pots containing soil from under canopies than when grown in soils from the open inter-canopy matrix. Not all experiments examining the effects of canopy-enriched soils on the performance of understory species demonstrate facilitation. Total nitrogen, phosphorus, and soil moisture in the soil under Sophora trees is much higher than in various other microhabitats on the island of Hawaii (Walker and Powell 1999); however, juvenile Argyroxiphum sandwicense (silversword) had lower relative growth rates under these canopies than in more nutrient- and moisture poor microhabitats. They observed that dense swards of grasses developed under Sophora canopies and hypothesized that poor Argyroxiphum regeneration was due to indirect effects (see Chapter 3), with Sophora facilitating grasses and then the grasses out competing Argyroxiphum.
Chapter 2
73
2.5.8. Nitrogen fixation and nutrient facilitation Facilitative canopy effects on soil nutrient pools and fluxes have been shown for many different types of trees and shrubs, but as suggested in examples above, nitrogen-fixers can have particularly strong effects on their microhabitats and the species associated with them (Bellingham et al. 2001). For example, field plots with Lupinus perennis, a common perennial in North American grasslands, have 32% higher plant N concentrations and 26% higher total plant N pools than plots without Lupinus (Lee et al. 2003). Leaf nitrogen concentration of cooccurring non-fixing plants increased 22% and mass-based net photosynthetic rates increased 41% in plots containing Lupinus compared to those without Lupinus. After glacial retreat in southeastern Alaska, bare sands and gravels contain about 30 kg/ha of total nitrogen (Lawrence et al. 1967). After 25 years of occupation by Dryas drummondii, a nitrogen-fixing early colonizer, soil nitrogen averages 95 kg/ha, and by the time Dryas begins to disappear from the system soil nitrogen averages 400 kg/ha. In the alpine vegetation of the southern Rocky Mountains, Trifolium nitrogen-fixers are estimated to contribute 0.5 mg N m-2 year-1 to the nitrogen budget, which is substantial considering that estimates of the total mineralized N in these systems is 1.2 mg N m-2 year-1 (Bowman et al. 1996). Høgh-Jensen and Schjoerring (1997) measured seasonal variation in nitrogen fixation, nitrogen transfer from the nitrogen-fixing Trifolium repens to the non-fixer Lolium perenne, and soil nitrogen absorption in the field over three years in different mixtures of the species. They found that only small amounts of fixed nitrogen were transferred to Lolium in the first year, whereas in the second and third years of the experiment up to 21 kg ha-1 was transferred. Transfer accounted for 3%, 16%, and 31% of Lolium’s annual nitrogen budget, in the three years of the experiment, respectively. More importantly, Trifolium-Lolium mixtures absorbed much higher amounts of soil-derived nitrogen than the pure stands of either species (Figure 2.19). They concluded that positive interactions between Trifolium and Lolium produced a more efficient fixation of atmospheric nitrogen and absorption of nitrogen from the soil than occurred when the species were grown alone. The fact that mixtures enhanced soil nitrogen uptake is unusual, and was probably due to the positive effect of Trifolium-fixed nitrogen on Lolium growth and biomass, and then an increase in uptake of soil-derived nitrogen by Lolium.
Direct Mechanisms for Facilitation
A
80 60 40 20 0 -20
B
80 60 40 20 0
22 July, '96
14 June, '96
2 Oct, '95
3 July, '95
1 June, '95
2 Oct, '94
27 June, '94
-20
16 May, '94
Proportion of nitrogen derived from atmosphere (%)
74
Figure 2.19. Proportion of atmospherically derived nitrogen in shoots of Lolium perenne in mixtures with Trifolium repens under A) three different fertilizer regimes (solid triangle=24 kg ha-1, open triangle=48 kg ha-1, solid circle=72 kg ha-1); and B) three different species ratios (upside down open triangle=1Trifiolium with 20 Lolium, open square=6:20, and upright open triangle=12:20) with fertilizer added at 3 kg ha-1yr-1. Error bars are 1 SE. Reprinted from Høgh-Jensen and Schjoerring (1997) with permission from Plant and Soil.
Although a nitrogen fixer, Lupinus arboreus is an exceptionally poor facilitator for Quercus agrifolia seedlings, the poorest of four different shrub species tested experimentally (Callaway and D’Antonio 1991). This may be because the strongest effects of L. arboreus occur after the plant dies, due primarily to the fact that lupine canopies reduce available light from 1725 μmol m-2 sec-1 to 13 μmol m-2 sec-1 (Maron and Jefferies 1999). Canopies of dead lupine transmit 370 μmol m-2 sec-1. Repeated cycles of lupine encroachment and dieback in coastal prairie appear to double total soil nitrogen and lead to a largescale shift in community composition from native perennials to exotic annuals, which appear to be strongly facilitated by increased nitrogen accumulation. A very similar dynamic process occurs during primary succession on the volcanic soils of Mount St. Helens. Lupinus lepidus, an early colonizing nitrogen-fixer, increases soil nitrogen and facilitates colonization of other species
Chapter 2
75
in the nutrient-poor soils; however clear facilitative effects are only evident after the death of the lupine (Wood and del Moral 1987, Morris and Wood 1989, Wood and Morris 1990, del Moral 1993a,b). This pattern is found yet again on volcanic substrate in Hawaii where the introduced nitrogen-fixing tree Myrica faya enriches the otherwise nitrogen poor soil (Vitousek et al. 1987). Living Myrica trees appear to suppress virtually all understory species, including nonnative grasses. However, when Myrica trees are removed, other exotic species increase in abundance in the patches formerly occupied by Myrica (Aplet et al. 2001). Adler et al. (1998) studied the recruitment of native and non-native species into the nutrient-rich patches remaining after an outbreak of an introduced leafhopper, Sophonia rufofascia, killed large number of Myrica. The cover of non-native grasses increased 2.3 to 14.1% (depending on the species) more under dead Myrica than under live Myrica, and native shrubs and herbs increased 4.8 to 15.2%. As for subalpine lupines in Oregon and coastal lupines in California, these results suggest that nitrogen-fixing plants may leave a stronger facilitative legacy in the soil after their death than they create while they are living. Non-nitrogen-fixing species are positively associated with the nitrogenfixer Hedysarum boreale on alpine glacial moraines in the Canadian Rockies (Blundon et al. 1983, Dale et al. 1991). Similarly, in the central Rockies, nitrogen-fixing alpine Trifolium dasyphyllum adds large amounts of nitrogen to the soil, driving a doubling in total productivity compared to Trifolium-free patches (Thomas and Bowman 1998). However, this increase in productivity is mostly due to increases in abundance of T. dasyphyllum and both facilitative and competitive effects of Trifolium were observed. Forb species increased 44% in foliar nitrogen concentration and tripled in biomass (Figure 2.20). Graminoid species, however, only showed a 9% increase in foliar nitrogen concentration and their abundance was much lower in the presence of Trifolium. Species-specific spatial relationships with Trifolium patches were highly variable with 5 species showing significant positive associations, 4 species showing significant avoidance and 7 species showing no significant preference. Similar studies of interactions between Trifolium repens and Lolium perenne show that Trifolium enhances the growth of Lolium under some conditions, and that the magnitude of the positive effect declines with increasing soil fertility (Wright 1981, Weiner 1985, Menchaca and Connolly 1990).
76
Direct Mechanisms for Facilitation
in Trifolium patch away from Trifolium patch
Foliar N (%)
2
1
0
Graminoids
Forbs
Figure 2.20. Foliar nitrogen concentrations of graminoids and forbs with patches of Trifolium dasyphyllum and in surrounding tundra without Trifolium. Error bars are 1 SE. Redrawn from Thomas and Bowman (1998) with permission from Oecologia.
In Hawaii, Walker and Vitousek (1991) used a combination of greenhouse and field experiments to examine the relative importance of different mechanisms by which the invasive Myrica faya affected the regeneration and growth of the native tree, Metrosideros polymorpha. The nitrogen-fixing Myrica caused large increases in available nitrogen in this nitrogen-limited system, and nitrogen fixed by Myrica is used by other plant species (Vitousek et al. 1987, Vitousek and Walker 1989). Nitrogen-rich soils and shade under Myrica improved Metrosideros seedling growth, germination, and survival, but the overwhelming effects of root competition created a strong net negative effect. The grasslands of the “Flooding Pampas” in central Argentina have been invaded by exotic nitrogen-fixing legumes, along with a large number of other non-native dicots (Quinos et al. 1998). The native grass, Paspalum dilatatum, was subjected to interactions with a suite of exotic non-nitrogen-fixing dicots, the exotic nitrogen-fixing Lotus tenuis, or no competitors in a fully factorial arrangement. The effects of the non-nitrogen-fixing dicots were weakly competitive, but not significant for any variable; however, the effects of Lotus on the nitrogen in the soil and tiller survival of Paspalum were positive across both dicot treatments (Figure 2.21).
Chapter 2
77
Dead tillers
Number of live tillers per plant
50
A
**
40 + dicots - Lotus - dicots - Lotus + dicots + Lotus - dicots + Lotus
30
*
20 10 0 -10
B 0 days Spring
17 days 55 days End of spring Summer
***
104 days Autumn
Figure 2.21. A) Tiller survival and B) mortality for plants of Paspalum dilatatum in treatments with and without dicot neighbors removed, and with and without Lotus tenuis neighbors removed. Asterisks represent significance of positive Lotus effect. Redrawn from Quinos et al. (1998) with permission from Oecologia.
2.5.9. Canopy interception of nutrients Throughfall effects described above are generally attributed to leaching from leaves, but canopies also filter and capture atmospheric nutrients, and these processes can contribute large amounts of nutrients to the understory (Art et al. 1974, Schlesinger and Reiners 1974, Lovett et al. 1982). Martin Kellman (1986) estimated that canopy filtration of atmospheric nutrients provided 8.1% of the K, 37.9% of the Mg, 9.6% of the Ca, 17.4% of the Na, and 24% of the P in the bulk influx of nutrients to an unthinned stand of Pinus caribbaea. The contribution of canopy-filtration to nutrient budgets may be exceptionally important in tropical forests where rooting depths are shallow and nutrient capitals in soil are low compared to those in biomass (Kellman, 1989). Kellman (1979) found that the surface soils beneath five species of savanna trees in Belize were substantially higher in nutrients than the surrounding matrix of open savanna, and some cases exceeded that in nearby rainforest soil. Excavations demonstrated that only one species of these savanna trees had deep roots, indicating that there was little translocation of nutrients from deep soils to the surface. Subcanopy soils did not have higher cation-exchange-capacities nor increased soil moisture retention (this ecosystem receives high levels of precipitation), indicating that canopy capture of nutrients and incorporation into a larger plant-litter nutrient cycle was probably the major source. These same savanna species are highly associated with the establishment of the seedlings of trees found only in the nearby rainforest
78
Direct Mechanisms for Facilitation
(Kellman and Miyanishi 1982, Kellman 1985). Kellman and Miyanishi (1982) argued that the creation of such nutrient-enriched microsites provide nuclei for repeated invasions of infertile savannas by rain-forest trees during Quaternary climate fluctuations. Because weathering at depth is not a significant nutrient source in intensely weathered soils, facilitation through capture of atmospheric nutrients and local sequestration may also be crucial to the development of soils with the nutrient content suitable for rainforest regeneration after burning or logging (Kellman 1989).
2.5.10. Interactions between nutrients and shade The effects of shade on the nitrogen cycle may interact with the direct effects of canopy shade on soil moisture and temperature. For example, decomposition rates of litter in the shade may decrease due to lower temperatures, or increase due to higher soil moisture content. Decomposition of tree litter is often slower than that of understory herbaceous litter (Vetaas 1992), and the return of nitrogen from understory herb decomposition to the soil is often faster than that of tree litter (Bernhard-Reversat 1982, Escudero et al. 1985). In these studies it was concluded that subcanopy nitrogen enrichment was due to processes that increased the productivity of understory grasses and those that altered nitrogen cycling. Thus, it is possible that the primary facilitative effects of shade on understory species composition, productivity and decomposition may also have important effects on nutrient cycles. These may in turn create positive effects on other species that are difficult to separate from the direct effects of canopy nutrient deposition. As described above, many researchers have measured higher levels of nutrients in soils directly beneath the canopies of perennials than in the surrounding open spaces. Other studies have described spatial associations between perennials with high-nutrient subcanopy soil and the productivity and species composition of understory herbs, and many studies have shown that litter from different canopy species has different effects on different understory species. However, few studies have attempted to isolate nutrient addition from other facilitative mechanisms such as canopy effects on soil moisture and temperature. Since the majority of correlations between enhanced subcanopy soil nutrients and herbaceous communities come from arid and semi-arid environments, confounding effects of shade on these correlations are quite likely. For example, the direct effects of shade provided by nurse shrubs and trees on young cacti are well documented (see above) and shade keeping the tissues of beneficiaries cooler is considered to be the primary mechanism by which facilitation occurs. However, Carrillo-Garcia et al. (2000a,b) conducted a field experiment in which they planted young Pachycereus pringlei (a giant columnar
Chapter 2
79
A
90
80
Shoot dry mass (g)
70
Resource island soil Bare area soil
4
3
2
1
0
3
2
1
11
D
10 9 8 7 6 5 4
0.2
0.1
F
35
Root length (mm)
E
0.3
Root:shoot ratio
B
4
0
C
Shoot dry matter (%)
Survival (%)
100
Plant dry mass (g)
cactus) in soils collected under Prosopis articulata, a common nurse associate, and in soils collected from open areas around the mesquite resource islands. They then placed replicates with both soil types in either 50% (the mean attenuation of light measured under Prosopis in the field) or 100% sunlight. Unlike other nurse plant studies of columnar cactus (Turner et al. 1966, 1969), they found that soil from under Prosopis had stronger positive effects on the survival and whole plant dry mass than reducing insolation (Figure 2.22). More importantly, they found strong interactions between soil and shade effects. In open soil, shade increased the final biomass of Pachycereus by 1.2 g, but in soil from beneath Prosopis shade increased the biomass of Pachycereus by 2.3 g. Interactions like this are observed frequently when researchers investigate more than one mechanism simultaneously, and emphasize the importance of considering synergistic processes when studying plant-plant interactions.
30 25 20 15 10 5
Half sun
Full sun
0
Half sun
Full sun
Shading
Figure 2.22. Shading and soil effects on survival and growth of seedlings of Pachycereus pringlei. Resource island soils are from under Prosopis articulata trees. All main effects are significant for each variable (shading and soil source) with the exception of survival (A) and root length (F). Redrawn from Carrillo-Garcia et al. (2000a) with permission from Plant and Soil.
80
Direct Mechanisms for Facilitation
Like Carrillo et al. (2000a,b), Franco-Pizaña et al. (1995, 1996) conducted experiments in a greenhouse with soil collected under a Prosopis species, P. glandulosa, and found evidence for facilitative effects. There were many other species that were more abundant under P. glandulosa than in the open and this tree appears to play a crucial role in the establishment of many woody species in southern Texas, USA and the conversion of grassland to shrub savanna to contiguous thorn woodland (see Archer et al. 1988) Lars Walker et al. (2001) manipulated shade and the soil resources that develop under shrubs by removing their canopies in the Mojave Desert. Seven months after removing the canopies the ex-shrub understories still had significantly higher organic matter and total and mineralizable nitrogen than open soils, but soil water had decreased. Transplanted Ambrosia dumosa seedlings had the highest survivorship where shrubs had been removed and in the open treatment and lowest under living shrub canopies. They interpreted their results as demonstrating the overriding importance of root and shoot competition over the potential of nurse shrubs to facilitate other plants. This certainly was the case in their experiment, but A. dumosa rarely seems to behave as a beneficiary species (but see Miriti 2006) and therefore is not a good species from which to generalize. Others have shown that A. dumosa and other Ambrosia species are primary nurse species themselves for the seedlings of many other species in deserts and rarely occur under the canopies of other species (Franco and Nobel 1989, McAuliffe 1986, 1988, Holzapfel and Mahall 1999). Parker and Muller (1982) tested shade and soil nutrients as complementary mechanisms for the distinct differences in species distributions under Quercus agrifolia versus grassland in California. In field transplants, they found that Avena fatua, the dominant species in the open grassland, consistently outperformed the understory dominants in full sunlight, regardless of the soil conditions. For the two understory species, Bromus diandrus and Pholistma auritum, the effect of soil was greater than that of light, and these species performed better in soils from under the trees than soils from the open. Quercus douglasii (which benefits from the shade of shrubs when it is a seedling - see above) also has been shown to have important facilitative effects on the biomass of grass species in its understory and this appears to be due to soil nutrients (Callaway et al. 1991). The percent of full sunlight reaching the subcanopy floor beneath individual trees does not correlate with the facilitative effect of the tree on understory biomass, but the effect of trees on soil fertility is highly correlated with its facilitative effect. In greenhouse experiments Bromus diandrus, the dominant understory species, grew better in oak soil or with oak litterfall and throughfall than in open soil with ambient precipitation (Figure 2.23). In the
Chapter 2
81
field however, the manifestation of the positive effects of soil nutrients depended on the abundance of fine oak roots in shallow soil depths. In contrast to Callaway et al. (1991) and Parker and Muller (1982), Weltzin and McPherson (1999) found that nutrient enhancement in the understory of Quercus emoryi was not a mechanism driving higher survival and growth in the understory. Nitrogen and phosphorous concentrations were about 2 times greater under canopies than in grassland, but nutrient amendments did not improve conspecific seedling growth nor did reciprocal transplants of soil indicate that soil characteristics were important. Instead, providing artificial shade increased rates of seedling emergence and survival as much as 19 times and recruitment rates 30-60-fold. From these results they argued that shade from adult canopies provided a “self-enhancing feedback mechanism” (see Wilson and Agnew 1992) that stabilized the lower tree line and ecotone between oak woodlands and grasslands at their sites in southern Arizona. A similar disproportionate importance of shade was reported by Gómez-Aparicio et al. (2005d), who found that microclimatic amelioration due to canopy shading was much more important that nutrient addition in the facilitative effects of shrubs on tree seedlings in Spain, and this effect was particularly pronounced in drier conditions (see Chapter 4). 1.4
1.2 "positive" tree soil "negative' tree soil OPEN SOIL
Biomass (g)
1.2
1.0
a
"positive" tree litter & throughfall "negative' tree litter & throughfall OPEN THROUGHFALL
1.0
a
a
a
0.8
ab
a
0.8
b 0.6
0.6
a
ab c
b
c 0.4
0.4 0.2
0.2
0.0
WOODLAND
SAVANNA
0.0
WOODLAND
SAVANNA
Figure 2.23. Biomass of individual Bromus diandrus plants (a) grown in soil from under ‘positive’ Quercus douglasii trees, ‘negative’ trees (see above) or open grassland and (b) grown in soil from open grassland and watered with precipitation collected in the open, open soil with 0.25 g of ‘negative’ tree litter and watered with ‘negative’ tree throughfall, and open soil with 0.25 g of ‘positive’ tree litter and watered with ‘positive’ tree throughfall. Error bars are 2 SE, different letters represent significant differences among treatments. Reprinted from Callaway et al. (1991) with permission from Ecology.
82
Direct Mechanisms for Facilitation
Walker et al. (1986) demonstrated that nitrogen-rich soil collected from under Alnus tenuifolia growing in riparian soils in central Alaska had positive effects on the biomass of Salix alaxensis and Populus balsamifera seedlings in greenhouse conditions. However, this facilitative mechanism was only weakly manifest in field experiments designed to compare the effects of shade and nutrients because of the strong effects of root interference. Similar positive effects of soils, but negative effects of canopies, were measured in an early successional, Salix-dominated community in the Cascade Mountains of the northwest United States (Jumpponen et al. 1998). There, soil developing beneath the Salix canopies had strong positive effects when removed from the influence of the canopies on shade. In Glacier Bay, Alaska, Chapin et al. (1994) found strong, facilitative effects of Alnus sinuata and Dryas drummondii, both nitrogenfixers, on the growth of Picea sitchensis. These effects came primarily from the inputs of organic matter and associated nitrogen. Naturally occurring and transplanted Picea seedlings had higher nitrogen and phosphorus uptake and tissue concentrations when with Alnus or Dryas, and the addition of Alnus litter enhanced the nutrient status of Picea seedlings. The relative effects of nutrient addition (versus shade or root competition) were much stronger than previously shown on floodplain soils (Walker et al. 1986), and the net effects were positive. This difference was attributed to the much lower N-availability after glacial retreat than after floodplain deposition, and therefore a greater potential for Alnus to facilitate later successional species by ameliorating a highly limiting resource.
2.5.11. Nutrients and conspecific facilitation Plants can modify their nutrient environment in ways that favor conspecifics (Wilson 1992). For example, Wedin and Tilman (1990) and Tilman and Wedin (1991a) found that when plots were planted with different species of perennial grasses the plots developed different levels of available soil nitrogen. Other experiments showed that the two species that produced the lowest nitrogen availability in their soils were the best competitors at low levels of soil nitrogen (Tilman 1991b), suggesting that these species may create soil nutrient conditions that favor conspecifics. Sphagnum bogs are a good example of how plants can “ecosystem engineer” nutrient conditions (Jones et al. 1994, 1997) in a way that favors conspecifics but inhibits other species. Ugolini and Mann (1979), Klinger (1990), and van Breeeman (1995) challenged the traditional bog-succession paradigm in which open water is colonized by Sphagnum and other mosses which ultimately creates habitat suitable for trees. They argued that forest floors are slowly invaded by Sphagnum, which gradually engineers an environment that
Chapter 2
83
is too acidic, anaerobic, and nutrient-poor for forest conifers. They marshaled evidence showing that Sphagnum inhibits conifer roots, Sphagnum increases in small-scale plots in forests over time, forest mortality increases as the area covered by Sphagnum bogs increases, and that Sphagnum detrital remains increase in substrate cores over time while forest detrital remains decrease. The process by which Sphagnum accomplishes this takeover is related to its unusual morphology which allows it to retain tremendous amounts of water proportional to its mass, and its chemical composition. Sphagnum is rich in phenolic chemicals, including a specific acid which only occurs in the genus. This acid appears to virtually eliminate decomposition and creates unusually low soil pH, which is why Sphagnum bogs are famous for preserving prehistoric human and animal remains. Low pH reduces the availability of phosphorus and other nutrients, primarily by decreasing microbial activity, and low phosphorus availability strongly favors Sphagnum (Wilson and Fitter 1984). Living Sphagnum is about 98% pore space due to the presence of hyaline cells and the way the successive generations of plants layer on top of each other. This pore space and the constantly accreting layers of dead Sphagnum store water, which creates highly anaerobic environments. Iron pans and organically clogged soils form under Sphagnum mats. These also contribute to the anaerobic conditions that Sphagnum prefers and that kill conifers. Sphagnum mats also have higher albedos than conifer-dominated vegetation which decreases absorption of solar energy and substrate temperature. Sphagnum species engineer low nutrient and other conditions that either have direct positive feedbacks to their own success and dominance or conditions that are inhibiting to competitors and indirectly facilitate their own success.
2.5.12. Nutrients and the chemical signature of neighbors The most apparent direct positive effect of trees on epiphytes is to provide access to light. However, species-specific spatial associations among host trees and epiphyte species have been interpreted as evidence for other kinds of facilitative effects such as host-specific canopy chemistry (Benzing 1981, Dejean et al. 1995, Kernan and Fowler 1995, Hietz and Briones 1998). Species-specific hostepiphyte relationships may be produced by differences in available nutrients in the throughfall washing through the canopies of different tree species and differences in the nutrient status of soil mats that accumulate in tree canopies. For example, Schlesinger and Marks (1977) demonstrated that the preference of Tillandsia usneoides, “Spanish moss”, for Taxodium distichum, Quercus spp., and Carya glabra and the avoidance of pines by Tillandsia, was correlated with the amount of calcium, potassium, and magnesium in the throughfall under these
84
Direct Mechanisms for Facilitation
trees in the field. In controlled experiments, they found that the foliage of Taxodium, the most common host for Tillandsia in the study area, released 4-20 times the levels of cations, total nitrogen, and total soluble phosphorus than the non-host species Pinus elliottii and Cinnamomum camphora. Lesica (1991) sampled nitrate, ammonium, and a large number of cations in the “soil mats” on different species of tropical forest trees in Costa Rica. They noted that the variation in nutrient content in the canopy varied much more that reported for surface soils in other tropical forests and argued that this variation could be an important mechanism for host-specificity among epiphyte species. In old-growth Picea abies forests in Norway the distribution of Usnea longissima, an epiphytic lichen, was correlated with the mineral composition of foliage (Gauslaa et al. 1998). Other studies have not found important effects of host chemistry on epiphytes. Steve Pennings and I investigated species-specific relationships among Tillandsia usneoides and Polypodium polypodioides epiphytes and 10 host tree species in a coastal plain forest in the southeastern U.S.A. (Callaway et al. 2002). Our results showed that throughfall nutrient content was correlated with the abundance of epiphytes, but experiments demonstrated that nutrient and chemical effects were relatively weak, and were secondary to water availability in determining overall patterns of host preference. In no case did Tillandsia perform better when watered with throughfall from the different host trees than when watered with rainfall collected in the open. In fact, Tillandsia tended to perform worse when watered with the throughfall of some species than when watered with rainfall, suggesting that allelopathic effects might have been more important than fertilization effects. In contrast to the airepiphyte Tillandsia, throughfall nutrients were marginally related to the host preference of the rooted epiphyte Polypoidium. Preferred hosts were “leakier” for 7 of the 11 nutrients we measured, and germination was ≈15% higher when watered with the throughfall of preferred hosts. The concentration of the highly toxic element cadmium was 3 times higher in leachates from the leaves of “bad” hosts than “good” hosts. However, the small positive effect of throughfall changed with life stage; there were no differences in the effects of throughfall from good versus bad hosts on mature Polypodium ramets. The most overlooked area for potentially paradigm-changing facilitative (and competitive) research may be understanding interactions among species with complementary root exudates. We know that plants exude a very large number of chemicals from their roots and that different plant species exude different chemical cocktails. These exudates are comprised of “a complex mixture of organic acid anions, phytosiderophores, sugars, vitamins, amino acids, purines, nucleosides, inorganic ions, gaseous molecules, enzymes and root border cells which have major direct or indirect effects on the acquisition
Chapter 2
85
of mineral nutrients required for plant growth” (Dakora and Phillips 2002). Many of these chemicals also detoxify compounds in the soil (Kochian et al. 2004). Since different species produce such strikingly different assortments of biochemicals, it is reasonable to hypothesize that some species may benefit by the presence of other species that release exudates that alter the environment in favorable ways. There is growing evidence for these sorts of interactions. Trees in the eastern deciduous forest appear to have species-specific effects (also see Chapter 5) on soil calcium due to the particular organic acids exuded from their roots. Dijkstra et al. (2001) measured organic acidity and its degree of neutralization in the forest floor under six North American tree species. They concluded that the “quantity, nature and degree of neutralization of organic acids differ among the different tree species.” Leaching with organic anions was greatest under Acer saccharum and Fraxinus americana where soils contained more exchangeable cations than did the acidified soil under Tsuga canadensis. Variation in organic acids leached from the roots of different species appeared to create species-specific differences in calcium availability (Dijkstra and Smits 2002, Dijkstra 2003). A good example of root exudate-mediated facilitation appears to occur with chemicals that are released for phosphorus acquisition. Plant roots release phosphatase to make organic P available, and roots release a bewildering array of different chelators that separate elements from inorganic P, a process that makes P available for root uptake. The effect of a chelating chemical depends on what element binds P in a particular soil. If a species does not exude a chelator that functions well in a certain soil type, it might benefit that species to grow next to a neighbor that does. Fusuo Zhang and Long Li (2003) of the China Agricultural University reviewed the processes involved in intercropping and found facilitation between different crop species that may have been related to shared effects of root exudates. Maize improved iron uptake for intercropped peanut plants, faba beans enhanced nitrogen and phosphorus uptake by intercropped maize, and chickpeas facilitated P uptake by associated wheat. In the last case, Zhang and Li hypothesized that P depletion in chickpea rhizospheres induced the release of substances that mobilize organic P. Centaurea diffusa, a highly invasive weed in the northwestern US, exudes an allelopathic compound from its roots called 8-hydroxyquinoline (8HQ, Vivanco et al. 2004). 8-HQ is an excellent chelator in volcanically derived soils increasing P availability by 2-3 times (T. DeLuca and R.M. Callaway, unpublished data). In earlier experiments on the allelopathic effect of C. diffusa Erik Aschehoug and I found that adding activated carbon as an adsorbant of organic molecules almost eliminated the uptake of labeled P by allelopathically tolerant European grass species (Callaway and Aschehoug 2000). While not
86
Direct Mechanisms for Facilitation
quite the appropriate experiment, these results suggest that European grass species might benefit substantially from the exudation of a chelator from C. diffusa neighbors. Such P facilitation might be more common than we realize. Li et al. (2003) found that total P uptake by Triticum aestivum (wheat) plants was 68% greater when its roots were allowed to mingle with the roots of Cicer arietinum (chickpea) and 37% greater when exudates were allowed to pass through a nylon mesh than when root and exudate were separated with a solid barrier. Cicer arietinum roots facilitated P utilization of T. aestivum. The potential for root-exudate mediated facilitation gets even more complicated and fascinating with recent evidence for cross talk among roots (Weir et al. 2006). They found that that Lupinus sericeus and Gaillardia grandiflora increase root exudation of organic acids in response to a phytotoxic compound released by the invader Centaurea maculosa. Oxalic acid reduces oxidative damage generated by the allelotoxins exuded by C. maculosa. Furthermore, native grasses are highly associated with Lupinus in communities invaded by C. maculosa (Figure 2.24) and field experiments show that Lupinus indirectly facilitates native grasses in Centaurea stands (Figure 2.25, also see Chapter 3). This facilitation is correlated with the presence of oxalic acid in the soil. Addition of exogenous oxalic acid to native grasses alleviated the allelopathic effects of catechin, indicating that root secreted oxalic acid may act as a chemical facilitator for plant species that do not produce the chemical. 20
Cover (%)
Lupinus in plot No Lupinus in plot 15
10
5
0
Pseudoroegneria
Festuca
Figure 2.24. Cover of native grasses in communities invaded by Centaurea maculosa in plots containing Lupinus sericeus and in plots without Lupinus. Error bars show 1 SE. Festuca cover and Pseudoroegneria cover is significantly different between treatments (t-test, P<0.05). Drawn from data reported in Weir et al. (2006) with permission from Planta.
Chapter 2
Biomass (g)
4
87
In Centaurea stand
Not in Centaurea stand
With Lupinus Without Lupinus
3
2
1
0
Pseudoroegneria Festuca
Pseudoroegneria Festuca
Figure 2.25. Biomass of native grasses experimentally transplanted next to Lupinus sericeus in communities invaded by C. maculosa and communities not invaded. In a two-way ANOVA for Pseudoroegneria, FCentaurea x Lupinus effect = 24.50, df =1,59, P<0.001. In a two-way ANOVA for Festuca, FCentaurea x Lupinus effect = 12.04, df =1,56, P<0.001. Reprinted from Weir et al. (2006) with permission from Planta.
In summary, there are many examples of plants facilitating other plants by improving nutrient availability. This is important because competition for nutrients is often seen as a fundamental determinant of community composition. If plants can facilitate each other by improving the availability of fundamental resources such as water and nutrients, the rules of interactive engagement that affect the composition of communities are based on much more complicated and fascinating processes than simple rates of uptake and utilization.
2.6. WIND Wind can damage plants by decreasing leaf and meristem temperatures, increasing vapor pressure differences between leaf and air, abrading tissue with particulates, or simply by battering leaves and branches against themselves and nearby objects (Wardle 1968, Marchand 1972, Marchand and Chabot 1978, Tranquillini 1980, Hadley and Smith 1983, 1986, 1989, Grace 1989, Arno and Hammerly 1990). Wind also increases soil erosion (Armbrust and Bilbro 1997). For example, they estimated that a canopy cover of only 4% reduced the transport capacity of a 16 m s-1 wind by
88
Direct Mechanisms for Facilitation
50%. High wind speeds can reduce the net carbon assimilation rates of alpine plants for one to two days after wind events are over (Hadley and Bliss 1964). In environments where wind is the limiting factor for growth and survival, plant species that do not adapt to the wind through mechanisms such as leaf area reduction, thickened cuticles, or stronger root bases must avoid wind by sheltering beneath or behind other plants or topographic structures (Osmond et al. 1987). The benefits of sheltering are illustrated by models describing the relationship between plant cover and the movement of air. These models demonstrate that plant cover is highly correlated with reduction in threshold wind velocity and transport capacity of wind. Clearly, plants can benefit from being protected from wind by other species. The effect of established plants on the movement of windblown particles is also addressed below in the section on disturbance. The effects of wind are also connected to the effects of temperature. For example, summer winds in the Rocky Mountains have been estimated to induce a 3-6oC cooling at the plant surface in fellfield communities (Bliss 1985). Tree “islands” or “ribbon forests” (described earlier in this chapter) often develop in windy subalpine areas, with seedling regeneration restricted to the leeward side of the islands or ribbons (Billings 1969, Pontecorvo and Bokhari 1975, Marr 1977, Minnich 1984). Mature trees directly buffer seedlings from harsh alpine winds, but they also improve soil moisture by altering the patterns of windflow over the land’s surface and snow deposition. Meristem tissue temperatures are substantially higher within the branches of krumholtz and windward needles have lower winter water content, lower water potentials, lower cuticular resistance, and higher transpiration rates (Hadley and Smith 1983, 1986, Grace 1989). For example, in the Medicine Bow Mountains of Wyoming, needles of Abies lasiocarpa buried by snow had autumn and winter water potentials 1.0 to 1.5 MPa higher than those above the snow pack (Hadley and Smith 1986). Neighboring trees or packed snow also protect leaves from extensive photoinhibiton damage during the winter (see above, (Ball 1994, Germino and Smith 1999, 2000). Snowdrifts that accumulate in the lees of trees provide more soil moisture as they melt, adding to the direct buffering effect of nearby trees. Holtmeier and Broll (1992) found that snow accumulated inside and leeward of tree islands in Colorado, increasing soil moisture. Tree islands also affect tundra soils. Tim Seastedt and Gina Adams at the University of Colorado found that “mobile tree islands”, gradually moving across the alpine landscape due to leeward recruitment and windward mortality, left discernable differences in soil organic matter and nitrogen storage potential that may persist for centuries (Seastedt and Adams 2001).
Chapter 2
89
Similar processes may occur at the margins of some high-elevation meadows and at timberline. In the Oregon Coast Range, Abies procera establishes primarily at forest-meadow ecotones, but much less either in the open meadow itself or in the interior of the forest (Magee and Antos 1992). Similar patterns have been described for Pinus uncinata recruitment in the central Pyrenees where seedlings are associated with krummholtz vegetation which apparently buffer recruits from the harsh environment (Camarero and Gutiérrez 1999). Alftine and Malanson (2004) found that empirical and modeled community composition at timberline corresponded with protection from wind. Based on the accumulating evidence for wind-based facilitation at tree lines, Bill Smith and colleagues (2003) proposed “another perspective on altitudinal limits of alpine timberlines”. They argued that most studies of the causes of the tree-alpine ecotone have come from studies on older trees, and not the mechanisms by which seedlings establish away from the forest edge in the tree line ecotone. They cite evidence that protection from severe mechanical damage from wind is provided by other trees. Other facilitative processes appear to occur throughout the summer growth period which enhance root growth, ameliorate drought stress, and increase seedling survival. There have been few studies that have tried to experimentally manipulate the effects of wind independently of other factors. Dayna Baumeister and I examined the relative importance of several interacting mechanisms important to the facilitative effects of Pinus flexilis on Pseudotsuga menziesii and Ribes cereum at the ecotone between the Rocky Mountain forests and Great Plains grasslands in Montana (Baumeister and Callaway 2006). This environment is known for strong, unidirectional winds, highly variable temperatures, and frequent winter Chinook winds. Chinooks are catabatic winds that dramatically increase air temperature as the air compresses. Sixty-nine percent of Pseudotsuga seedlings and saplings and 91% of Ribes were located beneath P. flexilis. The effects of overstory P. flexilis were more important at a windward site than at a leeward site, suggesting that wind amelioration was a dominant mechanism. We manipulated wind with Plexiglass wind shields, shade, and snowdrift accumulation separately and after two years found that shade, not wind, was of overwhelming importance for the survival of both Pseudotsuga and Ribes. However, once shade was provided, the effects of wind amelioration significantly increased survival of Pseudotsuga and Ribes. In Swedish Lapland, Carlsson and Callaghan (1991) showed that Carex bigelowii increased in leaf length and culm height when growing within clumps of Empetrum hermaphroditum or Racomitrium lanuginosum. They experimentally manipulated the effects of wind by erecting artificial shelters and
90
Direct Mechanisms for Facilitation
found that the shelters elicited similar positive responses from Carex. Shevtsova et al. (1995) compared the effects of coexisting shrub species (Empetrum nigrum, Vaccinium vitis-idaea, and Vaccinium uliginosum) on each other’s growth, branching, survival, and reproduction in northern Finland and found that the growth of E. nigrum and V. uliginosum was positively correlated with the cover of V. myrtillis. Furthermore, the removal of V. uliginosum or V. myrtillis resulted in decreased growth of E. nigrum and V. uliginosum, respectively. They hypothesized that a primary mechanism for this may have been protection from wind provided by the more stress-tolerant species. Shevtsova et al.’s results were surprisingly similar to those reported earlier by Maillette (1988) for three Vaccinium species, including V. uliginosum, in the arctic-alpine tundra on the Laurentian Plateau of Quebec, Canada. She found that the biomass of individuals of each Vaccinium species was lower outside of its own “area of dominance” (where it was proportionally more abundant) unless a plant was in contact with the locally dominant congener. In other words, one of the best positive predictors of the biomass for less abundant Vaccinium species at a site was the biomass of the locally dominant Vaccinium species. Microsite effects could have caused these correlations, but as Shevtsova et al. thought, Maillette hypothesized that protection from wind from dominant congeners was important for the success of subdominant species at a particular site. Gerdol et al. (2000) experimented with some of these same Vaccinium species in the alpine tundra of the Apennines in Italy by reciprocally removing one of the three dominant shrub species, Vaccinium myrtillus, V. uliginosum, or Empetrum hermaphroditum. In some cases neighbor removal had no effect, in others neighbor effects were competitive, and for some species neighbor effects were facilitative (Figure 2.26). Jonasson (1992) experimented with the effects of Betula nana on Vaccinium in the tundra of northern Sweden and found that Vaccinium myrtillus, various lichen species, and diversity in general, declined after Betula removal. This corresponded with his observation that V. myrtillus was much more abundant and healthy under the canopies of Betula and Salix shrubs than in the spaces between the shrubs. However, it was not clear if wind was an important mechanism.
Chapter 2
91 30
Empetrum hermaphroditum
Growth of current year shoots (%)
20 10
40
Vaccinum uliginosum
** 30
0 -10
20
-20 -30 -40 -50
*
1995 1996 1997
** **
Fertilized
V. uliginosum removed
60
0
Fertilized
E. hermaphroditum removed
60
Vaccinum uliginosum 40
10
**
**
*
Vaccinum myrtillis 50
**
40
*
30
20
20 0
10 0
-20
-10 -40
Fertilized
V. myrtillis removed
-20
Fertilized
V. uliginosum removed
Figure 2.26. Differences in the lengths of current-year shoots between ramets in plots with and without fertilizer, and with and without neighbor removal. “F” designates the fertilizer treatment, and error bars = 1 SE. Species names at top of each panel indicate target species. Differences were calculated as [(treated – untreated)/untreated). –Vm = removal of Vaccinium myrtillus, -Vu = removal of Vaccinium uliginosum, -Eh = removal of Empetrum hermaphroditum. Redrawn from Gerdol (1998) with permission from the Journal of Ecology.
In subalpine environments of the Snowy Mountains of southeastern Australia winter wind appears to affect tree lines. Egerton and Wilson (1993) found that growth and stem elongation of Eucalyptus pauciflora, the dominant tree species at timberline, decreased over the winter in grassland but not in shrubland. Furthermore, the removal of neighbors had negative effects on stem elongation in both habitats, suggesting that the presence of neighbors was facilitative. They hypothesized that the erect growth form of Eucalyptus exposed it to winds and scouring snow and ice during the winter and that by associating with protective vegetation damage could be reduced. In other experiments in the same system, Wilson (1993) found that the presence of neighbors was positive for the survival of Poa costiniana, a mid-altitude grass, but negative for the growth of the grass. Many authors have reported strong spatial associations among species in extreme alpine environments where wind is consistent and intense (Whitehead 1951, Griggs 1956, Bonde 1968, Blundon 1983, Sohlberg and
92
Direct Mechanisms for Facilitation
Bliss 1984, Alliende and Hoffman 1985, Kikvidze 1993, 1996, Kikvidze and Nakhutsrishvili 1998, Anderson and Bliss 1998, Pokatzhevkaya 1998, Cavieres et al. 1998, Nunez et al. 1999, Choler et al. 2001, Cavieres et al. 2002, Arroyo et al. 2003, Badano and Cavieres 2006a,b). Most of these studies reported high degrees of clumping of some species inside cushions or tufts of other species. Most of these studies are correlational, but some experiments support the argument that these associations are due to facilitative interactions and have hypothesized that protection from wind may be a major factor (Callaghan 1987, Choler et al. 2001, Callaway et al. 2002). The importance of experimental separation of microsite effects from real facilitation in alpine vegetation, however, was emphasized in a study by Jon Moen (1993). He noted that many of the species found associated with cushion plants or with other vegetation can also be found away from them, and in some cases germination may be as high in the open as near putative nurses (Bliss 1971, Pysek and Liska 1991). Moen conducted experiments in the mountains of northern Norway (69o56’N) at 850-m elevation in an alpine “block field” by sowing seeds of Oxyria digyna in plots with Ranunculus glacialis and where R. glacialis had been removed. Instead of facilitation, he found higher germination in one year where neighbors had been removed (suggesting competition) and no effect of neighbors in another year. Other experiments conducted in alpine and arctic plant communities have also demonstrated net competitive interactions among species (Chapin and Shaver 1985, Theodose and Bowman 1997, Olofsson et al. 1999). Competition must surely occur in arctic and alpine communities, and a recent large-scale and multi-species experiment in the French Alps demonstrates the potential pitfalls that may occur when we attempt to generalize from studies of one or few species and when we do not recognize the conditionality inherent to plant interactions (also see Chapter 4). These pitfalls beset both studies of facilitation and competition. Choler et al. (2001, also see Callaway et al. 2002) conducted a neighbor removal experiment using large proportions of the species in natural alpine plant communities of the southwestern Alps. The purpose was to test for the relative importance of positive and negative interactions along elevational and topographical gradients. The effects of neighbors on the biomass, growth rate, survival, and reproduction of five target species were tested at each of six different experimental sites, half of which occurred on convex topography exposed to the wind, and the other half in concave, sheltered sites. They measured both strong competitive interactions and strong facilitative interactions, depending on the particular species involved and the environmental conditions. Competitive effects were stronger in sheltered sites and facilitation was more common in exposed sites. At high-elevation and exposed sites, experimental
Chapter 2
93
evidence for facilitation was coupled with small-scale spatial associations among species. However, spatial disassociation was not coupled to experimental evidence for competition at any sites. These experimental results, in contrast with those of Olofsson et al. (1999), indicate that strong spatial associations are often indicative of positive interactions in harsh, windswept environments. On exposed ridges near the timberline in the northern Rocky Mountains of Montana and Idaho, Abies lasiocarpa of all sizes are often highly aggregated around mature Pinus albicaulis (Callaway 1998b; also see Chapter 6). Tree ring analyses indicate that the growth rates of mature Abies at timberline decreased by 24% when adjacent P. albicaulis died, suggesting an important facilitative effect of P. albicaulis on A. lasiocarpa. These spatial associations and growth responses did not occur at lower elevations indicating that positive interactions among conifer species are only important in the highly stressful conditions that occur at the altitudinal limits of tree growth. In fact, at lower elevations the interactions were competitive. One hypothesis for this relationship is that the highly stress-tolerant Pinus albicaulis provides shelter from wind desiccation and abrasion. At high-elevation study sites, large Abies were 2-4 times more aggregated with Pinus albicaulis trees than were Abies seedlings, a pattern that would not be expected if shade were the important mechanism. Furthermore, growth rates of large Abies were significantly lower when they were more than 3 meters from a mature Pinus than when they were adjacent to either living or dead Pinus albicaulis, but the growth rates of small saplings did not differ among these microsites. Stronger facilitative effects on mature trees than on seedlings or saplings may develop because the winter snow pack protects small Abies from blowing ice and snow, but as trees grow above the snow pack, shelter from other objects may be crucial. Damaged foliage at the tops of small saplings on open ridgelines was common. Arno and Hammerly (1990) also hypothesized that ice abrasion and mechanical injury during high winds may determine timberline limits and krummholz formation in the northern Rocky Mountains, and exposure to wind has been shown to cause needle mortality due to cuticle degradation and subsequent desiccation for Abies lasiocarpa (Hadley and Smith 1983, 1986). Wind may also create misleading patterns that may be falsely interpreted as facilitation (see section 11 in this chapter). Perennial plants often act as filters or traps for seeds in windy environments, and the consequent collections of plants under canopies may have nothing to do with the canopies’ positive effects. In fact, wind may redistribute seeds into environments in which the germinating plant experiences even greater competition than it would otherwise.
94
Direct Mechanisms for Facilitation
2.7. SOIL OXYGENATION In 1958 Coult and Valance demonstrated that the roots of rice plants were able to oxidize the anaerobic media in which they grew. Later, Armstrong (1964) measured oxygen diffusion from the roots of three bog species in Great Britain and found that the roots of all three species released oxygen, but when leaves were sealed with Vaseline oxygen diffusion from the roots dropped immediately. These findings have very important implications for they way plants in these environments interact because low levels of soil oxygen often limit plant growth in wetlands (McDrew 1983). Since these early experiments a number of other experiments have shown that emergent wetland plants are able to passively transport oxygen from leaves to roots through aerenchymous tissue. This appears to alleviate the effects of low soil oxygen for the transporters themselves and on soil processes very close to the roots (Armstrong 1979, Dacey 1981, Armstrong and Armstrong 1990, Grosse et al. 1991). Some experiments indicate that oxygen transported belowground is consumed only by the plant itself or associated microbes (Bedford 1991), but other studies show that oxygen can leak out of submerged roots, oxidize nutrients and toxic substances in the rhizosphere, and oxygenate marsh sediments (Howes et al. 1981, Armstrong et al. 1992). Armstrong (1964) noted that the three species he worked with released oxygen to highly varying degrees, and hypothesized that the variation may relate to the highly anaerobic soils that the species are able to tolerate. Such variation may also contribute to differences in the ability of different wetland species to facilitate neighboring plant species. Experiments that directly demonstrate soil oxygenation by one species yielding a facilitative effect on another species, while controlling for other correlated effects, have not been conducted to my knowledge, but a number of studies provide strong evidence for the process. For example, Plantago coronopus and Samolus valerandi are tightly clumped around tussocks of the aerenchymous Juncus maritimus in dune slacks on the coast of Holland, where survival rates appeared to be enhanced by soil oxygenation and oxidation of iron, manganese, and sulfide (Schat and Van Beckhoven 1991). When grown with Juncus maritimus in the greenhouse, growth and nutrient uptake of Plantago and Centaurium were improved (Schat 1984). In southern Spain, Spartina maritima aerates surface sediments in coastal marshes, which creates conditions favorable for the growth of Arthrocnemum perenne (Castellanos et al. 1994). In eastern salt marshes in North America, Hacker and Bertness (1994, 1995) found that the aerenchymous Juncus maritimus increased the redox potential in its rhizosphere. When Juncus was removed experimentally the growth of Iva frutescens, an associated woody-stemmed perennial, decreased substantially. They thought that oxygenation by Juncus extends the distribution of Iva to lower elevations in the
Chapter 2
95 60
Salinty (ppt)
50 40 30 20 10 0
Redox (mV)
100
0
-100
-200
No removal of vegetation Removal of vegetation
-300
HIGH JUNCUS
MID JUNCUS
LOW JUNCUS
Figure 2.27. Effects of biomass removal on a) soil salinity, and b) soil redox potential at three different elevations in the Juncus gerardi zone. Error bars are 1 SE. Redrawn from Brewer et al. (1997) with permission from Oikos.
marsh. In the same marsh, Brewer et al. (1997) manually removed Juncus gerardi at three different elevations and measured soil responses. They found that soil redox potential decreased dramatically at all elevations in the marsh when Juncus was removed, but the magnitude of the difference was greater at low elevations (Figure 2.27). In the low marsh zone removal of Juncus and decreasing soil redox potential was associated with lower species richness. Similar results have been reported for plants from freshwater marshes. Typha latifolia has a highly effective whole-plant aeration system based on internal convective flow of gases (Bendix et al. 1994). Leah King and I found that undrained pots containing Typha latifolia had dissolved oxygen contents over 4 times greater than pots without cattails (Callaway and King 1996). Other marsh plants grown with Typha survived longer and grew larger than in pots without cattails when pot substrates were kept between 11 and 12oC (Callaway and King 1996; Figure 2.28). In the field, Myosotis laxa plants growing next to transplanted Myosotis were larger and produced more fruit that those isolated from cattails. However, the effective oxygenation of soils by aerenchymous
96
Direct Mechanisms for Facilitation
*
o
11-12 C
*
*
200
200
*
Mass (g)
0
0 o
18-20 C 200
WITH Typha WITHOUT Typha
200
Length (mm)
100
100
100
100
* 0
ROOT LENGTH
ROOT MASS
SHOOT MASS
0
Figure 2.28. Growth characteristics of Myosotis laxa when grown for 40 days with and without Typha latifolia in the greenhouse at two different substrate temperatures. Error bars represent 2 SE. and * represent significant differences between treatments as P < 0.01, Student’s t-test). Reprinted from Callaway and King (1996) with permission from Ecology.
plants may depend on particular soil conditions. Jespersen et al. (1998) found that Typha latifolia substantially improved soil redox potential compared to soils with no Typha, but that the effect of Typha depended on the carbon content and quality of the substrate. When acetate was added to the soil (to saturate carbonreducing or methanogenic bacteria) the positive effect of Typha on soil redox potential was much less. When King and I increased soil temperatures to 18-20oC, the oxygenation of soils and the facilitative effect of Typha disappeared. Large stands of Spartina alterniflora appear to benefit from communal rhizosphere oxygenation. Spartina seedlings and small plants are generally not able to oxygenate marsh soils (but see Huckle et al. 2000), but large groups, which may be clonally interconnected, increase oxygen in the soil and enhance growth (Bertness 1991b). Increased growth may lead to a positive feedback loop between increased soil oxygen and plant productivity (Howes et al. 1981, 1986, de la Cruz et al. 1989). Substrate oxygenation via the roots of some plant species may also elicit positive effects on other plants indirectly by preserving nitrifying capacity and nitrate availability. Engelaar (1995) compared the effects of radial O2 loss from
Chapter 2
97
the roots of the flooding-resistant Rumex palustris, which has aerenchyma in its roots, and the flooding-intolerant, non-aerenchymous Rumex thyrsiflorus in welldrained and waterlogged soils with an initially high nitrifying capacity. In the well-drained soils nitrification was high for both species, and both species had high levels of nitrate reductase in their leaves (indicative of an adequate supply of nitrate). In waterlogged soils with R. thyrsiflorus, nitrification was inhibited and ammonium accumulated. Waterlogged soils planted with R. palustris had redox potentials high enough to keep O2 replenished and maintained high nitrifying capacities (Figure 2.29). However, oxygenation does not consistently cause higher nitrification rates. In field experiments that showed evidence for the positive effects of oxygen loss from Typha latifolia on Myosotis laxa growth and reproduction, soil nitrate and ammonium did not differ between sites with planted Typha versus those without Typha (Callaway and King 1996). Experiments conducted by Levine et al. (1998) indirectly support the complex and intertwined roles of root-mediated soil oxygenation and soil nitrogen in positive interactions. They manipulated Iva frutescens-Juncus gerardi combinations (see above) in the field by adding nitrogen and found that nitrogen had strong mediating effects on the positive effects shown previously for Juncus on Iva (Figure 2.30), (Bertness and Hacker 1994, Hacker and Bertness 1995, Brewer et al. 1997). Iva plants that grew without nitrogen fertilization, but in the presence of Juncus, were much smaller and produced many times fewer flowers than Iva plants in plots where Juncus had been removed. Increasing
Redox potential (mV)
500
400
300
200
100 20
30
40
50
60
Days Figure 2.29. Redox potentials in well drained soils (open symbols) and water-logged soils (closed symbols) planted with either Rumex thyrsiflorus (triangles) or R. palustris (circles). Water logging started on day 21. Each point is the mean of two replicates. Redrawn from Engelaar et al. (1995) with permission from Biology and Fertility of Soils.
Direct Mechanisms for Facilitation
Whole plant leaf biomass (g)
98
0.6 0.5 0.4 0.3 0.2 0.1 0.0 not fertilized fertilized Plot 1 Zero
Flower number
400
300
200
100
0
Juncus present
Juncus removed
Figure 2.30. Characteristics of Iva frutescens in fertilized and unfertilized treatments and with and without Juncus gerardi. Reprinted from Levine et al. (1998) with permission from Oecologia.
nitrogen availability caused Iva to increase in size both with and without Juncus, but interestingly, nitrogen addition allowed Iva to fully compensate, both in leaf biomass and flower number, for the effect of removing Iva’s benefactor Juncus. Importantly, nitrogen addition did not alter the ability of Iva to tolerate salt stress. However, they did not measure proline differences in leaves, nor did they consider the possibility that substrate oxygenation and facilitation by Juncus may have functioned indirectly by increasing available nitrogen (e.g. Engelaar et al. 1995). Even so, nitrogen addition appeared to eliminate the need for a benefactor that oxygenated the soil rather than just increasing the salt tolerance of Iva. Plants also may oxygenate the soil in ways other than transport of oxygen through their tissues. Berendse and Aerts (1984) reported that Molina caerulea did not appear to transport oxygen belowground, but transpired rapidly enough to reduce the depth of water in waterlogged containers. Lower water
Chapter 2
99
levels permitted aeration of the shallow root system of Erica tetralix which grew better in mixtures of Molina than alone. In summary, facilitation through soil oxygenation is often overlooked; but many plants adapted to anoxic soils appear to ameliorate conditions for less adapted plants by providing oxygen or altering soil chemistry.
2.8. SUBSTRATE Trees and shrubs often facilitate the growth and survival of other species by providing a substrate for them to grow on. For example, as described in detail above epiphytes benefit by gaining cheap access to light, or by escaping superior competitors. Vascular epiphytes are important features of subtropical, tropical, and temperate rainforests, making up about 10% of the world’s total flora and up to 25-35% of the plant diversity in tropical forests (Gentry and Dodson 1987). Many epiphytic species possess only rudimentary root systems and as a consequence are obligately dependent on trees for their substrate, providing a rather obvious example of facilitation. Epiphytism is ubiquitous in temperate forests as well; however, the beneficiaries of trees in these systems tend to be lichens and mosses. Epiphytes in marine seaweed and seagrass communities also constitute high proportions of the community diversity and productivity. Many authors have shown that terrestrial epiphyte-host species combinations are not random associations (Went 1940, Johansson 1974, Benzing 1981, Bennett 1986, Jernakoff and Nielson 1988, ter Steege and Cornelissen 1989, Migenis and Ackerman 1993, Dejean et al. 1995, Kernan and Fowler 1995). Others have shown that some apparently suitable host species are virtually unoccupied by some species of epiphytes (Schlesinger and Marks 1977). Lianas, like epiphytes, depend on the structural biomass of other species for substrate, but acquire nutrients and water from the soil. As for vascular epiphytes, patterns of liana host specificity have been described in tropical and temperate systems (Daniels and Lawton 1991, Talley and Lawton 1996, Talley and Setzer 1996). Despite the large amount of literature that describes correlations between host and epiphyte species, there has been little experimental work in terrestrial systems. This work has been primarily limited to the comparison of the effects of bark and leaf leachates and extracts on germination and growth under controlled conditions (Frei and Dodson 1972, Schlesinger and Marks 1977, Talley et al. 1996). For example Talley and Lawton (1996) found that the vines of Toxicodendron (Rhus) radicans, the climbing liana commonly known as poison ivy, are more common on some host tree species than on others. Experiments with seeds demonstrated that bark extracts from various hosts all decreased germination relative to controls, but that the extracts from
100
Direct Mechanisms for Facilitation
preferred hosts had weaker negative effects than extracts from avoided hosts. Transplant experiments with other species demonstrated that two dominant epiphytes had higher growth rates in some tree hosts than in others (Callaway et al. 2002). The basic need for substrate may dominate the interactions among woody hosts and epiphytes, but studies cited below show that other mechanisms are also important. Epiphyte-host relationship may be commensal, with only the epiphyte benefiting and the host left unaffected; however, epiphytes may damage host tissues as well establishing a somewhat parasitic interaction. On the other hand, epiphytes may return the favor to their hosts by creating a nutrient pool in the tree canopies that some tree species can access through aerial roots (Nadkarni 1986). Additionally, although probably not much of a payback, epiphytes also reduce branch temperatures by 2-5oC, perhaps reducing respirational costs for their hosts (Freiburg 2001). Exotic plant species generally have negative effects on native species. However, the effects of Celastrus orbiculatus, a partially epiphytic vine introduced from Asia, on developing forest communities in New England are complex (Fike and Niering 1999). They observed that Celastrus dominates some parts of the forest canopies “like a carpet being unrolled”, but in an unusual twist the exotic vine appears to provide important substrate for a native vine, Vitis labrusca, by serving as a “ladder” to carry Vitis into the tree canopies. In the presence of Celastrus, Vitis appears to occur in greater abundances, and have stronger effects on other native species. Although epiphytism and hemi-epiphytism are the most intuitive and obvious example of plant species providing substrate for other species, plants can also enhance substrate conditions in other ways. Perennials can facilitate other species by altering the physical characteristics of the substrates beneath them. In California woodlands, litter deposition and root penetration under Quercus douglasii appears to develop soils with lower bulk densities, which presumably also increases soil water storage capacity (Kay 1987, Callaway et al. 1991). In dry lakes of the Mojave Desert, Kochia californica establishes in cracks in very fine textured soil and eventually collects small mounds of coarser textured soil that Atriplex torreyi can colonize (Vasek and Lund 1980). Atriplex torreyi accumulates large mounds of sandy soil, which coalesce across the landscape and provides habitat suitable for Atriplex confertifolia and Haplopappus acradenius. Large shifts in soil chemistry are correlated with these soil textural and microtopographic changes. In similar desert environments of the Great Karoo in South Africa, Yeaton and Elser (Yeaton 1990) proposed a similar cycle of substrate development, facilitation, and succession (see Chapter 6). In this system, open exposed soil is colonized by species in the Mesembryanthemaceae family which accumulate soil and
Chapter 2
101
build mounds. These mounds are colonized by several species of woody shrubs which eventually outcompete the early mound builders. Trichglochin maritima occurs commonly in “pannes” in New England salt marshes where the microtopography creates unusually high levels of waterlogging. Trichglochin roots and rhizomes build raised rings in these pannes that are occupied by other species. Fogel et al. (2004) studied the effects of raised Trichoglochin rings on other panne species in field experiments with natural Trichoglochin rings, experimentally lowered rings, artificially raised mud, and bare mud. They transplanted four perennial species, followed naturally occurring seedlings, and monitored edaphic conditions to assess treatment effects over two growing seasons. Raised substrate treatments showed a decrease in waterlogging and an increase in reductive potential (greater oxygenation) than non-raised treatments and all transplanted species showed increased success in raised treatments. There was no significant effect of the presence of Trichglochin without raising the substrate. Incompletely decomposed remains of plants may also facilitate the recruitment of other species by providing substrate (see review by Harmon et al., 1986). In Sitka spruce-western hemlock (Picea sitchensis and Tsuga heterophylla) forests of northwest America 94-98% of tree seedlings are found on “nurse logs” that cover only 6-11% of the forest floor (Franklin and Dyrness, 1973; Grahm and Cromack, 1982; McKee et al., 1982; Harmon, 1985). Logs are also important seedbeds in other Pacific Northwest forests (Christy and Mack, 1984). Griffith (1931) and Smith (1955) found much higher growth of Abies lasiocarpa and Picea englemanii seedlings on nurse logs in subalpine forests in British Columbia than on the forest floor. In a series of elegant experiments, Harmon and Franklin (1989) determined that the primary role of these nurse logs during recruitment of P. sitchensis and T. heterophylla was to provide a substrate free from competition with herbs and mosses on the forest floor (Figure 2.31, see discussion of indirect interactions in Chapter 3). First, the survival of tree seedlings was highly negatively correlated with the abundance of mosses in test plots. Furthermore, in sites either sterilized for pathogens and cleared of moss competitors, or just cleared of mosses, survival of seedlings of P. sitchensis, T. heterophylla, Thuja plicata, and in some cases Pseudotsuga menziesii, was significantly higher than sites with mosses. Similar patterns involving nurse logs have been documented in the temperate rain forests of Chile, where all seedlings of some species are found on nurse logs (Christie and Armesto 2003).
102
Direct Mechanisms for Facilitation
Seedlings per 100 seeds
5 Picea sitchensis Pseudotsuga menziesii
4
Thuja plicata Tsuga heterophylla
3
2
1
0
0 0
0
0
On ground On nurse log On ground not protected protected
Figure 2.31. Survival of conifer seedlings for >1 year on logs, or on soil protected from mosses not protected from mosses. Error bars show 1 SE. Redrawn by estimating presentation in Figure 2 and results presented in text in Harmon and Franklin (1989) with permission from Ecology.
At 5,000 m altitude in the Annapurna Range of the Himalaya, the hummock forming species Chesnya nubigena occurs on glacial moraines and appears to facilitate other species (Jacquez and Patten 1996). Soil in Chesnya hummocks is higher in organics, the canopy accumulates snow, and daily temperature fluxes are moderated compared to surrounding gravels. The mounds developed by Chesnya are occupied by several other species that are rare on the open gravel. These beneficial effects of Chesnya on the associated grasses were not demonstrated experimentally, but because Chesnya is known to develop the mounds, the strong association of herbaceous species with mounds is likely a facilitative by-product of Chesnya. Similarly, Empetrum rubrum, a dominant species in early succession in recently deglaciated valleys in southern Chile, facilitates Nothofagus antarctica by developing substrate (Henríquez and Lusk 2005). They found that seedlings with an intact cover of E. rubrum showed significantly higher survival, growth and leaf number than seedlings which had been exposed experimentally. Partially decomposed remains of wetland plants and the sediment trapped in their remains provides habitat that is colonized by different plant species that are less tolerant of submersion (Barko and Smart 1986, Bertness 1988). In the Santee Swamp of South Carolina, Dennis and Bateson (1974) found that 11 plant species were restricted to floating logs and stumps. Similarly, Clethra alnifolia, a widely distributed wetland shrub in eastern North America,
Chapter 2
103
regenerates more commonly on logs, tree roots, and stumps than on more available soil substrates (Jordan and Hartman 1995). Agnew et al. (1993) reported that virtually all seedling regeneration of two dominant tree species at a forest-mire ecotone on the south island of New Zealand occurred on nurse logs. No nurse logs and no seedlings occurred in the mire. They argued that the presence of dead and decaying trees maintained the forest ecotone, creating a “positive vegetation switch” (Wilson and Agnew 1992) in which forest species created conditions that favored themselves. As noted above for nutrient effects, species that dominate bogs may also “engineer” positive feedback switches with respect to substrate building (Jones et al. 1994, 1997). The unique chemical properties of Sphagnum mosses contribute to highly unfavorable conditions for other species of plants, but allow Sphagnum to produce environments that favor its own growth (van Breeman, 1995). Sphagnum contains high concentrations of phenolic compounds, including genus-specific acids. The accumulation of many centimeters of dead Sphagnum, the tissue of which is one of the most water-absorptive materials known, also creates a water-logged, anoxic, nutrient-poor environment that impedes other species without substantially limiting Sphagnum. Sphagnum further appears to engineer its own preferred substrate by forming impervious iron pans and clogging soil pores under the peat debris. Crowding is a common state of affairs in plant communities, and is typically studied from the perspective of competition for limited resources. Harley and Bertness (1996) took a different approach and studied how crowding leads to morphologically modified individuals that become dependent on their neighbors for structural support. They created crowded and isolated treatments for four different salt marsh species, and found that crowded plants were typically taller and thinner, producing less aboveground biomass, and were more susceptible to breaking under stress. Then crowded plants were thinned to test the hypothesis that crowding leads to dependence on neighbors. They found that plants grown in crowded conditions were dependent on their neighbors to remain upright, in their words, “once plants have grown in a crowd, it is important that they remain in that crowd”. Structural interdependence has also been suggested for trees in tropical and temperate forests where thinning often leads to increased tree fall (Skorupa and Kasenene 1982, Holbrook and Putz 1989). Positive interactions in which plants provide other plants with vital substrate or structural stability is an overlooked process in ecology, and these kinds of interactions are good examples of ecosystem engineering (Jones et al. 1994, 1997). Understanding how sessile and crowded organisms perform in groups provides interesting insight into the nature of communities.
104
Direct Mechanisms for Facilitation
2.9. DISTURBANCE Some plant species are able to tolerate disturbance and create stable environments that benefit other species that cannot handle disturbance. This facilitative process is conceptually similar to the positive effects of some plant species on substrate development, but not identical. For example, it has been known for a long time that some plant species stabilize shifting sand dunes (Bagnold 1954, Bressolier and Thomas 1977) which creates habitat that can be colonized by other species (Whitfield and Brown 1956, Hewett, 1970, Toft and Elliot-Fisk 2002). However, the effect of established plants on the movement of moving sand is not simple. Franks and Peterson (2003) measured sand accumulation under mature plants and in the open and found that less sand accumulated under the plants than in the open, and therefore the mature plants protected seedlings from burial and mortality. Interestingly, he attributed this to plant stabilization of substrate, but noted that substrate instability may lead to either the accretion or loss of material. Martinez (2003) and Martinez et al. (2004) found similar patterns in Mexico, with the shrub, Chamaecrista chamaecristoides, ameliorating wind and protecting young grasses that grew beneath the shrub from burial. El-Bana et al. (2002) found that stable “rekbah” soil mounds built by Retama raetam protected desert annuals from disturbance in the Sinai Desert. The general effect of root systems on preventing soil erosion is obvious, but some species bind soil better than others, and therefore maybe better at facilitating other species. For example, Cerdà (1997) found that Stipa tenacissima, a large Mediterranean bunchgrass, was unusually good at stabilizing eroding soil (also see Alados et al. 2006). Such stabilization of substrate is important for buffering the disturbance created by wave impact on coastlines (Chapman 1974). The roots of tolerant species bind marsh and beach sediments and aboveground biomass slows water movement. This increases the deposition of sediments and protects other species from the impact of the waves (Pennings and Bertness 2001). Some of the most ingenious experimental work on the role shoreline plants play in protecting others from physical disturbance has been conducted in the last few years. Whole communities of halophytic forbs that occur on “cobble beaches” appear to depend on protection from beds of Spartina alterniflora in places where the impact of waves is high along the New England coast (Bruno and Kennedy 2000, van de Koppel et al. 2006). Cobble beach communities and Spartina beds form parallel bands, 3-10 m in width, along large sections of the shoreline. Species that form the cobble beach communities are highly restricted to the zones directly behind the Spartina. Spartina beds reduce the velocity of water movement during tidal events and stabilize the gravel substrate (Bruno 2000). Not all Spartina beds have equally facilitative effects.
Chapter 2
105
When seeds of two cobble beach species, Suaeda linearis and Salicornia europea, were planted behind Spartina beds of different sizes the seedlings of both species emerged and established in greater numbers behind large beds (Bruno 2000, Bruno and Kennedy 2000). When the substrate was mechanically stabilized behind small Spartina beds, seedlings emerged and established much as they did behind large Spartina beds (Figure 2.32). Other experiments showed seed supply and soil characteristics to be much less important than the presence of large, stable Spartina beds. Similar positive effects of neighbors occur on lakeshores where waves disturb sediments and vegetation (Wilson and Keddy 1986, Twolan-Strutt and Keddy 1996). Seedling emergence
Seedling establishment
Percent
A) S. europaea 100
100
80
80
60
60
40
40
20 0
20 0
0 0 80
100 No. seedlings/plot
B) S. linearis
C) S. europaea 80
0
D) S. linearis
60
60 40 40 20
20 0
0
0
Small
Small + Large stabilization
0
0 Small
Small + Large stabilization
Treatment
Figure 2.32. Emergence and establishment of Salicornia europea and Suaeda linearis seedlings behind small and large stands of Spartina alterniflora and in small stands where substrate was mechanically stabilized to mimic the effects of the large stands. Error bars are 1 SE. “Percent” refers to the frequency of the plots of each treatment that contained one or more seedlings. Reprinted from Bruno and Kennedy (2000) with permission from Oecologia.
106
Direct Mechanisms for Facilitation
The general facilitative role of Spartina alterniflora appears to vary with genotype. Proffitt et al. (2005) experimented with different clonal genotypes of S. alterniflora and found that different genotypes varied in their competitive effect on a codominant species, Salicornia bigelovii, and their facilitative effects on many less common marsh species. Spartina alterniflora also appears to have a facilitative role in stabilizing substrates at lower elevations in eastern marshes. Several authors have described an “obligate co-occurrence of fucoid algae, primarily Ascophyllum nodosum, with S. alterniflora on the eastern coast of North America” (Chock and Mathieson 1976, Gerard 1999, Chapman and Chapman 1999). One hypothesis for this relationship is the stabilization of algal mats, which have been described as “stapled in position by grass shoots,” (Chapman and Chapman 1999). Gerard (1999) experimented with removals of S. alterniflora and found that almost all Ascophyllum from Spartina-removal plots was lost during fall and winter storms, when disturbance tore the algae from the intertidal bed. Wilson and Keddy (1986) manipulated eight different species in a lakeshore community in southern Ontario. The experiment was designed to investigate “diffuse” competition (see Chapter 3), which is the combined effect of several species interacting together inclusive of the indirect interactions among the species, and found that diffuse effects on three of the eight species were facilitative. In other words, individuals grown with neighbors accumulated more biomass than conspecifics grown alone. They hypothesized that these positive responses may have been due to neighbors providing shelter from wave action or stabilizing sediments. In many streams and rivers of northern California, large tussocks of the aptly named “torrent sedge”, Carex nudata, are exceptionally tolerant to disturbance generated by high spring runoffs (Levine 2000). Levine found that most of the 60 plant species growing in this riparian habitat occurred almost exclusively within the thick root mats and vegetation of the Carex tussocks. In experiments designed to test the hypothesis that C. nudata provides stable substrate for other species during extreme winter floods he found that transplants of five species grew as large or larger in the open streambed as they did in C. nudata tussocks, but winter mortality was much higher outside of the tussocks (Figure 2.33). In streams in the Sonoran Desert, Dudley and Grimm (1994) observed that Cynodon dactylon (Bermuda grass) formed mats over sand and gravel substrates that were typically scoured of natural vegetation during annual floods. They found that five other plant species were associated with Cynodon and hypothesized that the invasive grass provided resistance to disturbance and thereby modified community structure in the streams. Johnson et al. (1985) argued for a similar function for Salix nigra, which appears to stabilize alluvial
Chapter 2
107
0.8
Mimulus cardinalis
Conocephalum conicum 2 X = 40.83, P<0.001
2
X = 32.95, P<0.001
0.6
0.4
0.2
0.0 1.0
Mimulus guttatus 2 X = 27.49, P<0.001
Brachythecium frigidum X2 = 13.32, P<0.021
0.6
0.4
0.6
ns e De
Th in ne d
Pi nn ed
ip pe d
Juncus covillei 2 X = 82.76, P<0.001
0.8
Cl
0.0 1.0
am be d
0.2
St re
Mortality (%)
0.8
Carex
0.4
0.2
ns e De
Th in ne d
Pi nn ed
ip pe d Cl
St re
am be d
0.0
Carex
Figure 2.33. Winter mortality of five transplanted target species in open streambed or in various manipulations of Carex nudata tussocks. Bars represent the proportion of transplants that died during the winter of 1997-1998. Redrawn from Levine (2000) with permission from Ecology.
soils deposited in the deltas of southern Louisiana, enhancing the development of many other species. While not directly a modification of disturbance regimes, seagrass communities dominated by Halodule wrightii and Thalassia testudinum appear to enhance the efficiency of nitrogen uptake by slowing the velocity of water moving around them. Large, dense stands slowed water velocity more, with apparent positive feedbacks to nitrogen uptake rates (Thomas et al. 2000). Using a field flume, they found that the effect of seagrass communities on water velocity also resulted in higher ammonium uptake.
108
Direct Mechanisms for Facilitation
Small-scale disturbance may also play a role in the facilitative effect of some dominant species. Heilbronn and Walton (1984) proposed that graminoids might initially be favored as colonizers of steep slopes because of their fibrous, highly branching roots. These roots stabilize sediments by limiting microsolifluction and thereby create stable islands for forb colonizers. Ikeda and Okutomi (1992) conducted experiments in which plant interactions were studied in the context of human trampling. They found that Plantago asiatica was sensitive to the drought conditions created by heavy trampling, and when in monocultures its abundance and survival decreased sharply as the soil compaction increased. However, when Eragrostis ferruginea was planted with Plantago soil moisture levels stayed higher and Plantago performance increased because of protection provided by the tougher species.
2.10. POPULATION SIZE AND POSITIVE DENSITY-DEPENDENCE Positive density-dependent processes within populations are known as Allee effects (Allee 1931). Allee hypothesized that population growth and survival may decrease at low densities because of reproductive limitations, the inability of small numbers to exploit a particular resource, or the greater potential for a predator to wipe out an entire population. For plants, the Allee effect may be manifest most explicitly as pollen limitation (Davis et al. 2004) or the inability of small populations to modify the environment through positive feedback processes. Most examples of positive density-dependence that I know of for plants involve animal pollinators. Interactions mediated by a third organism are indirect, and these kinds of density-dependent phenomena are discussed more thoroughly in Chapter 3. However, some density-dependent processes may be direct, for wind-pollinated species for example, and I address them in this chapter. An interesting twist on positive density-dependence in plant populations was demonstrated by Wied and Galen (1998) who found that seedlings of Frasera speciosa and Cirsium scopulorum, alpine herbs in the central Rocky Mountains, were highly associated with conspecific adults. They conducted experiments and found that survival was consistently higher when seedlings were covered by the infructescences of parental plants than when they were exposed (Figure 2.34). They attributed this “parental care” to higher soil moisture and lower evaporation rates beneath the decaying infructescences. Similar Allee effects have been reported for the shrub Ceratoides lanata in western Utah, but were based on spatial association (Freeman and Emlen 1995).
Chapter 2
109 100
Seedlings surviving (%)
80
60
40
20
A
B
0 170
180
190
200
210
170
180
190
200
210
100
80
60
EXPOSED COVERED
40
20
D
C 0 160
170
180
190
200
210
170
180
190
200
210
Days since planting Figure 2.34. Survival of Frasera speciosa seedlings in plots covered by conspecific infructescences or exposed. A and B show the responses of planted seeds at two sites, and C and D show the responses of natural seedlings. Error bars show 1 SE. Redrawn from Wied and Galen. (1998) with permission from Ecology.
Diettart Matthies at Philipps-University in Marburg, Germany studied the effects of a broad range of densities on the population ecology of the root hemiparasite, Rhinanthus alectorolophus (Matthies 2003). He found that the proportion of seeds making it to maturity increased linearly with seed density, suggesting positive interactions among seeds or seedlings. Matthies hypothesized that recruitment was intraspecifically facilitated through resource sharing among parasites through connected haustoria and host roots. Interestingly, overall fitness was not improved by increased recruitment because dense populations of Rhinanthus were smaller and produced fewer seeds than sparse populations. Increased survival, however, may “indirectly increase fitness because it will increase the genetic diversity of offspring and thus, for instance, reduce the impact of pathogens”. Allee affects can be also driven by group competitive effects. Naomi Cappuccino at Carleton University examined Allee effects of the invasive Vincetoxicum rossicum (pale swallow-wort). She planted patches of the weed composed of 1, 9, and 81 plants and found that biomass and seed set of
110
Direct Mechanisms for Facilitation
individual target plants in the patches were the largest in the 81-plant patches than in the 1- and 9-plant patches. Unlike most other examples of Allee effects in plants, the effect in Vincetoxicum was not due to differences in pollinator visitation rates. Vincetoxicum appeared to gain competitive advantage against the native when growing in larger patches. Allee effects may also be caused by interactions among seeds. Working with annual communities of the Negev Desert, Lortie and Turkington (2002) sowed seeds of Erodium laciniatum into natural seed banks. They found that Erodium seeds increased emergence of other species from the seed bank, and Erodium seeds and seedlings had positive effects on the density and biomass of annual community. The mechanism for this effect was not clear, and may have been related to group shading among annuals, but another similar annual species, Erucaria pinnata, did not have positive effects on the same seed banks suggesting that chemical mechanisms may have been involved. Roughly parallel to the idea of Allee effects, community attributes may also depend on the composition and diversity of nearby neighboring communities. In mountain grasslands of central Argentina, Cantero and Partel (1999) found that the diversity of different communities was correlated with the larger-scale pattern of the communities on the landscape. For example, the species richness of shortgrass prairie communities was highly correlated to the distance of that community to a tallgrass community. This relationship was not due to any modification of the abiotic or biotic environment of the shortgrass communities. The proximity of an additional species pool appeared to contribute to the movement of propagules among communities.
2.11. SEED SHADOWS Trees and shrubs can create distinct heterospcific seed shadows and develop large seed banks by trapping wind-dispersed seeds, protecting seeds from predators, providing perches for frugivorous birds, or creating microenvironments in which seed output is higher than in the matrix of open space. Trapping wind-dispersed seeds is a direct effect, whereas providing protection from predation and foci for bird-dispersal are better considered as indirect effects (see Chapter 3). Seed shadows and the plant distributions that develop from them are not necessarily cases of facilitation. For example, aggregation of propagules under canopies may be disadvantageous for species that grow and reproduce better in the open or for those that subsequently suffer greater effects from increased intraspecific competition. Furthermore, aggregation under canopies may simply redistribute populations or restructure
Chapter 2
111
communities and not provide overall gains in the fitness of community members. Therefore, the role of aboveground plant parts in collecting and concentrating propagules of other species is ambiguous in its overall positive effect. If propagules are moved to a favorable microsite or prevented from being destroyed, the effects of plants as seed traps are clearly facilitative. On the other hand, if differences in aboveground architecture act only to alter spatial patterns of distribution without providing more favorable sites, then seed trapping may confound real facilitation. In order to demonstrate that seed shadows constitute or lead to a facilitative effect, the benefits of being dispersed under other plants must be shown. Many studies have described the importance of propagule concentration under mature plants, but few have actually quantified seed abundance under canopies versus in the open. Hutto et al. (1986) found 5.33±2.58 (S.D.) saguaro seeds per 0.01 m3 of soil under the canopies of Cercidium (paloverde) trees compared to 0.17±0.41 seeds per m3 just one m from the canopy edge. Birds disperse saguaro seed, but in this case the aggregation of propagules under Cercidium results in crucial facilitative interactions as described earlier in this chapter (Turner et al 1966, Steenberg and Lowe 1977). Day (1989) found that seeds of many plant species were trapped in the canopies of Eriogonum ovalifolium on volcanic cinder cones, leading to a highly skewed preferential establishment of many species under Eriogonum. Another good example of propagule concentration is in the Patagonian shrub-steppe, where communities are dominated by scattered shrubs (primarily Mulinum sinosum, Adesmia campestris, and Senecio filaginoides) and interspersed Stipa spp., Bromus pictus, and Poa speciosa bunchgrasses. The bunchgrasses often form rings around the shrubs (Aguiar and Sala 1994, Aguiar et al. 1992). Aguiar found that the number of seeds of Bromus pictus trapped in the shrub-grass ring patches was 20 times that in scattered tussock patches, and attributed this difference primarily to the wind blowing grass seeds into the shrub canopies. Experimental manipulations of roots and shoots demonstrated strong root competition, but that the aerial parts of the shrubs protected germinating B. pictus seedlings. Positive spatial relationships among plants have been attributed to seed aggregation under adult perennial species in other systems (Fuentes et al. 1984). In an analysis of published data on seed banks in the Great Basin, Mojave, Sonoran, and Chihuahuan Deserts, Guo et al. (1998) found that the total number of seeds per unit area was highest under shrub canopies and decreased in surrounding intershrub areas.
112
Direct Mechanisms for Facilitation
300
DCA axis
a
r2 = 0.64
250 200 150 100 50 0 300 250
DCA axis
b
r2 = 0.46
200 150 100 50 0 0
10
20
30
40
Age (years) Figure 2.35. Detrended correspondence analysis (DCA) axis 1 (representing understory community composition, versus the age of overstory Retama sphaerocarpa shrubs. A) The relationship for the extant plant community. (B) The relationship for the seed bank. Reprinted from Pugnaire and Lazaro (2000) with permission from the Annals of Botany.
Few studies have tried to separate the effects of canopies as dispersal foci from their effects as facilitators. But in an excellent example, Pugnaire and Lazaro (2000) examined the possibility that the plant community developing under canopies of Retama sphaerocarpa does not reflect the seedbank trapped under the shrubs. They examined this “facilitative filtering” in southern Spain by comparing the composition of seed banks with the composition of extant plant communities under shrubs, and found that the composition of the seed banks and the composition of the plant communities were not correlated. The species composition of the seed bank was less correlated with shrub age (5-35 years) than the extant plant community (Figure 2.35). More importantly, many species
Chapter 2
113
occurred as seeds under both young and old shrubs, but only recruited to maturity under old shrubs. These results demonstrate that the effect of Retama as foci for dispersal is important to the structure of the plant community, but more importantly, that Retama canopies also function as filters that sort the species that eventually comprise the understory. This filtering process contributes to the strongly contrasting plant communities that occur under Retama shrubs versus the intershrub spaces (Pugnaire et al. 1996a,b). Processes leading to higher levels of organic matter, total nitrogen, phosphorus, higher nitrogen mineralization rates, and climate amelioration (Moro et al. 1997) were the most important filtering effects of Retama canopies. Similar filtering processes have been described by de Viana et al. (2001) for shrubs in the central desert of Argentina. Trichocerus pasacana cacti were more common than expected from seed density beneath two species of shrubs, but less frequent than expected for six other understory species.
2.12. COMMUNICATION The potential for plants to interact, and facilitate each other, through communication was suggested by Rhoades (1983) and Baldwin and Schultz (1983). Rhoades argued that Salix species that were not yet attacked by herbivores induced defenses against tent caterpillar and budworms after receiving pheromonal signals from neighboring Alnus and Salix that were under attack (see Chapter 3, Indirect mechanisms). Such signaling, whether via aboveground or belowground exchange, may constitute a whole suite of unusual positive interactions among plants. However, communication remains controversial and poorly understood, and therefore is an area with tremendous potential for future research. Communication among plants may take a number of forms including signaling among roots, resulting in avoidance and reducing root overlap (Mahall and Callaway 1992, 1996, Krannitz and Caldwell 1995, Schenck et al. 1999, Falik et al. 2005), stimulation of root growth (Gersani et al. 2001), pollen-stigma communication that promotes the germination and growth of pollen from unrelated neighbors and inhibiting pollen from close relatives, root-root chemical signals between parasites and hosts, oxidizing gases in smoke or acids from burned plants that cue germination of other species, and neighbor-altered light wavelength ratios that stimulate growth responses (Davis and Simmons 1994). Not all of these signals necessarily promote facilitation among plants, but most have the potential to be facilitative and the release of a chemical signal from a species under herbivore attack that is used by others to prepare for oncoming attacks constitutes a remarkable example of positive interactions among plants. Like other facilitative mechanisms, communicative interactions are not evidence
114
Direct Mechanisms for Facilitation
for co-evolved or intentional altruism among plants. For example, the positive effects of chemicals released by wounded plants to unwounded plants may simply be the by-products of the physiological wound response. Furthermore, as documented for the nurse-plant phenomenon in which beneficiaries may end up competing with and killing their nurses, signal-induced responses by unwounded beneficiaries may actually be bad for the wounded benefactor if the herbivores then have fewer susceptible hosts to attack and concentrate on the original victim. Some of the more remarkable examples of chemical communication among plants involve indirect interactions with herbivores, and these indirect interactions are presented in Chapter 3. Mediterranean-climate vegetation has many species that germinate closely following fire and appear to be stimulated by fire. The germination of many of these species is not stimulated by heat shock, but is induced by chemical ‘signal-mediated’ stimulation from smoke or charred wood (Keeley 1991, Baldwin et al. 1994, Keeley and Fotheringham 1997a,b). Keeley and Fotheringham (1998a, 1998b) demonstrated that nitric oxides from smoke and pH-dependent nitrate effects stimulated the germination of Emmenanthe pendulflora. To my knowledge, there is no evidence that charred wood or smoke from different source species elicits specific responses from sensitive species. Plants often exhibit growth responses to declines in the ratio of red:farred wavelengths of light that occur in the shade of other plants (Morgan and Smith 1979). Leaves reduce red:far red ratios by the selective absorption of red wavelengths by chlorophyll and adding far red by the reflection and transmission of these wavelengths. Reduction in red:far red ratios is detected by phytochrome systems within the leaf and can result in large adjustments in a plant’s biomass allocation and development. Smith et al. (1990) found that changes in these wavelengths could be detected as far as 30 cm from experimental canopies, a distance where total PAR was not affected. Because a neighboring plant can change red:far red ratios without affecting total light, researchers have hypothesized that these changes in wavelength ratios may constitute a signal for other plants, which may respond by altering allocational patterns. Ballaré et al. (1987) tested the hypothesis that a plant may detect and respond to a neighbor through the perception of changes in spectral composition. When “green fences” were planted in the vicinity of target Sinapsis alba plants and sensors, they found that sensors detected changes in red:far red ratios to the north side (not shaded and consistently sunlit) of the green fences before any changes occurred in PAR. Sinapsis alba plants on the north sides of fences produced longer internodes and had a lower leaf:stem mass ratios than those grown on the same side of fences that had been bleached to remove their greenness. These classic shade responses occurred without shade, and were interpreted by Ballaré et al. as evidence for communication of an “early warning signal of oncoming competition”.
Chapter 2
115
Although I may be stretching the point to a facilitative interpretation, the signal from one species may benefit another by providing it with a competitive advantage. Davis and Simmons (1994) performed field experiments with Hordeum vulgare (barley) in which they manipulated borders of conspecific plants to influence local red:far-red ratios without influencing other important light, microclimatic, or competitive factors. They planted Hordeum in the field in rows spaced 18 cm apart, with the center row designated for target plants and separated from the outside “border” rows by Plexiglas barriers in the soil to prevent root competition. This treatment was controlled by plots without border rows, and another treatment without border rows or soil barriers was used to assess the effect of the soil barrier. Border rows were clipped to 20-cm heights so that no direct shade fell on the targets. Target plants were planted at high densities (2-cm spacing) and low densities (16-cm spacing). Border rows consisting of conspecific Hordeum plants did not reduce PAR levels experienced by target Hordeum, nor did border rows affect air temperature or soil temperature. However, border rows did reduce red:far red ratios beginning only two days after emergence of the target plants. Borders reduced red:far red ratios of diffuse light by 16-34% and diffuse plus direct light by 7-8% over the first 12 days of the experiment. Corresponding with these changes in wavelength, Hordeum targets increased leaf and internode elongation and began reproductive development earlier. Furthermore, borders increased elongation rates more for closely spaced plants than for plants spaced farther apart, indicating that plantplant interactions affected allocational and developmental responses to declines in red:far red wavelength ratios. Studies of other plant species in which red:far red ratios have been manipulated directly have demonstrated similar growth and reproductive responses (Casal et al. 1985, 1987). Increased leaf and internode elongation and earlier reproductive allocation are classic shade responses; however, in Davis and Simmons’ experiment no reduction in the resource, PAR, was experienced by plants expressing these responses. They interpreted their results as due to signal detection, and a response that promoted the fitness of the receiver. Shadeintolerant species that elevated leaves more rapidly within a matrix of competing species before they were shaded by their competitors would confer a significant advantage. Ballaré et al. (1990) transplanted seedlings of the crop weeds Datura ferox and Sinapsis alba into conspecific field populations that were of similar heights. Control transplants, which were not manipulated in any way, responded with an increase in stem elongation within three days of transplanting and before direct shading from neighbors was important. When the internodes on transplanted seedlings were “blinded” to far-red radiation scattered by the
116
Direct Mechanisms for Facilitation
surrounding seedlings using cuvettes filled with CuSO4, internode elongation decreased, as did the total elongation of the plant. When cuvettes were filled with water, which blocks far less far red than CuSO4, internode elongation was not reduced. Ballaré et al. (1990) emphasized the importance of their work for understanding how plants acquire information about their position relative to other competitors; however, such forms of communication have the potential to function in facilitative interactions as well. Studies of airborne and other forms of chemical signaling among plants are in their early stages, but communication may entail a wide range of possibilities for complex interdependent interactions in which a signal from one species increases the performance of another. As stated by Aphalo and Ballaré (1995), understanding the functions of signals may “free us from the preconception of competition as the dominant type of interaction between plants…., as one could assign a positive or negative sign to the part of the interaction mediated by each individual signal.”
2.13. CONCLUSION Amassing over 100 pages of review for direct facilitative mechanisms is a bit mind numbing, but sometimes science by siege can make a point. By cataloging mechanisms by which plants directly facilitate each other we can demonstrate how amazingly varied and complex interactions among plants really are, probably more variable than the way that plants compete with each other. Furthermore, these positive interactions occur in virtually all biomes on Earth and often have exceptionally strong effects on the diversity, productivity and species composition of plant communities. In the next chapter, we will see that explicit inclusion of “intermediary” species in indirect facilitative interactions creates even more ways that plants can benefit from their neighbors.
CHAPTER 3 INDIRECT MECHANISMS FOR FACILITATION
As covered in Chapter 2, direct facilitative effects such as shading, adding nutrients to the soil, and protecting other plants from disturbance are important facilitative mechanisms. However, plants may also have major indirect positive affects on other plants. Indirect interactions require intermediate species, such as herbivores, pollinators, mycorrhizal fungi, soil microbes, or other competing plant species in order to occur (Strauss 1991, Wooton 1994, Miller 1994). For example, one of the most important positive effects one plant can have is to protect a neighbor from herbivory. By associating with an unpalatable neighbor, a tasty species may avoid being eaten and increase in size and reproductive fitness. In this chapter I review herbivore, pollinator, disperser, microbial and competitor driven indirect interactions. The seminal paper on how plants can protect one another from herbivores was published by Attsat and O’Dowd (1976). They argued that many plant species were “functionally interdependent with respect to their herbivores” (emphasis added). They described associated plants as “defense guilds”, in which some members functioned as anti-herbivore beneficiaries for other species in three major ways: 1) as hosts for insect predators that attack the herbivores of neighboring plants, 2) as repellant neighbors that make it difficult for herbivores to locate their prey, and 3) “attractant-decoy” neighbors that draw herbivores to them so that they leave the beneficiary alone. Not long after Attsat and O’Dowd’s paper, Sam McNaughton published a paper in which he showed that mortality rates of the highly palatable grass Themeda triandra decreased as the abundance of associated unpalatable species increased (McNaughton 1978). These studies paved the way for a flood of other research on positive defense interactions and other indirect forms of facilitation. Indirect interactions are complex and cryptic, and their effects are difficult to tease apart experimentally. Because of these characteristics indirect effects are often overlooked as alternative hypotheses for direct interactions. For example, direct interactions among plant species are more often competitive than facilitative. However, direct effects of competition between palatable and unpalatable plant species may be highly altered in the presence of herbivores. Many studies have shown that palatable species may 117
118
Indirect Mechanisms for Facilitation
benefit by being near unpalatable species. If the interactions between palatable and unpalatable species were studied in the absence of herbivores, one would conclude that competition was the most important interaction. Only when the intermediate species is included would the indirect interaction be detectable. Indirect interactions involving consumers have been understood conceptually by ecologists for a long time (e.g. Andrewartha, 1954a), and more recent research has repeatedly shown them to play important roles in the structure of ecological communities (Strauss, 1991, Wooton, 1994, Miller, 1996, Pennings and Callaway 1996, Levine 1999). But despite early experimental examination of food web-based indirect interactions in marine systems (Paine 1966, Lubchenko 1978) recognition of the broader importance of indirect interactions has developed more recently. Most indirect interactions involve a consumer; however, Levine (1976) demonstrated that the addition of a third competitor to a model of two competing species could change the cumulative effect of a species from competitive to facilitative because of the suppression of shared competitors. When the direct competitive effect of one species on another is weaker than the indirect positive effect via competitively suppressing a third species, net facilitative effects can occur in a community of competitors; and importantly, without the influence of consumers. One reason for the importance of understanding strong indirect effects that occur within plant communities is that they catalyze new thinking about old theories, specifically the theory that plant communities are individualistic. If “webs”, “hierarchies”, or “loops” of strong competitive effects within communities can elicit strong positive indirect effects, then plant species may be quite interdependent even though all pairwise interactions are competitive. In this chapter, I will focus on several types of indirect positive interactions; those mediated by herbivores and animal parasites of plants, those mediated by parasitic plants, the effects of mycorrhizae and other soil microbes on plant interactions, and last, those that develop as a consequence of competitive loops and webs.
3.1. HERBIVORE-MEDIATED FACILITATION The most widely studied type of indirect facilitation among plants, by far, is that which occurs when one plant species protects another species from herbivory. Plants protect other plants from herbivores in two general ways. Benefactors may have anti-herbivore characteristics such as spines or toxins, and other species benefit by being near heavily defended plants (McAuliffe, 1984b). I refer to this sort of effect as “shared defense”. Alternatively, palatable species may benefit from being hidden in a crowd of other species
Chapter 3
119
HERBIVORE
BENEFACTOR
+
BENEFICIARY
Figure 3.1. Conceptual diagram for how sharing defenses with a benefactor may indirectly facilitate a beneficiary. Solid lines represent direct effects and dashed lines represent indirect effects.
that are not necessarily well defended, and by taking advantage of diverse neighbors the palatable species may simply be more difficult for predators to locate (Rausher 1981, Brown and Ewel 1989). This type of facilitative interaction has been referred to as “associational resistance” (Root 1972, 1973; Milchunas and Noy-Meir 2002, Stiling et al. 2003, Smit et al. 2005). Both shared defense and associational resistance can be depicted in the same general model in which the presence of the benefactor reduces herbivory on the beneficiary (Figure 3.1). As will be obvious throughout this chapter, clearly separating the mechanisms that define shared defense from those that define associational resistance is not easy (see Hambäck et al. 2000). Hambäck and Beckerman (2003) reviewed the literature in order to estimate the prevalence and potential for indirect interactive effects between herbivores and competitors on plants. They pointed out that our perspectives on such interactive effects have been affected by contrasts between the two general approaches ecologists use to study herbivore-competitor interactions. One group of studies excludes plant neighbors and herbivores in factorial experiments and scores effects on plant biomass. Other studies observe herbivore abundance or quantify herbivory on plants with or without plant neighbors and identify mechanisms underlying interactive effects. Indirect effects have been commonly observed in studies using the second approach and only rarely in studies using the first approach. Studies using the first approach have primarily focused on mammalian herbivory while studies using the second approach have focused on insect herbivory. Furthermore, studies using the first approach have typically been very small-scale manipulations
120
Indirect Mechanisms for Facilitation 14
Neighbor effect Herbivore effect Combined effects
Number of studies
12 10 8 6 4 2 0
0
0
0
<(-4) (-4 to -3) (-3 to -2) (-2 to 0) (0 to 1)
00 0-1
1-2
Effect size Figure 3.2. Effect sizes in factorial experiments examining the effect of removing plant neighbors and herbivores on plant biomass. Effects are determined from studies of plant neighbor and herbivore removals using the log response ratios calculated from each study (only studies measuring plant weight were included). The log response ratio for the interaction is positive when one factor tends to enhance the effect of the other, i.e. when the plant size is larger in the interactive treatment than expected from each treatment alone, and it is negative when the two factors act in synergy. Reprinted from Hambäck and Beckerman (2003) with permission from Oikos.
which make it difficult to identify indirect effects in systems in which mammalian herbivory predominates. Despite the methodological problems identified by Hambäck and Beckerman, their analysis of the literature suggested that direct effects of neighboring plants are mostly competitive, the direct net effect of herbivores are mostly negative, but when both plant neighbors and herbivores are considered together, positive effects are much more common (Figure 3.2).
3.1.1. Shared defenses Shared defenses are those that occur when a palatable beneficiary is protected by a nearby unpalatable species. Plant defenses may be directly targeted at repelling herbivores, such as spines, toxins or odors, or defenses may occur as a by-product of other characteristics such as shade or accumulation of salts in tissues. Shared defenses are much more widely documented in the literature than associational resistance. However, this may
Chapter 3
121
be due more to the sampling bias that troubles studies of facilitation in general than to the greater ecological importance of shared defenses. It is much more tempting, and easier, to study species with either negative or positive spatial associations or that possess intriguing facilitation-promoting traits than to randomly sample the community. Attsat and O’Dowd (1976) described “insectary plants”, species that attract the predators or parasites of herbivores, as having beneficial effects on their less defended neighbors. Parasitization of tent caterpillar herbivores is up to 18 times higher on trees near nectar producing plants than on trees without nearby nectar sources, and observations suggest that insects attracted to the nectar are responsible for attacking the tent caterpillars. Species of Phacelia planted in orchards appear to increase the parasite load on the crop pest Prospaltella perniciosi. Caterpillar herbivores on alfalfa are more likely to be parasitized when adjacent weeds are blooming. Attsat and O’Dowd also reported examples of plants functioning as alternate hosts for parasites of herbivores, and by doing so maintaining regionally higher populations of insects that benefit plants by reducing herbivore pressure. In southern New Jersey strawberries are often grown near peach orchards because they act as alternate hosts for over wintering parasites of the oriental fruit moth that attacks the peaches. In California, Vitis californica (wild grape) is damaged by a leafhopper, Erythroneura elegantula. An egg parasite, Anagrus epos, attacks the leafhopper but the parasite requires an alternate host to complete its lifecyle. This alternate host is another leafhopper, Dikrella cruentata, which lives on Rubus ursinus (blackberries) (Doutt and Nakata 1973). The two plant species occur in the same riparian habitat and are often intertwined. Rubus leaves emerge first in February, stimulating the rapid population growth of its Dikrella leafhoppers and their Anagrus parasites. In March, Vitis leaves develop and initiate populations of its Erythroneura leafhoppers. However, this timing insures that large numbers of leafhopper parasites are waiting and ready. Attsat and O’Dowd argued that although Vitis derived maximum protection when it was intertwined with Rubus, the dispersal of Anagrus parasites in the spring may provide indirect facilitative control of leafhoppers on Vitis plants up to four miles away. If detecting these kinds of relationships among plants requires analysis at a scale of four miles, reports of such facilitation among alternate host plants will be much less common in the literature than it is in nature. Attsat and O’Dowd (1976) also described interactions among “repellant plants” and their less repulsive neighbors, probably a far more common interaction than insectary type indirect interactions. For example, grasses such as Agrostis and Festuca are eaten less by livestock when densities of the noxious Ranunculus bulbosus are also high in the same
122
Indirect Mechanisms for Facilitation
Under shrubs 60
Mortality (%)
40
20
0
acorn removed shoot eaten shoot dried in place
60
40
20
In the open 0 March 1989
June 1989
September 1989
February 1989
Figure 3.3. Apparent causes of mortality of Quercus douglasii propogules planted under Salvia leucophylla and Artemisia californica shrubs and in open grassland 1 m from the shrub canopies. Reprinted from Callaway (1992) with permission from Ecology.
community. In the southwestern United States poisonous Happlopappus tenuisectus protects nearby perennial bunchgrasses from herbivory. Pure stands of Trifolium fragiferum are completely defoliated by rabbits, but rabbits ignored nearby stands of T. repens. But in mixed stands rabbits also ignore T. fragiferum suggesting that the non-preferred species protects the preferred species. In some cases the effect of neighbors on insects can have complex negative effects on a plant species. Monteith (1960) demonstrated that parasitism by Bessia harveyi on sawflies eating Larix laricina was reduced when understory shrub odors masked the odors of the Larix. In this case the parasite beneficial to the tree was kept from finding its target because of neighbors. As described in Chapter 2, Salvia leucophylla and Artemisia californica shrubs facilitate the recruitment of Quercus douglasii by providing shade (Callaway 1992), but shade only appears to be part of the story. Analyses of the fates of individual acorns and seedlings under shrubs and in the open showed that the causes and timing of mortality differed significantly under shrubs and in the open grassland and that herbivory was an important factor (Figure 3.3). Under
Chapter 3
123
the shrubs, predation on acorns was the primary cause of mortality and appeared to be caused by rodents. In contrast, very little acorn predation occurred in grassland only 1 m from the grass-shrub ecotone. In the grassland, high mortality was caused by physical factors (probably drought and temperature) and deer browsing but not by acorn predation. The virtual absence of acorn predation in the grassland was probably because acorn predators find refuge from their own predators in the shrubs and therefore rarely venture into the open. By increasing acorn predation, but decreasing shoot herbivory, shrubs provide lifestagedependent refugia for Q. douglasii. Although risks for acorns are high at first, once shoots emerge in shrubs they are relatively safe. In the open, acorns are relatively safe, but seedlings and saplings are exposed to herbivory and harsher conditions for many years. Salvia and Artemisia shrubs are highly aromatic and seldom if ever browsed by deer and cattle, suggesting shared defense. However, the benefits gained by Q. douglasii seedlings could be simply due to concealment and if so the defense is associational, not shared. The effects of shade on herbivores may frequently confound the direct effects of overstory shade on plants beneath. This phenomenon is well illustrated by the suppression of Hypericum perforatum, a Eurasian species that has become an important invasive weed in North America, with biological control insects (Huffaker 1959, also see Singer 1971, Hicks and Tahvanainen 1974). Before biocontrols, Hypericum was most common in open, treeless environments and much less so in the shade. However, two leaf-eating beetles introduced to control Hypericum, Chrysolina quadrigemina and C. hyperica, prefer to lay eggs in the sun. In a short time Hypericum was virtually eliminated from open areas and now is more common under trees canopies than in the open. In southern France, the natural invasion of calcareous grasslands by Quercus humilis (downy oak) is facilitated by two shrubs, Buxus sempervirens and Juniperus communis (Rousset and Lepart 1999, also see Kunstler et al. 2006). Virtually all Q. humilis recruit under these shrubs (Figure 3.4), and as was the case for Q. douglasii (also see Russell and Fowler 2004), the facilitative mechanisms are complex. Rousset and Lepart found that most Q. humilis occur in the northern portion of the shrub canopies, suggesting that shade in this mediterranean climate is important to the seedling survival. However, the spatial association between shrubs and Q. humilis is much stronger in sites utilized by livestock than in an ungrazed site (Figure 3.4). Moreover, prior to the release of livestock, first-year seedlings were not spatially associated with shrubs. In experiments, germination was significantly higher in the northern portions of shrub canopies, and no Q. humilis germinated on the south-facing edges of shrub canopies. Thus in addition to the direct effects of shade, sheep grazing had powerful indirect effects, causing 44% mortality of Q. humilis seedlings in uncaged areas outside of shrubs compared to 1% in cages (Figure 3.5).
124
Indirect Mechanisms for Facilitation Under canopy Near canopy Canopy removed
Germination (%)
100
50
0
North of Buxus
South of Buxus
North of Junip.
South of Junip.
Mortality (%)
Figure 3.4. Mean germination rates of Quercus humilis under the canopies of Buxus sempervirens and Juniperus communis, near the shrub canopies, and where canopies had been removed at cardinal compass directions. Ftreatment=36.6; df=1,32; P<0.0001, Forientation x treatment=6.1, df=1,32, P=0.194. Reprinted from Rousset and Lepart (2000) with permission from the Journal of Ecology.
80
G ra z in g
60 40
N o g ra z in g
20 0 C
NC
C
NC
Figure 3.5. Causes of mortality for Quercus humilis seedlings in grazed sites and non-grazed sites. Open bars are for sheep, hatched bars are for rodents, and black bars are for drought. Reprinted from Rousett and Lepart (2000) with permission from the Journal of Ecology.
Tree seedlings benefit from sharing defenses with other species in other ecosystems. Joe McAuliffe (1986) found that Cercidium microphyllum (paloverde), which is a common nurse plant for cacti and other species as an adult, recruited almost exclusively under shrubs. This facilitation was determined to a large degree by herbivory. Of 84 Cercidium seedlings in the open, 92% were consumed by herbivores; however, only 64% of Cercidium seedlings under shrub canopies experienced the same fate. Protection provided by the shrubs increased with physical contact and proximity as well. Only 14% of the seedlings touching shrubs were consumed in comparison to 92% when sheltered by the canopy but not touching. McAuliffe hypothesized that the shrubs provided concealment and that the unpalatability of Ambrosia deltoidea
Chapter 3
125 200
Cirsium
150
Y Axis
100
50
Open
0
-50
-100
Veratrum
-150
-200 -150
-100
-50
0
50
100
150
200
X Axis Figure 3.6. Means and 95% C.I. for sample vegetation plots under Cirsium obalatum, Veratrum lobelianum, or in open meadow in the Central Caucasus. Reprinted from Callaway et al. (2000) with permission from Oikos.
and A. dumosa, deterred browsing. In the same deserts, McAuliffe, (1984b) found that young barrel-cacti, Mammillaria microcarpa and Echinocereus englemannii,were much more common under live and dead Opuntia fulgida cacti where they were protected from herbivores by the accumulation of spine-covered stem joints falling from the nurse-plant. Indirect facilitation has also been demonstrated in rangelands (see McNaughton 1978, Oesterheld and Oyarzábal 2004). In subalpine meadows of the central Caucasus Mountains, extensive overgrazing by sheep during the Soviet era coincided with dramatic increases in the abundance of two unpalatable invaders, Cirsium obalatum and Veratrum lobelianum. Cirsium and Veratrum are avoided by livestock because they have spines and toxins, respectively. Zaal Kikvidze, David Kikodze and I found that plant communities associated with these unpalatable species were different in composition than the open meadows (Callaway et al. 2000, Figure 3.6). Forty-four percent (15/34) of all species were found at only “trace” (<1.0%) cover values in the open (no Cirsium or Veratrum) meadow, but at significantly higher covers under Cirsium or Veratrum. Of the 38 species that were reproducing sexually at the study site, eight were found only under the unpalatable invaders. Communities associated with Cirsium and Veratrum had 78-128% more species in flower or fruit than open meadow communities. Furthermore, the composition and
126
Indirect Mechanisms for Facilitation
reproductive output differed substantially between the community associated with Cirsium and the community associated with Veratrum, indicating some degree of species-specificity in the effects of these species. Several years later, I collaborated with Liana Khetsuriani and David Kikodze on experiments using experimental exclosures, removal of unpalatable species, and transplants of palatable and unpalatable species in the same subalpine meadows (Callaway et al. 2005). We found that Cirsium obalatum and Veratrum lobelianum only had facilitative effects on community composition when livestock were present, and when livestock were excluded the positive effect disappeared. Removing Cirsium and Veratrum outside the exclosure decreased the richness of associated communities, but inside the exclosure removal of these species increased community richness. Experimentally transplanted palatable grass species (Anthoxanthum odoratum and Phleum alpinum) grew larger inside the exclosure, and there Cirsium and Veratrum had no effect on their growth. However, outside of the exclosure, Cirsium and Veratrum had strong positive effects on the growth of A. odoratum and P. alpinum. Excluding livestock decreased the growth of L. pseudosudetica, another unpalatable species, and Cirsium and Veratrum had no effect on L. pseudosudetica outside the exclosure. In contrast, inside exclosures Cirsium and Veratrum had competitive effects on L. pseudosudetica. Our results indicate that Cirsium and Veratrum, which are in some ways undesirable rangeland weeds, may also play an important facilitative role in maintaining species and functional diversity of overgrazed plant communities in the Caucasus. Similar results have been demonstrated using exclosure experiments in southeastern Spain (Gómez 2005). Much like the story for Cirsium and Veratrum (also see Smit et al. 2006), Weaver and Albertson (1956) claimed that prickly pear cacti provided refugia for grasses during drought and overgrazing, establishing “local sources of seed supply on many ranges almost devoid of vegetation.” This was carefully investigated by Rebollo et al. (2002) in the shortgrass steppe of Colorado. They found that Opuntia polycantha provides refugia from cattle grazing and increases cover and seedhead production of associated plants. In grazed treatments Bouteloua gracilis doubled seedhead production in Opuntia clumps. However, when grazers were not present Opuntia had no effect on Bouteloua. Similar patterns were observed for community diversity. These findings support Canfield (1948) who argued that range recovery was often due to the improvement of grass cover under established shrubs. In 1978, McNaughton found that the highly palatable grass Themeda triandra experienced 80% mortality when not associated with high densities of less palatable species in eastern Africa. However, as the abundance of associated unpalatable species increased, mortality rates of Themeda decreased. Themeda
Chapter 3
127
also appears to have the potential to benefit from the direct effects of neighbors. In a later study, artificial shade was found to increase Themeda seedling recruitment sevenfold by enhancing both germination and survival, which was probably attributable to an increase in soil moisture (O’Connor 1996). Oesterheld and Oyarzábal (2004) demonstrated that a highly palatable Patagonian grass (Bromus pictus) was spatially associated with the less palatable tussock grasses and in an exclosure experiment showed that this positive relationship diminished in the absence of grazing. Oliver Rousett and Jacques Lepart (2002) found that the particular composition of a plant’s local community appears to play an important role in determining herbivory intensity. They observed that Buxus sempervirens, a shrub that is rapidly increasing in abundance in calcareous grasslands in France, was much more likely to experience herbivory from sheep in some plant “neighborhoods” than in others. Herbivory on Buxus was lower in local plant associations that were composed primarily of unpalatable species, and higher in associations that were composed of palatable species. Communication among plants (also see Chapter 2) may also play an important facilitative role in response to herbivory. As pointed out by Sabelis et al. (2001), plants release airborne chemical information when attacked by herbivores. These chemicals may benefit attacked plants in several ways. First, if the plant possesses inducible defenses, they may communicate to other herbivores that the induced system is functioning and therefore the plant is not a vulnerable target. Second, chemicals may attract the predators of the herbivores that are damaging the plant. Chemicals released by plants under attack may also benefit their neighbors, a function that is clearly facilitative and indirect in nature. Plants respond to cues produced by damaged neighbors even when they are not damaged themselves (Baldwin and Schultz 1983, Rhoades 1983, Bruin et al. 1992, Kost and Heil 2006). The volatile compound, methyl jasmonate, has been identified as a chemical signal which increases defense production in cultivated tomato plants exposed to the signal in airtight jars (Farmer and Ryan 1990, McConn et al. 1997, Farmer et al. 1998). Until recently, communication of “warnings” of approaching insect attacks to plants not yet attacked has been met with skepticism because of methodological problems and the absence of field results. Recently however, well designed field experiments and careful attention to alternative hypotheses have provided very strong evidence for communication and indirect facilitation. For example, Rick Karban et al. (2000) experimented with Nicotiana attenuata (wild tobacco) growing in proximity to Artemisia tridentata in the field on the east side of the Sierra Nevada Range of California. These species occur together naturally, and seedlings of Nicotiana were transplanted to places 15 cm downwind of the canopies of established Artemisia plants. They found that the concentration of one form of methyl jasmonate (the
128
Indirect Mechanisms for Facilitation
3R, 7S epimer) increased over 10 times in the air near Nicotiana canopies when <1% of the leaves of an Artemisia neighbor were clipped with scissors. Nicotiana plants, which were never clipped in the experiments, near clipped Artemisia neighbors had higher levels of polyphenol oxidase in their leaves than Nicotiana plants with unclipped Artemisia neighbors (Figure 3.7A). Polyphenol oxidase is an inducible defense chemical. Furthermore, Nicotiana plants with clipped Artemisia neighbors experienced less natural herbivory than Nicotiana with unclipped Artemisia neighbors (Figure 3.7B,C). 20
PPO Activity
A 15
10
5
0
Proportion damaged
1.0
1996
1997
1998
0.8
0.6
0.4
0.2
0.0 0.5
Proportion damaged
Unclipped Clipped B
Uncl Clip
Uncl Clip
Uncl Clip
C 0.4
0.3
0.2
0.1
0.0
Unclipped Clipped
Figure 3.7. A) Polyphenol oxidase (PPO) activity for tobacco plants near clipped and unclipped sagebrush neighbors. Bars show least squares means and 1 SE adjusted for the covariates leaf size and amount of foliage clipped. B) Maximum proportion of leaves damaged by grasshoppers over three seasons on tobacco plants near clipped or unclipped sagebrush (means and 1 SE). C) Maximum proportion of leaves that were damaged by cutworms on tobacco plants near clipped or unclipped sagebrush (means and 1 SE). Reprinted from Karban et al. (2000) with permission from Oecologia.
Chapter 3
129
Karban and colleagues eliminated the possibility that clipped neighbors allowed more light to reach Nicotiana neighbors by conducting clipping experiments in the shade (50%) light reduction) and in the open and finding no differences between these treatments. They also found no evidence in feeding trials that clipped sagebrush was directly repellent to herbivores or caused them to reduce feeding rates in general. In experiments where soil borne communication was blocked Nicotiana showed the same strong induced defense in response to clipping Artemisia neighbors. Finally, experiments in which airborne communication was blocked by putting plastic bags over the clipped stems of Artemisia showed that Nicotiana near clipped, bagged Artemisia did not show the induced defense response. In sum this experiment provided very good evidence that interspecific communication from one species benefited a second species, the recipient of the signal, by inducing its defense responses and reducing herbivore damage. In the same year as Karban’s work with Nicotiana, Rainer Dolch and Teja Tscharntke (2000) showed that communication may induce defense and mediate facilitation in whole populations. They studied how defoliation by the specialist leaf beetle, Agelastica alni, of Alnus glutinosa affected subsequent herbivory on other, non-attacked plants. At each site, one tree was manually defoliated to simulate herbivory. After defoliation, damage by Agelastica on other conspecifics was measured for a growing season. After defoliation, herbivory by Agelastica increased with distance from the experimentally defoliated tree. They then tested leaf consumption of non-defoliated trees near and far from the experimentally defoliated treatment individuals in feedingpreference tests and found that the number of eggs oviposited per leaf were positively correlated with distance from the defoliated tree. Resistance was induced not only in defoliated alders, but also in their undamaged neighbors, suggesting intraspecific facilitation via the communication of induction signals. A similar communicative phenomenon was described by Kost and Heil (2006) who found that volatiles emitted from herbivore damaged Phaseolus lunatus (lima bean) stimulated the secretion of extrafloral nectar from nearby undamaged conspecifics, which attracted wasps and ants that attacked the herbivores. These studies provide clear examples of how plants can indirectly facilitate each other through shared defenses.
3.1.2. Associational resistance When predators search simultaneously for different prey species possessing different characteristics, the probability of finding a particular prey
130
Indirect Mechanisms for Facilitation
species is often reduced (Plaisted and MacIntosh 1995, Langley 1996), apparently because of conflicting search images. Conversely, prey species that have similar characteristics may increase a predator’s overall search efficiency for both species (Holt 1977, Holt and Lawton 1994, Plaisted and MacIntosh 1995, Langley 1996) and increase predation (Pusenius and Ostfeld 2002). Vandermeer (1974) found that individual shrubs of Calliandra grandiflora immersed in forested areas in the highland of Guatemala showed far fewer signs of insect attack than shrubs in dense intraspecific stands. Although the effect of neighbors was not separated from the effect of isolation in this case, indirect interactions among plants that are derived from the positive or negative effects of being camouflaged in mixed assemblages are common. The positive effect of being in a mixture of different species, where herbivores can be inhibited by visual or olfactory complexity, has been called “associational resistance” (Root 1972, 1973) and “associational plant refuges” by Pfister and Hay (1988). Root (1973) found that herbivore loads on collards (Brassica oleracea) grown in pure stands were consistently higher than on collards grown in mixtures with other species (also see Dempster, 1969; Bach, 1980; Risch, 1981). Morrow et al. (1989) reported that the specialist goldenrod leaf beetle (Trirhabda canadensis) avoided preferred host species when they were grown in high densities of non-host plants. In natural Costa Rican forests, rates of herbivory on many species were reduced in diverse species mixtures (Brown and Ewel 1987). Solomon (1981) compared infestation rates of the herbivorous moth Frumenta nundinella on Solanum carolinense in monospecific patches versus in patches with Phleum pretense. He found that Solanum in “highrelative abundance” patches had 10 times more Frumenta eggs laid on its leaves. Solomon did not think that Phleum inhibited herbivores mechanically, but by masking olfactory cues emitted by Solanum. The response of herbivorous insects to specific host odors may be camouflaged by non-host chemistry near the host (Andow 1986, Kareiva 1983, May and Ahmed 1983, Stanton 1983). Tahvanainen and Root (1972) demonstrated that the ability of flea beetles to find collards was limited by chemical stimuli from non-host tomatoes and ragweeds (Ambrosia artemisiifolia) that confused the beetle’s search images. In a review of 209 agricultural studies on the effects of crop diversity on the abundance of various insects, Andow (1991) found that 52% of the studies documented lower herbivore abundance in mixed crops than in monocultures. Only 15% of the studies found higher herbivore abundances in mixed crops. Insect herbivores that specialized on only one crop species were typically found at lower densities in mixed crops, whereas the response of generalist herbivores was much less predictable.
131 0.8 0.6 0.4 0.2 0 # eggs/plant
150 100 50
15 10 5 0
20
4
# larvae/plant
0
15 10 5
3 2 1
0
0
2000
0.08 Proportion leaf damage
# seeds/plant
# fruits/plant
# flowers/plant
Proportion meristem damage
Chapter 3
1500 1000 500
0.06 0.04 0.02 0
0 outside in Myrica Myrica
outside in Myrica Myrica
Figure 3.8. Number of eggs and larvae of Galerucella calmariensis observed per plant; proportion of leaf area removed and proportion of top meristem damaged by G. calmariensis, and the number of flowers, fruits, and seeds produced by naturally occurring Lythrum salicaria in and outside of thickets of Myrica gale (means and 1 SE). Reprinted from Hamback et al. (2000) with permission from Ecology.
On the Gulf of Bothnia in northern Sweden, flower and seed production of Lythrum salicaria (purple loosestrife) is higher in thickets of Myrica gale (sweet gale) than outside of the thickets (Hambäck et al. 2000). Based upon the characteristics of Myrica, two fundamentally different alternative hypotheses were possible. First, Myrica is a nitrogen-fixer, and higher soil nitrogen in thickets may have facilitated Lythrum. But Hambäck and colleagues also noticed that Lythrum growing in Myrica thickets were less damaged by insects and had lower abundances of the specialist chrysomelid beetle, Galerucella calmariensis on its leaves (Figure 3.8). Therefore they proposed a second general hypothesis, that the aromatic Myrica reduced herbivory on associated Lythrum either by making it difficult for Galerucella to locate Lythrum or by increasing predation on Galerucella. In order to test for the indirect effects of Myrica on Lythrum they compared herbivory intensity on potted Lythrum in Myrica patches to that on potted plants outside the patches. At peak abundance of the Galerucella
132
Indirect Mechanisms for Facilitation
herbivores the potted Lythrum outside Myrica patches held approximately 14 times the number of adult beetles, and far more eggs and larvae, than Lythrum inside Myrica patches (Figure 3.9). Not only did the herbivore populations suffer inside Myrica, the effects of Galerucella on Lythrum decreased as well. The proportions of leaf and meristem damage were highly reduced and the number of flowers, fruits per plant, and seeds per plant were much higher on potted Lythrum inside Myrica thickets than outside (Figure 3.10). Hambäck et al. examined the abundances of other insect herbivore species and found that populations of the seed-eating specialist Nanophyes marmoratus were slightly, but significantly higher inside Myrica patches, the reverse of the pattern found for the beetle Galerucella. They hypothesized that the increase in seed production stimulated by lower numbers of Galerucella provided more resources for Nanophyes. Of five insect predator taxa examined, Coccinella spp. was the only group that was significantly reduced on Lythrum growing in Myrica thickets. The results of Hambäck et al. demonstrated that at least some of the positive effects of Myrica were indirect and due to the associational reduction of herbivory rather than differences in plant quality. They concluded that the indirect associational resistance gained by Lythrum was due primarily to the negative effect that Myrica had on the ability of Galerucella to locate host plants. Myrica may have concealed Lythrum either physically within its leaves, or by olfactory masking. Myrica has a strong odor that could have interfered with host finding by specialist herbivores.
eggs
10
larvae 1
28-Jul
21-Jul
7-Jul
0.01
14-Jul
adults
0.1
30-Jun
# Galerucella/plant
100
Figure 3.9. Abundance of adults (diamond symbols), eggs (triangle symbols), and larvae (circle symbols) of Galerucella calmariensis herbivores on potted Lythrum salicaria plants placed inside and outside of thickets of Myrica gale (means and 1 SE). Reprinted from Hamback et al. (2000) with permission from Ecology.
Chapter 3
133
Proportion leaf damage
0.2
0.1
Proportion meristem damage
0 0.8 0.6 0.4 0.2 0
# flowers/plant
800 600 400 200 0
# fruits/plant
80 60 40 20
# seeds/plant
0 4000 3000 2000 1000 0 outside in Myrica Myrica
Figure 3.10. Proportion of leaf area removed and proportion of top meristems damaged by Galerucella calmariensis, and the number of flowers, fruits, and seeds produced by potted Lythrum salicaria placed inside and outside of thickets of Myrica gale (means and 1SE). Reprinted from Hamback et al. (2000) with permission from Ecology.
In their first study, Hambäck et al. (2000) did not clearly separate “associational resistance” from “shared defense” in the indirect facilitative effect of Myrica on Lythrum. The protection gained by Lythrum when mixed with Myrica would be “associational” if the olfactory masking was not produced by Myrica’s own herbivore defenses. If the chemicals that masked Lythrum were part of Myrica’s defense system the interaction would be more accurately labeled “shared defenses”. Despite the odiferous nature of the Myrica beneficiary, in a later study Hambäck and colleagues (2003) found that other, less scented but morphological similar species, provided as much protection for Lythrum as Myrica. These results support associational resistance as a facilitative
134
Indirect Mechanisms for Facilitation
mechanism and not shared defense and suggest a lack of species specificity in the facilitative relationship (see Chapter 5). In northern Sweden, Hjältén et al. (1993) conducted experiments in which they exposed branches of Betula pubescens, a less-preferred browse species, to moose herbivory either alone, or mixed with highly preferred species, Sorbus aucuparia or Populus tremuloides. They found that Betula experienced higher herbivory from moose when associated with these plants of higher palatability. Their results suggest that Sorbus and Populus have negative indirect effects on Betula because they attract herbivores to a species that otherwise might be left alone. In contrast, Betula experienced lower rates of moose herbivory when mixed with Alnus incana a species of lower palatability. In later experiments, Hjältén and Price (1997) transplanted potted Salix lasiolepis clones into a “matrix” of conspecific plants of different palatabilities in the field. They found that sawfly densities on the potted clones correlated significantly with the estimated palatability of the matrix plant. In other words, Salix clones appeared to benefit from reduced sawfly attacks if they were near other plants that were not as attractive to the herbivore. This pattern was not significant for two other herbivores, petiole gallers and leaf folders. In some cases plants appear to create physical environments in which their neighbors escape herbivores. Pedicularis densiflora is a hemiparasite that normally grows in the shade of its host trees where it attaches to their roots. Michael Singer (1971) conducted a large-scale investigation of the ovipositing preference of the nymphaline butterfly, Euphydryas editha, with P. densiflora as one of the focal host plants, and recorded its searching behavior in a number of different habitats in California. In the lab, Euphydryas butterflies were as likely to oviposit on P. densiflora as any other of five offered host plant species, but in the field there were no instances of ovipositing on P. densiflora. The avoidance of P. densiflora in the field had nothing to do with preference; rather Euphydryas butterflies avoided the shade in which the plant tended to occur. Plant species in the Brassicaceae or mustard family have remarkable relationships with flea beetle species in the genera Psylliodes and Phyllotreta, which are attracted to different Brassicaceae species by the glucosinolate chemicals in their tissues (Tahvanainen and Root 1972). Glucosinolates are part of an intricate induced chemical defense system in Brassicaceae, but for the specialist beetles they are an advertisement. However, the preferences of the flea beetles, apparently for particular blends of glucosinolates and related chemicals, can be deterred by other plants associated with their hosts. Hicks and Tahvanainen (1974) found that six related flea beetle species had distinct preferences for both host species and for the particular microhabitats in which their preferred host may or may not exist, clearly setting up a potential conflict of interests. Beetles appeared to resolve this conflict in favor of the
Chapter 3
135
physical environment; regardless of the presence of the host plant species, any given species of flea beetle was always found more commonly in either open or shady environments. Hicks and Tahvanainen followed up their measurements of spatial patterns with experiments in which five host plant species were transplanted either into open fields or shady woodlands in order to determine how important niche modification by other plant species was for the host choice of the herbivores. Again they found that each flea beetle species exhibited a distinct habitat preference for either shady or open sites. In other words, preferred hosts that had been transplanted into the shade created by other species did not experience the level of attack that would have been predicted by simple host preference for food. For example, the flea beetle Psylliodes napi was found ≈90% of the time on the host plant Barbarea vulgaris. However, P. napi was also found ≈90% of the time in open sunny habitats. Any B. vulgaris plant occurring in the shade was far less likely to experience herbivory than its conspecifics in the sun. Not only are these results a good example of indirect facilitation, they have profound implications for how a benefactor species, by modifying the niches of both beneficiary and predator of the beneficiary, may powerfully drive selection and speciation of both. Barbarea vulgaris plants avoiding its attackers in the shade are subject to very different selection pressures than those in the open. At first glance one might assume that an “arms race” has driven co-evolution between flea beetles and Brassicaceae species and their speciation; however, Hicks and Tahvanainen’s results show that indirect positive interrelationships among plant species have played a major role in the process. The effect of neighborhood species composition on plant-insect interactions, and thereby on indirect positive effects, may also involve more complex food webs. Kruess and Tscharntke (2000) studied plant-herbivoreparasite interactions in fragmented meadows in southwestern Germany by manipulating the isolation of Vicia sepium experimentally. When Vicia individuals were isolated from other conspecifics, the phytophagous enemies of Vicia were affected less than the parasitoids of the phytophagous insects. In other words, isolation from conspecifics decreased the benefits that conspecifics provided by maintaining higher populations of parasitoids. Neighbors may be good or bad. Associational defense has received a large amount of attention in the literature, but associational susceptibility (Brown and Ewel 1987) has not. Studies from the agricultural and ecological literature demonstrate that some plant species may experience much greater herbivory when associated with certain other species than when they are alone. This phenomenon has also been referred to as “associational damage” (Thomas 1986) and “shared doom” by etymologically liberated researchers in marine and intertidal systems (Wahl and Hay 1995). White and Whitham (2000)
136
Indirect Mechanisms for Facilitation 1.8
B 50
1.4
Defoliation (%)
Cankerworm density (larvae/shoot)
1.6
60
A
1.2 1.0 0.8 0.6
40
30
20
0.4 10 0.2 0.0
Cottonwood Cottonwood Cottonwood under under in open boxelder cottonwood
0
Cottonwood Cottonwood Cottonwood under under in open boxelder cottonwood
Figure 3.11. A) Mean maximum cankerworm numbers (1 SE) and B) mean maximum percentage defoliation of potted cottonwoods placed at different distances from source box elders. Both cankerworm numbers and percent defoliation decreased significantly with distance. Reprinted from White and Whitham (2000) with permission from Ecology.
documented associational susceptibility between two host species of Alsophilia pometaria (fall cankerworm). Populus angustifolia x fremontii saplings growing under Acer negundo (box elder) experienced 2-3 times higher defoliation and 2-3 times higher cankerworm infestation than Populus saplings growing under mature Populus trees or in the open. This shared doom appeared to be due to cankerworm’s much higher preference for Acer. When potted Populus seedlings were placed near Acer trees, the number of cankerworms and cankerworm damage decreased with increasing distance from Acer (Figure 3.11). In a similar situation in California, bare zones are common around shrubs where herbaceous species are either absent or are much smaller than they are in the open several meters away from the edges of the shrubs. Bartholomew (1977) investigated this pattern by comparing plant growth in cages at the margins of the shrubs to that outside the cages. He found that protection from rabbit herbivory near the shrubs allowed much higher growth and concluded that shrubs concealed the herbivores, and the herbivores created the bare rings around shrubs. Clearly herbs near the shrubs suffered from associational susceptibility. Other indirect interactions in chaparral among plants and herbivores appear not to be driven by associational susceptibility, but rather by associational resistance. Bullock (1991) studied the demography of Ceanothus greggii in a community dominated by a second shrub, Adenostoma fasciculatum. Six years after fires, Bullock found that survivorship of C. greggii was many times higher in plots that
Chapter 3
137
had been fenced than in plots that had been left open to herbivores. In fenced plots the height of the obligate seeder C. greggii was not affected by proximity to Adenostoma shrubs. In contrast, in non-fenced plots C. greggii that were growing within the branches of Adenostoma were significantly taller. They attributed this difference to repeated browsing from rabbits and other mammals. Jon Keeley (1992) also reported a possible example of associational susceptibility in which seedling establishment of several chaparral shrubs was higher under closed canopies and when covered by litter.
3.2. OTHER HERBIVORE-MEDIATED POSITIVE EFFECTS As Attsatt and O’Dowd pointed out in 1976, neighbors that are attractive to another plant’s herbivorous pests can provide some facilitative relief from the impact of the pests. However, these interactions are exceptionally complex because “decoys” also provide herbivores with good nutrition and thereby actually increase total herbivore populations. If larger populations stimulated by such decoys eventually spread to their neighbors the short-term benefits of a decoy may be swamped. However, some attractive decoys provide exceptionally poor nutrition for herbivores and thereby increase their facilitative potential. A superb example of how decoy neighbors can act as pest sinks was published by Hsiao and Frankel (1968) on potatoes and the beetles that eat them. The primary target of the Colorado potato beetle (Leptinotarsa decemlineata) is, as might be expected, the potato (Solanum tuberosum). However, Leptinotarsa is easily confused by several other species in the same family as Solanum (Solanaceae), and oviposits on their leaves as well. Hsiao and Frankel took advantage of the beetles’ poor judgment and designed an experiment to test for the possibility that these alternative hosts might act as decoys and indirectly facilitate Solanum tuberosum. When the primary host, S. tuberosum, was planted with equal numbers of conspecific control “decoys”, targets were infested with 50.3% of the egg mass in the treatments (49.7% of the egg mass went to the other S. tuberosum) and the average larval weight was ≈200 g (Figure 3.12). Other species, S. dulcamara and Lycopersicon esculentum, were good decoys and attracted 39% and 43% of the egg mass, respectively, but beetle larvae grew just as well on these two decoys as on the “normal host” S. tuberosum. Another species, Capsicum annum, neither attracted beetles to lay eggs on it nor provided substrate on which the larvae could grow. However, S. nigrum and Datura meteloides captured large numbers of eggs when in mixture with S. tuberosum and possessed tissues on which the larvae could not grow. Datura meteloides was a particularly efficient decoy, capturing 67% of the egg mass. Larvae that developed on D. meteloides showed no growth 60 hours after being deposited,
138
Indirect Mechanisms for Facilitation
Average larval mass (g)
250
Solanum tuberosum
200
EGG MASS CAPTURED (%)
Solanum dulcamara
150
50.3 39.0 43.0
Lycopersicon esculentum
100
Capsicum annum
50
0 67 48
Datura meteloides & nonfed larvae
0 0
10
20
30
40
50
60
Time (hours) Figure 3.12. Growth of fourth-instar potato beetle larvae fed on different Solanaceous plants. Redrawn from Hsiao and Frankel (1968) from the modification in Atsatt and O’Dowd (1976).
similar to control larvae that had not been fed at all (Figure 3.12). How the decoy effects of Datura played out for neighboring S. nigrum was not documented, but D. meteloides both reduced the mass of the eggs deposited on S. nigrum and eventually killed the larvae that grew from these eggs. This strongly suggests that S. nigrum would benefit from having D. meteloides as neighbors. If the sort of decoy-target effect on herbivores demonstrated in the potato-potato beetle system is controlled in any way by the relative abundance of neighbors the interactions may vary in function over time and space. For generalist herbivores the amount of an unpalatable species that is eaten depends somewhat on how abundant palatable species are in the same area. For example, when only toxic phenotypes of Lotus corniculatus are available, the slug Agriolimax reticulatus consumes it. However, when nontoxic phenotypes are mixed with toxic phenotypes, Agriolimax avoids the toxic phenotypes that it readily ate when there were no other choices. The potential ecological and evolutionary variation that these kinds of complex, multifaceted interactions could produce is overwhelming (Thompson 2005). Just as ecologists and agriculturalists have often suffered the ecological consequences of seeking answers in direct linear relationships among species, evolutionary biologists may have missed potentially important driving forces in selection and speciation, such as suggested by Attsat and O’Dowd’s research, by focusing on direct, linear interactions.
Chapter 3
139
The facilitation literature is rich with accounts of direct effects of tree and shrub canopies on the productivity and community composition of understory herbs. Although rarely considered, herbivores may also indirectly affect the way in which overstory species interact with understory species. Some of the most common and intense direct positive effects of perennial canopies involve increasing the nutrient quality of the throughfall and stem flow reaching the ground (Chapter 2). However, herbivory in the canopy can alter the nutrient quality of throughfall and stem flow and change the facilitative effect of canopies. Stadler and Michalzik (1998) experimentally manipulated the infestation of aphids (Cinara pilicornis and C. costata) on Picea abies (Norway spruce) in Germany. They found that trees with high levels of aphids, which excrete high levels of sugar-rich honeydew onto leaves, had much higher concentration of dissolved organic carbon (DOC) in throughfall. In contrast, dissolved nitrate and ammonium in throughfall were 28-46% lower from aphidinfested trees. Although the effects of throughfall chemistry due to the aphids were not tested on understory species, positive effects would likely change with such large differences in throughfall chemistry. Very few studies have considered indirect positive interactions among annual species. One of these, designed to determine the effects of seed foraging on annual plant communities in the Negev Desert of Israel, found significant positive correlations between the establishment of large-seeded species and small-seeded species; but only when seed-eating gerbils were excluded (Lortie et al. 2000). When experimental seedbanks were exposed to gerbil foraging there was a change from positive associations among large- and small- seeded plants to random associations. Gerbil herbivory apparently restructured the annual community because gerbils selectively foraged for larger seeds and eliminated these species, thereby shifting the balance of net interactions from facilitation to competition. Despite the contributions of studies like that of Lortie and colleagues, we still have a relatively poor understanding of how herbivore-mediated positive effects balance with competitive interactions. Rob Brooker and colleagues at the Banchory Research Station in Scotland found that Calluna vulgaris (heath) protected Pinus sylvestris (Scots pine) from herbivores during the first winter after planting (Brooker et al. 2003). However, this facilitation did not lead to greater growth for P. sylvestris, suggesting that long term competitive effects of Calluna were as or more important than the indirect facilitative effect. Brooker et al.’s research is a good start, but there is great potential for future research on how herbivory-driven indirect facilitation interacts with competitive and allelopathic processes and how combined interactions produce total neighbor effects.
140
Indirect Mechanisms for Facilitation
3.3. REPRODUCTIVE FEEDBACK, POLLINATORS, AND POPULATION SIZE
Inverse, or positive intra-specific density-dependent population growth responses, have been recognized theoretically since 1931 when Allee proposed models to account for the potential positive effects of population size as well as already established negative effects (Allee 1931, also see Chapter 2). Allee hypothesized that the growth rates of small populations in particular may be reduced by the difficulty of finding mates, the inability to exploit a particular resource (probably most relevant to animals), or by more efficient predator behavior. For plants, the functional mechanisms that may produce the “Allee effect” include low pollen availability in small populations, the inability of small populations to favorably modify the environment, or as also for animals, low genetic variability in small populations. For example, low quality habitat inhibits the development of large patches of Senecio integrifolius (Widen 1993) and Gentiana pneumonanthe (Oostermeijer et al. 1994), and seed set for both species is limited by small population size. Most considerations of Allee effects have focused on single populations; however Allee effects are also likely to substantially inhibit the survival and expansion of metapopulations (Amarasekare 1998) and interspecific guilds (Schemske 1981). It is no surprise that conspecific neighbors have positive effects on the pollination of out-crossing species, and much of the evidence for density-dependent facilitation of this sort involves intraspecific interactions. These examples are less relevant to community issues; however, they illustrate a process that can be extrapolated to communities, and there are examples of indirect positive interspecific effects involving pollinators. Intraspecific density dependence certainly falls within the definition of facilitation; one individual benefiting from the presence of another. Plant species that rely on animal vectors for pollination may be particularly susceptible to Allee effects (Olesen and Jain 1994, Byers 1995, Matthies et al. 1995, Groom 1998), and plants can have strong indirect facilitative effects on each other through mutual attraction of pollinators (Feldman et al 2004). In an elegant experiment with Clarkia concinna, a patchily distributed annual plant species that grows in the coastal mountains of California, Martha Groom (1998) demonstrated strong Allee effects in small populations and evidence that these effects fed back to patch extinction. Roll et al. (1997) investigated the effects of plant density on reproduction of Lesquerella fendleri, a desert plant pollinated by insects. They found that individual reproductive success, measured as seeds per fruit, proportion of flowers setting fruit, and total seed production, increased with the density of conspecifics within 1 meter (Figure 3.13). Sih and Baltus (1987) found that very small patches of flowering plants were less attractive to pollinators and received less pollen than larger
Chapter 3
141 14
Mean seeds/fruit
12 10 8 6 4 2
Arcsin (square root of the proportion of fruit set)
0 1.4 1.2 1.0 0.8 0.6 0.4 0.2
dense populations sparse populations
0.0
Ln (1 + total seeds)
8
6
4
2
0
0
1
2
3
4
5
6
Ln (1 + number of plants within 1 m
Figure 3.13. Effects of absolute density on three measures of reproductive success of Lesquerella fendleri. In an ANCOVA for the effect of relative density on total seed set (the three measures combined) Fdensity=5.75, df=1,32, P=0.02. Redrawn from Roll et al. (1997) with permission from Conservation Biology.
patches of flowering plants. Pollen limitation in small patches of flowering plants has been suggested as the cause of low reproduction for a number of species; (Lamont et al. 1993, Petanidou et al. 1993, Aizen and Feinsinger 1994, Byers 1995, Fischer and Matthies 1998ab, Fischer et al. 2000). Kéry et al. (2000) studied the reproduction and offspring performance of two grassland plants whose populations are declining in many parts of Europe due to habitat fragmentation. Reproduction of both species was much lower in small populations and plants in those patches produced smaller and fewer seeds. Plant
142
Indirect Mechanisms for Facilitation
size was not correlated with the relationship between fecundity and population size, indicating that pollen limitation or inbreeding depression were the probable causes for the negative effects of small population sizes. Further evidence for the effect of small population size on inbreeding depression was found by Diane Byers (1995), who experimentally manipulated both pollen supply (quantity) and flower compatibility (quality) relationships for the rare plant Eupatorium resinosum. Pollen addition increased seed set in a small population of E. resinousum, but not in a large population. Also, plants from the small population were two times less cross-incompatible than plants from the large population. This relationship between small population size and pollen incompatibility has also been described by others (Whisler and Snow 1992, Shore 1993). Cassia biflora is a common shrub in savannas and abandoned pastures of Central America. Silander (1978) found that the reproductive success of Cassia was positively density-dependent and indirectly mediated. Seed set was determined by visitation from carpenter bees and the proportion of ovules set as seeds was inversely related to the distance to the nearest conspecific neighbor. Observations of isolated and clumped Cassia individuals determined that isolated individuals received fewer visits per hour. Shorea siamensis, a widespread dipterocarp in Southeast Asia, has experienced decreases in population density after years of selective logging (Ghazoul et al. 1998). Movement among trees of the small Trigona bees that pollinate Shorea declines substantially with increased distance among flowering trees, and correspondingly seed set at the most disturbed (least dense) site is lower than at other sites. Overall, fruit set is significantly inversely related with an index of neighbor proximity. Research on Cassia and other species indicates that population size, per se, may not be as important as population density for reproductive success. Van Treuren et al. (1993) investigated the effects of population size and density on outcrossing rates of Salvia pratensis, a locally endangered perennial sage in the Netherlands, and found no correlation between outcrossing rate and population size. However, high densities of conspecifics in both natural and experimental populations promoted outcrossing. Differences in foraging behaviors of the bumblebees that are the primary pollinators of S. pratensis may bring about the effect of density on outcrossing. In dense populations of S. pratensis bumblebees move between flowers of different individuals more often than they do in sparse populations (Heinrich 1979). For species of Viola the occurrence of interplant flights by insects is also greater in high-density populations (Beattie 1976). Outcrossing rates have also been found to correlate with plant density for other animal-pollinated species (Valdeyron et al. 1977, Wolff et al. 1988). In other words, the presence of a neighbor may positively affect the genetic variation of a plant’s offspring, and if genetic variation in a species affects fitness, neighbors may positively affect the fitness of a nearby conspecific.
Chapter 3
143
One of the more extreme cases of the importance of population size on fertility involves Banksia goodii, an Australian shrub that is pollinated by honey possums and honeyeater birds. Even though the individual plants in large and small populations did not differ in size, populations with very small numbers (2-12 individuals) produced no seeds at all, and increasing population size above this minimum was highly correlated with increasing seed production (Lamont et al. 1993). The complete absence of reproduction in small populations was not because of lower cone production, but because a much smaller proportion of cones on the plants were fertile. Five of the nine smallest populations produced no fertile cones over the last 10 years. It is not clear why small populations were so severely reduced in fertility; however, if plants have very specific pollinator requirements (and Banksia goodii does) and these pollinators have host densitydependent foraging requirements below which visitation to a small population is not cost effective, these populations may be completely ignored and thereby cease to reproduce. Populations of Silene regia, a perennial native to the Midwestern North American prairie, show high rates of numerical increase after fires; however, the ability of populations to respond favorably to fire is positively density-dependent (Menges and Dolan 1998). Large populations increase in size faster after fire than small populations. This response is probably linked to higher rates of seed germination from large populations (Menges 1991). Most studies of positive density dependent effects on reproduction have compared existing populations that vary naturally in size or that vary due to anthropogenic disturbance. In contrast to this, Bill Platt and colleagues (Platt 1974) experimentally manipulated population densities of Astragalus canadensis, made observations on pollinator and seed predator activity, experimented with insect access to flowers, and measured seed productivity. Exclosure of pollinators with netting virtually eliminated reproduction, indicating the need for cross pollination through insect vectors. Seed production was 31.7±0.2 pods per 5 cm of raceme for high density populations versus 25.2±0.5 for low density populations, a difference that was attributed to pollination success and host detection by the pollinator. They observed that bumblebees, the predominant pollinator, foraged within clumps and visited a large number of flowers in a patch before moving on to the next patch. Interestingly, the density of seed predator snout weevils on Astragalus stalks and racemes was much lower in the high density clumps and the proportion of seeds predated by snout weevils was lower in high density clumps. They attributed this pattern to a “type II predator response” in which the percent predation declines as prey densities increase. In other words, without a numerical response in predator populations, high density patches may simply over-saturate the ability of predators to attack all of the plants. With a numerical response by the predator, however, high density
144
Indirect Mechanisms for Facilitation
populations might be in worse shape than low density populations. For Astragalus canadensis, however, other conspecifics can have strongly facilitative effects on reproduction and intensity of predation. Dianthus deltoides, maiden pink, is a perennial herb distributed throughout grasslands of Europe where habitat fragmentation has been extensive. Jennersten (1988) compared pollination success of Dianthus in smaller fragmented populations to that in larger continuous populations. The fragmented area had a lower diversity of flowering plants and flower-visiting insects. Dianthus flowers received fewer visits in the fragmented populations and seed set was much lower. In the fragmented populations hand pollinated flowers increased seed set by over four times; however, hand-pollinated flowers did not increase seed set in the non-fragmented population, indicating substantial pollinator limitation. These results could be explained by site-effects, but they suggest that the pollinators can drive indirect, density-dependent positive interactions among individuals in a population. The effects of pollen limitation may be exacerbated by positive densitydependent interactions among the pollen grains themselves while on the stigma. Positive density-dependent germination of pollen grains was first observed by Brink in 1924 and since then has been shown for many species (Brewbaker and Majumbder 1961, Cruzan 1986, Holm 1994). Negative density-dependent interactions (competition or allelopathy) among pollen grains have also been observed (Galen and Gregory 1989, Thomson 1989, Murphy and Aarrsen 1995), but on the whole negative interactions appear to be less common than positive interactions among pollen grains and usually only occur between different species (Pasonen and Kapyla 1998). Both positive and negative interactions among pollen grains may be caused by chemical or physical mechanisms. Some evidence indicates that positive effects are due to the presence of a diffusible, water-soluable growth factor that is released from the pollen grains, and Brewbaker and Kwack (1963) identified calcium ions as one such factor. Other evidence suggests that enhanced germination of larger pollen populations might be produced by accelerated breakdown of cellular structures in the style, making the environment more favorable (Cruzan 1986). The fact that positive densitydependent responses are more common among conspecific pollen grains raises an interesting question about the ultimate function of the phenomenon. Why would you stimulate the growth of another pollen grain that has the potential to beat you to the seed? The most likely explanation is that pollen grains with the ability to detect the presence of other conspecifics and respond by germinating or growing tubes more quickly are more likely to fertilize the egg. In other words, like for most positive interactions, the benefactor is probably not behaving altruistically, but the beneficiary is simply taking advantage of an essential characteristic of the benefactor. This perspective on positive pollen grain
Chapter 3
145
interactions is supported by experiments that show mixtures of pollen grains derived from different clones have stronger positive density-dependent effects on germination and pollen tube growth than the same densities of pollen grains derived from a single clone (Pasonen and Kapyla 1998). This is a convincing argument that the signal is much more complicated than ionic concentrations. Detecting and outracing a genetically different conspecific may have more important evolutionary consequences than detecting and outracing a genetically identical conspecific. Intraspecific positive density-dependent pollination success is important, interesting, and relevant to the general issue of positive interactions among plants – it is not always bad to have even family as neighbors. However, interspecific effects are much more relevant to fundamental questions in community ecology such as coexistence, interdependence versus independence, the determinants of biological diversity, and the very nature of communities. As noted above, interspecific indirect interactions among plants involving pollinators are often competitive (Levin 1970, Reader 1975, Wasser 1978), with interspecific neighbors either reducing visitation in a heterogeneous mixture or reducing “carry over”, the amount of pollen reaching the second conspecific after a pollinator stops over at an intermediate interspecific. However, much like spatial aggregation often points toward facilitation, convergence in flowering time and space suggests potential positive benefits of mutual and cooperative attraction of pollinators (Poole and Rathcke 1979, Rabinowitz et al. 1981, Schemske 1981). Macior (1971) hypothesized that floral mimicry and pollinator sharing may be beneficial for species with restricted distributions. As described by Doug Schemske (1981), a rare species that provides no food reward may benefit by mimicking the flowers of a more abundant provider of rewards. Alternatively, species that offer rewards may benefit by converging in floral characteristics and increasing their “effective densities” and thereby increasing pollination probabilities. Different species may also facilitate each other simply by extending the flowering season and maintaining viable local populations of pollinators (Moeller 2004). Schemske worked on pollinator sharing in the context of evolutionary convergence; however, indirect facilitative effects may occur when pollinators are attracted to spend more time in interspecific, flower-rich patches – as long as interspecific pollen transfer does not suppress fruit production, as described below. James Thomson (1982) measured the relative rates of visitation by pollinators during the flowering season for several plant species in the central Rocky Mountains, USA. He found that phenological overlap among similar species in flowering was not correlated with visitation rates (neither competition nor facilitation), but that neighbors influenced visitation both positively and negatively. For Draba spectabilis and Chrysopsis villosa, higher flowering
146
Indirect Mechanisms for Facilitation
neighbor densities were correlated with increased visitation rates, suggesting indirect facilitation effects, but competition appeared to occur between Delphinium barbeyi and Aconitum columbianum for pollinators. In another study of subalpine plants, Thomson (1981) measured visitation rates on Potentilla gracilis and P. fruticosa in different neighborhoods. He found that visitation rates on both species were correlated with the presence of other plant species that shared the same pollinators, suggesting a cooperative rather than a competitive process. Thomson also investigated the effects of pollinator visitation in mixed stands of the hawkweeds Hieracium aurantiacum and H. florentinum along roadsides (Thomson 1978). Visits to H. aurantiacum were correlated with increasing proportional density of conspecifics in the stand, indicating possible interspecific competition. However, H. florentinum received more visits in stands with higher abundances of H. aurantiacum than when blooming alone, suggesting the potential for H. florentinum to be facilitated by its congener. Thomson pointed out, however, that these species can self-pollinate; therefore pollinator visitation patterns provide only a model for how non-apomictic plants might benefit from the presence of neighbors with similar flowers. In a study of two understory herbs in Costa Rican rainforests, Costus allenii and C. laevis, Schemske (1981) argued that floral convergence increased pollinator sharing without a cost to interspecific pollen transfer. He found that these two species occupy the same habitats at very low densities, flower in a highly synchronous manner, are identical in flower color, morphology, and nectar secretion patterns, and share the same highly specific pollinator, but have strong barriers to hybridization. Genetic comparisons indicated that these two species, despite the exceptional similarity in flower characteristics, were not closely related within the genus (D. Schemske, personal communication). Schemske argued that such convergence could only be to enhance pollinator sharing by these species so as to increase overall visitation for both species simply by increasing the density of attractive flowers in an area. This was not explicitly quantified, but the high amount of interspecific pollination that occurred did not decrease fruiting success. Brown and Brown (1979) argued for similar indirect facilitative processes among hummingbird-pollinated plants in Arizona. The mayapple, Podophyllum peltatum, produces no nectar yet is selfincompatible. This species appears to depend solely on haphazard visits from naïve queen bees hoping to find nectar, but unrequited for their efforts (Laverty 1992). Laverty and Plowright (1988) found that Podophyllum plants that were near Pedicularis canadensis (lousewort) produced more fruits and seeds than plants far from Pedicularis and hypothesized that the showy and reward-producing flowers of the latter species acted as a “magnet” for
Chapter 3
147
pollinators that periodically also investigated the flowers of Podophyllum (Figure 3.14). Podophyllum stands <25 m from Pedicularis flowers were four times more likely to be visited by queen bees than stands >50 m from Pedicularis, and in all three years of his study Podophyllum reproductive success was influenced by proximity to Pedicularis. When Pedicularis flowers were experimentally removed, proximity to the site previously occupied by Pedicularis no longer enhanced Podophyllum fruit and seed set. Such “magnet species” effects are facilitative, but species can compete for pollinators as well; the presence of more attractive neighbors means less attention. Lammi and Kuitunen (1995) placed “attractive” garden violets near nectarless, but apparently physically unattractive, Dactylorhiza incarnata orchids in marshes in Finland. They found that the violets reduced the number of fruits produced per flower of Dactylorhiza, indicating competition. One cannot be sure about the issues related to using a garden species not found naturally with the target, but these results suggest that attractive neighbors may have either positive or negative effects one’s reproduction.
Figure 3.14. Podophyllum peltatum (right) and Pedicularis canadensis. Drawn by Wendy M. Ridenour.
148
Indirect Mechanisms for Facilitation
Effect of spatial intermingling of flowers and seed set and pollinator visitation
Using visitation rates to determine neighbor effects does not take into account the importance of visit quality. As many of us know, proximity to a more attractive neighbor may increase the potential for reproductive interactions, but it may reduce the quality of those interactions (R.M. Callaway, personal experience). If pollinators visit more attractive neighbors at disproportionally high rates between their visits to less attractive neighbors, then the pollen load from the latter may not “carry over” to other conspecifics and the pollen may be wasted. If pollen is actually lost because of attractive neighbors, this could completely confound the apparent positive effect of higher visitation rates in the presence of an interspecific and turn the benefactor into a competitor. In addition to wasting pollen, high proportions of interspecific visits could clog stigmas and foreign pollen could inhibit fertilization (Thomson, 1982). Therefore, Thomson proposed a general conceptual model for the balance of effects of interspecific flowering on species that share pollinators (Figure 3.15).
visitation
+
0 seed set Highly Intermingled intermingled clumps flowers
Isolated clumps
Figure 3.15. James Thomson’s conceptual model for the effect of spatial intermingling on the relationship between two plant species which share inconsistent pollinators. When the lines are above the neutral points (0) the relationship is facilitative. When intermingling is low, isolated clumps may compete for pollinators. When intermingling is exceptionally high, the effect of neighbors may be competitive because of heterospecific pollen transfer. Redrawn from Thomson (1982) with permission from Oikos.
Chapter 3
149
Steven Johnson and colleagues (2003) experimentally evaluated the potential for less attractive species to benefit from attractive neighbors using the non-rewarding bumblebee-pollinated orchid, Anacamptis morio, and associated nectar-producing plants in Sweden. Pollen receipt and pollen removal for the “unattractive” A. morio was significantly greater for individuals translocated to patches of nectar-producing plants (Geum rivale and Allium choenoprasum) than for individuals placed outside (>20 m away) patches (Figure 3.16). Their results provide strong support for the existence of facilitative “magnet species” effects. Magnet: Geum rivale
Visited flowers
0.6
Magnet: Allium schoenoprasum
t = 3.7 df = 19 P = 0.002
0.4
Pollinated flowers
t = 2.8 df = 26 P = 0.01
0.2
0.2
0.1
0.2
0.2 t = 3.5 df = 19 P = 0.002
0.1
t = 2.2 df = 26 P = 0.04 0.1
0.6
Pollinaria removed
0.3
t = 3.3 df = 19 P = 0.003
0.4
0.3
t = 2.8 df = 26 P = 0.01
0.2
0.2
0.1
0.0
In
Out
0.0
In
Out
Treatment
Figure 3.16. Mean pollination success (proportion of flowers visited, proportion of flowers pollinated, and proportion of flowers with pollinia removed) of the deceptive orchid Anacamptis morio that had been experimentally transplanted inside (In) or outside (Out) patches of nectar producing species. Differences in mean pollination success were examined with paired t tests of arcsine-square-root transformed data (G. rivale, n=21 groups of translocated orchids; A. schoenoprasum, n=28 groups). Reprinted with permission from Johnson et al. (2003) with permission from Ecology.
150
Indirect Mechanisms for Facilitation
O’Connell and Johnston (1998) analyzed the relative importance of phenological traits of male and female Cypripedium acaule, an orchid that does not provide rewards for its pollinators, in two Nova Scotia populations. These orchids were highly pollen limited, with only 5 to 13% setting fruit naturally even though 100% set fruit when hand-pollinated. Particular floral traits were significantly correlated with female and male success; however, the presence of ericaceous shrubs were more closely linked to high pollination rates than floral traits. They attributed the important effect of shrubs to general attraction of pollinators. These studies indicate, that as for herbivores, pollinators can be drivers of important indirect facilitative interactions among plant species.
3.4. DISPERSERS When birds prefer some species for perches over others, and when birds prefer to consume the fruits of some species over others, strong spatial associations among perch species and defecated species can occur (Bleher and Bohning-Gaese 2001). This process may transport seeds to favorable sites for establishment or it may create a pattern that has nothing to do with true facilitation (See Chapter 2.11). The actual positive effect of the canopy species on the understory species may be blurred because perch trees may affect the spatial distribution of other species, but not facilitate them. If perch trees do not increase survival, growth, or reproduction of the species that are deposited beneath them then there is no clear facilitative effect. In some cases, the species composition of the seed shadows created by birds using shrubs and trees as perches appear to depend on the particular species of plant providing the perch. For example, Izhaki et al. (1991) found that Rhamnus palaestinus shrubs in Israel were visited more by birds than other species in Mediterranean scrub communities. Correspondingly, more seeds of different bird-dispersed species were found in the understory of Rhamnus. Jordano and Schupp (2000) found that strong microhabitat preferences for different species of frugivorous birds resulted in highly perch-tree specific seed shadows for Prunus mahaleb in Mediterranean-climate scrub in southeastern Spain. In Costa Rica, more seeds arrive under fleshy-fruited species than nonfleshy-fruited species, even when the fleshy-fruited species are not fruiting (Slocum and Horvitz 2000). This species-specificity has powerful effects on succession in this system (Slocum 2001). Species-specificity extends to benefactor gender in some cases. Juniperus sabina is a dioecious shrub found in the mountains in the Mediterranean region. Juniperus sabina appears to have
Chapter 3
151
strong modifying effects on its microhabitat and is associated with higher soil moisture, organic matter, and nitrogen and lower soil temperatures (Verdù and Garcìa-Fayos 2003). A large number of other species are spatially associated with J. sabina. However, Juniperus communis, which is bird-dispersed and shares the same species of bird dispersers with J. sabina, was highly partial to female J. sabina shrubs rather than males. This preference did not appear to relate to any greater environmental changes caused by females (but see Verdù et al. 2004), but to spatially biased seed dispersal meditated by birds attracted to the fruit-producing females of J. sabina, but unattracted to fruitless males. Interestingly, wind-dispersed Pinus nigra and ant-dispersed Helleborus foetidus, species that were spatially associated with J. sabina, were not gender-biased in their choice of benefactors. It would seem that female J. sabina shrubs play a much larger role in the diversity of their associated communities than do males. On the other hand, females may create conditions for themselves in which they experience greater competition and reduced fitness. Birds deposit the seeds of Capsicum annuum, the wild chili, preferentially under other shrub species that have red fruits like Capsicum itself (Tewksbury et al. 1998), and Capsicum has much stronger associations with red-fruited shrub species than others. All nurse shrub species reduce soil temperatures to a similar degree and there is no correlation between the effects of the different nurse shrub species on soil temperature and Capsicum in shrub understories. The effect of shrub beneficiaries on Capsicum may have other direct facilitative components, but the species-specificity of the positive effects are primarily due to indirect interactions with bird dispersers that chose perch trees on the basis of similarity in fruit color. Rosa canina (wild rose) is often found growing through and intertwined with the branches and canopies of Crataegus monogyna (hawthorn) in the Sierra de Cazorla of southern Spain (Herrera 1984). When alone, the morphology of Rosa is shrub-like, but when with Crataegus (or other species) Rosa plants form a vine-like climbing habit with the terminal foliage and fruits emerging intermixed with the outer canopy of the Crataegus shrubs. Both species produce similar appearing fruits at the same time and are well defended from herbivores by thorns. Rosa shrubs are three times more commonly intermixed with Crataegus based on the proportional cover of Crataegus, suggesting positive effects among these species. However, the combined effects of direct and indirect interactions are exceptionally complex. Rosa associated with Crataegus were larger and produced many times more fruits, apparently due to reduced browsing by ungulates. However, strong preferences of the avian fruit dispersers for Crataegus fruits substantially reduced the percent of Rosa fruits that were removed from the canopies. For Rosa growing alone 71% of the fruits produced were dispersed; whereas, only 5-8% of the fruits on Rosa growing in Crataegus
152
Indirect Mechanisms for Facilitation 30
20
10
0
0
30
20
40
Basal diameter (cm)
60
20
10
0
0
200
400
600
800
Canopy width (cm) 30
20
10
0
0
200
400
600
800
Height (cm)
Figure 3.17. Bird-dispersed woody species beneath Juniperus virginiana of different sizes on a Virgnia barrier island. Reprinted from Joy and Young (2002) with permission from Plant Ecology.
were dispersed. This tradeoff was further complicated by the observation that most Rosa fruits were deposited by the birds under Crataegus where predation from mice was much higher than in open microsites that reduced the probability of establishment. For Rosa, complex combinations of indirect interactions, both facilitative and negative, appear to be important. As for wind-driven collection of seeds under shrubs and trees (direct effects, Chapter 2.6 and 2.11) few studies have tried to separate the effects of canopies as foci for indirect animal dispersal and their effects as facilitators. One exception has been found on barrier islands off the eastern coast of North America. There, Joy and Young (2002) found that the richness of woody species was higher under Juniperus virginiana trees than on the surrounding exposed dune grassland. Additionally, fleshy-fruited seeds were much more abundant in the seed bank below Juniperus, and the abundance of bird-dispersed woody species increased with Juniperus size (Figure 3.17), suggesting that species
Chapter 3
153
richness patterns may have been due to Juniperus functioning as a dispersal focus. However, Juniperus also modified temperatures and increased soil moisture content; processes which corresponded with a positive effect of the canopy on two fleshy-fruited species in out-planting experiments. Their results indicated that the facilitative effects of Juniperus were due to the integrated functioning of this species as a dispersal focus and its amelioration of a harsh abiotic environment; however, the relative importance of these factors was not quantified. In a similar study, Kunstler et al. (in press) found that the relative importance of the facilitative effects of shrubs on the tree recruitment was much more important than the role of the shrubs as dispersal foci. García and Obeso (2003) documented a facilitative syndrome for the yew Taxus baccata in northern Spain, which was very similar to that described by Joy and Young (2002) for J. virginiana. Seed rain for Taxus was much higher under mature conspecifics and the holly Ilex aquifolium, but survival was also much higher under Ilex canopies where Taxus seedlings were protected from herbivory. In a unique and elegant study, Rey and Alcántara (2000) quantified the relative importance of dispersal and facilitative habitat modification by studying the processes affecting different stages of regeneration for Olea europea (wild olive) in southern Spain. Olea seeds are dispersed by birds, and Rey and Alcántara found that recruitment under adult conspecifics was very low, demonstrating that dispersal is crucial. However, the particular microhabitat to which an Olea seed was dispersed was also crucial. There was no recruitment from seeds reaching open interspaces, even less recruitment under Pistacia lentiscus than under Olea, and much higher recruitment under Quercus coccifera and Phillyrea latifolia. Although dispersal was not the most critical process, the ultimate recruitment success of Olea in these different shrub-defined microsites was a product of both dispersal probabilities and microsite effects on survival. For, example, Pistacia and Phillyrea had similar effects on the probability of seedlings becoming saplings (17.0% and 17.5%, respectively), but dispersal rates of Olea seeds were over six times higher under Phillyrea than under Pistacia. Microsite effects were most apparent in the comparison of seedling-sapling transition rates under Olea to rates under other species. Only 3.9% of seedlings became saplings under Olea canopies, whereas 12.5% to 17.5% of seedlings survived to sapling-hood under other species. Rey and Alcántara did not examine the mechanisms driving seedling recruitment directly, nor did they measure differences in abiotic factors among canopies species such as shade or soil nutrients, but they observed that most mortality was due to drought and that amelioration of abiotic conditions was more important than biotic safe sites. However, abiotic amelioration seems an unlikely explanation for the much lower seedling-sapling transition rates measured under Olea, and alternative
154
Indirect Mechanisms for Facilitation
possibilities such as allelopathic conspecific inhibition (Webb et al. 1967) or accumulation of host-specific pathogens (Packer and Clay 2000) are intriguing. Rey and Alcántara’s (2000) hypothesis that drought and moisture amelioration is more important than seed dispersal for the success of Pistacia lentiscus is supported by other research. Verdú and García-Fayos (1996) also found that Pistacia was more abundant under “perches” (mainly Ceratonia siliqua trees) than away from perches, but discovered that soil compaction was lower and soil moisture was higher under perches than in the open. Interestingly, they found a strong relationship between soil water potentials found under perches and the water potential necessary for effective germination. Pistacia germination was reduced at soil water potentials below –0.10 MPa, conditions that were reached 3 days after rainfall in open soils. Under perches, however, water potentials remained far above minimum requirements for 7 days. They also found that soil from beneath perches remained much less compacted after rains than soil from the open and was much easier for seedling radicles to penetrate. Considered together, their results indicate that Pistacia seedlings have a longer window in which to germinate and grow under bird-attracting perches and better conditions in which to establish once germinated. Although many studies demonstrate species-specificity in indirect facilitative interactions due to dispersers and the direct facilitative effects of beneficiaries once seeds are dispersed, other research suggests that the species of the perch tree is not particularly important. Toh et al. (1999) found that succession on abandoned farmland in Queensland, Australia was focused similarly around different mature isolated tree species where clusters of many different forest species aggregated. Most of the species in the clusters appeared to have dispersed there by bird or bats. Perches acting as dispersal foci may often simply drive changes in spatial pattern, but many examples in the literature demonstrate that dispersal interacts with other facilitative effects of benefactor species. Once nurse species collect wind or animal dispersed propagules, other facilitative mechanisms can then operate.
3.5. MYCORRHIZAE Mycorrhizae, mutualisms that arise between plant roots and some taxa of soil fungi, can modify or even reverse competitive interactions among plant species. For example, species that have an obligate relationship with mycorrhizal fungi acquire disproportionally more resources in the presence of the fungi than non-mycorrhizal or weakly mycorrhizal species and therefore gain competitive advantages (Caldwell et al. 1985, Allen and Allen 1990,
Chapter 3
155
Hetrick et al. 1989, Hartnett et al. 1993). However, for mycorrhizae to mediate indirect facilitation they must make the presence of some plant species beneficial for other plant species. This may happen in two ways. First, one plant species may cultivate certain mycorrhizal taxa that are beneficial for a second plant species (see section on plant soil feedbacks below). Second, mycorrhizae may mediate strong indirect facilitative effects by the transfer of resources and fixed carbon between individuals (Chiarello et al. 1982, Francis and Read 1984, Grime et al. 1987, Moora and Zobel 1996, Walter et al. 1996, Watkins et al. 1996, Simard et al. 1997, Marler et al. 1999a, Kennedy et al. 2003, Nara 2005). The presence of “donor” benefactors can increase the biomass and fitness of “receiver” beneficiaries, indirectly through mycorrhizae. However, clear proof of carbon transfer has been elusive (see Robinson and Fitter 1999). Direct and indirect effects of mycorrhizae can vary with resource availability, (Allen and Allen 1990, Hetrick et al. 1990, 1992, Johnson et al. 1997, Simard et al. 1997), the size of neighboring plants (Marler, 1999a), and the composition of the fungal community (van der Heijden et al. 1998). The direct effects of mycorrhizae are dependent on the plant species involved (Hartnett et al. 1993), but we know relatively little about species-specificity in interactions that appear to involve resource or carbon transfer (but see Francis and Read 1984, Grime et al. 1987, Simard et al. 1997). Some species may establish mycorrhizal communities that are beneficial to other species (Nara 2006ab, Nara and Hogetsu 2004). For example, CarrilloGarcia et al. (1999) investigated AM fungal colonization of perennial plants in disturbed and undisturbed plots in the Sonoran Desert near La Paz, Baja California Sur in Mexico to determine if arbuscular mycorrhizal (AM) fungi contribute to resource-island stability and plant establishment. They found that plant species that were not densely infected with AM fungi (cacti of the tribe Pachycereae: Pachycereus pringlei, Machaerocereus gummosus, and Lemaireocereus thurberi; and Agave datilyo) established preferentially in association with nurse trees. These pachycereid cacti grew preferentially under Prosopis articulata, whereas A. datilyo was disproportionally abundant under Olneya tesota canopies. They also found that the densities of AM fungal propagules were lower in soils where there were no plants than under plant canopies (40 vs. 280 propagules/kg soil). Furthermore, soils under shrub and tree “islands” were much more “enmeshed” with AM hyphae. In experiments, seedlings of P. pringlei were ten times larger when grown in soil collected under P. articulata than plants growing in soil from bare areas. Ideally Carrillo-Garcia et al. would have also included a comparison of the relative effects of the two soils after sterilization (if the two soils remained so dramatically different then mycorrhizae would not likely be the mechanism),
156
Indirect Mechanisms for Facilitation
but their results circumstantially support the hypothesis that enhanced development of AM fungi in soils associated with some species may increase resource uptake and help to stabilize windborne soil, in turn enhancing the establishment of other plant species. Hasselquist et al. (2005) found that seedling establishment by mature trees at the alpine treeline may be enhanced by a belowground, mycorrhizal component in addition to other direct facilitative effects of trees on seedlings (see Chapter 2). They found that colonization by Cenococcum mycorrhizal fungi was four-fold greater for young conifers that were adjacent to adult trees than for young conifers at least 7 m away from mature trees. Greater colonization enhanced seedling water potential, but not phosphorus concentrations or photosynthesis. In one of the earliest, yet most thorough, investigations of the potential for transfer of carbon from one plant species to another through AM fungi, J.P. Grime and colleagues used “microcosms” to manipulate the transfer of fixed carbon among plants (Grime 1987). Large Festuca ovina grasses were planted in the center of these microcosms and were surrounded by 18 much smaller subdominant species. To test for carbon transfer, Festuca root fragments containing primarily the AM fungal species Glomus constrictum, were added to some microcosms and uninfected fragments were added to controls. Afterwards the large Festuca plants were labeled with 3,700 kBq of 14C. Neighboring subordinate species, which were not labeled with 14C, were harvested 72 hr after inoculation and analyzed for fungal infection and 14C concentration. In the AM-inoculated treatment the percent root length of different subordinate species that was infected by AM fungi ranged from 0 to 89%, but in the no-AM treatment there was no sign of fungal hyphae in the roots. Correspondingly, shoot radioactivity (presence of 14C) for species with AM fungi in their roots ranged from ≈10 to 100 times more than that of conspecifics that were not inoculated with AM fungi. “Subdominant” species that were in microcosms with Festuca “donors” and AM fungi were approximately 3-10 times larger than conspecifics in microcosms with Festuca but without AM fungi. Importantly, there was no difference in shoot radioactivity for the non-mycorhizal Rumex acetosa in either of the two AM treatments. The discovery of radioactivity in the shoots rather than only in the roots of subdominant plants is also important because one of the primary arguments against carbon transfer is that carbon acquired by AM fungi remains in the tissues of the fungi (Graves et al. 1997, Fitter et al. 1998). If carbon labels are only found in the roots, it would not be possible to know if carbon had actually moved into the plant because label in fungal tissues cannot be distinguished from label in the root tissue in which the fungi are embedded. If this was the case, carbon transfer could be
Chapter 3
157
important to fungal carbon budgets but would be less likely to have important effects on host plant fitness. One of the subdominant species in the experiment by Grime and colleagues was Centaurea nigra. Centaurea nigra plants without AM fungi averaged 1.70 mg/plant, but 10.90 mg/plant in microcosms with AM fungi. Radioactivity of these same plants increased from 1,338±140 (95% C.I.) kBq with no AM fungi to 45,081±140 kBq with AM fungi suggesting that C. nigra was the recipient of fixed carbon from F. ovina. Interestingly, many Centaurea species from Europe and Asia have become some of the most devastating weeds in the world (Maddox and Mayfield 1985, Maddox et al. 1985, Roché and Roché 1988, Mueller-Scharer and Schroeder 1993) and several are heavily colonized by AM fungi in the soils of their new habitat (Marler, 1999, Callaway et al. 2001, Callaway et al. 2003). One of the worst of these invaders is Centaurea maculosa (spotted knapweed). Centaurea maculosa invasion often results in the development of dense, monospecific stands of C. maculosa and the competitive exclusion of virtually all native species - among which are species of Festuca. Inspired by Grime et al.’s work, Marilyn Marler and colleagues designed experiments with one of the Festuca species native to North America, F. idahoensis, in which both the putative donor F. idahoensis was manipulated as well as the fungi. Soil fungi had no direct effect on either species when they were grown alone. However, when the plant species were grown together, F. idahoensis plants were 170% larger when fungicide was added to the soil than when soil fungi were abundant. In contrast, when larger F. idahoensis were grown with C. maculosa, the C. maculosa plants were 66% larger when fungicide was not added to the soil (Figure 3.18). These results indicate that F. idahoensis facilitated C. maculosa through the indirect mediation of soil fungi, but whether or not the specific mechanism was carbon transfer is not clear. However, other experiments designed to measure transfer of carbon by taking advantage of natural isotopic differences between F. idahoensis and C. maculosa found that the stable carbon isotope concentration of C. maculosa was significantly more similar to that of F. idahoensis in the presence of soil fungi (Carey et al. 2004). In other experiments in which 13CO2 was used as a label there was no evidence for carbon transfer even though putative recipient C. maculosa were larger when mycorrhizae shared the rhizosphere with F. idahoensis (Zabinski et al. 2002).
Indirect Mechanisms for Facilitation
Biomass (g)
158 0.8
with AM fungi without AM fungi
0.6
*
0.4
0.2
0.0
Festuca grown with Centaurea
Centaurea grown with Festuca
Figure 3.18. Total biomass of Festuca idahoensis when grown with Centaurea maculosa, and C. maculosa grown with F. idahoensis, either with or without AM fungi in the soil. Error bars show 1 SE and the asterisk denotes a significant difference in a t-test, df=2,14; P=0.025. Reprinted from Marler et al. (1999) with permission from Ecology.
Sylvain Lerat and colleagues at The Université Laval in Quebec found strong evidence for carbon transfer in the field from the spring ephemeral, Erythronium americanum, to Acer saccharum through AM fungi (Lerat et al. 2002). In this case, a source-sink relationship appeared to develop because of differences in phenology of the receiver and donor plants. They found that 14C labels applied to the leaves of Erythronium showed up in the newly expanding leaves of Acer within a few days. There was far less (13 times) radioactivity detected in the leaves of Betula alleghaniensis, an ectomycorhizal species used as a control. They estimated that each nearby Erythronium plant had the potential to provide the carbon for 1.6% of the cost of Acer leaf expansion in the spring. In a similar experiment, Simard et al. (1997) measured net transfer of carbon in the field between seedlings of Betula papyrifera and Psuedotsuga menzesii, two ectomycorrhizal species. Over five times less carbon label appeared in the AM fungal species Thuja plicata. The amount of carbon received by plants was correlated with the shade in which it was growing, suggesting that shade-suppressed plants were facilitated by subsidies while waiting for an opportunity to reach the canopy. Some of the most interesting evidence for facilitative carbon transfer among plants is the lifestyle of over 400 species in 87 genera of “heteromycotrophs”, parasitic plants that do not have the capacity for
Chapter 3
159
photosynthesis, but do not attach themselves directly to the roots of heterotrophic plants (Leake 1994, Figure 3.19). Instead these species are mycorrhizal and are connected to heterotrophic species by the fungi in these mutualisms (Leake 1994, Cullings et al. 1996, Taylor and Bruns 1997). Heteromycotrophs appear to acquire fixed carbon through the fungi in the mutualism. Although some experiments point to carbon transfer from autotrophs to heteromycotrophs (Bjorkman 1960), the possibility remains that the carbon is actually being taken up by the fungi from organic matter in the soil. Regardless, the only options that seem to explain the means of existence for heteromycotrophs are 1) root connections have been missed, 2) these plants can turn atmospheric carbon into organic carbon without chlorophyll, 3) they acquire carbon from organic matter, or 4) they acquire carbon from other plants. The first two seem quite unlikely and the latter two require the movement of fixed carbon from fungi to plant tissues. Carbon transfer may play a role in indirect facilitation, but alternatively, facilitative effects may arise as different combinations of plant species change the composition of the microbial community by shifts in the
Figure 3.19. The mycoheterotroph, Monotropa uniflora. Drawn by Wendy Ridenour.
160
Indirect Mechanisms for Facilitation
composition of the total fungal community (see Bever 1994, Bever et al. 1997, Stephan and Meyer 2000), or by shifts in the composition of AM fungal communities (Johnson et al. 1997, Egerton-Warburton and Allen 2000). A third possibility is that putative donor benefactors have strong positive effects on the growth of soil fungi, but the feedback of soil fungi to the benefactor is less positive than the feedback to the beneficiary. I found that plants of Centaurea melitensis (a European annual that has invaded grasslands around the world) grown alone were over 50% smaller when grown in North American soil with an intact resident microbial community than when fungicide was applied (Callaway et al. 2003). However, when grown with the California native, Nassella pulchra, the effect of fungicide was reversed. Centaurea melitensis grew almost 5 times larger with the resident microbial community intact than in the fungicide treatment. These results suggest a tripartite facilitative interaction in which Nassella benefits C. melitensis, but only in the presence of soil fungi. Many other studies have documented hyphal connections between plants (Hirrel and Gerdemann 1979, Heap and Newman 1980, Newman et al. 1994), and provide evidence that carbon or phosphorus moves between plants via these connections (Chiariello et al. 1982, Whittingham and Read 1982, Francis and Read 1984, Read et al. 1985), but fungal mediation of plant-plant interactions remains a fascinating and mysterious aspect of ecology.
3.6. PLANT-SOIL MICROBE FEEDBACKS Mycorrhizae typically alter interactions among plants by providing an advantage to infected plants. The advantage may be an extended resource acquisition network or protection from harmful microbes; however, other mutualistic and pathogenic soil microbes may indirectly meditate facilitative processes through plant-soil community feedbacks. Soil microbes affect plants, but feedback occurs because plants can also change the microbial communities in their rhizospheres (Westover et al. 1997, Stephan and Meyer 2000). Positive feedback occurs when plant species accumulate host-specific, or at least host-favored, beneficial microbes over time, such as mycorrhizae (Bever 1995, Bever et al. 1996, 1997). Positive feedbacks are thought to lead to a loss of community diversity because they favor long-term persistence of a species in a particular place. Negative feedbacks occur when plant species accumulate host-specific pathogenic microbes (van der Putten et al. 1993, Bever et al. 1996, 1997) and are thought to enhance community diversity by increasing species turnover rates. Although to my knowledge such negative feedbacks within communities have never been explicitly considered in the context of
Chapter 3
161
0.6 0.4
*
*
*
*
0.2 0.0 -0.2 -0.4
Exotic invasives
* *
Aletris
Liatris
* Agaliris
Gentiana
Cirsium
Polygonum
Lythrum
* Euphorbia
-0.8
* Polygala
-0.6
Alliaria
Growth of plants in their own soil relative to that in soil of other species
indirect facilitation, indirect facilitation occurs when the presence of one species creates beneficial, relatively pathogen-free, habitat for other species. In other words, when one species establishes a soil biota more favorable for a neighbor than the soil biota established by the neighbor itself, indirect facilitation occurs. John Klironomos of Guelph University demonstrated that positive and negative plant-soil microbial feedbacks have the potential to elicit very strong effects on the abundance of species in plant communities (Klironomos 2002). In a series of experiments using species from old fields in southern Ontario, Canada, he showed that rare plants exhibited negative feedbacks, a relative decrease in growth on ‘home’ soil (soil trained by conspecifics) in which pathogens had had a chance to accumulate (Figure 3.20).
Rare natives
Figure 3.20. Soil feedback responses of five invasive and five rare plant species. N=10 per treatment. Error bars represent 1SE. Asterisks represent significant differences in plant growth when grown in home versus foreign soil after t-test analysis. Reprinted from Klironomos (2002) with permission from Nature.
162
Indirect Mechanisms for Facilitation
In support of Klironomos’ findings, colleagues at University of Montana and I found that the invasive exotic plant, Centaurea maculosa, cultivates soil biota with increasingly negative effects on its growth in soil from its native Europe (Figure 3.21, Callaway et al. 2004). However, the opposite occurred in soils where C. maculosa has invaded in North America. In soils from North America the weed develops positive feedbacks with soil biota, an interaction which may contribute to the success of this exotic species in North America Kurt Reinhart and colleagues (2003) explored the soil microbePrunus serotina relationships discovered by Packer and Clay (2000) in its native North America in the context of P. serotina invasions in northwestern Europe. Contrary to the highly suppressive effect of soil biota in its native North America, soil microbes in Europe facilitate P. serotina. In the native range, the soil community that develops near black cherry inhibits the French soil pre-cultured by: Centaurea F. ovina
Centaurea biomass (gm)
1.0
0.5 P=0.095
0.0 Montana soil pre-cultured by: 1.0
Centaurea F. idahoensis P=0.032
0.5
0.0
NOT STERILE
STERILE
Figure 3.21. Total biomass of Centaurea maculosa plants grown alone in European soil and North American soil that had been pre-cultured by either C. maculosa or a Festuca species native to the place of soil origin. Plants were grown in soils either sterilized or not sterilized to reduce the effects of soil biota. Reprinted from Callaway et al. (2004) with permission from Nature.
Chapter 3
163
establishment of neighboring conspecifics in the field and reduces seedling performance in experiments. In contrast, in the non-native range young P. serotina are abundant near conspecifics, and soil biota from near P. serotina adults enhances the growth of conspecific seedlings. Sarig et al. (1994) measured the biomass of a dominant understory bunchgrass species (Stipa capensis), soil moisture, salinity, and nitrogen under the shrub Hammada scoparia in the Negev Desert of Israel. Under the outer edges of shrub canopies S. capensis biomass increased by 4 times, soil moisture decreased, and soil salinity and nitrogen increased (Figure 3.22).
200 150 100 50
Biomass (g m-2)
250
0
Soil moisture (% WHC)
35 30 25
*
20
20
Nitrogen (ug.g-1)
+
DEPTH
*
*
*
16
+ +
Microbial carbon (ug g-1)
12
140
+
+ +
120 100
+ + +
80 1/3 2/3 radius radius
4/3 radius
Relative distance between trunk and canopy edge (r = radius)
Figure 3.22. Distribution of Stipa capensis biomass, soil moisture, soil nitrogen, and soil microbial biomass beneath and outside the canopies of Hammada scoparia shrubs. * represent significant differences among depths (ANOVA), + represent significant differences among locations (ANOVA), and error bars represent 1 SD, n=6. Redrawn from Sarig et al. (1994) with permission from Acta Oecologia.
164
Indirect Mechanisms for Facilitation
Although soil nitrogen may have played a direct role in the facilitation of the bunchgrass, Sarig et al. also found that microbial biomass was almost twice as high under shrubs as in the open. If fact, across all samples, the biomass of S. capensis was much more strongly correlated with microbial biomass than other characteristics. Sarig et al.’s results suggest that soil microbes may play an important indirect role in the positive effect of shrubs on understory species; however, no experiments were conducted to explore the biological role of microbes. Furthermore, data were only collected on microbial biomass, and the effect of shrubs on the particular species of microbes may have had important effects as well.
3.7. POSITIVE INDIRECT INTERACTIONS AMONG COMPETING PLANTS So far in this chapter I have focused on the indirect positive effects that occur when animals or microbes alter interactions between two plant species. But indirect interactions can also occur within the plant community via two different mechanisms. First, competitive interactions between two species can be altered by simultaneous competitive interactions with additional species (Buss and Jackson 1979, Lawlor 1979, Stone and Roberts 1991, Miller 1994, Wooton 1994). Second, indirect interactions can arise through the cumulative “diffuse” effects that occur when numerous species have different kinds of direct effects that act on a single species (Davidson 1980, Wilson and Keddy 1986, Vandermeer 1990). In both of these categories of indirect interactions the facilitative effect is produced by something analogous to an alliance – an enemy of my enemy is my friend. If a minor enemy hurts your major enemy more than it hurts you then the minor enemy can be your facilitator. Indirect effects will not occur when competitive interactions among plants within a community are hierarchical (transitive patterns); where all species of higher competitive rank outcompete all species of lower rank. However, when at least one species of lower rank outcompetes one or more species of higher rank (networks, or intransitive patterns) indirect effects can have dramatic effects on communities (Karlson and Jackson 1981). For example, Levine (1976) modeled the effects of adding a third competitor in a system with two competing species and found that under some conditions the additional species could change the cumulative effect of one species from competitive to facilitative because of how it suppressed a shared competitor (Figure 3.23). The third species may have competitive effects on both species, but as long as the third species indirectly relieves the total competitive pressure on a neighbor more than it directly hurts that species, facilitation can occur. Other models have been developed to examine direct and indirect interactions among invading exotic and native
Chapter 3
165
SPECIES A
SPECIES B
- + -
SPECIES C
Figure 3.23. Conceptual model for indirect effects among three competing plant species. Solid lines represent direct effects and dashed lines represent indirect effects. Thickness of lines is proportional to strength of effect.
species. A model developed by Case (1991) suggested that natives may be completely displaced by exotics in pair-wise interactions, but can coexist with exotics in diverse communities, apparently because of the greater number of indirect interactions. Despite the modeled importance of indirect interactions in plant communities, only a few experiments have attempted to experimentally demonstrate the existence and intensity of indirect interactions among competing plant species within communities (Miller 1994, Li and Wilson 1998, Levine 1999, Callaway 2000, Rice and Nagy 2000, Maestre et al. 2004, also see Chapter 4.4, 4.5). However, because of a current awareness of the importance of indirect interactions among competitors, the number of experimental studies is increasing rapidly. Some studies have interpreted spatial patterns to be suggestive of indirect effects among plants, but to my knowledge the first experiment designed explicitly to quantify indirect effects among interacting plants was conducted by Tom Miller (1994) using old-field species in the Midwest of the USA. Miller proposed a model of plant interactions designed to estimate direct and indirect effects occurring between all possible pairs of species in a community and based his model on the mean responses of focal individuals to variation in the abundances of different associate species. The model assumed “competitor equivalence” (all species have the same per-gram competitive effect on a focal species regardless of their identity) therefore the yields of different associate species can be added for a net effect. Competitor equivalence is controversial (see Chapter 5), but Miller integrated his model with an elegant two-year field experiment in which he selectively removed species from experimental plots and compared the growth of remaining species to their potential growth without competitors. In this way he
166
Indirect Mechanisms for Facilitation
Ambrosia -.41
-.38 +.43 -.73 -.30
Agropyron
+.13
+.41
Plantago
-.38 -.27
+.22
-.15
+.23
-.26
+.13 Chenopodium
Trifolium
Figure 3.24. Flow diagram of direct (solid lines) and indirect (dashed lines) effects of five different old field plant species on each other. Redrawn from Miller (1994) with permission from The American Naturalist.
quantified direct and indirect effects among six old-field plant species (two species were substituted for each other between the years) over two years. Miller found that direct and indirect effects were common and strong, but interactions among the five species in the second year did not occur in a fully balanced web and interaction strengths were not equivalent among species. But interactions sorted themselves out so direct negative effects among particular species were balanced by positive indirect effects (Figure 3.24). Total effects were always neutral to negative, but indirect positive effects appeared to strongly ameliorate direct competitive effects. Ambrosia artemisiifolia was the competitive dominant and had strong direct suppressive effects on all other species, whereas no other species affected the biomass of Ambrosia. Agropyron repens and Plantago lanceolata had significant direct competitive effects on other species, but were significantly suppressed by Ambrosia. Such direct competitive relationships are not unusual, but Miller found that strong positive indirect effects largely offset these negative direct effects. The strongest example of offsetting direct and indirect effects was for Ambrosia and Agropyron. The proportional direct effect of Ambrosia on the potential growth of Agropyron was –0.41, yet the proportional indirect effect was +0.43, apparently because Ambrosia also highly suppressed other competitors. However, no other species had significant direct competitive effects on Agropyron so there was no clear link among species interaction that
Chapter 3
167
0.5
Biomass (g/plant)
a
2.5
Mimulus guttatus
b
0.4
2.0
0.3
1.5
Conocephalum conicum
1.0
0.2
thinned Carex 0.1
0.5
Carex not thinned
0
0.0 0.8
c
0.14 Mimulus cardinalus
d
Brachythecium frigidum
0.12
0.6
0.10 0.08
0.4 0.06 0.04
0.2
0.02 0.0
0.00
No Mimulus guttatus
Mimulus guttatus added
No Mimulus guttatus
Mimulus guttatus added
Figure 3.25. (a) The effect of Carex nudata on the biomass of Mimulus guttatus. (b, c, d) The effects of Carex on the biomass of Conocephalum conicum, Mimulus cardinalus and Brachythecium frigidum with and without M. guttatus present. Reprinted from Levine (1999) with permission from Ecology.
could explain Ambrosia’s strong positive indirect effects. It is possible that Ambrosia altered strong intraspecific competition among Agropyron individuals. A second example of balanced direct and indirect effects occurred among Ambrosia, Agropyron, and Trifolium (Figure 3.24). Ambrosia plants suppressed Agropyron (-0.41) and Trifolium (-0.71), but Agropyron also suppressed Trifolium (-0.38). Therefore, the suppression of Agropyron contributed to Ambrosia’s significant positive indirect effect on Trifolium (+0.41). Three less abundant species, Lepidium campestre, Chenopodium album, and Trifolium repens, were all highly suppressed in the full community treatments and had no significant effects on any other species. As described in Chapter 2, Jonathan Levine (2000) showed that large tussocks of Carex nudata are tolerant to scouring river flows and that many other species rely on C. nudata to survive floods. In addition to this direct facilitative effect, C. nudata has indirect facilitative effects on the species growing within its root mats and tussocks. Levine (1999) manipulated the presence of C. nudata and Mimulus guttatus in a factorial design. He demonstrated that thinning the leaves of C. nudata resulted in much bigger M. guttatus, indicating that the tussocks strongly suppressed M. guttatus (Figure 3.25a). Second, in the absence
168
Indirect Mechanisms for Facilitation
of M. guttatus, which he had removed, dense C. nudata strongly suppressed the liverwort Conocephalum conicum, whereas thinning the tussocks released the liverwort from direct competitive effects. However, in a crucial treatment for understanding indirect effects, Levine found that the effect of C. nudata was reversed in the presence of M. guttatus (Figure 3.25b). In other words, when the liverwort had to cope with an exceptionally aggressive neighbor, M. guttatus, it helped to be buried inside dense tussocks of C. nudata. Without the aggressive neighbor, the same dense tussocks were only negative. Similar, but weaker patterns were found for a second target species, M. cardinalis, but for a third species, Brachythecium frigidum, no indirect effects occurred. For Conocephalum, field patterns corresponded with the experimental results. The cover of M. guttatus was negatively correlated with C. nudata stem density, and where stem density of C. nudatum was low (and M. guttatus high) there was less Conocephalum. Levine did not investigate all possible mechanisms behind these indirect interactions, but M. guttatus was probably suppressed by the low light levels within the tussocks. Liverworts are well-known for shade tolerance, and such tolerance may have made tussocks a competition-free refuge. How M. gutttatus suppressed Conocephalum is less clear, but Levine hypothesized that the tightly packed stolons of the former physically resisted the spread of the liverwort. The widely described “nurse” effects of many tree and shrub canopies on the regeneration of other species may also operate in an indirect way. These indirect effects occur when the canopy species reduce or eliminate particularly competitive species, and by doing so create a low-competition environment in which other species can survive. Canopies may have no direct positive effects what-so-ever on species that recruit beneath them, but by eliminating competitors stronger than the nurses themselves, nurses can be facilitative. For example, several different shrub species occur at high densities in the ecotones or “tension zones” between tallgrass prairie and deciduous forest in central Oklahoma. These shrubs marginally increase soil moisture and nutrients, but are highly associated with the establishment of tree seedlings (Petranka and McPherson 1979). They found that trees from river terraces did not recruit in tallgrass prairie except under the shrub Rhus copallina. Establishment of upland tree species was not as obligate, but upland tree species were also nursed by Rhus. Petranka and McPherson noted that grass densities under shrubs were only 25% of that in surrounding prairie, and their experiments implicated both the deep shade of Rhus canopies and allelopathic chemicals from shrub leaves as factors suppressing the grasses. Although not explicitly stated in their paper, the concomitant suppression of grasses and increase in tree seedling establishment raises the possibility of important indirect effects of shrubs on
Chapter 3
169
trees. To show this, measurements of the performance of tree seedlings in the open, but with grasses removed, would have had to be compared to the performance of tree seedlings under shrubs. Werner and Harbeck (1982) were more explicit in their case for the indirect effects of a congener, Rhus typhina, on tree seedling establishment in old fields in the midwestern USA. The density of late successional trees (wind and bird dispersed) was over 3 times higher under Rhus than in the open grassland. However, herbaceous understories also changed dramatically under Rhus. The sod-forming grassland dominant, Agropyron repens (quackgrass) died out almost completely under shrub canopies and was replaced by smaller and perhaps less competitive Poa species. Werner and Harbeck argued that the effects of Rhus on tree recruitment were not likely due to direct effects, but through the “creation of openings required for successful tree seedling establishment…” Similarly, Veblen et al. (1979) found that different species of Nothofagus trees in southcentral Chile had strikingly different understory communities, even when the individuals were close together and intermixed. They argued that the higher summer light levels and earlier snow melt under evergreen (in contrast to deciduous) Nothofagus favored tall, rapidly growing Chusquea tenuiflora (bamboo) which excluded other herbs, shrubs, and tree seedlings. However, like Petranka and McPherson and Werner and Harbeck, Veblen and colleagues did not conduct experiments to separate direct positive effects from indirect positive effects mediated through the suppression of superior competitors. The indirect process by which a benefactor facilitates a neighbor by keeping more aggressive competitors at bay may have reverse cases as well. Aguiar and Sala (1994) found that shrubs in Patagonian steppe facilitated tussock grasses beneath them until competition from the exceptionally dense grass understories overshadowed facilitation (also see Maestre et al. 2004). Indirect interactions among plants may occur at the margins of some high-elevation meadows. In the Oregon Coast Range the recruitment of Abies procera occurs primarily at the ecotones between forests and subalpine meadows (Magee and Antos 1992). Abiotic site conditions may produce these patterns, but age analyses of seedlings indicate that the invasion of meadows over time by Abies is a steady, consistent process. Abies seedlings may benefit from the way that tree canopies at the ecotone collect snow, provide shade and protection from wind, or some other direct factor. However, the cover of dominant grasses also declines dramatically at the ecotone, perhaps due to the suppressing effects of acidic conifer litter and canopy shade. Competition from grasses and forbs suppresses tree seedling establishment in many environments (Billings and Mark 1957, Dunwiddie 1977, Vale 1981, Li and Wilson 1998), and it may be that the direct negative effects of trees on grasses at the ecotone are more important than any of their direct effects on the seedlings themselves.
Indirect Mechanisms for Facilitation
Growth rate ([ln(g.g-1)].d-1)
170
0.008 0.006 a
0.004
a b
0.002 0.000
a
a
b b
b
Density
1 5
1 5
1 5
1 5
Grass
Pre
Rem
Pre
Rem
Species
Symphoricarpos
Picea
Figure 3.26. Growth rates of Symphoricarpos and Picea grown at densities of one or five plants per plot, with grass intact or removed (pre = present, rem = removed). Error bars show 1SE and letters show differences from means contrasts between density and grass removal treatments for each target species. There was a significant three way interaction between density, grass treatment and species. Reprinted from Li and Wilson (1998) with permission from Ecology.
Li and Wilson (1998) presented more conclusive evidence for indirect plant-plant interactions at forest-grassland ecotones in central Canada. They tested whether the presence of conspecifics enhanced the growth and survivorship of Picea glauca (white spruce) and Symphoricarpos occidentalis (snowberry) seedlings either with or without the perennial grass Bromus inermus. After two growing seasons, woody seedling survivorship was reduced in the presence of Bromus. For Picea, conspecific neighbors significantly decreased growth rates in all cases and grasses decreased Picea growth rates in non-manipulated plots (Figure 3.26). However, when shadecloth was used to simulate the effects of forest canopies the competitive effects of grasses on Picea were eliminated. For Symphoricarpos, conspecific neighbors significantly decreased growth rates when no grasses were present, but conspecifics increased growth rates when growing in stands of Bromus. In both cases, and primarily the latter, indirect effects were important. For Picea, simulated forest shade did not have strong direct effects on seedling survivorship or growth, but shade impeded the ability of grasses, presumably
Chapter 3
171
adapted to high light, to suppress seedling growth. Unlike Picea, which did not benefit from the presence of other small conspecific neighbors, Symphoricarpos seedlings grew larger when they were planted in a group with other Symphoricarpos seedlings, but this intraspecific facilitation only occurred when grasses were present. When grasses were removed Symphoricarpus seedlings competed with each other. Jean-Philippe Pages and Richard Michalet (2003) pursued indirect facilitative effects in a mature deciduous forest in the northern French Alps by comparing the responses of seedlings of five tree species to tree canopy removal in the presence and absence of a dominant herbaceous neighbor, Molinia caerulea. Molina becomes exceptionally dense in forest openings and may hinder the regeneration of tree species after forest cutting. Pages and Michalet observed more tree seedlings within the forest than in clear-cut areas dominated by Molina, suggesting possible indirect facilitation by tree canopies on seedlings through the suppression of Molina. They compared the performance of seedlings of five tree species planted within the forest and in experimental gaps, with and without the herbaceous competitor Molinia caerulea. Within the forest, competition between Molinia and tree seedlings was very weak. However, the gap treatment produced a strong increase in biomass for Molinia, which resulted in greater competitive suppression of the tree species. Together, these results indicate a facilitative indirect effect. However, as discussed in general in Chapter 4, this facilitative effect interacted strongly with competitive effects. Despite the competitive effect of Molinia, the growth of target tree seedlings was higher in the light gaps than under the forest canopies because the adult trees were exceptionally strong competitors for light, and thus had a greater direct negative effect on tree seedling growth than Molinia. In sum, adult trees suppressed an important understory competitor, but the direct negative effect of trees on the seedlings was larger than this indirect positive effect. Kunstler et al. (2006) used a protocol similar to that of Pages et al. (2003) to study the indirect effects of Buxus sempervirens and Juniperus communis and shrubs on the recruitment of Quercus pubescens and Fagus sylvatica in southern France. But in contrast to Pages et al. (2003), they found that the shrubs had strong indirect facilitative effects on tree recruits through the suppression of understory grasses. In the upper zones of coastal salt marshes in southern California, the two dominant perennial species have strikingly opposite direct effects on most co-occurring winter annual species. Arthrocnemum subterminale, a succulent sub-shrub, facilitates two annual species, Parapholis incurva and Hutchisinia procumbens (Callaway 1994); whereas Monanthecloe littoralis, a clonal grass, appears to eliminate virtually all annuals by establishing a thick
172
Indirect Mechanisms for Facilitation
vegetative mat. Furthermore, long-term community monitoring shows that Monanthechloe has increased in abundance and has displaced both Arthrocnemum and winter annual species. Steve Pennings and I tested the hypothesis that Monanthechloe might directly outcompete annual species, but that the positive effect of Arthrocnemum might buffer annual species from the full negative impact of Monanthechloe (Pennings and Callaway 2000). Over a 13 year period we correlated changes in the relative abundance of Monanthechloe with the abundance of other species and measured patterns of abundance of the annual species associated with Arthrocnemum in the presence and in the absence of Monanthechloe. We also conducted field experiments to investigate the direct effects of the perennials on each other and on annual species, and to measure the effects of Monanthechloe on other species in the presence and absence of Arthrocnemum. The results showed strong direct competition, facilitation, and indirect facilitation among plant species in the salt marsh. Monanthechloe, which had been increasing in abundance in the marsh, outcompeted annual species and the herbaceous perennial Limonium californicum. Winter annual species were able to coexist with Arthrocnemum, but not with Monanthechloe, and the presence of Arthrocnemum shrubs in mats of Monanthechloe was correlated with higher abundances of annual species (Figure 3.27). Density (individuals per 100 cm2)
300
a
Arthrocnemum only Monanthechloe only Athrocnemum & Monanthechloe
250 200 150 100
a c a
b
a
50
b
b
c
b a
3 2 1 0
b 0
Spergularia Parapholis
Lasthenia
Juncus
a bb 00 Hutchinsia
b a
0
Limonium
Figure 3.27. Densities of annual species in plots containing the perennials Monanthechloe littoralis, Arthrocnemum subterminale, or both in Carpinteria Salt Marsh, California. Bars represent 1 SE, and shared letters for a species indicate no significant difference between means (ANOVA, post-ANOVA Tukey, P < .05. Reprinted from Callaway and Pennings (2000) with permission from The American Naturalist.
Chapter 3
173
The perennial herb, Limonium californicum, was absent in plots with only Monanthechloe, uncommon in plots with only Arthrocnemum, and most abundant where Monanthechloe and Arthrocnemum were both present. These patterns suggested that Arthrocnemum buffered the competitive effects of Monanthechloe. In field experiments, when Arthrocnemum was left intact, but Monanthechloe was removed, the density of Parapholis was over five times higher than when both perennials were removed (Figure 3.28). In intact patches of Monanthechloe, Parapholis survived only when Arthrocnemum was present. Similar patterns were observed for Lasthenia and Juncus, but the direct effects of Arthrocnemum were not nearly as strong. The most striking indirect effects in the factorial field experiment occurred for Spergularia and Limonium. In the absence of Monanthechloe, Spergularia was much more common when Arthrocnemum was removed, indicating a competitive effect 80
10
Parapholis
Juncus 8 6
60
P<0.001
P=0.388
Density (ind./100cm2)
40 4 20 2
P=0.080 0
0
0 60
P=0.175
Lasthenia
0
0
80
P=0.008
Spergularia
60
40
40 20 20
P=0.003
P=0.062
0 0
-A +A -M
-A +A +M
0
-A +A -M
-A +A +M
Figure 3.28. Densities of the four most common annual species after the factorial removal of Monanthechloe littoralis and Arthrocnemum subterminale at Carpinteria Salt Marsh, California. Bars represent 1 SE. In ANOVAs for each species the Arthrocnemum x Monanthachloe interaction effect was significant for Parapholis and Spergularia. Reprinted from Callaway and Pennings (2000) with permission from The American Naturalist.
174
Indirect Mechanisms for Facilitation
Density (ind./10cm2)
6 5
P=0.0778
4
P<0.001
3 2 1
0 -A
+A -M
-A +A +M
Figure 3.29. Density of Limonium californicum after the factorial removal of Monanthechloe littoralis and Arthrocnemum subterminale at Carpinteria Salt Marsh, California. Bars represent 1 SE. In an ANOVA the Arthrocnemum x Monanthachloe interaction effect was significant. Reprinted from Callaway and Pennings (2000) with permission from The American Naturalist.
of Arthrocnemum. But when the dominant competitor Monanthechloe was present, Arthrocnemum strongly facilitated Spergularia. Limonium was indifferent to the removal of Arthrocnemum when Monanthechloe was absent, but in the presence of Monanthechloe, Limonium recruits were only found in plots in which Arthrocnemum was also present (Figure 3.29). Two different types of indirect interactions may explain these results. First, webs of indirect effects (sensu Lawlor 1979, Miller 1994, Wooton 1994) probably best explain the results with Spergularia and Limonium. In the absence of Monanthechloe, Arthrocnemum competed with Spergularia and had no effect on Limonium, but in the presence of Monanthechloe, Arthrocnemum facilitated Spergularia and Limonium, probably because Arthrocnemum had a competitive effect on Monanthochloe. Second, “diffuse” interactions (sensu Wilson and Keddy 1986) may best explain the effects of Arthrocnemum on Parapholis and Lasthenia (i.e., a direct positive effect of Arthrocnemum on these annual species canceling out a direct negative effect of Monanthochloe) because Arthrocnemum had positive effects on these species with or without Monanthechloe. The competitive effect of Monanthechloe on Arthrocnemum, Limonium, and all of the annual species was stronger and more consistent than
Chapter 3
175
the positive effects of Arthrocnemum on other species or the competitive effect of Arthrocnemum on Monanthechloe. However, weak or “feeble” indirect effects among consumers have been shown to have important overall effects in communities (Berlow 1999). The mechanisms by which Monanthechloe outcompetes annuals and Limonium were not clear, but heavy heavy mats of Monanthechloe created very dense shade which may have eliminated annuals. In grasslands of the northern Rocky Mountains invaded by Centaurea maculosa, Lupinus sericeus appears to play the role of Arthrocnemum. Tiffany Weir and colleagues found that sample plots containing Lupinus were much more likely to also contain native grasses, and native grasses had much greater cover with Lupinus than without Lupinus (Chapter 2, Figure 2.25). In transplant experiments Festuca idahoensis and Pseudoroegneria spicata grew larger when next to Lupinus than when far from Lupinus, indicating a facilitative effect of Lupinus. However, this facilitative effect occurred only in dense stands of Centaurea (Chapter 2, Figure 2.26). In the absence of Centaurea, the effect of Lupinus was neutral to competitive, suggesting that the facilitative effect may have been indirect. The mechanism for the indirect effect appeared to be mediated by chemicals exuded from the roots of these plants. Centaurea exudes a potent allelopathic compound (±)-catechin (Bais et al. 2003). When Lupinus is exposed to catechin it increases exudation of organic acids, primarily oxalic acid, from its roots. Experiments showed that oxalic acid reduces oxidative damage generated by catechin. When oxalic acid is added to media infused with catechin, the effects of the allelotoxin were reduced. Similar indirect facilitative and interfering interactions may occur among Quercus agrifolia, Pholistima auritum, and annual grasses in Californian woodlands. Parker and Muller (1979) found that P. auritum, a large native herb in the Hydrophyllaceae, sometimes occurred in virtually “pure stands” directly beneath the canopies of some Q. agrifolia individuals, with a strong shift to dominance by annual grasses at the edges of the tree canopies. Why Pholistima is so strongly associated with some Q. agrifolia trees is not known, but the absence of annual grass species in the understory is clearly not due to the direct effects of the oaks. In fact, if Pholistima is not present some of these grass species perform far better under oak canopies than in the open grassland (Parker and Muller 1982, Callaway et al. 1991). Parker and Muller found that litter and leachates from Pholistima were highly inhibitory to understory grasses. In controlled conditions fresh Pholistima litter reduced the germination of Bromus diandrus and Avena fatua from 96% and 93%, respectively, to zero. However, when experiments were conducted with Pholistima litter that had been leached (litter was placed in running
176
Indirect Mechanisms for Facilitation
deionized water for 48 hr) over 92% of seeds germinated in every treatment. Pholistima leachate reduced germination of these species by 32-26%. In field experiments, fresh Pholistima litter reduced B. diandrus germination by 73% and A. fatua by 96%. The abundance of Pholistima under some oaks, and its rarity in the open, indicates that the oaks are facilitating Pholistima. However, by facilitating Pholistima the oaks indirectly inhibit annual grasses. All true epiphytes benefit directly from their hosts. However, interactions among epiphyte species may create linked indirect positive interactions that are somewhat unique. In mixed evergreen-deciduous forests of the southeastern USA, the vascular epiphytes Tillandsia usneoides and Polypodium polypodioides are much more common on some host species than on others Chapter 2, (Callaway et al. 2001b). Furthermore, transplant experiments showed that the growth rates of Tillandsia strands are higher on the host species on which they occur most frequently in nature. Not only do host trees have species-specific direct effects, they also harbor unique communities of non-vascular epiphytes which correlate highly with the relative abundance of the vascular epiphytes. In experiments in which the foliose Parmotrema lichen species were removed from branch segments of Q. virginiana (a preferred host of Tillandsia on which Parmotrema was abundant) growth rates of Tillandsia were 20% less than on branches for which Parmotrema was not removed. Furthermore, Tillandsia seedlings that were watered with extracts from Cryptothecia rubrocincta, a lichen species common on poor Tillandsia host tree species, had lower growth and survival than those watered with extracts from Parmotrema, Pyxine caesiopruinosa, “green algae”, or rainwater. Although far from conclusive, these results raise the interesting possibility that different epiphytic lichen species occurring on different host tree species may indirectly affect, in positive and negative ways, the distribution and abundance of vascular epiphytes.
3.8. CONCLUSION Positive indirect effects within groups of interacting plant species may be one of ecology’s most overlooked phenomena and one that could transform our understanding of the mechanisms that maintain coexistence and diversity. Ecologists have tried for decades to understand coexistence and diversity in the context of direct competition - niche partitioning, variation in particular resource requirements and uptake, shifts in competitive hierarchies in different microenvironments - and in the context of nonequilibrium processes such as herbivory and disturbance, but little attention has been paid to how species interact indirectly with each other. The demands of sorting
Chapter 3
177
through indirect interactions among groups of species are daunting, but the rewards may be great. For example, comparing the performance of Miller’s (1994) species in two-way interactions to the performance of the same species in communities is fascinating. In two-species experiments, Ambrosia reduced the biomass of Chenopodium by 94-98%. In communities, however, the total effect of Ambrosia on Chenopodium was to reduce its biomass by only 17%. Overall, Miller found that the indirect effects from his experiments were larger than predicted by theoretical studies and argued that such strong indirect effects should be the rule rather than the exception in plant communities. The mechanisms by which plants can indirectly facilitate each other appear to be at least as diverse as direct facilitative mechanisms. Although much more difficult to determine experimentally, the wide variety of outcomes that could be produced by different combinations of direct and indirect interactions play important roles in structuring and maintaining diversity in communities. Furthermore, deeper understanding of indirect interactions involving consumers, mutualists, and plant-plant interactions is certain to shed light on exotic invasion and dominance in natural communities as well as attempts to stem invasions with biological controls. As we will see in the next chapter, understanding the balance of positive and negative interactions suggests yet more interesting perspectives on the processes that organize plant communities.
CHAPTER 4 INTERACTION BETWEEN COMPETITION AND FACILITATION
During the 1950’s and 1970’s range managers in California advocated the clearing of Quercus douglasii (blue oak), a widespread Californian endemic, because the tree reduced the production of forage for livestock (Johnson et al. 1959, Murphy and Crampton 1964, Murphy and Berry 1973). Now protected as one of California’s hallmark species, the competitive effect of Q. douglasii caused it to be eliminated or thinned over thousands of hectares of land by hand cutting, bulldozing, and aerial spraying (Leonard 1956, Murphy and Crampton 1964, Murphy and Berry 1973). However, while Q. douglasii trees were being managed in this manner, a graduate student at Fresno State University, and later at the University of California, Berkeley, V.L. Holland, demonstrated that many Q. douglasii trees and stands did not suppress understory productivity, and in fact some canopies were associated with a large increase in the biomass of grasses and other herbs in the understory (Holland 1980). This facilitative effect on productivity beneath canopies was attributed to higher levels of soil nutrients (Holland and Morton 1980), a hypothesis supported in later experiments conducted by others (Kay 1987). This odd variation in the effect of Q. douglasii has been confirmed in other studies. In 1989, McClaran and Bartolome (1989) measured the effect of Quercus douglasii on understory productivity at five sites over two years, finding that Q. douglasii canopy effects were primarily neutral at xeric sites and negative at mesic sites in both years. They found no evidence for positive effects. But Ratliff et al. (1991) found that the effect of Q. douglasii on understory productivity varied with habitat and was often facilitative. In mesic swales oaks had significant negative effects, while in uplands oaks had significant positive effects. Connor and Willoughby (1997) also found that the effect of Q. douglasii on understory productivity tended to vary among years that differed in precipitation, but that the variation was not significant over a five-year period. In 1986 I was collecting data in central California while developing a Ph.D. proposal at the University of California Santa Barbara. At the Hastings Natural History Reservation, which has been protected from livestock grazing for
179
180
Interaction between Competition and Facilitation
decades, I noticed that trees within meters of each other could have exceptionally different understory productivity. Some trees had far more grass beneath them than was growing immediately outside the edges of the canopies, whereas others had virtually bare understories. After quantifying these patterns (Callaway et al. 1991, Figure 4.1), I collected soil from under trees that represented these two extremes and measured nutrient content. However, while collecting soil, it became immediately obvious that digging underneath the trees with little subcanopy biomass was much more difficult than from anywhere else because of the dense network of Quercus roots in the shallow horizons. It turned out that root biomass was ≈5x times greater under trees with low understory biomass (“negative” trees) than under trees with high understory biomass (“positive”) trees. Furthermore, the predawn xylem pressure potentials of “positive” trees were much higher at the end of the long dry summer than those of “negative” trees, indicating that the former utilized the ground table or some other more permanent soil water source (Figure 4.2). Root exclosures reduced the competitive effects of Q. douglasii roots under “negative” trees, but not under “positive” trees (Figure 4.3).
“ negative” trees “ positive” trees
-2
-1
Productivity (g m yr )
400
300
200
100
0
tree
canopy
Figure 4.1. Annual productivity of understory plants in savanna woodland under Quercus douglasii trees with dense surface roots (negative trees), without dense surface roots (positive trees, see Figure 4.4) and in the open grassland matrix surrounding the trees. Data are for 0.125 m2 quadrats sampled along transects aligned due north from the trunks. Error bars are 2 SE. Redrawn from Callaway et al. (1991) with permission from Ecology.
Chapter 4
181
Predawn xylem pressure potential (MPa)
Further experiments suggested that allelopathic mechanisms might contribute to the overall negative effects of the roots (Callaway 1990). However, greenhouse bioassay experiments and cross transplants of soil in the field indicated that both “positive” and “negative” trees produced similar strong positive effects by enriching the nutrient content of the soils beneath them (see Chapter 2). The crucial difference was that the positive effect of “negative” trees on their soil was not manifest because of the negative competitive effects of the dense surface roots. Figure 4.4 illustrates the general idea for the complex way in which Quercus douglasii trees interact with understory species. First, tree canopies deposit many times more total nitrogen, phosphorus, calcium, potassium, and magnesium onto the soil surface than open precipitation (Chapter 2.3), resulting in much higher levels of almost all of these nutrients in the surface soils under both “positive” and negative” trees. These soils have facilitative potential. Second, Q. douglasii trees vary in root architecture, with some tapping the ground table and not developing an extensive surface root network, whereas those that do not tap the ground table develop a much higher biomass of surface roots. The competitive effects of these surface roots override the positive effects of canopy litter and throughfall. These facilitative and competitive mechanisms operate simultaneously, and the balance of both
-0.5
“ negative” trees “ positive” trees
-1.0 -1.5 -2.0 -2.5 -3.0 -3.5 -4.0 -4.5 APRIL
JUNE
AUGUST
OCTOBER
Figure 4.2. Predawn xylem pressure potentials for “positive” and “negative” Quercus douglasii trees in woodland and savanna sites. Error bars show 2 SE. For both sites repeated measures one-way ANOVA, Fhabitat=40.25, df 1,10, P<0.001, Fhabitat x date=16.00, df=6,60, P<0.001. Redrawn from Callaway et al. (1991) with permission from Ecology.
182
Interaction between Competition and Facilitation
interactions determines the overall effect of the overstory tree on its understory. A re-examination of Kay’s (1987) large scale removals of Q. douglasii supports these conclusions (Chapter 2, Figure 2.15). Over a 15-year survey of productivity under live Q. douglasii, open grassland, and patches where the tree had been removed, he found that productivity under Q. douglasii canopies was consistently lower than in surrounding open grassland; suggesting competitive effects. However, removing the oaks did not simply restore productivity to that of open grassland, but removal raised productivity to significantly higher levels in many years, because the understory soils were more fertile. By the end of the experiment the higher productivity at sites where oaks had been killed had disappeared, because the facilitative effect of oaks on soil nutrients had been used up by the greater grass productivity. Further support of the hypothesis that Q. douglasii facilitates understory productivity comes from Ratliff et al. (1991) who found that fertilization improved productivity in the open, but not under Q. douglasii canopies.
Biomass (g)
1.5
“positiv e” tree - “positiv e” soil “positiv e” tree - “neg ativ e” soil “neg ativ e” tree - “positiv e” soil “neg ativ e” tree - “neg ativ e” soil
1.0
0.5
0.0
WITH EXCLOSURE
WITHOUT EXCLOSURE
Figure 4.3. Biomass of individual Bromus diandrus plants grown in soils reciprocally transplanted between ‘positive’ and “negative” Quercus douglaisii trees and in PVC root exclosures and controls without exclosures in savanna woodland. Error bars show 2 SE. Shared letters designate means that were not significantly different in post-ANOVA tukey tests, Twoway ANOVA, Ftree type x exclosure-soil source=6.25, df=7,197, P<0.01. Redrawn from Callaway et al. (1991) with permission from Ecology.
Chapter 4
183
Figure 4.4. Illustration of Quercus douglasii trees varying in root architecture and with positive (right) and negative effects on understory productivity. Drawn by Wendy M. Ridenour.
Understanding these kinds of simultaneous effects of competitive and facilitative mechanisms is prerequisite to understanding the shifting roles of competition and facilitation on abiotic gradients, conditional facilitative effects, and species-specificity in facilitative effects, all of which are important components of this and other chapters. Furthermore, understanding this sort of complexity is fundamental to understanding how interactions among species affect community properties. In this chapter I address such simultaneous effects with an emphasis on abiotic stress. Addressing all of the relevant conceptual aspects of this problem while coordinating all appropriate references in the same sections is very difficult, and therefore I have provided links to other sections and pages as much as possible. Sometime it takes the death of canopy trees to unmask facilitative effects that are operating at the same time as competitive effects (Maron and Connors 1996). Ludwig and colleagues (2004) at Wageningen University studied the complex and interacting effects of Acacia tortilis trees on nutrient, water and light availability and the understory vegetation in eastern Africa. They found that soil nutrient content increased with tree age and size, relative
184
Interaction between Competition and Facilitation
to the open grassland, and was the highest under dead trees. Soil moisture under large Acacia trees was lower than in the open, suggesting a competitive effect of trees on grasses for water. This competition was reflected in a different species composition beneath Acacia trees than in open grassland. There were no differences in above-ground biomass, which was similar under living trees and in the open grassland. However, under dead trees herbaceous production was 60% higher than under living trees indicating that competition for water when trees were alive did not allow the expression of facilitative nutrient enhancement. About the same time that range scientists were beginning to study the effects of Q. douglasii on ecosystem productivity in California, ecologists from the Intermountain Research Station in Utah conducted an experiment on the effects of Populus tremuloides on herbaceous understory productivity in stands ranging from 2440 m to 2730 m in elevation (Ellison and Houston 1958). They established plots in open areas away from the influence of aspen canopies, and then set up trenched and untrenched plots under the aspen canopies. In each of the plots they planted native herbaceous species in monocultures at fixed densities. After three years they harvested the plots. Three of the four species, Bromus carinatus, Elymus glaucus, and Rudbeckia occidentalis, were much smaller under P. tremuloides canopies than in open meadows near the trees, indicating that the trees had competitive effects on these understory species. However, when plots were trenched to exclude the P. tremuloides root systems the biomass of Bromus carinatus and Elymus glaucus exceeded that in the open. The original results are difficult to interpret statistically, but a pair-wise analysis of the data for Bromus carinatus using the means reported for each site in the paper provides a conservative indication that these results were significantly different (Popen vs. untrenched<0.01; Puntrenched vs. trenched<0.01; Figure 4.5). Much like the early work for Q. douglasii, Ellison and Houston interpreted their results in terms of competitive mechanisms – “trenched plots were much more productive that untrenched plots under aspen, which suggests that the principal factor in depressing yields under the aspen is root competition” - even though several of the herbaceous species did not even survive in the open at some of their sites. It is easy to casually observe striking community differences beneath aspen stands throughout the northern and central Rocky Mountains. With hindsight, it appears that strong facilitative and competitive effects were functioning at the same time in their system, with root competition suppressing strong facilitative canopy effects.
Chapter 4
185
Biomass (g/plot)
150
100
50
0
Open
Untrenched
Trenched
Under Populus tremuloides Figure 4.5. Aboveground biomass of Bromus carinatus growing in open meadow surrounding Populus tremuloides, and beneath P. tremuloides either in plots that had been trenched to eliminate P. tremuloides roots or in plots that had not been trenched. Redrawn from Ellison and Houston (1958) with permission from Ecology.
Shifts in the manifestation of positive and negative effects of a canopy species also occur with position on the landscape. Schade et al. (2003) studied the landscape-dependent influences of Prosopis velutina trees (mesquite) on soil moisture, nitrogen availability, and understory vegetation along a topographic gradient in the Sonoran Desert. The effect of Prosopis canopies depended on landscape position, with canopies in more xeric locations having stronger positive effects on soil moisture, biomass, soil N, and species richness than canopies in mesic riparian areas. In fact, canopies in riparian vegetation reduced soil moisture. The xeric habitats were much less productive than the riparian habitat. Such substrate-specific differences in spatial relationships have also been described from the Mojave Desert (Schenk et al. 2003). Similar facilitative processes may occur among Acacia trees and understory plants in the Negev Desert of Israel. Munzbergova and Ward (2002) found that species diversity was higher under three native Acacia species than in the open and that there was a “suite of species with higher occurrence under the trees”. In an interesting twist, understory plant species composition also differed between sites where Acacias had died versus that under living Acacias. While not clearly separating sites effects from tree effects, they suggest that these results may have been due to shifts in the strengths of facilitative and competitive mechanisms.
186
Interaction between Competition and Facilitation
4
Effect size on soil moisture
3
Net effect (Tussock - Open)
4 3
2
2
1
1
0
0
-1
-1
-2
-2
-3
-3
4 3
Microclimatic amelioration (Tussock - No shade)
4 3
2
2
1
1
0
0
-1
-1
-2 -3
2001
2002
J F M A M J J A S O N D J F M A M
-2 -3
Belowground competition (Tussock - Herbicide)
Water inputs from runoff (No run-off - Tussock)
2001
2002
J F M A M J J A S O N D J F M A M
Figure 4.6. Dynamics of the positive, negative, and net effects of Stipa tenacissima on soil moisture at Aguas. Data represent mean effect size (Hedges’ d index) and 95% confidence intervals. Reprinted from Maestre et al. (2003a) with permission from Ecology.
Although a great deal of circumstantial evidence and logic indicates that positive and negative mechanistic operate at the same time, little research has been designed to experimentally separate specific facilitative and competitive mechanisms and evaluate their strengths. In one of the few such studies, Fernando Maestre et al. (2003a) evaluated spatial and temporal variation in the positive, negative, and net effects of Stipa tenacissima on the shrub Pistacia lentiscus in southern Spain (also see Alados et al. 2006). They incorporated three manipulative treatments (removal of belowground Stipa competition, removal of shade from the canopy of Stipa, and preventing runoff water collecting at Stipa tussocks) into an elegant experimental layout with Pistacia planted upslope and adjacent to isolated Stipa tussocks or in the open in the intertussock areas and leaving the dead canopies intact. They removed shade but left strong root competition by bending Stipa canopies away from target Pistacia, and diverted runoff by inserting metal sheets in the ground upslope of experimentally manipulated Stipa. The elimination of Stipa canopy shade and root competition significantly decreased and increased, respectively, Pistacia seedling performance as compared to the tussock treatment. The net effect of Stipa on soil moisture and Pistacia was always facilitative (Figure 4.6),
Chapter 4
187
and the magnitude of the net effect increased with the harshness of the environmental conditions in both space and time (Figures 4.7 and 4.8). The survival rate of Pistacia ranged from 0-36% in at the dry experimental site (Ballestera) and 32-96% at the more mesic experimental site (Aguas). Seedlings with shade, but no competition survived at a far higher rate than those in any other treatment (Figure 4.7). Interestingly, the net effects was not determined by the weakening in intensity of any effect - the magnitude of the competitive effect, the facilitative effect, and the net effect all increased over the summer but differently, indicating that these mechanisms responded differently and unequally to environmental variation, but balanced to establish a net effect that was calibrated to a specific place or time (Figure 4.8). 100
Aguas 80
60
Survival rate (%)
40
20
Tussock No run-off No shade Herbicide Open
a a b c ab
0
2001
2002
100
Ballestera 80
60
40
20
Tussock No run-off No shade Herbicide Open
a a b c ab
0 Ja Fe Ma Ap My Ju Jl Au Se Oc No De Ja Fe Ma Ap My
Figure 4.7. Survival of Pistacia lentiscus seedlings planted in different treatments. Shared letters after the key to symbols indicate no significant differences between treatments (log-rank test, P . 0.05). Reprinted from Maestre et al. (2003a) with permission from Ecology.
188
Interaction between Competition and Facilitation 1.0
Aguas
0.5 0.0
Effect size on survival
-0.5 -1.0 -1.5 -2.0 1.0
Ballestera
0.5 0.0 -0.5 -1.0 Microclimatic amelioration Water inputs from run-off Belowground competition
-1.5 -2.0 Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Figure 4.8. Dynamics of the positive and negative effects of Stipa tenacissima on Pistacia lentiscus survival. The effects of microclimatic amelioration, water inputs from runoff, and belowground competition are estimated as (SNS 2 ST)/ST, (SNR 2 ST)/ST, and (SH 2 ST)/ST, respectively, where ST, SNS, SNR, and SH are the survival rates for the tussock, no-shade, norunoff, and herbicide treatments, respectively. The sign for the effects of microclimatic amelioration and belowground competition has been changed for intuitive interpretation. Reprinted from Maestre et al. (2003a) with permission from Ecology.
In an experiment explicitly designed to tease apart the facilitative effects of canopies and the competitive effects of roots, I manipulated the canopies and root systems of potential facilitators (Callaway 1994) in a western salt marsh. The upper zones of many California salt marshes become extremely saline during the summer and fall when tides are low and salt concentrations increase with the seasonal drought. In the winter, rains wash salts from the soils and produce a brief, but spectacular flush of ephemeral species. Very few perennial species are able to tolerate the hypersaline conditions of the summer, with one of the exceptions including Arthrocnemum subterminale, a small shrub in the
Chapter 4
189
Chenopodiaceae. Arthrocnemum decreases soil salinity under its canopy by ≈30% and PAR by 65%, and increases soil moisture by 10-15% (also see Espinar et al. 2002). The distributions of several species of winter ephemerals are affected by Arthrocnemum, but some in positive ways and others in negative ways. For example, Hutchinsia procumbens and Parapholis incurva were found primarily under Arthrocnemum canopies and not in the open, whereas Spergularia marina was found primarily in open spaces. Emergence of the three ephemerals was not affected by root competition or canopy facilitation from Arthrocnemum. However, survival of Hutchinsia was 4-7 times higher under Arthrocnemum canopies, with no effect of root exclosures (Figure 4.9). Survival of Parapholis was two times greater under canopies when roots were not excluded, but four times greater when roots were excluded. In contrast to the Hutchinsia and Parapholis, survival of Spergularia doubled when canopies were removed regardless of whether or not roots were excluded. The response of ephemeral biomass to canopy removal treatments was similar to that of survival. For Hutchinsia, which was spatially associated with Arthrocnemum and survived at much higher rates under the perennial, root exclusion had no effect on biomass, but canopy removal reduced biomass by more than 80% (Figure 4.10). In contrast to the effects of Arthrocnemum roots on survival, root competition had stronger effects on biomass under intact canopies and weaker effects when canopies were removed. canopy & roots
80
Survival (%)
a
canopy & no roots
no canopy & no roots
a
b
b
60
a
40
20
no canopy & roots
a
a
a
b
ab
b
b
0
Hutchinsia
Spergularia
Parapholis
Figure 4.9. Survival of three upper salt marsh annuals in four experimental treatments manipulating the canopies and roots of the dominant perennial, Arthrocnemum subterminale. Error bars represent 2SE and bars that share a letter within a species are not significantly different. Redrawn from Callaway (1994) with permission from Ecology.
190
Interaction between Competition and Facilitation 180 160
canopy & roots
a
canopy & no roots
no canopy & roots no canopy & no roots
a
Biomass (%)
140 120 100 80 a
60
a
40 20
b
b
a
a
a a
b
ab
0
Hutchinsia
Spergularia
Parapholis
Figure 4.10. Biomass of three upper salt marsh annuals in four experimental treatments manipulating the canopies and roots of the dominant perennial, Arthrocnemum subterminale. Error bars represent 2SE and bars that share a letter within a species are not significantly different. Redrawn from Callaway (1994) with permission from Ecology.
In sum, facilitative and competitive mechanisms operated simultaneously in the highly stressful saline conditions of the upper salt marsh. Facilitation by Arthrocnemum was crucial to the survival and growth of other species in the community. Competition occurred, but was weak relative to facilitation, and like is often the case in deserts, net effects were positive (Franco and Nobel 1989, Holzapfel and Mahall 1999). Weak root competition, which was evident only for Parapholis, may have been due to minimal root overlap between the ephemerals which typically rooted in the upper 10 cm and Arthrocnemum which rooted much deeper and rarely near the soil surface. Root interference also may have been minimized by low phenological overlap, much like in Quercus douglasii-grass understory systems. Arthrocnemum does not begin to develop green photosynthetic tissues until the late spring or early summer, generally after the ephemerals have produced seeds and died. Studies by Jose Facelli and Amanda Temby (2002) in southern Australia have also demonstrated a complex array of facilitative and competitive mechanisms working simultaneously in interactions between shrubs and the annual plants that grow around them. Shrubs altered soils, the seed bank, the effects of large vertebrates, and the subcanopy microclimate. But in a very interesting contrast to many other studies in similar systems, canopy effects were negative, reducing the success of annuals, whereas the
Chapter 4
191
roots of one shrub species, Atriplex vesicaria, facilitated annual species growing on the edges of its canopy (Figure 4.11). Trenching around Atriplex shrubs decreased the abundance of annual plants. Facelli and Temby attributed the positive effects of Atriplex shrubs to hydraulic lift (see Chapter 2.1), and the decrease in the productivity of nearby annuals after root severing to impeding the flow of water from the roots into the shallow soils. Both positive and negative effects occurred in this system, but the typical effects of mechanisms were reversed, with canopies inhibiting and roots facilitating understories. One of the best examples of interacting simultaneous competitive and facilitative mechanisms is that for Lupinus arboreus (bush lupines) and invading exotic annual grasses on the northern coast of California. The nitrogen-fixing Lupinus adds large amounts of nitrogen to the soil resulting in 60
A
p < 0.01
50
50
B
p < 0.05
40
p = 0.053
40
p < 0.01
30
plants/m 2
plants/m
2
p < 0.01 30
p < 0.01
20
20 10 10 0 0
18
removed
C
shrub
open
removed
trenched
p < 0.05
30
D
shrub
open
trenched
NS p < 0.05
15
2
plants/m
NS
9
p < 0.01
6
plants/m 2
20
12
NS
10
3 0 0 removed
shrub
open
trenched
removed
shrub
open
trench
Figure 4.11. Abundance of annual plants in plots with Atriplex vesicaria removed, shrub present, unmodified open spaces, or trenched open spaces. (a) total numbers; (b) Carrichtera annua; (c) Tetragonia tetragonioides; and (d) other species. (a) Error bars represent 1 SD. (b-c) Points represent individual replicates. Probability values from planned orthogonal contrasts shown (Dunn tests except for (a), Tukey test). Reprinted from Facelli and Temby (2002) with permission from Austral Ecology.
192
Interaction between Competition and Facilitation
total nitrogen pools over twice that of surrounding lupine-free grassland (Maron and Jefferies 1999). However, the facilitative benefit of this nitrogen amendment is not available to colonizing annual grasses because of the competitive effect of dense shade from the lupine’s canopy. Lupines reduced available light from 1725 μmol m-2s -1 to 13 μmol m-2s-1. The importance of interacting facilitative and competitive mechanisms became clear after intense insect herbivory killed the lupines (also see del Moral and Bliss 1993). Lolium multiflorum, Bromus diandrus, and other annual grasses, rapidly took up 5070% of the available nitrogen and rapidly colonized the Lupine-enriched soil that was previously off limits due to light competition. In sum, perceiving facilitation in the context of other interactions, primarily competition, is a fundamental prerequisite for understanding how interactions among plant species affect natural community organization.
4.1. COMPETITION, FACILITATION AND ABIOTIC STRESS The case studies from Quercus douglasii savannas and subalpine Populus tremuloides forests provide good examples of how very different mechanisms interact to control the way that a plant species affects its neighbors. However, synthesizing these kinds of examples into conceptual theory is crucial if we are to predict when and where different facilitative and competitive mechanisms will predominate – but a consistent synthesis has been elusive. This section focuses on the idea that facilitative effects tend to be stronger, relatively speaking, in abiotically stressful, low productivity environments; whereas competition predominates when conditions are relatively benign. The most widely utilized conceptual model is that proposed by Bertness and Callaway in 1994. In part derived from Grime’s hypotheses about the relative importance of competition in plant communities, and in part from extensive studies by Mark Bertness and colleagues along gradients of salt stress in marshes, we postulated that competitive interactions would be most important to the structure and composition of plant communities when abiotic stress does not strongly limit the ability of plants to acquire and exploit resources. Such non-stressful conditions result in high productivity, but only when consumer pressure is low to moderate (Figure 4.12). In contrast, we suggested that facilitative interactions would become more important when abiotic stress is high or when consumer pressure is intense. This very basic conceptual model has been supported by mathematical modeling (Travis et al. 2004), but the logic behind the idea is based on the fundamental nature of competition, which is by definition the struggle to preempt limiting resources
193
Importance of facilitation
Importance of competition
Chapter 4
Increasing abiotic stress Increasing consumer pressure Figure 4.12. Conceptual model for variation in the importance of competition and facilitation in communities along gradients of abiotic stress and herbivory. Redrawn from Bertness and Callaway (1994) with permission from Trends in Ecology and Evolution.
such as light, water, and nutrients that determine rates of carbon acquisition. Under relatively benign abiotic conditions that permit rapid resource acquisition competition can be intense. However, if severe physical conditions restrict resource acquisition, amelioration of severe stress by a neighbor may be more likely to favor growth than competition with the same neighbor is to reduce growth (Bertness 1991b). The Quercus douglasii-dominated ecosystems described earlier in the chapter provide a good example of the abiotic stress hypothesis. Ratliff et al. (1991) measured herbaceous productivity under various tree species and in open grassland for eight years and over many different habitats. Analysis of their results supports the general abiotic stress hypothesis for the relationship between productivity and the relative importance of facilitation and competition. Over all times and sites the overall effect of Q. douglasii was positive; however in mesic swales, where annual productivity was by far the highest, the effect of Q. douglasii was negative. Regressing the relative effect of the tree against mean open habitat productivity suggests that facilitation was stronger where productivity was lower (more xeric upland habitats) and competition was stronger where productivity is high (Figure 4.13).
194
Interaction between Competition and Facilitation
Proportional effect of Quercus douglasii
60
40
20
Facilitation 0
Competition -20 2000
3000
4000
5000
Open herbaceous production (kg/ha) Figure 4.13. The relationship between productivity in open grassland sites in central California grassland and the proportional effect of the savanna canopy tree, Quercus douglasii. Calculated from data in Ratliff et al. (1991).
Rob Brooker and Terry Callaghan (1998) also examined the relationship between stress, environmental gradients, and the balance of positive and negative interactions. Because of the unresolved debate over the relationship between the role of competition and stress (Grime 1979, Taylor et al. 1990) they concentrated their discussion on the role of disturbance in moderating the balance between facilitation and competition. Based on their research on interactions among arctic plants and reviewing evidence in the literature for facilitative plant-plant interactions they developed a conceptual model that was very similar to that proposed by Bertness and Callaway (1994). They suggested that variation in environmental severity (nominally taken as disturbance), through either space or time, might influence the observable outcome of interactions between plants by shifting the balance between competitive and facilitative interactions. This may mean shifts at either end of stress gradient. They also proposed that, because of the dominant role played by abiotic interactions in high altitude or high latitude environments, these sites might prove (as for deserts) excellent testing grounds for the development of theory on plant interactions.
Chapter 4
195
The large majority of empirical studies of facilitation and competition on stress-productivity gradients in the literature support the general stress model proposed by Bertness and Callaway. However, there are many exceptions and results reported in the literature are by no means unambiguous. Attempts to articulate fundamental organizing concepts for positive interactions in plant communities are hampered by the same multivariate, system-inherent, methodological and semantic problems that plague the rest of ecology. Conceptual models for interaction-productivity relationships do not fit all empirical studies for a number of reasons. First, despite recent attempts to force “functional” categories onto species, the idiosyncratic characteristics of species resist simple functional classification. Second, the environments in which species interact differ in many different ways. Even though we can quantify similarities in the timing and amount of precipitation and energy load in Mediterranean-climate deserts such as the Negev and the Mojave, we cannot easily quantify how these rough similarities interact with differences in geology, topography, ecological history, and local fauna. Third, definitions of “stress” for organisms that have adapted physiologically and morphologically to different environments are inherently problematic (Lortie et al. 2004). An alpine herb may not be particularly stressed at high elevations in the Rocky Mountains (Körner 1999), but may be in serious trouble in the Great Basin Desert. If we insist on defining stress relative to the adaptations of particular species we should probably drop any hopes of getting anywhere on the issue of how stress affects plant communities (see the debate between Lortie et al. 2004 and Körner 2004). However, even within this contextual fog, there is some hope to express in a practical way how stressful a particular environment is, as argued skillfully by Grime (1977), by defining it in terms of productivity. While not providing perfect calibration for variation in stress intensity in different parts of the world, this approach defines stressful environments as those in which the local producers (their composition admittedly limited in global generality by species pool limitations) are constrained in their ability to convert energy to biomass.
4.1.1. Stress gradients and the importance of productivity Community ecologists should attempt to examine stress in the context of productivity. Biomass is a decent alternative to productivity (see Chapter 6.6), but measurements of particular abiotic variables or resources as a proxy for stress are not. This is because the chosen abiotic variable may misrepresent the stress gradient (see pages 215-219 in this chapter) because the chosen
196
Interaction between Competition and Facilitation
variable may not be the fundamental driver of stress or the target taxa may be adapted to low level of the particular stress variable (see Lortie et al. 2004 and Körner 2004). The fourth cause of idiosyncrasy in studies of stress and plant interactions is the fact that gradients in studies of competition and facilitation almost always consist of just two points, and these two points may be quite similar or poles apart. For example sites may differ in precipitation, temperature, or productivity by a factor of 0.5 or 5.0, but in a particular study these end points are presented as “stress” and “no stress”. As we will see in a few pages (Figure 4.15, and discussion there), such variation in the endpoints of stress gradients creates serious problems for meta-analyses which attempt to combine them. Fifth, as discussed by Deborah Goldberg et al. (1999), most relevant studies in the literature are inconsistent in either experimental design or the metric used for the intensity of the interaction. For example, neighbors may be defined as particular species sought for a treatment or a random pick from the plant communities in an area. Indices used to quantify interaction intensity also vary among studies and may yield different relationships along gradients. For example, Suding and Goldberg (1999) found that significant plant interaction-productivity relationships were more likely to be found when the log response ratio (ln[removal/control] was used to quantify interactions than when relative competition index ([removal-control]/removal) was used (also see Brooker et al. 2005). In different studies targets may be transplants or naturally occurring individuals, and may be measured for total biomass, leaf length, or flower number. However, despite the lack of methodological standardization and consistent results in the literature for studying conditionality of competition and facilitation, the general trends are quite easy to observe.
4.1.2. Stress gradients and meta-analysis Considered separately, most studies that have incorporated some aspect of the abiotic stress-gradient hypothesis (only one part of the Bertness and Callaway model) have found that competitive effects are stronger and facilitative effects are weaker where productivity or biomass is high, and that competitive effects are weaker and facilitative effects are stronger where productivity or biomass is low. Many of these studies are discussed in detail below. However, in an early meta-analysis of a number of studies Katherine Suding and Deborah Goldberg (1999) found the opposite: competition intensity often significantly declined with productivity, and facilitation was often found in the
Chapter 4
197
more productive sites (Figure 4.14, also see Gurevitch et al. 1992). Interestingly, these results also conflict with a rich and much more extensive body of literature on competition (see Michalet et al. 2006). However, these relationships depended on the response variable used. For example, competitive neighbor effects on the final biomass and survival of target plants decreased with decreasing standing biomass (a surrogate for productivity) - which corresponds with the Grime (1977) and Bertness and Callaway (1994) models but neighbor effects on growth rate did not. Facilitative interactions were more common when the standing biomass of a community was low if targets were being measured for final biomass or growth rate (again corresponding with Grime and Bertness and Callaway), but more common in communities with high standing biomass if survival was the metric. Despite the caveats (also see the next paragraph), Suding and Goldberg’s meta-analyses did not demonstrate significant positive relationships between competition and biomass and frequently found the opposite. Importantly, this differs from a more qualitative, “vote counting” interpretation of the literature. Based on my admittedly subjective and highly biased interpretation of the literature incorporated into this book, 167 papers report results that are relevant to the abiotic component of the stress hypothesis (Bertness and Callaway 1994) in some way, and 131 of these indicate that facilitation is more important or more intense in stressful conditions than in benign conditions.
Emergence response to neighbor removal (%)
400
300
200
100
0
-100 0
50
100
150
200 2
Average vegetation biomass (g/m ) Figure 4.14. The relationship between community production and neighbor effect in six different communities. Biomass was taken from Table 1 in the paper and the emergence response was calculated as the emergence with neighbors divided by the increased measured when neighbors were removed, as estimated from Figure 7. All responses above zero indicate facilitation. Redrawn from Suding and Goldberg (1999) with permission from the Journal of Ecology.
198
Interaction between Competition and Facilitation
One of the more weighty caveats to consider when interpreting Suding and Goldberg’s results is how the criteria for choosing studies for the metaanalysis may have biased the results. To conduct a meta-analysis, they were faced with the serious problem of finding studies that were reasonably homogenous in methodology. Their conditions were that the chosen studies had to 1) include data on standing biomass, 2) experiments must be in the field, 3) the studies must provide a quantitative measure of, or the potential to calculate, relative competition index or log response ratio, and 4) manipulation had to be the removal of above and below ground parts of naturally occurring neighbors around targets (or studies in which aboveground removal was thought to substantially reduce belowground effects). To find studies with these qualifications they used studies from four earlier reviews on competition. Although these choices may have satisfied the requirements for a technically sound meta-analysis, they excluded hundreds of studies that actually reported finding facilitation and dozens that explicitly focused on competition, facilitation, and abiotic stress. The exclusion of important facilitative studies can be attributed to how much facilitation research fails to meet the sophisticated analytical approaches required for the meta-analysis. For example, most studies of facilitation examine only one or a few species and much too often these species are restricted to those that show obvious spatial relationships with other species that suggest facilitation. If the studies included in Suding and Goldberg’s meta-analysis were conducted in such a way that allowed either facilitation or competition to be expressed, which they appeared to be, why would they not provide a thorough look at the relative importance of facilitation and competition in communities? First, all studies incorporated into the meta-analysis were conducted on herbaceous species in herbaceous communities. Facilitation has been reported in herbaceous communities (Chapter 2), but most detailed examples of facilitation involve large benefactor species modifying the environment for smaller or morphologically dissimilar species. Other potential problems involve the use of interaction indices; neither the relative competition index (RCI) nor the log response ratio (LRR) are symmetrical around 0, and RCI might theoretically vary from - infinity to +1. Although the problems inherent to RCI and LRR are unlikely to alter the meta-analyses in any qualitative way, they are worth considering for quantitative accuracy. Suding and Goldberg’s meta-analysis brought new ways to address the stress-gradient hypothesis, but a more recent meta-analysis demonstrates how this approach can be problematic. Maestre et al. (2005a) used meta-analyses to evaluate studies of facilitative and competitive plant interactions within semiarid environments and concluded that “the magnitude of the net effect provided by neighbors, either positive or negative, is not higher under high abiotic stress
Chapter 4
199
conditions”. However, careful examination of the studies chosen for the analysis and the assumptions about the nature of the gradient revealed problems (Lortie and Callaway 2006, Maestre et al. 2006b). First, Maestre et al. (2005) used a random statistical model in their analysis, which is certainly appropriate but no more appropriate than a fixed model. In a reanalysis of the total literature data set Chris Lortie and I found that a fixed effect analysis of survival yielded a significant difference between the two levels of stress, supporting the stress gradient hypothesis if the model is limited to analyzing extant data (fixed model) rather than using extant data for global predictions (random model). More importantly, many studies included in the meta-analyses clearly were not conducted along stress gradients (but simply reported results from two different sites), did not identify a stress gradient within the study, or focused on highly competitive non-native invasive species. Another crucial issue with using meta-analysis to examine gradientbased data is that gradient length and gradient steepness among studies can vary dramatically. This crucial source of variation is never accounted for statistically or in interpretations. Chris Lortie and I hypothesized that the mixing of extremely variable gradient length and steepness (resulting in highly variable effect sizes among studies) in the same analysis would be highly unlikely to demonstrate gradient effects even in technically appropriate metaanalyses. Our concern was that the very nature of forcing together of experiments conducted at different points along different stress gradients (one person’s low might be much higher than another person’s high for example) would yield insignificant results simply due to the variation in effect size. Figure 4.15 is a cartoon that illustrates this concern. In Figure 4.15 we have displayed four hypothetical “empirical” studies, represented along each of four different gradients. An arbitrarily chosen effect size is above each bar. In this cartoon I am assuming (pretending actually) that each of four examples represents a study in which high and low sites showed significant differences along the lines of the Bertness and Callaway stress gradient model using traditional statistics. Each individual “empirical” study shows highly statistically significant effects of the stress gradient in the direction suggested by Bertness and Callaway (1994). As drawn, the high stress points of different “empirical” studies are often below the low stress points of the other “empirical” studies, something that undoubtedly occurs repeatedly when combining ecological studies from the literature. Therefore, combining the two studies into a “meta-analysis” yields no significance. This cartoon actually underemphasizes the problem because I compared the different effect sizes with t-tests rather than a random model meta-analysis. It is crucial to note that this lack of significance is not at all what the actual biological studies demonstrated;
200
Interaction between Competition and Facilitation
they each show significant differences between competition at the low stress end of the gradient and facilitation at the high-stress end of the gradient. Metaanalysis applied to studies in which the end points are highly variable can lead to conclusions that are directly at odds with the original studies. If one were to add in a study or two for which there is no true gradient, there is little chance for a random model meta-analysis to arrive at a significant difference. In other words, excessive variation derived from combining ecological studies into meta-analyses makes non-significant results (rejection of a hypothesis) highly questionable. On the other hand, significant differences in such meta-analyses should be reliable as they have overcome this excessive variation. To test the idea that combining studies conducted at different points along gradients can be misleading, Chris Lortie and I also conducted an analysis on what we considered to be the 10 strongest and statistically significant studies in the literature that clearly showed facilitation at high stress sites and competition at low stress sites. Even this meta-analysis did not support +10.0 +0.5
Effect size
-0.5
-1.0 +40.0
+0.5
-5.0 -40.0
Figure 4.15. Illustration of how combining effect sizes from four different studies of gradient length and steepness can lead to a false rejection of significant effects occurring in each individual study. Bars represent effect sizes for different high and low sites and numbers above bars are the arbitrarily chosen numbers that were combined for high and low sites. The four low site numbers (-11.6±9.5, 1 SE) and four high site numbers (12.8±9.4) were compared with a t-test yielding a p value of 0.117).
Chapter 4
201
the stress gradient hypothesis when tested with a random model, empirically confirming that mixing studies with different gradient lengths and different species renders a simple meta-analysis functionally unable to test this hypothesis. Strong support for the stress gradient hypothesis comes from a study conducted in semi-arid Spain, interestingly enough tested with meta-analysis. The results from this study were not a part of Maestre et al.’s meta-analysis. Lorena Gómez-Aparicio and colleagues at the University of Granada in Spain conducted the largest-scale study yet of facilitation and competition in semiarid environments using 18,000 replicates of 11 different potential beneficiary species with 18 different species of potential nurse shrubs (Gómez-Aparicio et al. 2004, Chapter 5, Figures 5.6 & 5.7). Additionally, the work was carried out over four years at many different sites. These results convincingly demonstrated that “pioneer shrubs facilitate the establishment of woody, late-successional Mediterranean species” and that “nurse shrubs had a stronger facilitative effect on seedling survival and growth at low altitudes and sunny, drier slopes than at high altitudes or shady, wetter slopes. Facilitation in the dry years also proved higher than in the one wet year. Interestingly, meta-analysis was used to compare sites and despite its highly conservative nature (see Figure 4.15, and discussion) still demonstrated significant gradient effects. Gómez-Aparicio and colleague’s results demonstrating stronger facilitation in higher stress conditions in the semi-arid shrublands of Spain provide powerful generality for the stress gradient hypothesis, and recent research in other semi-arid parts of the world provides similar support (Kitzberger et al. 2000, Ibáñez and Schupp 2001, Pugnaire and Luque 2001, Tewksbury and Lloyd 2001, Maestre et al. 2003, Peek and Forseth 2003, Pugnaire et al. 2004, Lloret et al. 2005, Riginos et al. 2005, Padilla and Pugnaire 2006, Sthultz et al. 2006). There are many other studies that have incorporated either large numbers of species or sites into the experimental design, and show support for the idea that facilitative interactions become stronger or more common in communities as abiotic stress increases (see Brooker et al. 2005). In one of the studies incorporated into Goldberg et al.’s (1999) meta-analyses, Scott Wilson and Paul Keddy (1986) transplanted ramets of three different species along a gradient of productivity represented by eight sites on the shoreline of Lake Ontario, Canada, in order to study variation in the intensity of diffuse competition. The productivity gradient was related to differences in effects of wave action as distance from the lake increases. As hypothesized, diffuse competition was much stronger where the standing crop of natural vegetation was higher, but at three of the eight sites the index of diffuse competition was negative, indicating that the transplants were facilitated by their neighbors (Figure 4.16).
202
Interaction between Competition and Facilitation Arcsine diffuse competition
25 20 15 10 5 0 -5 -10 -15 0
2
4
6
8
10
12
14
16
18
Standing crop (g per 0.04 m ) -2
Figure 4.16. The relationship between diffuse competition and standing crop biomass on the shoreline of Axe Lake, Ontario, Canada. Reprinted from Wilson and Keddy (1986) with permission from The American Naturalist.
Philippe Choler and colleagues also used large numbers of species in experiments on stress gradient relations in alpine vegetation in the French Alps (Choler et al. 2001). They explicitly investigated shifts in competition and facilitation using along altitudinal abiotic severity gradients and conducting a simple neighbor removal experiment using 6 sites and 5 species per site. Sites were chosen so that they represented three different elevations ranging from ≈2,000 to 3,000 m in elevation and either concave, mesic topography or convex, xeric topography. Elevation and topography are considered to be the two main ecological gradients determining the composition of these communities, and the experimental sites encompassed most of the floristic diversity observed in alpine plant communities in the western Alps. Over all species at all sites, there was a strong shift from a negative effect of neighbors (competition) at low elevations and sheltered, concave microsites to positive effects of neighbors at high elevations and exposed, convex microsites (Figure 4.17), supporting the basic conceptual model proposed by Bertness and Callaway (1994). See Olofsson (2004) for similar shifts in competition and facilitation in sheltered and exposed alpine microsites. At the low elevation and sheltered (concave topography) site all five experimental species produced more leaves and more total biomass when neighbors were removed than when neighbors were not disturbed. At the intermediate elevation but still on sheltered topography, the competitive effect of neighbors was much weaker, and for one target species, Festuca violaca, the removal of neighbors resulted in
Chapter 4
203
lower leaf growth; evidence for facilitation. At the high elevation-sheltered site (2900 m), most neighbor effects tended towards facilitation, but only one species showed a significant effect of neighbors, Poa alpina, and that effect was positive. Facilitation was by far the most evident at the high elevation site on exposed topography. 2
Low & Sheltered
Low & Exposed
1
*** NS
0 -1 ***
***
A al gr pi os na t i (1 s se 0) m C pe a r rv e x ir e ns H (1 nu elia 0) m nt m he ul m ar u iu m m co L (9 ) r n ot ic us ul at us P (1 0) v i ol y vi g pa on r u um m (1 2)
A vu lch lg em ar i se C is ( lla 11 m ar pe e ) rv x ire ns (1 m 1) on Ge t a um nu m (1 co 2) r n Lo ic tu ul s at us (1 P 1) vi oly vi g pa on ru um m (1 0)
***
Intermediate & Exposed
Intermediate & Sheltered
**
**
1
NS
NS
*
NS
**
0 NS NS
-1
NS
2 1
High & Sheltered
High & Exposed *
*
**
A hy l c h br e m id ill a a se (7 m C pe a r ) rv e x ir e ns Le (1 u 2) ad can us th tu e m m um ( co L 6) rn ot ic us ul at us (6 or Phy ) bi t e cu u la ma re (1 0)
-2 A vi nd ta os lia a na c e (1 2) Fe vi s ol t u ac ca ea (1 m G 1) on e u ta m nu m (1 2) V al ero lio n ni ica i( 11 A ) vu lch lg e m ar il is la (1 2)
Log response ratio (final biomass)
*
**
-2
2
NS
**
***
NS
NS
** **
***
NS NS
0 -1
he Ox lv ytr et op ic is a (1 0) Pe py tr re oc na all ic is a pe Ph (8 ) de y t m eu on m ta a na (7 ) al Po pi a na (9 ) P vi olyg vi pa on r u um m (7 )
F vi est ol uc ac a ea (8 m G ) on e u ta m nu m (1 0) al Po pi a na (1 0) R ku anu ep nc f e ul r i us ( pr Sib 10) oc b um ald be ia ns (5 )
-2
Figure 4.17. Mean log response ratio (LRR) of each species in each site (sample sizes are given in parentheses). Positive values reflect facilitation; negative values reflect competition. Error bars represent 1 SE. P values above bars indicate the probability of LRR 5 0 in a one-sample t test: *P =0.05; **P=0.01; ***P=0.001; NS = not significant at P=0.05. Reprinted from Choler et al. 2001, with permission from Ecology.
204
Interaction between Competition and Facilitation
CA 2nd-axis (Elevation)
CA 2nd-axis (Elevation)
2
Carex sempervirens
Lotus corniculatus
1 F*
F**
0 -1
C*
NS
-2
C***
C***
2
Polygonum viviparum
F**
Geum montanum
1 F**
0 -1
NS F***
-2
C***
C***
-2
-1 0 1 2 CA 1st-axis (Mesotopography)
3-2
-1 0 1 2 3 CA 1st-axis (Mesotopography)
Figure 4.18. Dual graphical display of species and relevés after reciprocal scaling of CA scores. In each graph, the distributional mean (61 SE) of one target species (open square) and the distribution mean (±1 SE) of the experimental sites (solid circles) in which the species was manipulated are shown. Response to neighbor removal is indicated near the site in which the species was manipulated. C indicates competition, and F indicates facilitation: *P=0.05; **P=0.01; ***P=0.001; NS = not significant at P=0.05. Reprinted from Choler et al. 2001 with permission from Ecology.
Choler and colleagues also used ordination analyses (reciprocal scaling) to quantify the positions of several species that occurred broadly along their elevational and topographical gradients in order to provide a large-scale spatial and community context for the experiments. By merging ordination and experimental analyses they were able to: 1) quantify the relationship between abundance of neighbors and the strength of their effects, and 2) examine the effects of neighbors on particular species and at particular sites at different points along the natural distribution of species on abiotic gradients. Significant responses to neighbor removal were related to the position of particular species along the elevational gradient (Figure 4.18). In other words, when removal experiments were conducted on species at elevations above their distributional optimum, the loss of neighbors was detrimental to the biomass of the target
Chapter 4
205
plant, indicating facilitation. Conversely, when neighbors were removed from around a species occurring below its distributional optimum, the loss of neighbors was beneficial to that plant, indicating competition. Because the abundance of a species is probably related to its success in a particular environment, they also tested relationships between the abundances of target species and the direction and strength of their responses to neighbor removal at the same site. On exposed topography, the effect of removing neighbors was more detrimental for uncommon species than common species. In summary, Choler et al. (2001) found that competitive effects of neighbors were much more prominent at lower elevations where temperatures were higher and wind speeds probably lower, where soils were more developed, and biomass was higher – and likely where abiotic stress was lower (see Grime 1977). However, the effects of neighbors on particular species were related to the location of the experimental site relative to the distributional optimum of the species. In experiments at sites higher in elevation than the distributional optimum of a species, the effect of neighbors was usually facilitative. In experiments at sites lower than the distributional optimum of a species the effect was usually competitive. With an international group of part-time ecologists, the Alpine Pals, I expanded Choler and colleague’s experiments with a series of similar experiments conducted in subalpine and alpine plant communities with 115 species in 11 different mountain ranges (Callaway et al. 2002). We found that competition generally, but not exclusively, dominated interactions at lower elevations where productivity was higher and abiotic conditions are less physically stressful. In contrast, at high elevations where abiotic stress is high the interactions among plants were predominantly positive (Figure 4.19). For nine locations where plants were harvested, Relative Neighbor Effect (RNE) for biomass was -0.20±0.05 (1 s.e.) at the low sites versus +0.27±0.05 at the high sites, with a positive RNE indicating a positive effect of neighbors. For the 10 locations where leaf growth rates were measured, RNE for total leaf growth was -0.33±0.04 at the low sites and +0.16±0.05 at the high sites. At 8 of the 11 locations, removal of neighbors had significantly different, and more facilitative, effects on aboveground target plant biomass at high elevations than at low elevations. At the other three locations RNE at the high site was significantly greater than zero, indicating facilitation, but because RNE was also positive at the low sites there was no difference between sites. Two of these three sites were in highly stressful arctic environments (the Kluane Mts. in western Canada and Abisko, Sweden), and the third was in the Sierra Nevada of Spain where both high and low sites were dry (see Arroyo et al. 2003,
206
Interaction between Competition and Facilitation
Cavieres et al. 2005). No significant effect of competition on biomass was found at any of the high elevation sites.
-0.5
-1.0
-1.0
LOW 1.0
0.5
HIGH
P=0.354
0.5
*
0.0
*
0.0 -0.5
*
-0.5
LOW
HIGH
P<0.001
1.0
-1.0
-1.0
LOW
LOW
HIGH
HIGH
P<0.001
0.5
RNE
*
0.0
-0.5
1.0
1.0
RNE
RNE
0.0
RNE
*
P=0.991
0.5
LOW
RNE 1.0 0.5
LOW
HIGH
RNE
-1.0
-1.0
P<0.001
*
1.0
0.0
1.0
-0.5
0.5
**
-1.0
LOW
** LOW
HIGH
P=0.029
0.0 -0.5
**
0.0 -0.5
*
-1.0
-1.0
LOW
HIGH
P=0.405
0.5
RNE
** **
0.0 -0.5
0.0 -0.5
HIGH
P<0.001
0.5
HIGH
P<0.001
0.5
RNE
1.0
** LOW
1.0
**
-1.0
RNE
-1.0
**
0.0 -0.5
**
0.0 -0.5
P<0.001
0.5
RNE
P=0.009
0.5
RNE
1.0
1.0
HIGH
LOW
HIGH
Figure 4.19. Relative Neighbor Effect (RNE) at the 11 experimental sites. Error bars represent one standard error, P values denote significance of differences between the two sites (ANOVA with site and species as main effects), and asterisks denote a site effect that was significantly different from zero (P>0.01). Reprinted from Callaway et al. (2002) with permission from Nature.
Survival (%)
100
90
CONTROL NEIGHBOR REMOVAL
80
LOW ELEVATION
HIGH ELEVATION
Figure 4.20. Proportion of surviving target species in controls and neighbor removal treatments at high and low elevation experimental sites for all 11 locations combined. Error bars show one standard error. (ANOVA, Ftreatment=8.47, df=1,206, P=0.004; Fsite=2.10, df=1,206, P=0.149; Ftreatment x site=22.13, df=1,206, P<0.001). Reprinted from Callaway et al. (2002) with permission from Nature.
Chapter 4
207
For all species and locations combined, neighbors had weakly competitive effects on mortality of target plants at low elevations, but highly facilitative effects on mortality at high elevations (Figure 4.20). Similarly, flowering and fruiting was reduced by neighbors at low elevations but enhanced by neighbors at high elevations. Considered together, the results of these experiments demonstrate that competition and facilitation are both important determinants of plant reproduction, survival and growth, community composition, and community diversity. The general warming trend in global temperatures may affect competitive and facilitative plant interactions in alpine communities (see Klanderud 2005, Brooker 2006). Wipf et al. (2006) at the University of Alaska conducted snow manipulation and neighbor removal experiments and found that under ambient conditions and after delayed snowmelt, positive and negative neighbor effects were “generally balanced”. However, when snowmelt occurred earlier in the summer, overall facilitative effects were stronger on the survival, phenology, growth and reproduction of Empetrum nigrum. Earlier snowmelt was correlated with colder spring temperatures and more frosts, leading Wipf et al. to conclude that plants experienced harsher environmental conditions after early snowmelt, and that facilitation was more important for ameliorating this harsher physical environment at the beginning of the growing season. Other, non-experimental studies support the findings of Choler et al. (2001) and Callaway et al. (2002) in alpine plant communities (see Klanderud and Totland 2004, Olofsson 2004). Arroyo et al. (2003, also see Cavieres et al. 1998) measured large numbers of species associated with Bolax gummifera cushions in the Patagonian Andes at 700 and 900 m elevation. They found that other species were disproportionally associated with the densely packed cushions of Bolax at both elevations, but that the strength of the spatial association varied with elevation. At 700 m many species were found within the cushions, but only 14% showed significant association. Furthermore, speciesarea relationships suggested that species richness accumulated faster outside of the cushions. In contrast, at high elevations 30% of the species were significantly associated with Bolax cushions and species-area curves increased more rapidly inside cushions than outside cushions. Arroyo et al. (2003) did not measure wind, but they argued that one of the strongest facilitative mechanisms may have been how the unusually dense cushions of Bolax sheltered beneficiaries from high winds. In sum, their observations supported the general thesis that facilitation is more important when stress increases. In a second study conducted by the same research group in the Patagonian Andes, Arroyo et al. (2003) found similar patterns of association for many alpine species at the same elevations with Azorella monantha. At 700 m
208
Interaction between Competition and Facilitation
only one species (3.1% of the total number) was preferentially associated with Azorella, but at 900 m 17 species (48.6%) were found more often in Azorella cushions than in open microhabitats. In a similar study, but in a much drier and hotter alpine climate, Cavieres et al. (2005) found the patterns indicative of facilitation were stronger at lower elevations than at higher elevations; but their results also suggested that the most intense stress was also at the lower elevations. The shift from competition at low elevations to facilitation at high elevations is probably based on fundamental physiological limitations of the different plant species. As conceptualized by Grime (1977) for shifts in competitive intensity on stress gradients, non-resource factors such as temperature, wind, and soil disturbance may have been less limiting to plant growth at low elevations, permitting plants to grow to the point where further growth or reproduction is limited by resources. At high elevations, temperature, wind scouring, or soil instability may limit plant growth more than resource availability. Amelioration of these severe stresses by neighbors may favor growth more than competition with the same neighbors impairs growth. Other experiments with arctic plants have indicated concurrent facilitative and competitive interactions. Dormann and Brooker (2002) took two approaches to measure interactions between Luzula confusa and Salix polaris. In the first experiment the removal of Luzula led to a 30% decrease in Salix shoot weight indicating facilitation. In contrast, the growth of Salix was higher in pure stands than in mixed stands suggesting interspecific competition. Similar contrasts were seen for the effects of Salix on Luzula. They argued that the competition effect was not detected in the removal experiment because manipulation of neighbors inevitably alters microclimate, creating a more stressful situation for the remaining target. Evidently, both facilitation and competition are occurring and both are important. Such mixed approaches have a lot of potential, but in this case comparisons of mixed and pure stands were confounded by site effects – perhaps pure stands of Luzula or Salix occurred in microsites that were preferred by these species and this is why they performed better there. Totland and Esaete (2002) found that 9 of 15 species in arctic Norway had greater mass beneath Salix lapponum canopies than outside, but Salix appeared to have negative to neutral effects on population density.
Chapter 4
209
Annual plant density under Cercidium as a proportion of density oustide
The most stressful ends of elevation gradients do not always have to be at high altitudes or latitudes. In deserts the reverse may be true with positive effects more important at hotter temperatures at low elevations. For example, a re-examination of Halvorson and Patten’s (1975) results suggests that the facilitative effects of Cercidium microphyllum on annual species are stronger at lower elevations than at higher elevations (Figure 4.21; also see Cavieres et al. (2005). Experimental studies of interactions such as facilitation and competition seldom provide unbiased, neutral estimates of the relative occurrence or importance of these interactions in communities. This is because focal species are chosen based upon their spatial association or disassociation with other species or a very small proportion of the community is examined. However, in the experiments of Callaway et al. (2002), we made no effort to choose species that showed particular spatial relationships with any other species, or that occupied any particular position on elevational or topographic gradients. We experimented with 15-25% of the species found at the sites, which supports the generality of stress-gradient relationships than studies that have focused on few species at one or two locations. Interestingly, our demonstration of common, strong facilitation is in line with the general neutral model constructed by 2.0
1.5
1.0
0.5
0.0
Low elevation (622-634 m)
High elevation (655-715 m)
Figure 4.21. Proportional effect of Cercidium microphyllum on annual plant density in the Sonoran Desert. Drawn from results in Halvorson and Patten (1975).
210
Interaction between Competition and Facilitation
Dodds (1997) for community organization, and the empirical model developed by Miller (1994) to incorporate both direct and indirect effects among plants. They found that positive interactions among species were as likely as negative ones in communities as long as relatively large numbers of species and connections were considered. Although the general pattern of the alpine global experiment in Callaway et al. (2002) provides compelling evidence for generality in the shift from competition to facilitation with increasing abiotic stress, the overall effect of neighbors on target species varied substantially among the geographical locations where we conducted experiments. Regression analyses of different combinations of temperature and precipitation variables against RNE values yielded only one significant relationship, a negative correlation between the RNE and maximum June (December in the Andes) temperatures estimated for each of the 22 sites. This correlation has important implications for predicting the response of alpine plant communities to climate change. Increased temperatures may alter the current balance of facilitation and competition in alpine plant communities and changes in these interactions may affect species composition and diversity differently than predicted by physiology-based models. In a study of interactions at the southern range limit of Pinus sylvestris (Scots pine) Castro et al. (2004) analyzed processes affecting seedling establishment (also see Castro et al. 2002). They monitored emergence, survival and growth for four years and found that emergence was highest under the canopy of shrubs. Survival was also highest under shrubs and extremely low (often zero) under pines or in bare soil. Environmental measurements indicated that shrubs buffered summer drought without reducing radiation to levels critical for growth (see discussion of Holmgren et al. (1997) in Chapter 2). In addition to these results from field experiments, they found that naturally occurring saplings were positively associated with shrubs indicating that positive interactions affect the ultimate spatial pattern of P. sylvestris. Interestingly, the facilitative processes controlling seedling establishment in these more xeric southern forests are very different that from those operating in the more mesic north where facilitation appears to be rare. Mulder et al. (2001) conducted another study designed explicitly to examine the role of positive interactions on a stress gradient using bryophyte species from New Zealand (this study is discussed in more detail in Chapter 6). They found no relationship between bryophyte species diversity and community productivity in humid, shaded conditions, but when communities were exposed to drought and high light diversity significantly increased community productivity. The responses of individual species – survivorship increased for almost all species with diversity in drought and the least drought
Chapter 4
211
resistant of species showed the greatest increase biomass with diversity in drought – indicated that facilitative interactions rather than sampling effects or niche complementarity were the cause of their results. In other study using a climatic gradient, Mark Bertness and Patrick Ewanchuk (2002) examined links between climate and interspecific plant interactions in New England salt marshes. As described in Chapter 2, Bertness and his colleagues clearly demonstrated that the harsh edaphic conditions in marshes can be ameliorated by neighboring plants, and that plant neighbors can have net competitive or facilitative interactions, depending on ambient physical stresses. In their system, high soil salinities can be ameliorated by plant neighbors under stressful conditions leading to facilitative interactions. In less stressful salinity-edaphic conditions, these same neighbors may be competitors. They applied this mechanistic understanding of marsh plant interactions to examine the hypothesis that latitudinal and inter-annual variation in climate can influence the nature and strength of marsh plant species interactions at regional scales. They transplanted marsh plants into vegetated and unvegetated bare patches at sites north and south of Cape Cod, a major biogeographic barrier on the east coast of North America. They hypothesized that the cooler climate north of Cape Cod would lead to fewer positive interactions. They found both latitudinal and inter-annual variation in the interactions among marsh plants that were associated with latitudinal differences in temperature and salinity. As hypothesized, south of Cape Cod, interactions tended to be more facilitative, whereas north of Cape Cod, plant neighbor interactions were more competitive. At all sites, soil salinity increased and plant neighbor interactions were more facilitative in warmer versus cooler years. As Callaway et al. (2002) found in alpine ecosystems, Bertness and Ewanchuck showed that variation in the direction and strength of interspecific interactions can be linked to climate, and that facilitation and competition shift in importance along gradients of abiotic stress. However, when Steve Pennings, Mark Bertness and colleagues conducted parallel experiments in the even hotter climates of Georgia and Alabama, where they expected to find even stronger facilitation than in southern New England, they found that removing neighbors increased porewater salinities but plants generally performed best in neighbor-removal treatments, indicating stronger competitive interactions (Figure 4.22, Pennings et al. 2003). Three of their experimental species were the same as those used in the Maine-Rhode Island experiment. Pennings et al. suggested several mechanisms that might explain the differences among the two studies. First, for widely distributed plants, southern conspecifics may be more tolerant to saline soils than northern conspecifics. Confirming this, of the three species that occurred in both studies, southern ecotypes of two species were much more salt
212
Interaction between Competition and Facilitation
tolerant than the northern conspecifics. However, most species were not studied in both geographic regions., indicating that he most likely explanation for the lack of impressive facilitation in the south is that northern marsh floras are dominated by salt-sensitive species that are likely to be facilitated by tougher neighbors, whereas southern marsh floras are dominated by salt-tolerant species that do not need neighbor amelioration of soil salinity. The bottom line, however, is that facilitation did not increase with their particular measurement of abiotic stress, but unfortunately comparisons of productivity were not reported. If productivity is higher in the southern marshes, as would be expected from general moisture and growth season length, then the results of Pennings et al. (2003) would be more in line with the stress gradient hypothesis. 2
Aster Clip Control
1
0 70
4
Batis
60 40
2
30 20
1
10 0
0
Mass (g)
16 14 12 10 8 6 4 2 0
Borrichia
2.5 2.0 1.5 1.0 0.5 0.0 40
Salicornia
0.30
Distichlis
Solidago
0.25
30
0.20 0.15
20
0.10
10
0.05
0
0.00
8
Juncus
Spartina
40
6
30
4
20
2 0
Limonium
3
50
10
J
JM MS
Georgia
J
JM MS
Alabama
0
J
JM MS
Georgia
J
JM MS
Alabama
Figure 4.22. Final biomass of transplanted plants with and without neighbors in Georgia and Alabama. Data are means 1±1SE Abbreviations: J, Juncus; JM, Juncus/meadow; MS, meadow/Spartina. Reprinted from Pennings et al. (2003) with permission from Ecology.
Chapter 4
213 Dicots
High 1.0 o oo o
0.5
oo oo
o
o
o
1992-1993 growing season Grasses
o
o o
o o oo o
o
oo o oo o o
o
o oo
o o
o o o
o o
0.0
Proportion surviving
Intermediate 1.0
o o o o o o o o o oo o o oo o o
0.5
o
o
oo o ooo o oo o o o o o oo o o
o
o
o
0.0
Low 1.0
ooo o oo o oo o o
o
ooo o oo o oo oo o
o
o o o
o
0.5
o
o
o
o
o
o
0.0 0
2500
5000
7500
10000
0
2500
5000
7500
10000
2
Cumulative no. emergents/m )
Figure 4.23. The effect of density of emerging seedlings on survival to the end of the growing season for dicots and grasses from two source communities (desert, solid lines and filled circles; semiarid, dashed lines and open circles) under three irrigation regimes (high, intermediate, and low) in the first year of the experiment. Curves represent exponential fits to the data where P<•0.10. ANCOVAs of these data can be examined in the original publication. Reprinted from Goldberg et al. (2001) with permission from Ecology.
In addition to the this work from eastern salt marshes, there are other clear, well designed, and elegantly conducted studies that do not support the stress-gradient hypothesis for facilitation and competition, and a suite of these have been conducted on annual plants in the Negev Desert (also see pages 226230). Studies conducted on large proportions of the species in single communities or community types can be illuminating because they lack the flaw inherent to meta-analysis of combining highly different endpoints. In one
214
Interaction between Competition and Facilitation
of the most complete community experiments ever conducted, Goldberg et al. (2001) studied density dependence among whole communities of desert annuals from arid and semi-arid environments in the Negev Desert (Figure 4.23). To control for habitat by community interactions, they created giant “sandbox” experiments in which seed banks from source communities were 1/16 1/8, ¼, ½, 1x, 2x, 4x, and 8x their natural densities. Furthermore, each of these source community seed banks was treated with either high or low water inputs by irrigation. They found strong evidence of community-level density dependence, but both the direction (positive versus negative) and strength of density dependent effects varied substantially depending on whether they considered emergence, survival or growth responses. The strongest facilitative responses to increasing community density was at the survival stage, where curves could be fit to the data at a significance of P<0.10 in 9 of 12 possible cases (Figure 4.23). Other studies have shown that annuals can facilitate each other, but usually annuals tend to be facilitated by larger woody plants (Went 1942, Muller 1953, Mott and McComb 1974, Halvorson and Pattern 1975, Patten 1978, Schmida and Whittaker 1981, Yavitt and Smith 1988, Callaway et al. 1991, Callaway 1992, Pugnaire et al. 1996a, Moro et al. 1997, Holzapfel and Mahall 1999, Pugnaire and Luque 2001, Tewksbury and Lloyd 2001, Robinson 2004, Schiffers and Tielbörger 2006). Therefore strong facilitation among annuals may not be typical. However, the bottom line is that Goldberg et al. (2001) did not find strong support for increasing facilitation with decreasing community density (contrasting with Bertness and Callaway 1994). But it is also not clear how density correlated with productivity. If the growth of individual annuals was density dependent (bigger plants in low density treatments), then this design may not have represented a true stress-productivity gradient (see pages 216-219). In another experiment with annuals in the Negev Desert, Lortie and Turkington (2002) found evidence for strong facilitative effects of two annual species on the rest of the annual community. However, they found facilitative affects at both high and low irrigation levels, rather than higher facilitation under the stressful conditions of low water. They hypothesized that increased water availability increased the size of the primary benefactors disproportionally in comparison to beneficiary species, thereby increasing the facilitative effects of the benefactor rather than causing a shift to competition. Clearly, when a focal benefactor species gains more from reduction in general abiotic stress than beneficiaries the predicted relationship between community productivity and plant interactions will be muddied; however, in many situations primary benefactors are those that are the most stress tolerant and
Chapter 4
215
provide ameliorated conditions to less stress tolerant species (McAuliffe 1988, Bertness and Hacker 1994, Callaway 1994a, 1998b, Choler et al. 2001). A field experiment with rhizomatous perennial plants also contradicts the general stress-facilitation model. Franz Rebele (2000) conducted a species replacement experiment with three species that lasted for five years. His results suggested that root competition, shoot competition, and positive interactions varied independently along a productivity gradient, resulting in positive interactions of maximum strength at intermediate productivities. The strongest facilitative interactions occurred where root competition was declining and where shoot competition had not yet maximized. However, root and shoot competition were not measured but inferred from changes in vegetation processes and species abundances in the plots. As for Rebele (2000), the discrepancy between Bertness and Callaway’s theory and the experiments from North American marshes and Negev Desert annual communities (Goldberg et al. 2001, Bertness and Ewanchuk 2002; Pennings et al. 2003) questions a simple linear relationship between increasing stress and the intensity of facilitation (see Michalet et al. 2006). This lack of linearity was examined in detail by Fernando Maestre and Jordi Cortina (2004) from the University of Alicante in Spain. They evaluated variation in the net effect of the tussock grass Stipa tenacissima on the shrub Pistacia lentiscus across a gradient of abiotic stress by fitting the relationship between accumulated rainfall at different sites with and the relative neighbor effect (RNE) to a quadratic model. In other words, they reported that competitive interactions dominated at both extremes of the gradient and facilitation was the most important at the middle of the gradient (Figure 4.24). They argued that these results suggest that a shift from facilitation back to competition under unusually high abiotic stress is likely to occur “when the levels of the most limiting resource are so low that the benefits provided by the facilitator cannot overcome its own resource uptake”. This may be true in some extreme conditions, but only if the primary facilitative mechanism and the primary competitive mechanism are operating on precisely the same resource. For example, if the most important facilitative effect of Stipa on Pistacia is to ameliorate drought through shade, but soil water is the most limiting resource for which these species compete, it is reasonable to expect facilitative intensities to decrease at the extreme ends of the stress gradient.
216
Interaction between Competition and Facilitation 100
Tussock Open
Seedling survival (%)
80
60
40
20
0 Ve
Ft
Ji
Pa
Al
Ln
RNE index (-1, relative units)
0.6
Re
Ma
Pe
Fi
R2 = 0.95, P = 0.002
0.4 0.2 0.0 -0.2 -0.4 -0.6 -0.8 90
100
110
120
130
140
150
160
170
180
Accumulated rainfall (mm)
Figure 4.24. (a) Survival of transplanted Pistacia lentiscus seedlings with Stipa tussocks (filled bars) and in the open (open bars) after a growing season in the field, and (b) the relationship between accumulated rainfall during the experiment and RNE. The codes on the x-axis of this graph are the abbreviations of the name of the experimental sites, which are sorted by accumulated rainfall in ascending order. In (b), positive and negative values of the RNE indicate facilitation and competition, respectively; R2 = 0.95 and p = 0.002. Reprinted from Maestre and Cortina (2004) with permission from the Proceedings of the Royal Society of London B.
As noted above, and below, in the discussion of Tielbörger and colleagues work on desert annuals (Tielbörger and Kadmon 1995, 1997, 2000), measuring stress is not easy, and different metrics used to estimate stress gradients can yield very different relationships. For example, Maestre and Cortina (2004) used the total amount of precipitation occurring at each study site during the course of their experiment as in index of stress. Drought certainly seemed to be the predominant abiotic factor in their system (Maestre 2003), but realistic drought effects are hard to measure accurately because of
Chapter 4
217
temporal distribution of rain, the effects of temperature on precipitation, and soil characteristics. A far better metric of stress, (see page 195) is the response of the target species themselves to site conditions. Such a bioassay incorporates other factors that might affect targets such as temperature, soils, precipitation prior to the experiment and so on. I reexamined Maestre and Cortina’s results by using survival (productivity was not measured) of Pistacia that was reported from the open microsites is as a bioassay of stress in a reanalysis of Maestre and Cortina’s results and the outcome suggested a different possible interpretation (Figure 4.25). Although the regression analysis was not significant, the correlative pattern showed that as survival of Pistacia in the open decreased, suggesting more stressful growing conditions in general, RNE decreased, suggesting stronger facilitative and weaker competitive effects. The relationship is linear and does not indicate that facilitation shifts back towards competition as stress reaches extreme levels. Fernando Maestre generously provided a similar analysis of a large body of his research on nurse effects for this chapter. He combined the results from Maestre (2002), Maestre et al. (2001, 2003ab, 2004) and Maestre and Cortina (2004) and placed the net interaction effect of nurses (RII) on the yaxis, and survival of beneficiaries (as a bioassay of stress) on the x-axis (Figure 4.26). When Stipa tenacissima was the nurse, and effects were primarily direct, net effects were highly facilitative when survival in the open was low (high 0.6 2
r =0.30, p=0.20
0.4
RNE
0.2 0.0 -0.2 -0.4 -0.6 0
20
40
60
80
100
Survival of Pistacia in the open Figure 4.25. Re-analysis of Maestre and Cortina’s (2004) results using survival of Pistacia lentiscus as a bioassay for abiotic stress.
218
Interaction between Competition and Facilitation
Net effect (RII index)
1.0
Stipa as nurse, r2 = 0.19; p=0.054 Pinus as nurse
0.8 0.6 0.4 0.2 0.0 -0.2 -0.4 -0.6 -0.8 -1.0 0
20
40
60
80
100
Post-summer survival in the open microsite (8-10 months after planting) Figure 4.26. Relationship between survival of beneficiary species in open treatments (a metric of abiotic stress) and the net effect (RII) of nurse plants on beneficiaries in semi-arid habitats in Spain. Plots represent different sites, and points above the line show facilitative interactions. Points below the line indicate competitive interactions. Provided by F.T. Maestre.
stress) and generally neutral when survival in the open was high (low stress), supporting the stress-gradient model. When Pinus halapensis was the nurse, and effects were primarily indirect (see Chapter 3), there was no relationship between beneficiary mortality and RII. It is completely reasonable to expect the importance (see Figures 4.34 & 4.35, this chapter) of facilitative interactions to decrease in exceptionally stressful conditions. Conditions may deteriorate to the point where stressintolerant taxa cannot make it even with a beneficiary or stress-tolerant taxa may simply not be able to ameliorate the crucial limiting factors. Michalet et al. (2006) pointed out that recent experimental studies suggested that the role of facilitation itself may wane in exceptionally severe environments (Belcher et al. 1995; Kitzberger et al. 2000; see also Bruno et al. 2003). Furthermore, although Callaway et al. (2002) found an overall increased role of facilitation with increasing stress in their intercontinental study on biotic interactions along altitudinal gradients in alpine and arctic communities, they also found that facilitation was much more intense at temperate high elevation sites than at the
Chapter 4
219
most severe sites in high elevation arctic environments. This suggests that the importance or intensity of facilitation may actually decrease with increasing stress in the most extreme conditions. In the next section, I continue this discussion of facilitation, competition, and abiotic stress, but in the context of spatial and time scales.
4.2. SPATIAL SCALES, TIME SCALES AND THE BALANCE OF FACILITATION AND COMPETITION ON STRESS GRADIENTS
In the prairie of the Midwestern United States, Bryan Foster (2002) conducted removal and addition experiments along a topographic-productivity gradient in order to study shifts in competition and facilitation. He demonstrated a shift from facilitation at low productivity to strong competitive interactions at high productivity. Transplanted Schizachryium scoparium also were facilitated across all treatment types considered together with 56% survival in plots without neighbors and 88% survival in plots with neighbors, but facilitation was stronger in plots with low productivity and high stress. Neighbors strongly facilitated survival, but suppressed growth rates. Foster argued that the effects of neighbors on growth must be integrated over a beneficiary’s entire lifetime, and if resources are generally available most of the time, then integrated effects should be competitive. The effects of neighbors on survival, however, may be most crucial during relatively brief periods of time when water and temperature stress are extreme, and therefore may reflect the effects of the same species on growth rates. Individuals that are able to survive due to the positive effects of neighbors may not be able to take advantage of more common benign times because of the competitive effects of the same neighbors. Foster’s idea for the importance of time scale has rarely been considered in other systems and for other mechanisms, leaving this an excellent avenue of future research. Other stresses may have even more transitory effects and these may be very difficult to analyze. A plant may compete with an unpleasant neighbor for an entire growing season for resources, but if the unpleasant neighbor protects the life of the plant during a crucial 10-second search by a large herbivore then the most important interaction between the two is probably facilitation for the survival of seedlings. Buckley et al. (1998) observed similar temporal scales of positive and negative interactions involving Quercus rubra seedlings planted in different habitats. They found that mature tree canopies suppressed the growth of Q. rubra seedlings, but that canopies significantly increased seedling survival. They found that seedlings in clear-cut
220
Interaction between Competition and Facilitation
plots were often frost damaged and argued that mortality there was due to much lower temperatures and browsing. The benefit of removing neighbors on longterm competitive effects was compromised by loss of simultaneous short-term facilitative effects because competitors moderated periodic early frosts. Spatial scale is as important as temporal scale for understanding how competition and facilitation work together (Lortie et al. 2005). Schenk and Mahall (2002) found that seedlings of Ambrosia dumosa and Acamptopappus sphaerocephalus, two Mojave Desert shrubs, emerged predominantly on the northern side of other shrubs, indicating facilitative effects of canopy shading on emergence. However, survival of Ambrosia seedlings was much higher in open areas than at the edge of conspecific shrubs. Removal of conspecific neighbors increased the size of Ambrosia. Acamptopappus shrubs and seedlings were positively associated with mature Ambrosia, and were most common the on the northern side of Ambrosia. When Ambrosia neighbors to the north (these did not provide much shade) were removed the growth of Acamptopappus increased, indicating competition. Removing Ambrosia neighbors from the southern side Acamptopappus plants had positive effects. Apparently the negative effects of nearby Ambrosia on Acamptopappus were mediated if the encroaching neighbor was positioned to block enough sun. This combination of positive and negative interactions contributed to the spatial relationships between Ambrosia and Acamptopappus and to the specific orientation of neighbors, illustrating how complex plant interactions can shape community structure by affecting distances among neighbors and also the directions from which neighbors are welcome (see also Noble 1980). Schenk and Mahall showed the importance of including small spatial variation in study designs, and Tewksbury and Lloyd (2001) demonstrated how differences at much larger scales can change perspective on competitive and facilitative interactions and stress. They used three scales of resolution to examine the effects of a long-lived desert tree, Olneya tesota (ironwood), on the structure and diversity of understory plant communities in the Sonoran Desert. They examined the effects of Olneya canopies in both mesic and xeric sites throughout the central Gulf Coast subregion of Sonora, Mexico (for more detail see Chapter 6.4). In xeric sites, Olneya canopies had strong positive effects on plant richness and abundance, and small positive effects on the size of understory plants, emphasizing the role of facilitation in extreme environments (Figure 4.27). In contrast, Olneya canopies in mesic sites had very little effect on understory perennials and a negative effect on ephemeral richness, suggesting that competitive effects predominated in this less stressful environment. Overall, Olneya canopies increased biological diversity through facilitation where abiotic stress was high, but did not increase diversity in mesic areas.
Chapter 4
221
Richness
Perennials 7 6 5 4 3 2 1 0
A a
Ephemerals
5 B 4
a
3 b
2
a
a
D a 200
15
a a
a
150
10
a
100
b
50
5
0
0 Unique species (%)
a
0
20
50 40 30 20 10 0
a
b
a
1 C
Abundance
a
E
a
*** c c b
Xeric
Mesic
50 40 30 20 10 0
F *** a
a a
Xeric
a
Mesic
Figure 4.27. Richness (A, B), abundance (C, D) and the proportion of species in a site found exclusively under Olneya canopies or in control plots (E, F) for perennials (left) and ephemerals (right) for xeric and mesic sites. Gray bars are Olneya plots; white bars are control plots. Mean ± one standard error from estimated marginal means are shown. Letters above bars indicate differences significant at the P=0.05 level from GT2-method post-hoc tests. Reprinted from Tewksbury and Lloyd (2001) with permission from Oecologia.
Michalet et al. (2002) noted that studies of spatial correlations among overstory and understory species tended to show to increases in association on gradients from wet to dry climates. They measured canopy-understory associations along the strong rainshadow gradient that occurs from the wet external Alps to the dry inner Alps by analyzing the species composition of 290 relevés of forests dominated to different degrees by Abies alba and Picea abies. Using multivariate analyses, they found no significant correlations between canopy species and understory composition in the much wetter external Alps despite the fact that Abies and Picea occurred in substantially different physical environments. In contrast, Abies and Picea occurred in more similar physical environments in the dry inner Alps, but the composition of understory plant communities associated with either Abies or Picea was significantly different. This increase in canopy-understory association was in part determined by strong differences in the physical environment (moisture between southern and
222
Interaction between Competition and Facilitation
northern slopes in the inner Alps), which affected both canopy and understory species distributions. However, differences between the canopy effects of Picea and Abies also appeared to contribute to stronger associations between canopy and understory species, and consequently seemed to increase beta diversity in communities of the inner Alps. This suggests that canopy-understory associations in closed forests may be weak or nonexistence on the wet ends of moisture gradients but strong on dry ends. One of the most detailed studies of spatial scale and shifting facilitative and competitive interactions was conducted by Ian Dickie and colleagues along a continuum of grassland and oak woodlands in Minnesota (Dickie et al. 2005). They found that facilitative effects of canopy trees on tree seedlings was manifest when seedlings were 12 to 20 m from canopy trees, through the beneficial effects of increased mycorrhizal infection resulting in increased nitrogen uptake and increased growth. However, when seedlings were closer to canopy trees, strong competitive effects were manifest through canopy closure. As a result, seedling growth was maximized at intermediate distances. These nonlinear facilitative interactions resulted in a positive correlation of tree density and seedling growth at low densities of canopy trees, and a negative correlation when canopy trees were dense. In one of the most arid regions of Europe, Francisco Pugnaire and colleagues have conducted several important studies of facilitation and competition involving a dominant shrub, Retama sphaerocarpa (also see LòpezPintor et al. 2003). Their results show that Retama acts as a trap for seeds and also filters the seeds in facilitative and competitive ways that appear to determine which species appear in the understory. As a result, highly contrasting plant communities occur under Retama versus intershrub spaces (Pugnaire et al. 1996a,b). The facilitative mechanisms appear to include increased levels of organic matter, nitrogen and phosphorus, higher mineralization rates of nitrogen, and amelioration of the microclimate (Moro et al. 1997). However, landscape scale aspects of this relationship are important. The striking associations of herbs with Retama canopies that occur in the driest environments are not as conspicuous in less stressful and more mesic environments. At higher elevations where soils are less fertile, precipitation is on average 38% greater, and mean temperatures are slightly lower, the total understory plant biomass under shrubs is not different than that in the open spaces between the shrubs (Pugnaire and Luque 2001). In their study the basic pattern of higher understory productivity with Retama at low elevations, but not at high elevations, held true for all of the dominant understory species in each area and one particular species, Artemisia barrelieri, which occurred at both sites. In an experiment with Artemisia, total plant mass and shoot mass of Artemisia seedlings were larger under Retama than in gaps at the lower, drier site, but not at the higher wetter site. Interestingly, belowground exclosures
Chapter 4
223
designed to reduce root competition demonstrated that the competitive effect of Retama was also more intense at the dry site than at the wet site. It was not less intense competition at the dry site that caused the net positive effect of Retama; it was exceptionally intense facilitative mechanisms. For most systems, published evidence for facilitation is based on spatial patterns or experimental evidence from one or two year’s work. In systems where abiotic conditions fluctuate over time, clearly understanding variably over time scales can be important and there is potential to examine shifts in facilitative and competitive effects among years that vary in abiotic stress or the productivity of the community. The first experimental evidence that I know of for temporal shifts between net competition and facilitation was from studies that Jack Greenlee and I did in the intermountain prairie of the northern Rocky Mountains (1996). Our work also provided compelling evidence for abiotic stress as a determinant factor for the relative importance of competition and facilitation. On open, south-facing, calcareous soils, Lesquerella carinata (a small herb in the Brassicaceae) was up to three times more common under bunchgrass canopies than expected by chance. In 1993 we did an experiment in which Lesquerella seedlings were planted either under bunchgrass canopies, in the open, in plots with clipped canopies, and in the open with artificial shade. There were no differences in seedling germination among treatments; however in complete contrast to what we expected based on the positive spatial associations there was 100% survival in the open and only 68% survival under bunchgrass canopies (Figure 4.28A). This demonstrated competition, not facilitation.
Survival (%)
100
100
A
80
80
60
60
40
40
canopy clipped canopy artificial shade open
20
B
20 0
0 0
50
100
150
0
50
100
150
Days after planting Figure 4.28. Effects of bunchgrasses and artificial canopies on the survival of Lesquerlla carinata in a wet year (a) and in a dry year (b) in western Montana. Reprinted from Greenlee and Callaway (1996) with permission from The American Naturalist.
224
Interaction between Competition and Facilitation
However, the early summer growing season in 1993 was one of the coolest and wettest on record in the region. Mean temperatures at nearby Missoula, Montana in June and July were 14.9oC, total precipitation 88.1 mm, and the two months had a total of 37 overcast days. Because the weather in 1993 was so strange we repeated the experiment in 1994 and obtained the opposite results. The weather in 1994 was much warmer and drier with mean temperatures in June and July of 18.4oC, total precipitation of 60.7 mm, and only 14 overcast days. Overall seedling survival at the experimental site used in 1993 (Rattler Gulch) was only 6%, as compared to 78% in 1993. In 1993 only one seedling out of 120 survived either in the open or where bunchgrass canopies had been clipped (Figure 4.28B). However 16% of seedlings survived under canopies and 15% survived under artificial shade. In 1994, we also repeated the experiment at a second site (Bear Gulch), one that appeared to be less stressful because it was west facing and occupied by more mesic plant species. At Bear Gulch, survival of Lesquerella seedlings was much higher. However, even though treatment effects at the “low-stress” site were of very different magnitudes, they followed the same patterns with much higher survival under canopies and artificial shade structure than in the open or where canopies had been removed. Treatment effects on Lesquerella biomass were much less clear-cut, but there were no effects of treatment on whole-plant biomass in 1993 (the general trends also suggested competition not facilitation), and significantly lower Lesquerella root biomass in the open and clipped treatments than under canopies and shade covers (suggesting facilitation). The strong spatial associations of seedlings and adults with bunchgrasses suggest that facilitation is predominant over the lifespan of Lesquerella plants. This research supports the general role of abiotic stress in plant-plant interactions proposed by Bertness and Callaway in 1994. Unlike the results of Pugnaire and Luque (2001) for Retama shrubs in Spain, the very strong shift to net facilitation in the dry, stressful year for Lesquerella appeared to be due to decreasing competitive intensity as well as increasing facilitation intensity. In the wet year of 1993, survival in the clipped canopy-treatment (no shade but still with root system intact) was very similar to the canopy treatment, and both were significantly different than either the open or artificial shade treatments. This indicates that the competitive effects of bunchgrass roots were strong. In the dry year and at both sites, however, effects of the canopy and artificial shade treatments were very positive and very similar, and the canopy-clipped and open treatments were very negative and very similar. In other words, root competition appeared to be weak because the presence or absence of canopies had the same effects whether or not roots of competitors were present.
Chapter 4
225
Most studies using stress gradients have been conducted within a single broad vegetation type. Considered together, they make a powerful argument for the generality of increasing facilitation with abiotic stress. The strength of experimenting within a vegetation type is that the potential to confound different mechanisms and different stresses that might operate in completely different vegetation types with shifts in competitive and facilitative intensity are minimized. The weaknesses are that shifts in competition and facilitation within a vegetation type may not identify basic, consistent processes that operate above the constraints of community idiosyncrasies and cross-community generalization must be based on non-quantitative conclusions. The potential for idiosyncratic shifts in interactions along gradients is emphasized by experimental research by Sally Hacker and Mark Bertness on tidal gradients in the eastern U.S. They tested for the effects of competition, facilitation, and abiotic factors across vegetation zones that develop along the elevational gradient from the lower salt marsh to the upper marsh. They found that neighbors reduced leaf area and flower production of target plants in the high marsh more than in other zones. Strong competition and physical factors dominated target performance in the low marsh; however, direct positive interactions were crucial in between these two zones. While corresponding to the general “abiotic stress” model, they pointed out that several conditions had to be met in order for strong facilitation to emerge. The competitive dominant had to be absent, abiotic stress had to be intense, and a species with benefactor potential had to have the capacity to grow in those conditions. It stands to reason that if these conditions are not met in other systems, predictions about the relationship between stress and plant-plant interactions may fall short. For example, the rather tidy relationship between temporal differences in abiotic stress and facilitation found by Jack Greenlee and I (1996) was not apparent at all in longer term studies conducted by Tielbörger and Kadmon (1997, 2000, but see contrasting results in Schiffers and Tielbörger 2006) in the Negev Desert in Israel. In order to test the hypothesis that temporal fluctuations (testing the time scale) in the abiotic environment modify the relative importance of net positive and negative interactions “in a manner consistent with the prediction of Bertness and Callaway” they investigated the common association between desert annuals and shrubs (positive associations have been reported by Went, 1942, Muller 1953, Halvorson and Patten 1975, Patten 1978, Holzapfel and Mahall 1999, Facelli and Temby 2002, Robinson 2004) by comparing spatial relationships among a number of desert annual plant species and perennial shrubs over four years that varied dramatically in precipitation. Comparisons of densities under shrubs versus in the openings between the shrubs for four annual species over four years yielded 6 species-year combinations with significant positive shrub correlations and 1 with a
Interaction between Competition and Facilitation
226
Reproductive success
significant negative shrub correlation. Considered together with past results, they demonstrated that the overall spatial relationship of shrubs with annuals was consistently positive, but with substantial variation among the different annual species (Tielbörger and Kadmon 1995). However, the strength of the positive relationship was not consistent and temporal relationships between precipitation and association strength did not support the general hypothesis proposed by Bertness and Callaway (1994). The proportional density of annuals under shrubs versus openings ([density under shrubs – density in openings] / density in the open) along an increasing gradient of annual rainfall showed a general trend of increasing, rather than decreasing positive association. The same relationship held true for the proportional seed production of annuals over the four-year period. Based on these results Tielbörger and Kadmon (2000) proposed a conceptual model for the way rainfall alters the positive and negative effects of desert shrubs on the reproductive success of annual plants (Figure 4.29). First,
shrubs
e
openings
d
a
b
c
Rainfall
Figure 4.29. Conceptual model of the qualitative relationships between the amount of rainfall and the reproductive success of annual plants under shrub canopies and in open areas. The model predicts negative effects of shrubs on reproductive success in dry years (rainfall < c) but positive effects for wet years (rainfall > c). Point a = threshold value of rainfall that allows reproduction in the open habitat; point b = threshold value of rainfall that allows reproduction in the shrub habitat (higher than point a due to rainfall interception by shrub canopies); point c = critical rainfall threshold above which positive effects of shrubs on plant reproductive success predominate over negative effects; below point c water is the main limiting resource under shrubs, while above c nutrients are limiting in the open areas; point d = reproductive success in the open habitat when water is not limiting; point e = reproductive success under shrubs when water is not limiting (higher than d due to increased nutrient regeneration). Redrawn from from Tielbörger and Kadmon (2000) with permission from Ecology.
Chapter 4
227
they argued that a threshold level of rainfall is necessary before annual species can reproduce. The threshold is higher under shrub canopies because the canopies intercept rainfall and intercept proportionally more rainfall in drier years. Above the rainfall threshold under shrub canopies reproduction is positively correlated with rainfall until a point where water no longer limits reproduction. Reproductive output is higher under shrubs when rainfall is high because nutrient concentrations are higher under shrubs. Tielbörger and Kadmon’s model provides an important and empirically supported alternative (or modified) hypothesis to the model of a simple positive relationship between temporal variation in abiotic stress and facilitation strength (also see Michalet et al. 2006). However, several caveats should be considered for interpreting their studies (also see discussion in Robinson 2004 and contrasting results in the same system in Schiffers and Tielbörger 2006, Holzapfel et al. 2006). These same caveats should be considered when interpreting any study of interactions along gradients, whether the results suggest increasing facilitation with stress or not. As was the case in a similar study in Spain (Maestre and Cortina 2004), annual precipitation, which was used by Tielbörger and Kadmon as the metric for the stress gradient, did not appear to be correlated with productivity in the four years of the study, raising the issue of crucial confounding processes. Neither productivity nor biomass was measured, but the cumulative density of the four species in the wettest year (167 mm) was ≈195 plants per 635 cm2, whereas in a year with only 50 mm of rainfall cumulative density of annuals was ≈225 plants per 635 cm2. In a year with 38 mm of rainfall cumulative density of annuals was ≈70 plants per 635 cm2. If densities are even an approximation of productivity, then total annual rainfall would not have provided a reasonable surrogate for stress, and as for Maestre and Cortina (2004), may not have been the best gradient along which to develop a general model for shifts in positive and negative effects. Total precipitation in a year may not correlate simply with the productivity of desert annuals for several reasons. Because the lifespan of desert annuals is extremely short the timing of rainfall events is more important then the total amount. For example, large amounts of precipitation in the late spring may have minimal effects on plant growth. As such, proportional canopy interception of precipitation (see Figure 4.29) does not depend on annual rainfall; it depends on the intensity of individual rainfall events. The Negev receives precipitation in the winter, when cool temperatures may affect the productivity of annuals as much or more than precipitation. If annual density is used as a surrogate for the productivity gradient then patterns suggestive of facilitation tend to be stronger towards intermediate densities, as suggested by Maestre and Cortina (2004) and weakest at the highest and lowest densities (Figure 4.30) although these
Interaction between Competition and Facilitation 35
25 20 15 10 5
COMPETITION?
mean ratio Senecio Ifloga Rumex Erodium
30
FACILITATION?
Relative increase in density of annuals under shrubs compared to open areas
228
0
60
80
100
120
140
160
180
200
220
240
Estimate of total annual density (productivity) Figure 4.30. Relationship between the relative increase in density of annual plant species under shrubs relative to productivity in open areas around shrubs and estimate of total densities of annuals in different years in the Negev Desert. Redrawn from data in Tielbörger and Kadmon (2000).
patterns are not strong. Interestingly, two later studies along spatial gradients of precipitation in the same communities did find competition to be stronger at the wet end and facilitation stronger at the dry end (Schiffers and Tielbörger 2006, Holzapfel et al. 2006). Holzapfel et al. (2006) conducted an exceptionally thorough investigation of shrub-annual interactions at four sites along a 245 km aridity gradient in Israel. This study is probably the most complete whole-community study yet conducted on facilitation in deserts. They found that above-ground productivity, richness, seedling density, seed bank density, and fecundity of annual plants were higher, and usually substantially higher, under shrubs than in areas between shrubs at the arid end of the gradient, but significantly lower at the humid end. In their words, “net effects of shrubs on annuals expressed as relative interaction intensity indicated a steady and consistent shift from net positive or neutral effects in the desert to net negative effects in the mesic part of the gradient”, which is consistent with the stress-facilitation relationship proposed by Bertness and Callaway (1996). Studies of shrub effects on annuals in Chilean deserts also suggest that the use of density or biomass as metrics may provide different results. Gutiérrez et al. (1993, also see Holmgren et al. 2000) found that the density of annuals was 3.7 times higher (significant) in the open than under shrubs, suggesting general inhibition. Biomass, however, showed the reverse, and
Chapter 4
229
showed a non-significant trend of 1.3 times higher under shrubs compared to the open. Yet another concern involves beneficiary identity. Tielbörger and Kadmon combined different shrub species into a single analysis of “shrub effects”. While not an egregious problem, the benefactor-specific nature of some facilitative interactions among desert annual and shrub species (see Chapter 5) could confound other effects. The importance of direct measurements of productivity, and accounting for benefactor identity in shrub-annual systems, are illustrated in Duncan Patten’s study of shrub facilitation of annuals in the Sonoran desert. He harvested annual plants in a time series of samples under three species of woody perennials in a dry growing season (≈100 mm, 1973-1974) and in a wet growing season (≈500 mm, 1972-1973). He found marked differences in the effects of different shrubs, and between the years. In the wet season the average rate of productivity ranged from 1.71 to 11.80 kg-1ha-1day-1; whereas in the dry season productivity ranged from 0.19 to 1.61 kg-1ha-1day-1. Supporting the stress hypothesis, he found that the strong facilitative effects for Cercidium in the dry year shifted to neutral-competitive effects in the wet year. When precipitation was ≈100 mm for the season, productivity under Cercidium was 40% greater than in the open, when precipitation was ≈500 mm productivity of annuals was 5% less under Cercidum than in the open (Chapter 6.5, Figure 6.10). Interestingly the effect of two highly inhibitory shrub species, Larrea tridentata and Ambrosia dumosa shifted in the opposite direction with precipitation and total productivity. For these two species, their competitive effect was stronger in the dry year than in the wet year. Such species-specific idiosyncrasy makes it very unlikely that all study systems will behave well with respect to our conceptual generalizations. It is also worth considering whether or not winter annual species make very good representative beneficiary species for developing general stressgradient models. For example, Tewksbury and Lloyd (2001) found that facilitative effects of canopies were much stronger for perennials than annuals in the Sonoran Desert of Mexico (Chapter 5.2.4). Even though desert annuals are often highly correlated with shrub canopies, their life history makes them “stress avoiders”, not “stress tolerators”. They grow rapidly when moisture is available, complete their lifecycle in a few weeks, and then spend the really stressful times as seeds. This issue is emphasized in the driest year of Tielbörger and Kadmon’s experiments (1996, 1997, 2000). Total densities were very low in the driest year, when only one of the four species occurred at densities of >1-2 individuals per sample plot.
230
Interaction between Competition and Facilitation
A last issue of importance to all gradient studies is the use of natural gradients versus experimental gradients. Tielbörger and Kadmon’s findings are based on demographic measurements along a natural gradient. In general, studies of interactions along natural gradients are consistent with the concept that increasing productivity corresponds with stronger competitive interactions (e.g. Grime 1977; see del Moral 1983, Gurevitch 1986, Wilson and Keddy 1986, Reader and Best 1989, Pennings and Callaway 1992), whereas studies using experimentally created gradients do not. As suggested by Goldberg and Barton (1992), results from studies using natural gradients of productivity are generally consistent with the model of Grime (1974), and show an increase in competition intensity with increasing productivity (Kadmon 1995, Greenlee and Callaway 1996, Twolan-Strutt and Keddy 1996, Foster 1999, Gerdol et al. 2000, Choler et al. 2001), whereas results from studies using experimentallyinduced gradient of productivity show less variation in competition intensity with productivity or soil fertility (Wilson and Shay 1990, Di Tommasso and Aarssen 1991, Wilson and Tilman 1991, Campbell and Grime 1992). Kadmon (1995) tested both approaches (natural versus experimental gradients) within the same system, but established a productivity gradient by varying water availability, and found an increase in competition intensity with productivity using both approaches. The bottom line is that stress-interaction relationships may vary depending on the experimental approach taken, and interpretation of experimental results should take this into account. Tielbörger and Kadmon’s research in the Negev Desert provides a great deal of crucial insight into the conditionality of positive interactions. Furthermore the essential absence of a relationship between precipitation and facilitation strength in their results is supported by other studies. Casper (1996) found that positive or negative interactions between shrubs and the annual Cryptantha flava did not change with drought. In experiments in the coastal dunes of Florida and Georgia, Franks and Peterson (2003) tested the hypotheses that facilitative interactions should increase with increasing stress and disturbance. He transplanted Uniola paniculata and Iva imbricata individuals into plots either with conspecific neighbors, neighbors of the other species, or no neighbors, at three zones across an environmental gradient on the dunes. The plants were harvested after two growing seasons. Neighbors significantly increased the survival of target species, indicating facilitation. However, the growth of experimental plants was either unaffected or reduced by the presence of neighbors, indicating neutral and weak competitive interactions. Importantly, competition was greatest on the foredunes, which was the zone of earliest succession and appeared to be the most abiotically stressful. For these two coastal dune plant species, the outcome of interactions differed for growth and survival components of fitness and depended on position along the
Chapter 4
231
environmental gradient, but facilitation did not increase in importance with abiotic stress. Tielbörger and Kadmon’s long term studies provide valuable insight into the relative importance of facilitation and competition in communities as abiotic stress fluctuates over relatively small time scales; however, even longer temporal scale effects are recorded in the annual rings of trees. Kitzberger et al. (2000) used the annual tree rings of Austrocedrus chiliensis to assess temporal and spatial variation in establishment and growth patterns, specifically to understand the effects of climatic variation on these variables. Austrocedrus was highly significantly associated with shrubs in general, apparently due to the shade provided by the canopies, but recruitment under shrubs depended on the climatic conditions of particular years. Years in which Austrocedrus recruited in the open had higher than average amounts of summer precipitation, years in which Austrocedrus recruited under shrubs but not in the open had average precipitation in the summer, and years in which there was no recruitment anywhere had less than normal summer precipitation. These results suggested that net facilitative or competitive interactions between the shrubs and Austrocedrus correlated with a temporal gradient of drought stress. Net facilitative effects of nurse shrubs were only apparent during climatically suboptimal years, whereas during optimal conditions Austrocedrus was able to recruit irrespective of nurse shrubs. When conditions were really bad no establishment occurred at all. These results support Maestre and Cortina’s (2004) and Michalet et al.’s (2006) “humpbacked” model of facilitation. The response of Austrocedrus on this particular temporal gradient provides a great deal of insight into the difficulty ecologists have in producing general models of facilitation and competition on gradients. Not all gradients are the same. Abiotic stress can reach an extreme where recruitment, growth or reproduction simply cannot occur, and therefore interactions do not occur. If Kitzberger and colleagues experimented with nurse shrubs and Austrocedrus recruitment only in unusually wet years and unusually dry years; they likely would find evidence for competition in the former and no interaction in the latter (see Houle 1996). Facilitation occurred in “average” years in this very arid place, and based on the positive spatial associations among shrubs and Austrocedrus, facilitated recruitment in “average” years is crucial to sustaining viable populations of this species. Analyses of annual tree ring growth rates also proved highly useful for examining the relative importance of facilitation and competition along elevational gradients in the northern Rocky Mountains of Montana, USA. Subalpine and timberline forests there are often dominated by Pinus albicaulis and Abies lasiocarpa. Under some conditions the stress-tolerant P. albicaulis is early-successional and the shade-tolerant. A. lasiocarpa predominates in late
232
Interaction between Competition and Facilitation
succession. At lower elevations where stand biomass and lateral growth rates are much higher A. lasiocarpa seedlings, saplings, and trees are distributed randomly with respect to mature P. albicaulis. Along exposed ridges at timberline, however, A. lasiocarpa of all size classes are tightly clustered around large P. albicaulis (Callaway 1998b). At high elevations clustering is far stronger for the larger size classes of A. lasiocarpa. Many of the P. albicaulis trees around which A. lasiocarpa trees cluster are dead; killed in outbreaks of mountain pine beetles in the 1930’s. Because many of the A. lasiocarpa individuals near P. albicaulis were old enough to have lived through the death of their apparent benefactors, the demise of P. albicaulis established a “natural experiment” in which the effects of this species could be compared before and after death, and between stressful timberline habitats versus more benign habitats. Additionally, the response of A. lasiocarpa to P. albicaulis was compared between a wet year and a dry year at both high- and lowelevation sites. Growth increment ratios (post-death rates/pre-death rates) of large A. lasiocarpa trees at one high-elevation site (Bitterroot Mountains) were 20% and 28% lower when adjacent to dead P. albicaulis than when next to live P. albicaulis in the wet and dry years, respectively (Figure 4.31). These results suggest that the growth of high-elevation A. lasiocarpa decreased after the death of their neighbor, a response indicative of facilitation. At low elevations, however, growth increment ratios of A. lasiocarpa were 4% and 9% higher in the wet and dry years, respectively, when adjacent to dead rather than living P. albicaulis. Site by “treatment” interaction effects in the analysis of variance were significant. These results suggest that the growth of A. lasiocarpa increased after the death of their neighbors at low elevations, a response indicative of competition. Treatment-by-year interaction effects were not significant, indicating that the effects of the benefactor were similar in both the wet and the dry year. A second experiment in the Rattlesnake Mountains, however, showed no significant effects even though spatial associations between Pinus and Abies were as strong. Overall, taking into consideration the weaknesses of natural experiments, these shifts in spatial clustering and growth responses suggest that facilitation was more important in the low-productivity and high stress timberline environments than in subalpine forests at lower elevations. The shift in net effects over long time scales appeared to be due more to stronger facilitative effects than weaker competitive effects.
Chapter 4
Increment growth ratio
1.2
233
B
A adjacent to living P. albicaulis adjacent to dead P. albicaulis
1.0 0.8 0.6 0.4 0.2 0.0
Low-elevation High-elevation site site
Low-elevation High-elevation site site
Figure 4.31. Increment growth ratios of large (>20cm) Abies lasiocarpa trees at sites in the Bitterroot Mountains (Montana, USA), and adjacent to either living or dead Pinus albicaulis trees. Ratios in panel (A) were obtained by dividing annual growth increment in 1983 (a wet year occurring long after the death of P. albicaulis in the 1930’s) by annual growth increment in 1925, prior to the death of P. albicaulis. Ratios in panel (B) were obtained by dividing annual growth increment in 1988 (a dry year occurring long after the death of P. albicaulis) by annual growth increment in 1925. Error bars represent 1 SE. In a 3-way ANOVA, Fsite x treatment = 22.63, df=1,159, P<0.001. Reprinted from Callaway (1998b) with permission from Oikos.
Some of the most detailed experimental studies of facilitation and competition on stress gradients that vary in spatial scale come from work done by Mark Bertness and colleagues in the Rumstick Cove salt marsh on the coast of Rhode Island, USA. There, Bertness has utilized salinity gradients that develop over slight elevational changes from the lower marsh to the upper marsh to test the hypothesis that facilitation is more important when the substrate is drier and more saline. The seaward border of salt marsh habitats in New England is dominated by the grass Spartina patens, and stands of Juncus gerardii dominate the terrestrial border. A third perennial grass, Distichlis spicata is restricted primarily to disturbed patches within the marshes. Bertness and Shumway (1993) created experimental patches in which various species were removed in order to examine their effects on neighbors. Some plots were watered to alleviate the dominant stress perceived in the system. In the zone of lowest elevation no facilitation was apparent, competition was strong, and artificial watering had effects on plot colonization (Figure 4.32). However, at the Spartina-Juncus borders at mid-elevations the removal of S. patens reduced the cover of J. gerardii, indicating facilitation in this more saline, stressful environment. Watering plots at the border eliminated the
234
Interaction between Competition and Facilitation
positive effect of S. patens suggesting that this facilitation was dependent on stressful conditions. A similar effect was seen at even higher elevations in the J. gerardii zone where salinities were also high. Removal of D. spicata suppressed J. gerardii cover, and watering eliminated this facilitative effect. Bertness and Shumway clearly demonstrated competition in moderate conditions, facilitation in stressful conditions, and experimentally eliminated facilitation by eliminating the abiotic stress. CONTROL 100
50
50
0
Percent cover
WATERED
100
100
SP
SR
Spartina
SP
SR
Distichlis
50
0 100
100
SP
SR
Spartina
SP
SR
Distichlis
BORDER
50
SP
SR
Spartina
SP
SR
Juncus
50
0
0
SPARTINA ZONE
0 100
SP
SR
Spartina
SP
SR
Juncus
JUNCUS ZONE
50
SP
SR
Distichlis
SP
SR
Juncus
0
SP
SR
Distichlis
SP
SR
Juncus
Figure 4.32. Vegetation cover in experimental bare patches after two years of recovery. For each zone results are presented for each species that invaded (listed below the x axis) with (white bars) and without (black bars) perennial neighbors under natural patch salinities and reduced patch salinities (watering). Results are presented with and with seedlings in order to consider the effect of sexual reproduction on colonization (SP = seedlings present, SR = seedlings removed). Redrawn from Bertness and Shumway (1993) with permission from The American Naturalist.
Chapter 4
235
As described above, experiments at Rumstick cove have repeatedly demonstrated a gradient of decreasing salinity and redox stress with increasing elevation in the salt marsh. Brewer et al. (1997) experimented with the effects of total biomass removal on species diversity along this gradient (spatial scale) as the communities proceeded through successional (time scale) recovery. After two years they found that removal plots at high elevations (low salinity, high productivity, low stress) had almost twice the species diversity as control plots, whereas at low elevations (high salinity, low productivity, high stress) removal plots had significantly lower species diversity than controls (Figure 4.33). These results corresponded with the effects of biomass removal on redox potential and soil salinity. At low elevations removal had much stronger effects on these abiotic variables, increasing salinity and decreasing redox potentials. After three years species richness had returned to equal levels in treatments and
6
June 1995
Neighbors intact Neighbors removed
5 4 3 2 1
Species richness
0 6
August 1995
5 4 3 2 1 0 6
August 1996
5 4 3 2 1 0
High Juncus
Mid Juncus
Low Juncus
Figure 4.33. Effects of biomass removal on species richness in a) June 1995, b) August 1995, and c) August 1996 in transects at three different elevations in the Juncus zone in a New England salt marsh. Hatched bars are biomass-removal means and open bars are control means. Error bars show 1 SE. Reprinted from Brewer et al. (1997) with permission from Oikos.
236
Interaction between Competition and Facilitation
controls at all elevations. Brewer et al.’s results indicated that when abiotic stress was high facilitation alleviated the effects of abiotic stress on species richness, but when abiotic stress was low competition limited species richness. In alpine communities in Colorado, Wied and Galen (1998) found that seedlings of Frasera speciosa and Cirsium scopulorum were associated with conspecific adults, and the litter from these adults facilitated seedlings. However, naturally occurring seedlings of Frasera showed a greater positive response to litter in a dry site than in a wet site. Interestingly, much like Kitzberger et al. (2001) found for recruitment in exceptionally dry years, seedlings of Cirsium showed strong facilitative responses to parental litter cover, but in the dry year there was no seedling establishment in any microenvironment. These results emphasize the potential for exceptionally stressful conditions to override nurse plant effects when considering time scales. The effects of litter on seedling establishment also vary from facilitative to competitive in Brazilian savanna. Hoffman (1996) found that litter had a negative effect on the establishment of Micronia albicans, a subtropical tree, in densely wooded sites, but a positive effect in drier, hotter open grassland. Plants may experience stress because they do not get enough of an essential resource, or they may experience stress because they are exposed to too much of a resource, for example water, or if they experience extreme levels of non-resource factors such as temperature, wind, salinity, or disturbance. Productivity gradients may arise from variation in resources or from variation in factors that are fundamental to the fitness of plants, but that are not resources used for growth and reproduction. Shifts in interactions along productivity gradients that develop along resource gradients may differ completely from those along productivity gradients that develop due to non-resource factors. To my knowledge, resource-driven productivity gradients have never been compared to those caused by non-resource factors; however, there are good reasons to expect differences. If stress or productivity along a gradient is determined by a limiting resource such as water or nitrogen (e.g. Goldberg et al. 2001), plants may be much less likely to have positive effects on their neighbors because they require the same limiting resource and must compete for it. For a plant to facilitate its neighbor it must be able to enhance the particular limiting resource for its neighbor despite the limitations of the same resource for the potential benefactor. This would seem to be rare except in cases involving mechanisms such as hydraulic lift and nitrogen fixation (see Chapter 2). However, non-resource stress such as wind or temperature (e.g. Choler et al. 2001, Callaway et al. 2002) can be moderated at no additional resource cost to the benefactor. In these conditions it would be more likely that the positive effects of ameliorating limiting conditions would outweigh the negative effects of competition for resources.
Chapter 4
237
Throughout this section, I have assumed that facilitative and competitive interactions are responding to factors correlated to productivity, or more subtle factors inherent to productive versus unproductive ecosystems. However, facilitated augmentation of species diversity may often be the cause of increased productivity. This is explored in Chapter 6. Perhaps the best way to conclude this section on abiotic stress and shifts in the importance or intensity of competition and facilitation would be to return to the simplest perspective on the concept. Certainly, all sorts of variation on this theme are possible, but the fundamental issue is the effort to understand in what conditions we might expect to find strong facilitative effects. Common sense suggests facilitation would be more common in relatively stressful conditions because without some kind of stress, or limitation to growth or fitness, there is not much for a neighbor to facilitate.
4.3. FACILITATION AND STRESS: IMPORTANCE VERSUS INTENSITY So far, I have used the terms ‘importance’ and ‘intensity’ in a relatively imprecise manner, and the use of the terms in the literature is loose as well. However, as pointed out by Weldon and Slauson (1986) for competition, and Hastwell and Facelli (2003) and Brooker et al. (2005) for facilitation, ‘importance’ and ‘intensity’ denote different ways of measuring or describing interactions among species. To be precise, the ‘intensity’ of an interaction refers to the change in plant performance due to interactions with neighboring plants relative to the plant’s performance in the absence of neighbors. ‘Importance’ on the other hand, is the impact of an interaction relative to the impact of all the factors in the environment on plant success. These differences are crucial to our attempts to understand the role of interactions among plant species in community organization. Rob Brooker and his Alpine Pal colleagues (Brooker et al. 2005) argued that failure to distinguish between ‘importance’ and ‘intensity’ has hindered our ability to resolve key questions about the role of plant interactions in the organization of plant communities. In a revealing reanalysis of an important global data set published by Reader et al. (1994) they found that the intensity of competition showed no response to system productivity (Figure 4.34a, as reported by Reader et al. 1994 as evidence against J.P. Grime’s [1977] conceptual paradigm). However, in contrast to the interpretations of Reader et al. (1994) the importance of competition declined with decreasing system productivity in support of predictions by Grime (Figure 4.34b). There is no
238
Interaction between Competition and Facilitation
fixed relationship however between importance and intensity as comparisons of other data sets show that both the intensity and importance of competition shift toward facilitation with decreasing productivity (Figure 4.35ab). However, for Grime’s (1977) and Bertness and Callaway’s (1994) models of plant interactions along productivity gradients, importance should probably be emphasized over intensity. 1 0.5 0
C int
0
100
200
300
400
500
600
700
-0.5 -1 -1.5 -2 -2.5 -2
Neighbor biomass (g m ) 0.2 0
Cim p
-0.2
0
100
200
300
400
500
600
700
-0.4 -0.6 -0.8 -1 -1.2 -2
Neighbor biomass (g m )
Figure 4.34. (A) Competition intensity Cint and (B) competition importance Cimp vs. neighbor biomass (g m-2) recalculated from the data of Reader et al. (1994, Fig. 1). Fitted lines show results of generalized linear mixed model analysis performed using the MIXED procedure SAS version 8.0 (SAS Institute 1999). Cint = -0.00088; neighbor biomass =/-0.6957, df = 1,39; F =/2.76; P =/0.1047. Cimp = -0.00100; neighbor biomass = -0.09264; df = 1,42; F =/27.32; P = 0.0001. Site was specified as a random effect. Reprinted from Brooker et al. (2005) with permission from Oikos.
Chapter 4
239 1.5 1
C int
0.5 0 -0.5 0
0.5
1
1.5
-1 -1.5 Mass of removal plants (PT-N) 1.5 1
C im p
0.5 0 -0.5 0
0.5
1
1.5
-1 -1.5 Mass of removal plants (PT-N)
Figure 4.35. A) Competition intensity (Cint) and B) competition importance (Cimp) experienced by Artemisia barrelieri due to the presence of Retama sphaerocarpa vs. mass of A. barrelieri plants without neighbors (PT-N; used here as an index of system productivity in the absence of information on system neighbor biomass) recalculated from Pugnaire and Luque (2001), data from F. Pugnaire (pers. comm.). The Pearson correlation coefficient between Cimp and PT-N is -0.74, P<0.001, and between Cint and PT-N is -0.77, P<0.001. Reprinted from Brooker et al. (2005) with permission from Oikos.
Hastwell and Facelli (2003) took a different tack for exploring differences in the importance and intensity of interactions in Australian shrublands. They tested the relationship between facilitation and environmental severity in a field experiment by shading seedlings of the perennial shrub Enchylaena tomentosa during the cooler wetter summer and the hotter dryer summer and measuring plant growth and survival. Shade always increased the growth of Enchylaena seedlings, but the effect of shade on growth was the same in both the winter and the summer. The intensity and importance of shade facilitation on seedling survival, however, was enhanced more by shade in the dry summer than in the mesic winter conditions – shade had negative effects on
240
Interaction between Competition and Facilitation
survival between July and October in the cool austral winter, but positive effects between January and February when the weather was xeric. As Brooker put it, “hopefully by developing explicit conceptual and mathematical definitions of competition we can move away from a semantic debate and return to the original aims of the exploration of competition [and facilitation] as a driving force in nature”.
4.4. FACILITATION AND LIFE HISTORY STAGE Often the positive mechanisms that are so important to the recruitment of some species cease to exist, or become much less important when beneficiaries mature. This is obvious for nursed perennials that outlive their benefactors for centuries. For example, Quercus species last much longer than the shrubs that nurse them (Callaway 1992, Callaway and Davis 1993), mature Pinus monophylla can be found growing in the middle of the long-dead remains of Artemisia (Callaway 1996), and saguaro cacti and other long-lived desert perennials can often be found growing above their dead benefactors (McAuliffe 1984a, 1988). One of the best-studied examples of the potential long-term effects of nurses on perennial understory beneficiaries is from Prosopis-dominated savannas and woodlands in southern Texas. Steve Archer and colleagues studied successional processes in this system for two decades, finding that Prosopis glandulosa plays a central role in the encroachment of numerous woody species in grasslands and savannas by acting as a nurse plant (Barnes and Archer 1996, Brown and Archer 1987, 1989, Archer et al. 1988, Archer 1989, 1990, 1995, Scanlan and Archer 1991). Paul Barnes and Steve Archer (1999) tested for continued facilitation of two shrubs long after they had colonized the understory of Prosopis (Zanthoxylum fagara and Berberis trifoliolata) by conducting selective Prosopis removal experiments and monitoring the physiological responses and growth of the two understory species. Overall, they found that the founding overstory Prosopis tree may continue to facilitate understory shrubs after the shrubs are mature, but that these late-stage beneficial effects are small and transitory (also see Brooker et al. 2006). Prosopis was important for the establishment of the shrubs, but less so as the shrubs matured. These results were consistent with the findings of a previous study which showed little difference in the performance of shrubs growing where the Prosopis had naturally died versus where Prosopis were still alive. Recent work in this system by Zou et al. (2005) has shown that hydraulic lift (see Chapter 2) by Prosopis has facilitative effects on some understory
Chapter 4
241
species but not others, and these effects are dependent on the season. They concluded that “net interactions between overstory and understory elements in these woody patches can be positive, negative and neutral over an annual cycle” and that hydraulic lift facilitates shallow rooted species, but not all understory species. It is important to note that canopy removal experiments only manipulate some positive mechanisms. For example, the long-term effects of nurse perennials on soil bulk density, and therefore on soil water-holding capacity and nutrients, may persist even when the canopy is removed. Thus there is some possibility that the positive effects of Prosopis on soils, remaining after canopy removal, may have minimized the differences between treatments in Barnes and Archer’s experiments, leading to an underestimate of the facilitative effects of Prosopis on mature understory shrubs. Despite the potential underestimation of long term facilitation, it was interesting that these results did not show the development of strong competition, evidence for “seed-seedling conflicts”, between Prosopis and understory species as the latter matured (see Schupp et al. 1995, Miriti 2006). All things considered, the very weak late-stage positive effects indicate that the loss of Prosopis will be unlikely to result in the demise of the shrub species that were nursed when they were young. Other studies also suggest that the role of facilitators weaken as beneficiaries mature. Most research on dynamics and interaction in “old-fields”, fields once used for agriculture but abandoned to re-vegetate naturally, has pointed to the importance of competition in these communities. In an experiment in Mediterranean old-fields, Sans et al. (1998, 2002) examined how neighbors affected the colonization and persistence of Picris hieracioides, a biennial composite, 3 and 40 years after abandonment. They found that survival of Picris was higher when neighbors were present than in their absence for 30 weeks after transplanting. Furthermore, there was a shift from positive to negative interactions with life stage. Facilitation was evident during recruitment and early growth, whereas competition became dominant later in the lifecycle of Picris, and the removal of neighboring plants resulted in lower final growth and reproduction. Overall, however, Sans et al. (1998) found that the early fate of seedlings in response to desiccation and heat stress were the key to establishment and persistence of Picris in old-field communities.
242
Interaction between Competition and Facilitation A *
1.5
5
A *
A *
A
B *
1.0
0.5
C *C * C
C C C * *
soil organic C (%)
Total N (g/kg)
2.0
100
200
300
400
estimated age
B *
B *
25
D D CD
0
B *
B AB * CD * CD * D *
100
200
300
400
estimated age
total oil P (mg/kg)
available P (mg/kg)
A *
A *
D D D* *
1
500
C A *
BC *
2
DEATH
75
50
3
A *
0
0.0 0
4
BC * B *
700
A AB * *
550
400
500
DEATH
A * AB C BC ** BC
D
* CD *
D D D
250 0 100
200
300
estimated age
total soil S (mg/kg)
1000
BCD * CD
750
AB *
A *
DD
500
400
0
500
100
200
400
ABC * D CD * D
300
400
estimated age
DEATH
A *
E EC (dS/m)
0
300
AB * B
200
500
DEATH
F AB * B
B
B
B
B
B
100
250
0
0 0
100
200
300
estimated age
400
DEATH
500
0
100
200
300
estimated age
400
500
DEATH
Figure 4.36. Soil concentration of A) organic carbon, B) total nitrogen, C) total phosphorus, D) available phosphorus, E) total sulfur, and F) electrical conductivity under canopies of Acacia papyrocarpa of different age classes, around individuals dead for different times, and in soil from surrounding open spaces. Asterisks indicate a significant difference between the canopy and the open soil for a given age class (t-tests). Different letters indicate a significant difference among soils from canopies of different ages (ANOVA). Error bars are 1SE. Reprinted from Facelli and Brock (2000) with permission from Ecography.
Probably the most detailed examination of soil and plant community changes over the lifespan of a beneficiary species was performed by José Facelli and Daniel Brock in southern Australia. In arid deserts near Whyalla, South Australia, soils under the tree Acacia papyrocarpa undergo dramatic increases in total nitrogen, organic carbon, total and available phosphorus, total sulfur, and electrical conductivity as the tree ages (Chapter 2, Facelli and Brock 2000). Then the effects of the tree decline as it senesces and dies. Tree understories are associated with a dramatic change in community composition in comparison to the open desert around the trees (Figure 4.36). Furthermore,
Chapter 4
243
comparisons of the abundances of plant species in the understory of living, healthy A. papyrocarpa to that under trees dead for less than 10 years, under trees dead for more than 30 years, and in the open desert show a gradually changing species composition back to that of the open community type (Figure 4.37). To my knowledge, Facelli and Brock’s research is the only quantification of the development and decline of “islands of fertility”. The waning dependence of beneficiaries on other species may be tied to “seed-seedling conflicts” as argued by Eugene Schupp (1995, also see Miriti 2006). He analyzed several data sets that demonstrated substantial conflicts between what was good for a seed versus what was good for a seedling (but see Castro et al. 2002). In the first example, survival from seed to seedling for Faramea occidentalis, an abundant understory tree in tropical Panamanian forests, was shown to be much higher in the understory of other species than in gaps where survival was reduced virtually to zero by predators. However, the growth of Faramea seedlings was ≈50% higher in gaps compared to that in the 40
A A
20
B
10
C A N N
D
A
C
10
B
B
S
T
ER TH
O
N
A
N A C
SP
S VE
D
YS
IN
M O
C
B
SP R
ET
SP A T O TH ER S
VE S
A
20
C A VE S C A N N
A
M SE
OPEN
0 D
ET O M R SP IN SE R YS
0
AB B
C
C SE
BC B
C
30
SE
B
frequency (%)
B
10
C
SP A T O TH ER S
O
40
30
20
A
0
DEAD 3
M
frequency (%)
40
B
B
C
R
D SE
A VE S C A N N
M
ET O M R SP IN SE R YS
0
A B
A
D
20
M
10
B
A
M R SP IN SE R YS
B
DEAD 1
30
ET O
30
frequency (%)
CANOPY
SE
A
SP A T TH ER S
frequency (%)
40
Figure 4.37. Frequency of the dominant plant species under the canopies of Acacia papyrocarpa, in open spaces, and under A. papyrocarpa trees dead for <10 (Dead 1), and >30 years (Dead 3). Different letters for an understory species indicate significant differences (ANOVA following MANOVA, Tukey P<0.05). ETOM=Enchylaena tomentosa, RSPIN=Rhagodia spinescens, SEYS=Sisymbrium erysimoides, MSED=Mairenana sedifolia, AVES=Atriplex vesicaria, CANN=Carrichtera annua, SPAT=Sclerolaena patenticuspis. Error bars represent 1 SE. Reprinted from Facelli and Brock (2000) with permission from Ecography.
244
Interaction between Competition and Facilitation
understory (Figure 4.38). The second example, Welfia georgii, was similar and seed survival was much higher in understory habitats than in gaps and seedling growth was much higher in gaps than in the understory (Figure 4.39). The third example was of recruitment of Prunus mahaleb in arid scrubland of southern Spain. There, the proportion of seeds surviving was higher in grasses growing on deep soils than in spiny shrubs, but the survival of seedlings was reversed by habitat (Figure 4.40). In a similar study Ibáñez and Schuppp (2002) found evidence for seed-seedling conflict during recruitment of Cercocarpus ledifolius in the shrub steppe in Utah, USA. They found that the emergence of Cercocarpus was highest in the open spaces between both conspecific and Artemisia tridentata shrubs and experiments showed that litter from these species inhibited Cercocarpus emergence. However, litter beneath trees increased Cercocarpus seedling survival.
Survival (%) Dry mass shoot (g)
0.25
0.20
understory 0.15
0.10
0.05
0.00
gap
Survival
Growth
Figure 4.38. Survival of seeds and growth of seedlings of Faramea occidentalis in understories and gaps in tropical forests in Panama. Redrawn from Schupp (1995) with permission from the American Journal of Botany.
Chapter 4
245
Survival (%) Dry mass shoot (g)
1.4 1.2
understory
1.0 0.8 0.6 0.4 0.2
gap 0.0
Survival
Growth
Figure 4.39. Survival of seeds and growth of seedlings of Welfia georgii in understories and gaps in tropical forests in Costa Rica. Redrawn from Schupp (1995) with permission from the American Journal of Botany.
0.5
deep soil
Survival (%)
0.4
0.3
0.2
0.1
low, spiny shrubs
0.0
1st year seedlings
2nd year seedlings
Figure 4.40. Survival of seeds and growth of seedlings of Prunus mahaleb beneath spiny shrubs or in grassland (found on “deep soil”) in tropical forests in southern Spain. Reprinted from Schupp (1995) with permission from the American Journal of Botany.
246
Interaction between Competition and Facilitation
In all of these cases the habitat good for seeds was not good for seedlings, and the habitat good for seedlings was not good for seeds. Schupp (1995) examined 62 studies of habitat suitability for seeds and seedlings in the literature and found 19 studies from a wide range of habitats demonstrating seed-seedling conflicts in preferred habitats. In another example, Celtis pallida occurs almost exclusively underneath Prosopis glandulosa in savanna shrublands in southern Texas (Fulbright et al. 1997). However, relative growth rates, net assimilation rates, and seedling mass of Celtis are greater in the open than in the shade of Prosopis. Clearly the aggregation of Celtis under Prosopis is not because the nurse improves growth rates; however, just as clearly Prosopis is facilitating Celtis. Quantifying the subtle changes that may occur in plant interactions as they mature is difficult, but shifting effects of nurses from positive for the establishment of other species to competitive for those species when they mature may be quite common and important in plant communities (also see Rey and Alcantra 2000). Seed-seedling conflicts may explain “canopy filtering” effects such as described by Pugnaire and Lázaro (2000). They showed how Retama sphaerocarpa shrubs acquired increasingly more species diverse seedbanks over time; however, many species occurred as seeds under young and old shrubs, but only recruited under old shrubs. Shrub canopies had a filtering effect on the seedbanks beneath them, determining the composition of the understory community. Seed-seedling conflicts, such as described by Schupp, are due to the different kinds of tradeoffs seedlings experience under nurse plants. As stated above, facilitation may be important in conditions where the beneficial effects of a species, such as ameliorating fatal temperatures or shelter from lethal winds, outweigh the competitive effects of the species that provide the refuge. In other words, under some conditions the benefits of a neighbor for survival outweigh the costs of losing resources to that neighbor. At no time is this more evident than at the seedling stage. As noted in Chapter 2, in the Great Basin seedlings of Pinus monophylla are facilitated by Artemisia tridentata indirectly by reduction of herbivory and directly by reduced desiccation and heat stress (Phillips 1909, Everett et al. 1986, Drivas and Everett 1988, Welden and Slauson 1990, Callaway et al. 1996). The situation is complicated by the particular costs and benefits involved in being nursed. In experiments with planted seeds the survival of germinating P. monophylla seeds under shrubs was 100%, but only 6% in the open spaces surrounding the shrubs (Callaway et al. 1996). Mortality in the open appeared to be divided by ≈25% predation and ≈75% desiccation. In a second experiment with planted seedlings, survival of P. monophylla was 50% under Artemisia shrubs, 30% where shrubs had been cut, and 15% in the open spaces. In contrast,
Chapter 4
247
however, the total leaf area of P. monophylla seedlings was 30% greater, and the aboveground mass 35% greater, when growing in the open. This tradeoff appeared to be balanced toward shrub nurses in the region where the experiments where conducted because the large majority of P. monophylla seedlings were associated spatially with Artemisia canopies. A key aspect of the tradeoff, however, is that this balance can change, and in areas with more precipitation or cooler summer temperatures the benefits of shrubs for survival might be less than the costs of shrubs to growth. In other words, competitive mechanisms might outweigh facilitative mechanisms and seedlings might become spatially disassociated with shrubs. In Namaqualand of the northwestern Cape of South Africa the size and reproductive output of an annual forb, Goteria diffusa, is reduced when it grows next to the perennial shrub Leipoldtia constricta (Cunliffe et al. 1990). However, far more Goteria are found near the shrub than a meter away from the shrub. Culiffe and colleagues concluded that soils collected around the shrub provide a “safe site for germination and establishment”. This pattern suggests the interesting and very likely possibility that a shrub removal experiment with maturing annuals would find evidence for competition; whereas facilitation would have been the interaction responsible for the fundamental patterns of distribution and abundance. In Goldberg and colleague’s (Goldberg et al. 2001) giant “sandbox” experiments (described above) in the Negev Desert natural seed banks containing whole annual communities were either diluted or concentrated to establish a range of densities (1/16, 1/8, ¼, ½, 1x, 2x, 4x, and 8x natural densities). Using this range of densities they quantified density-dependent emergence, survival, and growth providing an exceptional look at variation in plant interactions at different life history stages. Emergence was negatively density dependent, suggesting unusual interactions among the seeds themselves. Seeds may have suppressed the germination of neighbors via allelochemicals leached from seed coats (Dyer et al. 2000). Furthermore, the emergence of dicots was more strongly inhibited by community density than emergence of grasses. In contrast to emergence, survival of the annuals was positively density dependent in the majority of cases. For survival, grasses showed stronger facilitative density dependent responses than dicots. Finally, the overall effect of community density on the growth of annuals was negative, indicating the importance of competition at this lifestage. As stated by Goldberg et al. (2001), “This variation in mechanism, direction, and magnitude of interactions among life history stages suggests that current models of plant community structure that are based largely on exploitation
248
Interaction between Competition and Facilitation competition as it influences growth (with mortality a simple function of growth) are inadequate for even this simple plant community.”
4.5. COMPETITIVE ADVANTAGES PROVIDED BY BENEFACTORS One species may be facilitated by another simply because the benefactor provides a favorable physical habitat for the survival, growth, and reproduction of its beneficiary. For example, I described in Chapter 2 how oak canopy creates shade in which Bromus diandrus thrives and Avena fatua declines (Mahall et al. 1981, Parker and Muller 1982). A subtle twist on this process is related to the indirect interaction described in Chapter 3.7, in which a beneficiary does not derive its primary advantage directly from ameliorated conditions, but from a competitive advantage in those conditions, which actually may be case for some species growing under oak canopies. This chapter began with a description of the effects of Quercus douglasii on understory productivity, but the effects of this tree on understory community composition are probably even stronger than on productivity (Holland 1973). In a biogeographical study of the canopy effects of Q. douglasii, McClaran and Bartolome (1989) found weak negative to neutral effects of canopies on total productivity, but certain species differed in biomass by over an order of magnitude between subcanopy and open habitats. In the coastal mountains of central California Q. douglasii may roughly double total grassland productivity, but the productivity of some species such as Bromus mollis or B. diandrus may increase by 4-7 times (Callaway et al. 1991). Ordination analyses of herbaceous communities show considerable differences in composition between understory and open (Figure 4.41). These effects appear to be produced primarily by the much higher levels of nutrients under oak canopies (see Chapter 2), but experiments by Rice and Nagy (2000) demonstrated that the causes of the distributional patterns for at least two understory species were more multifaceted. At their study sites, the distributions of two Bromus congeners, B. diandrus and B. hordeaceus, are highly discrete with respect to Q. douglasii canopy cover. They measured the cover of B. diandrus at 0.9-1.5% in the open versus 14.0-22.2% under canopies, whereas the cover of B. hordeaceus was 12.2-15.8% in the open versus 1.0-7.4% under canopies. The reproductive rates of these species followed the same pattern. They planted B. diandrus and B. hordeaceus at different intraspecifc and interspecific densities in a split-block design in order to test the relative importance of competition and the direct effects of the abiotic environment in determining these patterns. Soil nutrient differences were important, as the fertile soils from under oaks improved B. diandrus reproductive rate but not that of
Chapter 4
249
B. hordeaceus. However, they also found that the low abundance of B. hordeaceus under canopies was not simply due to the environment created by the canopy, but because of competitive exclusion by B. diandrus, as shown by significantly decreasing reproductive rate with increased B. diandrus density. Bromus diandrus, on the other hand, did not appear to be significantly influenced by competition from B. hordeaceus, suggesting that its close association with oak canopies is due in part to its inability to adapt to lower nutrients in the soils of the open grassland. In sum, Q. douglasii appears to create understory environments that favor B. diandrus. Even though understory environments may also be favorable to other species such as B. hordeaceus, B. diandrus maintains its dominance by outcompeting them. The evergreen oak, Quercus agrifolia, affects understory community composition in ways that appear to be very similar to Q. douglasii; however the 350
300
DCA Axis 2
250
200
150
100
50
0 0
50
100
150
200
250
300
350
DCA Axis 1 Figure 4.41. Ordination of vegetation samples (0.125 m2 quadrats) collected under the canopies of Quercus douglasii and in the open grassland in central California. Large symbols show the means and 95% C.I.s for each group of samples. Solid circles represent understory samples and open circles represent open grassland.
250
Interaction between Competition and Facilitation
mechanisms that drive the differences may be completely different. In the autumn, when most herbaceous plants in California are at the seed stage in the soil, Marañón and Bartolome (1993) reciprocally transplanted soil blocks between Q. agrifolia understories and open grassland. After the onset of the winter rains in the first year plant communities in these soil blocks were similar to those from where the blocks were collected – there were few changes. In the second year shifts in species composition were substantial, but in contrast to the findings of Rice and Nagy, competition was predominant in the open and not under canopies. Deep shade under the evergreen Q. agrifolia caused high seedling mortality of the transplanted species from the open - a direct negative effect of the overstory species. In the open, however, it was the competitive effects of the dominant grasses that appeared to limit the establishment of the species from the understory. The striking differences in the important mechanisms between evergreen Q. agrifolia and winter-deciduous Q. douglasii canopies illustrate the subtle complexities that determine the strength and direction of one species’ total effect on another species. The canopies of both oaks provide shade and increase soil nutrients, but to different degrees. Quercus douglasii does not limit light during the growing season of the understory (its leafless canopies reduce PAR by ≈10% in the winter), but provides much more fertile soils than the open grassland. Quercus agrifolia also provides fertile soil, but reduces PAR during the same time by over 90%. Because plants don’t grow well in the dark regardless of how much water and nutrients they get, Q. agrifolia creates stressful conditions where only tolerant species survive. Fertile and well-lit conditions under Q. douglasii, however, create an environment where species can perform maximally – and these species compete intensely in this environment. Understanding how vegetationdriven shifts in different abiotic conditions (such as light and soil nutrients) can radically alter the balance of plant interactions may provide a conceptual link to foundational studies of similar processes on gradients (Tilman 1991b).
4.6. INDIRECT EFFECTS AND THE BALANCE OF COMPETITION AND FACILITATION
As described throughout this chapter, the total effect of one species on another is often a baffling mix of competitive and facilitative mechanisms. Sometimes these mechanisms are direct, and combinations of direct interactions have been the focus of much this chapter. An example of combined direct interactions is when shade from an overstory tree benefits species beneath it, but the tree’s roots compete with the same understory species for nitrogen. But sometimes the competitive and facilitative mechanisms that determine total interactive
Chapter 4
251
effects are a mixture of direct and indirect effects, for example when a welldefended shrub protects seedlings of other species in its canopy, but also competes with those seedlings for water. In this second example the facilitative interaction is indirect, and mediated by herbivores. In fact some of the best examples of tradeoffs between the negative and positive effects are when positive effects are due to the protection that one species provides another from herbivores. Jonathan Levine found that large tussocks of Carex nudata provide critical stable substrate for many other species during the annual winter floods that scour the floodplains of the Eel River in northern California (Levine 2000). He also found that C. nudata protected Mimulus guttatus from herbivory by insect larvae, and Epipactis gigantea from deer, apparently reducing herbivory by more than 75%. In two years of experiments, nearly all Mimulus guttatus transplants in streambeds without C. nudata were completely defoliated by insects. However, these strong direct and indirect benefits did not come without costs. In order to assess the costs of growing with C. nudata, the performance of six transplanted species was followed in streambeds and in C. nudata tussocks where the leaves had either been pinned back to allow more light, thinned to allow more light and to reduce the use of other resources, completely clipped, or left in dense thatches. Competition during the growing season reduced the biomass of five of the six species by over 50% and reproductive performance by 60%. The most dramatic effects of competition were for Mimulus guttatus which grew up to 10 times larger and produced five times as many flowers when C. nudata tussocks were clipped. The moss Bracythecium frigidum was not affected by competition. In general the competitive effects of clipped C. nudata tussocks and tussocks with leaves pinned back to were similar, indicating that belowground competition was weak and that the primary direct competitive effects were for light. In experiments using cages to prevent deer browsing, Levine found that the positive effects of the indirect associational defense were equal in magnitude to the competitive effects of C. nudata on Epipactus. In other experiments he hand-picked insect larvae off of plants in his different treatments and found that protection from these herbivores in the open streambed resulted in a mean plant biomass for all species of 2.17±0.4 g (1 SE) versus 0.19±0.07 g for controls left with unmanipulated herbivores. In the C. nudata tussocks, manipulations of insect herbivory had no effect on plant biomass. This study provides a remarkable example of the multiple mechanisms driving the effects of one crucial species, Carex nudata, in streambed plant communities. Some mechanisms operated simultaneously, others were separated seasonally. During the winter floods, C. nudata protected all species examined from physical disturbance and reduced mortality, which ranged from 70 to 100 percent outside of the tussocks and 0-60% within the tussocks. During the summer the competitive effects of C. nudata reduced plant biomass and
252
Interaction between Competition and Facilitation
reproduction for all species but one, the highly shade tolerant moss. Also during the summer, C. nudata provided protection from generalist vertebrate and specialist invertebrate herbivores. A similar combination of facilitative and competitive interactions has been described for Spartina alternifolia and associated plant species in New England salt marshes. van de Koppel et al. (2006) found that Spartina facilitates a suite of forbs in the upper intertidal zones by stabilizing the shoreline environment. However, transplant experiments found that these forb species grew better when Spartina was removed, showing that competition was very important once the initial beneficial effect had been established. As described in Chapter 3, Cassia biflora shrubs benefit from being in conspecific clumps (positive density-dependence) because of increased pollinator visitation, seed set, and reproductive success (Silander 1978). However, this indirect facilitative mechanism exists in balance with indirect competitive mechanisms, those involving negative density-dependent effects of herbivores. Silander found that the proportion of Cassia seed pods that escaped bruchid beetle infestation increased with distance to the nearest conspecific neighbor, creating a situation in which the positive effects of density on seed set could be offset by the negative effects of density on seed mortality. However, further investigation indicated that clumping did not result in a higher loss of individual seeds to beetle attack, and Silander concluded that the overall indirect effect of clumping on reproductive success was positive. A similar balance between the indirect effects of neighbors on pollinators and herbivores has been demonstrated by Petit and Dickson (2005) for Xanthorrhoea semiplana (grass tree) and Caladenia behrii, an endangered orchid. They found that Xanthorrhoea protected the orchid from kangaroo herbivory, but at the cost of reduced herbivory. Hay (1986) found that several species of highly palatable seaweeds on the east coast of North America appeared to avoid local extinction by finding shelter from herbivores in the leaves and stems of unpalatable seaweed species. However, this facilitative effect came at a cost because he also found that the palatable species were competitively inferior to their unpalatable hosts; when herbivores were excluded the palatable seaweed species grew 14% to 19% less when in mixtures with superior competitors than when alone. The facilitative mechanism was overwhelmingly important; however, as in the presence of herbivores palatable species survived only when mixed with the competitively superior, but unpalatable species. Clearly, competitive and facilitative indirect effects co-occur in complex combinations, perhaps even more complex than direct competitive and facilitative effects.
Chapter 4
253
4.7. POLLUTION AND SHIFTS IN FACILITATION AND COMPETITION Zvereva and Kozlov (2004) from the University of Turku in Finland investigated the hypothesis that the importance of facilitative plant interactions will increase with environmental stress imposed by long-term aerial pollution. They compared the effects of canopy cover on the growth and reproduction of four ericacous shrubs, Vaccinium myrtillus, V. uliginosum, V. vitis-idaea, and Empetrum nigruma spp. hermaphroditum in industrial barrens and versus unpolluted forests in the Kola Peninsula of Northwest Russia. In unpolluted forests, the relative neighbor effect (RNE) of canopy species was always competitive, suppressing the reproductive output of the shrubs. However, in polluted barrens RNE was facilitative in three of the four cases and significantly less competitive in the fourth. While these patterns cannot exclude the possibility of confounding site effects, they raise a fascinating idea that facilitative interactions may ameliorate forms of stress caused by human impacts.
4.8. CONCLUSION A large number of recent studies indicate that facilitating and interfering mechanisms operate simultaneously and shift in importance as abiotic conditions change. Such conditionality of interactive mechanisms should not be surprising because virtually all other ecological interactions including mutualisms (Bronstein 1994, Gehring and Whitham 1995), competition (Pennings and Callaway 1992), herbivory (Maschinski and Whitham 1989), predation (Martin 2006), parasitism (Gibson and Watkinson 1992, Pennings and Callaway 1996), and allelopathy (Tang et al. 1995) are conditional. The conceptual model of Bertness and Callaway (1994) in which the relative importance of competition and facilitation are proposed to vary inversely along gradients of abiotic stress has been supported by most studies; however, some studies contradict the idea. It is very important to remember that at its most basic level facilitation cannot occur unless some stressful biotic or abiotic factor exists for a neighbor to ameliorate. If the growth of a plant is not limited by something, there can be no facilitation. The basic nature of competition, by definition, is a struggle to preempt limiting resources such as light, water, and nutrients that determine rates of carbon acquisition. When abiotic conditions are benign, rapid resource acquisition is possible and therefore competition will be intense. When abiotic conditions are severe enough to limit plant growth independently of limited resources, amelioration
254
Interaction between Competition and Facilitation
of severe stress by a neighbor should favor growth more than competition for resources with the same neighbor will reduce growth. This become even more apparent when the species-specificity of facilitative interactions is considered; which I will do in the next chapter.
CHAPTER 5 SPECIES-SPECIFIC POSITIVE INTERACTIONS
Whether or not benefactor species are species-specific in their effects is crucial for deciding how we should integrate facilitation into our concept of the plant community. Is a facilitator no different than shade cloth? Or can facilitation depend on specific traits of particular facilitator species, with some species eliciting strong positive effects and other similar species producing neutral or negative effects? Can narrow species traits such as root architecture, biochemistry, throughfall nutrient concentration, or light attenuation make some species better facilitators than others? Beneficiaries certainly can respond in species-specific ways, but to understand facilitation an appreciation of species-specific effects is much more important. Some species-specific positive effects are obvious; a species that is not aerenchymous cannot oxygenate soil, not all species are hydraulic lifters, species with poor defenses against herbivores cannot defend their neighbors, and non-mycorrhizal species cannot exchange materials through shared fungal networks. There is a great deal of evidence for subtle and fascinating species-specific facilitative effects, and I focus primarily on these effects in this chapter. The species-specificity of positive interactions is important for our general concepts of plant communities (see Gleason 1926, Goodall 1963, Shipley and Keddy 1987, Austin 1990, Collins et al. 1993, Callaway 1997, Lortie et al. 2004). But even if facilitation is not species-specific its importance is not negated; even broad and non-species-specific facilitative interactions create some degree of interdependence in plant communities. But if positive interactions are often species-specific then interdependence in plant communities would be more intense, the potential for interactions to drive biological diversity in communities would be much higher, and the mechanisms determining interactive outcomes would be more intricate and ecologically interesting (see Chapter 6). In this chapter I explore the specificity of positive interactions among plants by exploring the literature with a focus on the following general questions: 1) Are beneficiary species non-randomly associated with potential benefactors? 2) Are positive mechanisms produced by species-specific plant traits? 3) Can potential benefactors have similar positive effects, but different negative effects? 255
256
Species-Specific Positive Interactions
5.1. ARE BENEFICIARY SPECIES NON-RANDOMLY ASSOCIATED WITH POTENTIAL BENEFACTORS? Clearly, there is specificity to some positive associations simply as a consequence of differences in plant morphology and size. It would be difficult for desert ephemerals to shade columnar cacti. However, there are many examples of species-specific spatial associations even among benefactors that are similar in morphology. In 1955, J.D. Ovington published a striking, yet rarely cited, example of species-specific effects by different facilitator species on the productivity of their neighbors. He measured understory biomass and species composition beneath different tree species in forests that had been planted by the British Forestry Commission. Because the Forestry Commission planted large blocks of different tree species without regard to microhabitat, this ‘haphazard’ approach had the distinct advantage of measuring canopy effects that were not confounded by microsite effects. He found a tremendous range in understory productivity associated with different species, ranging from 25,196 kg ha-1 under Pinus sylvestris, 4,788 kg ha-1 under Quercus robur, to 0 kg ha-1 under Picea abies and Abies grandis. Understory species diversity did not correlate with productivity, and ranged from 1 species under A. grandis, 9 species under P. abies, 16 species under Q. robur, to 34 species under Pseudotsuga taxifolia. A perusal of Ovington’s list of species frequencies under different trees types shows that a number of different species are found to be consistently associated with some overstory species and never with others. In similar early study, Bray (1955) found that the species composition of the understory was correlated with the amount of shade (see below) cast by particular overstory species in Wisconsin savannas. He found that different understory species appeared to have three general patterns in response to canopy shade. Some species peaked in deep shade, others under canopies producing intermediate shade, and others in the full sunlight. While Bray did not specifically address species-specific effects of the overstory trees, the general responses he documented for understory species distributions have important implications for species specific facilitation. Jans-Christian Svenning and Flemming Skov (2002) described the distribution of sixty species in a temperate forest in Denmark with respect to many abiotic habitat factors and tree cover types. In ordination analyses the cover of Quercus-Alnus, Fagus, or conifer cover types were generally more strongly correlated with the ordination axes than abiotic factors such as aspect, altitude, slope, soil fertility, clay content, drainage, or soil water. Furthermore, the particular species forming the canopy was highly correlated with the community composition of the understory. Five understory species were associated strongly with Fagus stands, 10 species were positively associated with Quercus-Alnus stands, and seven species were positively associated with conifers.
Chapter 5
257
One of the first, and still most widely publicized (see Chapters 2 & 3), examples of facilitation is that involving Carnegia gigantea (saguaro) in Sonoran Desert communities. Shreve (1931) observed that saguaro seedlings were common under desert shrubs and trees, but not in open spaces. Patterns of positive spatial associations between the seedlings of one species and sheltering adults of another species became widely referred to as the “nurse plant syndrome” because of the work of Niering et al. (1963), Turner et al. (1966, 1969), and Steenberg and Lowe (1969, 1977). They found that saguaro seedlings were commonly sheltered by many different species of perennial plants, but predominantly by Cercidium microphyllum. Such spatial patterns suggestive of nurse plant relationships have since been reported for many species in arid and semi-arid environments around the world (Yeaton 1978, McAuliffe 1986, Franco and Nobel 1989, Yeaton and Elser 1990, ValienteBanuet et al. 1991, Arriaga et al. 1993, Flores-Martinez et al. 1994, Callaway 1995, Pugnaire et al. 1996ab, Suzán et al. 1996). Hutto et al. (1986) also reported that saguaros were distributed nonrandomly among potential nurse plants at two locations in Organ Pipe National Monument in the Sonoran Desert. They found that saguaros were far more abundant under shrubs and trees than in the open, but significantly more saguaros were associated with C. microphyllum trees and fewer saguaros were associated with Larrea tridentata than expected based on the proportional cover of these species (Figure 5.1). This preference is not geographically consistent. In deserts in central Argentina, Larrea divaricata more strongly nurses the columnar cactus Trichocerus pasacana than other shrubs despite large numbers of seeds in the soil beneath all shrub species (de Viana et al. 2001). Not all research on saguaro has found significant spatial associations between saguaro and specific nurse species (Steenberg and Lowe 1969), but it appears that in North and South American desert systems with large columnar cacti, some shrub species are better benefactors than others. 50
associated saguaros cover
Proportions
40
30
20
10
0
Ambrosia
Larrea Cercidium
Lycium
other
Figure 5.1. Proportional saguaro numbers and cover of nurse species in the Sonoran desert. Redrawn from data in Hutto et al. (1986) with permission from the Southwestern Naturalist.
Species-Specific Positive Interactions
258
Species-specificity in nurse effects on saguaros may be caused by the effects of the nurse on soil quality. Turner et al. (1966) found that saguaro seedlings survived better on soil collected from under Cercidium trees (the best nurse in Hutto et al.’s 1986 study) than on soils from under either Prosopis or Olneya concentrations (Figure 5.2). There may have been nutrient effects but soil albedo had strong effects on temperature, with dark soils from Prosopis and Olneya creating hotter conditions which corresponded with greater saguaro mortality. Carrillo-Garcia et al. (2000a) eliminated albedo and temperature effects by planting Pachycereus pringlei (cardon), a giant columnar cactus similar to saguaro, in soils from under old and young Prosopis and old Olneya. They found that soil from under old Prosopis trees increased Pachycerus survival and more than doubled biomass accumulation in comparison to soil from young Prosopis or Olnyea tesota, but soil from beneath all trees was superior for Pachycerus growth than soil from the open (Figure 5.3). These species-specific soil effects appear to have been caused by increased nutrients because N and P were generally higher in understory soils than in the open, but interestingly, soil that promoted the greatest growth was not the most nutrient rich, suggesting that soil biota may have been important for the facilitative effect.
100
Survival (%)
80
60 Cercidium, shade Prosopis, shade Cercidium, no shade Prosopis, no shade
40
20
12, '59 N ov
Feb 10, '59
8 2 , '5 July
'5 8 31, Jan
July
15, '57
0
Figure 5.2. Survival of saguaro cacti seedlings planted outdoors in shade or not, and in soils from different nurse species. Redrawn by estimating coordinates from Turner et al. (1966) with permission from Ecology.
259
Chapter 5
Total biomass (g)
5
a
4
3
b 2
c
1
0
d old young old Prosopis Prosopis Olneya
bare soil
Figure 5.3. Growth of Pachycerus pringeli in soil from beneath different nurse plants in Mexico. Different letters designate significant differences based on post-ANOVA contrasts. Redrawn from data in Carrillo-Garcia et al. (2000a) with permission from Restoration Ecology.
McAuliffe (1986) found that one of the best nurses for saguaro, Cercidium microflorum, itself benefits from species-specific facilitation. He found proportionally more Cercidium seedlings under Ambrosia than other shrub species, but this positive interaction was indirect (see Chapter 3, pages 124-125). The positive affect of Ambrosia on Cercidium was because the former’s dense low lying canopy provided excellent shelter from rabbits. McAuliffe (1988) also found strong, species-specific patterns of association in other more complex Sonoran Desert communities and in Mojave Desert plant communities. In both regions, seedlings and saplings of many species were much more frequently under Ambrosia dumosa canopies than under other shrubs. In similar communities in the central mountains of Baja California, young alluvial terraces are co-dominated by a rich assemblage of shrubs and cacti including the important nurse plant Viguiera laciniata and Larrea tridentata, which is generally a poor nurse plant (Figure 5.1, also see page 287), is absent (McAuliffe 1991, pers. comm.). On older, adjacent terraces Larrea is dominant, the widely preferred nurse plant, Viguiera, is absent, and there is no regeneration of Idria columnaris and Pachycerus pringli, which are the species primarily associated with the good nurse Viguiera. These correlations suggest that the preference of beneficiary species for particular nurse species can have very strong effects on desert community composition. Suzán et al. (1996) identified many aborescent, shrub, and cacti species that were highly associated with Olneya tesota (ironwood) in the Sonoran Desert and argued that, as a habitat modifier, Olneya is a “keystone species” for regional biodiversity (also see Burqúez and Quintana 1994). There are many other tree species in the system, but they do not play the same facilitative role. Suzán et al. described 30 species as “shade dependent”, with five preferring
260
Species-Specific Positive Interactions
Cercidium species, four preferring Prosopis species, and 22 preferring Olneya tesota. As in other studies described above, Franco and Nobel (1989) found that most saguaro seedlings in their Sonoran Desert study sites were associated with paloverde trees. But in contrast, a second cactus species, the exceptionally heattolerant Ferocactus acanthodes, was preferentially associated with a bunchgrass, Hilaris rigida. Valiente-Banuet et al. (1991) also found similar disproportional associations among many potential nurse plants and five different species of cacti in central Mexico. The striking species-specific effects of different desert shrubs on the recruitment of other perennials can also be seen in the effects of these shrubs on annual species. Duncan Patten (1978) measured the production of annuals under three different woody perennial species over two years in the Sonoran Desert and found that Cercidium microphyllum, the preferred benefactor for many perennials (also see Chapter 6.6), was associated with much higher productivity that either Ambrosia dumosa or Larrea tridentata. Not only did productivity differ strongly among different shrub species, community composition differed among microsites beneath shrubs. Festuca octoflora, Poa biglovei, and Eucrypta chrysanthemefolia were important components beneath Cercidium, whereas Filago spp., Pectocarya ssp., Schismus barbatus were more common beneath Ambrosia and Larrea. Species-specific patterns of annuals growing under different shrubs were quantified long before Patten’s study. In 1942, Went counted annuals under hundreds of shrubs in the Mojave Desert and found that a number of annual species were found almost exclusively beneath canopies. However, not all shrubs performed this facilitative function to the same degree. Encelia farinosa and Ambrosia (Franseria in Went’s day) dumosa, although similar in size and canopy density, were “very different in their aptitude for harboring annuals”. After analyzing shrub-annual spatial relationships he concluded that “This proves again, that although the presence of shrubs as such is essential for the occurrence of a number of desert annuals, there is also a strong specificity on the part of certain annuals for certain shrubs.” On barren volcanic soils in southern Idaho, two early colonizing species of Eriogonum differ strikingly in their positive associations with other species that appear to regenerate underneath them (Day and Wright 1989). Five other species were found to be strongly associated with Eriogonum ovalifolium, but no species were consistently associated with Eriogonum umbellatum. These differences were attributed to the different aboveground architecture of the two Eriogonum species. In Californian shrubland and woodland, Quercus agrifolia and Q. douglasii seedlings are much more common under shrubs than in open grassland, and experimental manipulations have demonstrated the importance of nurse shrubs for the survival of both species (Callaway and D’Antonio 1991, Callaway 1992). However, not all shrubs have the same positive effects on Q. agrifolia. In experiments, 43% of germinating seedlings survived under Ericameria ericoides, 34% under Artemisia californica, 5% under Mimulus
261
Chapter 5
auranticus, and 0% under Lupinus chamissonis (Callaway and D’Antonio 1991). In another experimental planting in tropical savannas in Belize, Kellman (1985) found that the survival of Xylopia frutescens (a forest tree) was five times higher under Miconia albicans than beneath four other woody trees or shrubs. His data suggested that potassium and phosphorus nutrition may have been more favorable under Miconia than the other species. Slocum (2001) did not conduct experiments, but found that the species assemblage of woody recruits in neotropical pastures depended on the tree species under which they grew. In general, nitrogen-fixing species are typically benefactors and nonfixers are typically beneficiaries. The nitrogen-fixing Lupinus arboreus does facilitate other species, but Jennifer Rudgers and John Maron (2003) found that seedlings of this shrub were themselves facilitated by the non-fixer, Baccharis pilularis, in coastal sand dunes in central California. However, not just any ecotype of Baccharis will do; only a prostrate architectural form of this species is positively associated with Lupinus (Figure 5.4). Prostrate Baccharis shrubs enhance seedling germination, survival, and growth of Lupinus. Apparently, low lying branches and leaves trap more moisture from condensation and fog and provide more consistent shade than the more erect ecotypes. Regardless of the mechanism, the preference of a beneficiary for a particular ecotype is a good example of how specific facilitative interactions can be. Similar ecotypespecificity in facilitative effect has been demonstrated for Spartina alterniflora (Proffitt et al. 2005). 40
X2=0.97 P=0.97
A 2
X =1.40 P=0.95
Frequency (%)
30 20
X2=12.69 P=0.0004
X2=1.05 P=0.99
10 0 40
observed expected
X2=1.57 P=0.54 X2=0.00 P=0.99
30 20
B
X2=17.63 P=0.0001 X2=2.55 P=0.33
10 0
A. arenaria
exposed sand
B. pilularis upright
B. pilularis prostrate
Figure 5.4. Observed and expected frequencies of L. arboreus in four microhabitats for (A) Lupinus arboreus seedlings and (B) L. arboreus adults. Redrawn from Rudgers and Maron (2003) with permission from Oikos.
262
Species-Specific Positive Interactions
In some cases, morphologically and physiologically similar beneficiary species cannot utilize the same nurse plant. In many parts of the Great Basin, Artemisia tridentata acts as a nurse plant for Pinus monophylla (Everett et al. 1986, Callaway et al. 1996). In contrast, when Pinus ponderosa (a congener with a similarly high water-use-efficiency) occurs in the same region as A. tridentata, this species is often restricted to outcrops of low-phosphorus and acidic soils where A. tridentata cannot grow (Billings 1950, DeLucia et al. 1988, Callaway et al. 1996). Experiments show that A. tridentata competitively excludes P. ponderosa from higher quality soils, but A. tridentata facilitates P. monophylla in the same system (DeLucia et al. 1988, Callaway et al. 1996). The species-specific mechanisms determining these two pine species respond to A. tridentata are not known, but they must be subtle. The ultimate community-scale effect of the contrasting facilitating and competing effects of A. tridentata on P. monophylla and P. ponderosa is not subtle. Normal desert soils support abundant populations of A. tridentata and P. monophylla, but the distribution of these species ends immediately and that of P. ponderosa begins where patches of hydrothermally altered soils occur. Many other species show a similar discontinuity in their distributions (E. DeLucia and R.M. Callaway, unpublished data). The abrupt border for the conifers is determined to a large degree by the competitive effects of A. tridentata on P. ponderosa and the facilitative effects of the shrub on P. monophylla. In the greenhouse P. monophylla grows just as well on altered soils as P. ponderosa (DeLucia et al. 1989), and P. monophylla seedlings are abundant on the altered soils (Callaway et al. 1996). However there are no Artemisia nurse plants and therefore no mature P. monophylla on altered soils. To most of us, all bryophytes look alike. However, morphological and physiological variation among bryophytes creates opportunities for speciesspecific facilitation because aggregation among species with different traits can alleviate moisture deficits for the more sensitive species (Bates 1988, Kosiba and Sarosiek 1993, Mulder et al. 2001, Fenton and Bergeron 2006). Zamfir and Goldberg (2000) explored the effects of initial density on interactions among seven bryophytes species using performance at different densities in monocultures and mixtures of equal proportions of all species at low densities. They did not find strong effects of initial density, but they did find large differences in interactive effects among bryophyte species, ranging from strongly positive to strongly negative (Figure 5.5). For proportional growth, three of the seven species showed facilitative effects (negative Relative Competition Intensity, RCI) at low initial community abundances, but all bryophyte species had competitive effects at high initial abundances. However, when biomass was used to calculate RCI, three of the four species had positive effects. These positive effects among bryophytes are thought to come from enhancing the whole-community boundary layer, maintaining high humidity around transpiring tissues, but different species had strikingly different interactive effects. Mulder et al. (2001, see Chapter 6.2) also explored interactions among bryophytes in replacement experiments and found that
263
Chapter 5
facilitative effects were more evident when conditions were harsher and drier. They also found that some species had stronger facilitative effects than others and also hypothesized that particular architectures created more complex community boundary layers in diverse communities. Alternatively, taller bryophytes may have protected shorter species from photoinhibition. Not only do bryophytes have species-specific effects on each other, different herbaceous vascular plant species have species-specific effects on bryophytes. Ingerpuu et al. (2005) tested the effects of Trifolium pratense, Festuca pratensis, Prunella vulgaris on the cover of two bryophyte species, Rhytidiadelphus squarrosus and Brachythecium rutabulum, and found Trifolium was an excellent facilitator of the mosses, Festuca had moderate positive effects, and Prunella did not facilitate at all. 0.8 0.6 Dic Hom Hyl Hyp Rac Rht Rhy
RCIG
0.4 0.2 0.0 -0.2 -0.4 0.6 0.4
RCIR
0.2 0.0
-0.2 -0.4 -0.6 21
35
49
70
84
98
140
Initial community abundance (mg) Figure 5.5. (a) Relative competition intensity based on proportional growth (RCIG), and (b) relative biomass on a gradient in initial community abundance ((RCIB). Species: Dic = Dicranum scoparium; Hom – Homalothecium lutescens; Hyl- Hylocomium splendens; Hyp – Hypnum cupressiforme; Rac – Racomitrium canescens; Rhy – Rytidiadelphus triquetrus; Rht – Ryytidium rugosum. From Zamfir and Goldberg (2000) with permission from the Journal of Ecology.
264
Species-Specific Positive Interactions
Many authors have shown that terrestrial vascular epiphyte-host associations do not occur at random in natural communities (see Chapter 2, pages 49-50, 83-84, Went 1940, Johansson 1974, Benzing 1981, Bennett 1986, ter Steege and Cornelissen 1989, Migenis and Ackerman 1993, Dejean et al. 1995, Kernan and Fowler 1995). Went (1940) noted that “certain epiphytes are found almost exclusively on certain host trees. This relationship is so constant that the trees can be identified by the epiphyte community which they harbor”. Other more quantitative studies have not found such cut and dry patterns, but many studies have recorded strong association and disassociation among epiphytes and hosts. Others have shown that some apparently suitable host species are virtually unoccupied by some species of epiphytes (Schlesinger and Marks 1977). Like vascular epiphytes, patterns of liana host specificity have been described in tropical and temperate systems (Daniels and Lawton 1991, Talley et al. 1996ab, Muñoz et al. 2003). Nonvascular epiphyte species also correlate with particular tree species, apparently due to factors such as bark pH, moisture, and roughness (Hale 1952, 1965, Callaway et al. 2002). On Sapelo Island off the coast of Georgia, USA, Steve Pennings and I observed that two common vascular epiphytes Tillandsia usneoides and Polypodium polypodioides, were much more common on some host species than others, and both epiphytes are the most common in the same three host species – Quercus virginiana, Celtis laevigata (hackberry), and Juniperus virginiana (Callaway et al. 2002). This host specificity appeared to be, at least in part, due to how differences in how host trees were suited for capturing dispersing seeds and vegetative fragments. However, when both epiphyte species were transplanted into the canopies of ten different host species, which varied in natural epiphyte abundances, the growth rates of the transplanted epiphytes correlated strongly with the epiphyte abundance rank of the host tree species. We tested several mechanisms including nutrients in throughfall (see Chapter 2, Figure 2.8), but found that moisture-holding characteristics of bark best correlated with host preference. Similar host preferences have been described for Ficus crassiuscula, a montane species of strangler fig common in Costa Rica. The juvenile form of this species, which develops as a vine, is associated evenly among potential host species in proportion to their abundance in the community (Daniels and Lawton 1991). However, mature F. crassiuscula are not distributed randomly, and are four times more abundant on Guarea tuisna, twice as abundant on Sapium pachystachys, and one-fifth as abundant on Ocotea spp. and a Conostegia sp. in proportion to host species stem densities. In deciduous forest of Alabama in the southeastern US, Talley et al. (1996a) integrated spatial patterns of tree hosts and a liana with experiments using leachates from the host. They found that the vines of Rhus radicans, the climbing liana commonly known as poison ivy, are more common on some host trees species than on others. Carya ovata and “hardbarked oaks” support twice as many Rhus lianas as expected from their stem densities and Acer sacharum, Quercus muehlenbergii, Sassafras albidum, and Juglans nigra support fewer
Chapter 5
265
than expected. Host size is important, but does not fully explain the preference of Rhus for Carya ovata. Experiments with Rhus seed germination demonstrated that bark extracts from various hosts all decreased germination relative to controls, but that the extracts from avoided hosts had stronger negative effects than extracts from preferred hosts. No positive effect of host chemistry on Rhus was found, suggesting that host preference may have been determined, in part by interactions between interfering and facilitating mechanisms (see Chapter 4). Given the option, Rhus may prefer telephone poles for its “beneficiaries”, but in the natural communities where it is found its abundance and distribution appears to be dependent on other species in the community - and on some species more than others. In a second paper Talley and colleagues (1996b) examined hostspecific relationships between two climbing vine species and various host trees in North Queensland, Australia. She found that that Freycintia excelsia and Piper caninum exhibited host associations and examined bark characteristics, allelopathy, and morphological features for causes of the host specificity. The relatively low abundance of lianas on some species, such as Austromyrtus shepherdii, was correlated with thin, peeling bark. Another species with low liana abundance, Alphitonia petriei, was found to have high levels of extractable phytochemicals in the bark. Unlike the results of Callaway et al. 2002, bark water-holding capacity was not as important to lianas as a rough stable surface. To my knowledge, the most ambitious and convincing study for the broad importance of facilitation in a particular system was conducted by Lorena Gómez-Aparicio and colleagues (2004, also see Gómez-Aparicio et al. 2005ac, 2006) at the University of Granada in southern Spain. In an effort to determine the general importance of shrubs as facilitators of forest regeneration they planted over 18,000 seedlings of 11 different woody species under 16 species of shrubs at seven study sites. Of their 146 different experimental cases, over 80% showed facilitation for survival and 76% showed facilitation for growth. However, this general effect showed a great deal of species-specific variation for growth and survival (Figures 5.6 & 5.7). For both survival (Figure 5.6) and growth (Figure 5.7), shrubby legumes were the best beneficiaries and rockroses (Cistus) were the worst. Across all combinations of nurse and beneficiary species temperatures were lower under shrubs and soil moisture was higher, leading Gómez-Aparicio et al. to conclude that this was the likely facilitative mechanism. However, it was not clear if variation in soil moisture among nurse species may have contributed to species-specificity in facilitative effects. The strong negative effects of the Cistus species (rockroses) were thought to have been due to the negative effects of allelopathic leachates (see Chapter 4).
266
Species-Specific Positive Interactions
3
3.0
Effect size (d+)
2.5
Target species P<0.00001
Nurse shrubs P<0.00001
2
2.0
1
1.5 0 1.0 -1
0.5
-2
0.0 -0.5
Evergreen Deciduous Pines
Shrubs
-3
Legumes
Small shrubs
Spiny
Rockroses
Figure 5.6. Effect size for survival of categories of different target species for all nurse shrubs combined (a) and for categories of nurse shrubs for all beneficiaries combined (b). Values reported are the mean effect size (d1) and the 95% CI. Significance (P) is for the Q statistic for the difference between groups in the effect of nurse shrubs. Reprinted from Gómez-Aparicio et al. (2004) with permission from Ecological Applications. 1.5
1.5
Nurse shrubs P<0.00001
Effect size (d+)
Target species P<0.00001 1.0
1.0
0.5
0.5
0.0
0.0
-0.5
Evergreen Deciduous Pines
Shrubs
-0.5
Legumes
Small shrubs
Spiny
Rockroses
Figure 5.7. Effect size for growth of categories of different target species for all nurse shrubs combined (a) and for categories of nurse shrubs for all beneficiaries combined (b). Values reported are the mean effect size (d1) and the 95% CI. Significance (P) is for the Q statistic for the difference between groups in the effect of nurse shrubs. Reprinted from Gomez-Aparicio et al. (2004) with permission from Ecological Applications.
Chapter 5
267
5.2. WHAT MECHANISMS CAUSE SPECIES-SPECIFIC FACILITATION? 5.2.1. Species-specific direct effects 5.2.1.1. Shade As described in detail in Chapter 2, the canopies of different species transmit very different quantities of light, which establishes the potential for preferential relationships between different canopy species and understory species that are best adapted to particular light environments. Charles Canham at the Institute for Ecosystem Studies in New York has produced a substantial body of work on how canopy species affect light and how understory species respond to variation in light. Canham and colleagues used fish-eye photography and light sensors to quantify photosynthetically active radiation (PAR) and the importance of sunflecks among six different canopy species in a mature hardwood forest. They found substantial variation among canopy tree species in their effects on light transmission (Figure 5.8) and that this variation correlated with survivorship of sapling species (also see Veblen et al. 1977, Kobe et al. 1995). In general, saplings performed best in the light conditions created by conspecifics. Their results provided clear evidence for how the interactions driving “secondary succession [are] driven by interspecific differences in resource uptake and tolerance”… and it might be added, also by species-specific differences in resource availability caused by different canopy species. In a second study, Canham et al. (1999) compared light transmission through the canopies of different conifer species in a cedar-hemlock forest in
6
7
PAR (% full sunlight) Sunfleck PAR (%)
6
5
5
4
4
3
3
2
2
1
1
0
Tsuga Acer Acer Fagus canadensis grandifolia saccharum rubrum
Quercus rubra
Fraxinus americana
Sunfleck PAR (% x 10)
PAR (% full sunlight)
7
0
Figure 5.8. Percent transmission of PAR through the canopies of tree species in the northeastern US and percentage of understory sunlight received as sunflecks (>100 μmol m-2s-1). Redrawn from Canham et al. (1994) with permission from the Canadian Journal of Forest Research.
268
Species-Specific Positive Interactions
British Columbia, Canada. Again they found substantial variation in “gap light index” among stands dominant by tree species (Figure 5.9), but less than interspecific variation in hardwood forests.
30 25 20 15 10
Betula tremuloides
Pinus contorta
Populus tremuloides
Picea (hybrid)
Abies lasiocarpa
0
Thuja plicata
5 Tsuga heterophylla
Gap Light Index (%)
35
Figure 5.9. Means and 95% confidence intervals for understory GLI (% of full sun) in stands dominated by the designated species. Redrawn from Canham et al. (1999) with permission from the Canadian Journal of Forest Research.
It is difficult to find studies in which variation among species in the shade cast by their canopies has been explicitly linked to facilitative effects. However, in one case, Jones (1995) measured the effects of different shrub species on soil temperatures and then compared the facilitative effects of the shrubs species on the recruitment of Pseudotsuga menziesii. He found that the shrub species with strong facilitative effects had lower soil temperatures beneath their canopies. For example, Ceanothus velutinus had the most Pseudotsuga recruitment and also had one of the shadiest canopies. GómezAparicio et al. (2005c, 2006) also argued that species-specific facilitative effects of shrubs on seedlings of different tree species were due in part to differential modification of light availability. In northern Botswana, Colophospermum mopane (mopane) and Acacia erioloba are common dominants in savanna woodlands. With Angus Beal, I sampled understory communities beneath these two species at the beginning of the wet season and near the end to determine whether or not canopy species correlated with understory species. At the beginning of the rainy season, we found that understory communities were dramatically different than open communities at both sampling times. However, at the beginning of the rainy
Chapter 5
269
season there were no differences between communities under Colophospermum and Acacia based on ordination analysis (data not shown). But after a season’s growth communities beneath Colophospermum and Acacia had diverged substantially (Figure 5.10). The greatest differences were still those between understories of both trees and the open, but separation on the Y-axis of the DCA between tree species were much more distinct than early season communities suggests that tree canopies acted as “filters” for the development and maturity of similar seed banks. The species-specific community patterns illustrated in Figure 5.10 may be caused by many factors, but sunlight reaching the understory was less under Colophospermum than under Acacia canopies during the growing season (498±24 versus 622±42 μmol m-2s-1; Fcanopy=6.57, df=1,24, P=0.014). Because of this, and the published differences in fungal root symbionts associated with these species, Elmar Veenendaal and I conducted an experiment in which shade and soil biota (also see section on soil biota below) were manipulated in a factorial design, and the effects of these treatments on the grass, Urochloa panicoides, were measured. Urochloa was common in all three microhabitats, but significantly more common under Colophospermum. Soil was collected from the open and from the understory of each tree species and then either 300
DCA Y axis
250
Colophospermum
200 150
open
100
Acacia 50
0
100
200
300
400
500
DCA X axis Figure 5.10. Ordination of vegetation plots beneath Colophospermum mopane, Acacia erioloba, and in the open at the Harry Oppenheimer Okavango Research Centre in February during the growing season. Grey symbols represent the means and 95% C.I. for each group. Data not published elsewhere
Species-Specific Positive Interactions
270
14
Height (cm)
12
*
Nonsterile soil Sterile soil
*
10
*
8 6
*
*
*
4 2
Deep shade
Light shade
Co lop h. Ac ac ia O pe n
pe n O
ph . Ac ac ia
Co lo
Co lo ph . Ac ac ia Op en
0
No shade
Figure 5.11. Height of Urochloa panicoides grown in three shade treatments (70%, 35%, and full) and in fresh and sterilized soil from under Colophospermum mopane and Acacia erioloba trees and in from the open grassland. Error bars show 1SE. In an ANOVA, Fshade=49.16, df=2,74, P<0.001; Fsoil source=11.71, df=2,74, P<0.001; Fsoil sterilization=41.22, df=1,74, P<0.001; Fsoil source x sterilization=9.47, df=2,74, P=0.001. Data not published elsewhere.
applied as fresh or sterilized. Shade was the strongest main effect (Figure 5.11) and increased the growth of Urochloa, but Urochloa grew the largest in soil from under Colophospermum in the shade. This shade and soil interaction may have been due to microbial differences as when this soil was sterilized the large increase caused by shade disappeared (also see section 3.6, Soil microbes). In addition to light quantity (as described above) canopy species can vary in their affect on light quality. Kurt Reinhart and colleagues (2006) measured light quantity as PAR and light quality as the ratio of red to far red wavelength ratios (R:Fr) under the native riparian canopies in western Montana dominated by the native Populus fremonti (black cottonwood) and Acer platanoides, Norway maple) an European invasive. They found that A. platanoides reduced PAR by an order of magnitude, but also reduced R:Fr ratios by 60% relative to the native canopy (Figure 5.12). In experiments manipulating PAR and R:Fr, A. platanoides seedlings outperformed native seedlings in the light conditions created by conspecific indicating intraspecific facilitation.
Chapter 5
271
Irradiance (mmol m-2 s-1)
7e+5
sun light light in understory
6e+5 5e+5 4e+5 3e+5
Forest type native riparian
2e+5 1e+5
Acer platanoides invaded
0 300
400
500
600
700
800
Light spectrum (nm) Figure 5.12. Light spectra in open patches, understories of native riparian forests, and understories of riparian forests invaded by Acer platanoides. Bars represent the maximum 95% CI for each curve. Vertical shaded bars represent the Red (656 to 664 nm) and Far-red (726 to 734 nm) portions of the spectrum. The visible portion of the spectrum is from 380 to 750 nm. Reprinted from Reinhart et al. (2006) with permission from Ecological Applications.
Vascular plants vary in their ability to attenuate UV wavelengths, and sometimes UV attenuation by a species is more dramatic than its affect on total PAR. Deckmyn et al. (2001) at the University of Antwerpen in Belgium found that Trifolium repens attenuated UV-B much more than grasses, but without also attenuating PAR to the same degree. This resulted in Trifolium providing a relatively light-rich microhabitat but with reduced UV irradiation. Because many organisms are sensitive to UV-B this effect has a lot of facilitative potential. However, whether or not the effects of Trifolium on UV light facilitate other plants is not known. Other photosynthetic organisms alter UV wavelengths in ways that facilitate neighbors. Some species of cyanobacteria are highly tolerant of ultraviolet wavelengths and protect other species that grow beneath them in “mat” communities. Richard Sheridan (2001) found that Nostoc and Scytonema species dominating the upper surfaces of cyanobacterial communities growing on mangrove trunks in the West Indies absorbed large amounts of ultraviolet energy. When irradiated with a light spectrum containing ultraviolet wavelengths these species remain on the surface above a number of other species without pronounced UV absorption characteristics. However, when UV light was removed from the spectrum experimentally, Sheridan found that mat community structure rapidly became disorganized with Nostoc replaced by species from below, and nitrogen-fixing capacity declined.
272
Species-Specific Positive Interactions
5.2.1.2. Nutrients Plants can have species-specific effects on nutrient cycling (Hobbie 1992). Sara Hobbie (1996) found that the root, stem, and leave tissues of Alaskan tundra species varied substantially in nutrient composition, structural biochemicals, and the content of phenolic and non-polar fats, oils, and waxes. Reasoning that these chemical differences should affect the decomposition of these species she experimented with the effects of temperature on the decay rates of litter. She found large differences among species in decomposition rates and N release and that lignin content was one of the strongest correlates with decomposition. Taking a different approach to species-specific effects on nutrients, Wedin and Tilman (1990) showed that that when different grass species were planted on similar soils, the characteristics of the soil diverged in the rate and timing of annual net nitrogen mineralization. Andropogon gerardi and Schizachrium scoparium reduced mineralization rates, while Poa pratensis and Agropyron repens increased mineralization rates. In this case, the conditions created by the grass species were favorable to themselves, a positive feedback, but other species in other systems might create negative feedbacks such as predicted in Tilman’s R* model of coexistence (Tilman 1985) in which some neighbors benefit from the nutrient status developing under other neighbors. There are many mechanisms, other than variation in decomposition rates, by which different plant species might drive species-specific soil nutrient characteristics. Plant phenology, nutrient uptake rates, litter-fall amounts, effects on microbial communities (see Chapter 3.6 and below), and chemical composition of root exudates can have powerful effects on soil nutrient cycles (Hobbie 1992, Schimel et al. 1998, Van Breemen and Finzi 1998, Wardle et al. 1998, Chen and Stark 2000, see Chapter 2, Nutrients). Plants release an astounding number of different biochemicals from their roots, and the purpose of many of these is to make nutrients available (Fan et al. 1997, Bais et al. 2004, Ryan et al. 2001, Dakora and Phillips 2002). Different species release different biochemicals which vary in their nutrient-acquiring or other functions depending on substrate characteristics. The potential for species-specific effects is endless. For example, Centaurea maculosa, a native of Europe that has invaded grasslands throughout the northwestern US, releases large amounts of (±)-catechin from its roots (Bais et al. 2003). (±)-Catechin is not found in soils where C. maculosa is absent, and (±)-catechin is an unusually powerful chelator of minerals (in addition to its allelopathic effects) that bind P and make this nutrient inaccessible to plants. The chelation process makes P much more available, and results in higher levels of available P in invaded grasslands (Thorpe et al. 2006). Tree canopies certainly modify microclimate and soil (see Chapter 2) and there are many other reasons why different tree species may have strikingly different effects on understory productivity and community composition. Different tree species also vary litter quality and quantity, the nutrient
273
Chapter 5
composition of throughfall, soil accumulation and permeability, and their effects on herbivores. For example, a comparison of two co-occurring conifer species, Pseudotsuga menziezii and Juniperus scopulorum, in southern Utah found that throughfall under Pseudotsuga was invariably more enriched in chemicals than throughfall under Juniperus (Hart and Parent 1974). Voight (1960) compared nutrient inputs beneath Pinus resinosa, Tsuga canadensis, and Fagus grandifolia and found that several different nutrients were higher in the stemflow of the former two species. These differences (also see section on nutrients, Chapter 2.5) may be due in part to variation in the nutrient content in throughfall and stemflow. Although they did not sample many trees, Gersper and Holowaychuk, (1970) reported differences in stemflow nutrient inputs beneath Fagus americana, Querus rubra, Acer saccharum, and Carya glabra trees. Similar patterns have been found in other deciduous forests (Crozier and Boerner (1986). In mixed evergreen-deciduous forests on Sapelo Island in southeast Georgia, USA, nutrients in the throughfall of many different tree species have different chemical signatures (Figure 5.13, Callaway et al. 2002). PCA axes 1 and 2 explained 27% and 18% of the variation in the nutrient matrix, respectively, Axis 1 correlated best with K, Ca and Cd (slopes of –0.48, –0.40 and +0.40, respectively); whereas axis 2 correlated best with Mg and Na (slopes of +0.62 and +0.55, respectively). Comparison of the nutrients from “good hosts” for Tilandsia versus “bad hosts” for Tilandsia (species with abundant epiphytes versus those without) showed strong differences between the groups (Figure 5.14), but experiments with the leachate did not show facilitation, epiphyte growth was always better in pure rainfall (Callaway et al. 2002). The strongest positive regression relationship between an individual macronutrient and epiphyte field abundance was for K (r2=0.78). Nitrogen leached from leaf samples was highest for Celtis, which was the best host for Tilandsia.
PCA Y-Axis Scores
4
P. taeda 2
Magnolia Celtis
Q. nigra
Ilex
0 Juniperus -2
Liquidambar
Q. virginiana -2
0
2
4
PCA X-Axis Scores Figure 5.13. PCA means and 95% confidence intervals for total N, K, Ca, Mg, Mn, Na, B, Al, Cd, Co, Cu, and Mo in leachates collected from artificial rain passed over branches of these 8 species. Each point is the mean of 7 branches. R.M. Callaway and S. Pennings (unpublished data).
Species-Specific Positive Interactions
PCA Y-Axis Scores
274
2
“ Good hosts” 0
“ Bad hosts” -2
-2
0
2
PCA X-Axis Scores Figure 5.14. PCA means and 95% confidence intervals for nutrients leached from leaves and branches collected from good hosts for Tilandsia usneoides (Celtis laevigata, Quercus virginiana, Juniperus virginiana) and bad hosts for T. usneiodes (Magnolia grandifolia, Q. nigra, P. taeda, Ilex opaca, and Liquidambar styraciflua). R.M. Callaway and S. Pennings (unpublished data)..
I do not know of studies showing that the effects of particular canopy species on throughfall nutrient concentrations correlate with similar effects on nutrients in the soil. However, many studies have shown species-specific soil nutrient concentrations among species within communities. Canopy effects within closed forests are often less obvious than those in savannas, but several studies have reported species-specific canopy effects on understory soils in forests. For example, total nitrogen, nitrogen fluxes, and a large number of total cations in subcanopy soil are correlated with particular tree species in natural deciduous forests in the northeastern U.S. (Finzi et al. 1998a,b). They found that net nitrogen mineralization and nitrification rates were roughly twice as high in soils under Acer saccharum, Fraxinus americana, and Acer rubrum as that in soils under Fagus grandifolia, Quercus rubra, and Tsuga canadensis and that these processes were highly correlated with C:N ratios. They argued that the lack of differences in soil texture among sites, and the lack of clear correlations between deep mineral soil calcium content and surface exchangeable calcium indicated that the correlations between tree species and soil properties within mixed-species stands were due to tree effects. Bernard Pelletier and colleagues at McGill University measured the correspondence between different tree species and a number of properties of the forest floor in eastern Canada, including total nitrogen, phosphorus, potassium, calcium, and magnesium. Using “trend surface analysis” and a “neighborhood matrix” they found that 30% of the variation in forest floor characteristics could be explained by species influence (Pelletier et al. 1999). Most of this was due to the effects of Fagus grandifolia on calcium and Acer species on organic matter and associated nutrients. At other sites in the northeastern U.S. Lovett and Rueth (1999) found that nitrogen mineralization and nitrification were higher under Fagus grandifolia than
Chapter 5
275
under Acer saccharum. “Single-tree influence circles” have been described in the Appalachian Mountains in which higher levels of pH, calcium, magnesium, potassium, and mineralizable nitrogen have developed at the surface under Liriodendron tulipifera than under Tsuga canadensis on soils with uniform texture and pH at depth (Boettcher and Kalisz 1990). Beatty (1984) also described large differences in soil nutrient status between Tsuga canadensis understories and those of deciduous species, and found that Tsuga cover was more highly correlated with understory composition than any other measured variable. In the absence of Tsuga microtopographic effects on understory composition were strong. In Queensland Australia, Ebersohn and Lucas (1965) also found that tree canopies had important effects on understory communities. Similar speciesspecific effects of tree canopies on soil characteristics have been reported from temperate deciduous forests in North America (Dijkstra et al. 2001, 2003, Dijkstra 2003, Dijkstra and Smits 2002). Correlative studies of species-specific effects in natural forests, no matter how dramatic the differences between deep soils and surface soils, cannot separate the effects of trees on the soil from the effects of the soil on the trees. However, some studies have been conducted in plantations where different species have been planted without regard to existing soil conditions. Alban (1982) measured larger quantities of exchangeable calcium in the surface forest floor and smaller quantities of exchangeable calcium in the mineral soil beneath Picea glauca and Populus tremuloides than under Pinus resinosa and Pinus banksiana. These differences were attributed to uptake from deep soils and deposition on the surface via litterfall and throughfall. Challinor (1968) sampled differences in soil nutrients under Picea abies, Quercus rubra, Pinus strobus, and Pinus resinosa that had been experimentally planted. Differences in nitrogen, potassium, and calcium were striking and the differences disappeared with soil depth, strongly implicating that the trees were the cause of these differences in soil nutrients. Dzwonko and Loster (1997) compared the species composition of plant communities under Pinus sylvestris and Robinia psuedoacacia that had recently invaded flat, sandy grassland in Poland and found that canopy species had stronger effects than any other aspect of the microhabitat. Soils under the different canopy species differed in the concentrations of some nutrients, cation exchange capacity, and pH. Lodhi (1977) compared soil nutrients under several different tree species in lowland deciduous forests in Missouri, USA. He found that virtually every soil nutrient he tested differed significantly among at least some pairs of species with concentrations of available phosphorus under different tree species ranging from 13 to 400 ppm and pH ranging from 5.5 to 8.1. Some of this range was certainly due to differences in microhabitat, and the study lacks statistical analysis, but the relative trend in tree effect ranged from a reduction of available phosphorus by about 50% to a four-fold increase. Such patterns of soil nutrients associated with different tree species within a forest may originate from difference in the nutrient content of litter and throughfall, microclimatic effects on nutrient cycling, or as hypothesized by Lodhi, differences in the way that chemicals in leaves affect nitrification.
276
Species-Specific Positive Interactions
Cross and Schlesinger (1999) compared the patch-scale distribution of nutrients in grasslands dominated by Bouteloua gracilis and B. eriopoda to that in shrublands dominated by Larrea tridentata. The spatial distribution of soil nutrients in the grassland was relatively homogeneous, with ordinations showing complete overlap among samples collected under the bunchgrasses and samples collected between the grasses. In shrublands, however, ordinations of all elements and plant-essential nutrients showed that below-shrub soils differed from intershrub soils. These differences were driven by higher concentrations of NO3-N, total N, K, organic carbon, and Cl under shrubs and higher concentrations of Ca in the inter-shrub spaces. These results indicate that shrub-driven concentrations of nutrients may contribute to long-term changes in the function of desert ecosystems as woody perennials have expanded in abundance. The influence of shrubs on landscape-scale nutrient patterns and concentrations was quite different that that of grasses and has probably increased over the last century as vegetation throughout the southwestern United States has changed from perennial grassland to woody shrubland (Archer 1989, Grover and Musick 1990). Similar, species-specific, plant induced fertile islands have been described in the succulent-dominated Namaqualand Desert of South Africa. Stock and colleagues at the University of Cape Town found that the longevity of shrub species correlated with higher nitrogen concentrations in the soil. In general, the species in Namaqualand are relatively short-lived and the rapid dynamics of the vegetation (see Chapter 6) leads to shifting patterns of fertilization. Rob Jackson and Martyn Caldwell (1993) used geostatistical techniques to compare the characteristics of soil near two different perennial plant species to that of soil far from the plants. They quantified the scale of soil variability around individual Artemisia tridentata and Pseudoroegneria spicata (bluebunch wheatgrass) and then compared that small-scale variability to that occurring across a much larger shrub-steppe landscape. In contrast to the shrub-grass comparison of Cross and Schlesinger (1999), interpolated contour plots of soil organic matter, phosphate, and potassium showed strong spatial patterning around Pseudoroegneria tussocks, but less consistent relationships around Artemisia shrubs. They attributed these differences to the sparse open canopies of the shrubs and pocket gopher activity, which was more common under Artemisia than other places. The current trend to lump species into functional groups for ecosystem analyses is likely to conceal interesting species-specific effects on soil nutrients and neighbors. Spehn et al. (2002) found that the presence of legume species significantly enhanced biomass and nitrogen pools in a cross-European collaborative study. In Germany, where the effect of legumes was the strongest, nitrogen that was fixed symbiotically by legumes was ultimately acquired by nonnitrogen fixing species, but the transfer of nitrogen depended on the particular legume species fixing the nitrogen and the recipient species. For example, Trifolium species were “efficient” nitrogen fixers with greater effects on ecosystem function than Lotus species.
Chapter 5
277
5.2.1.3. Oxygen Soil oxygen often limits plant growth in wetlands (McDrew 1983). As described in Chapter 2, many wetland emergent plants passively transport oxygen from leaves to roots through aerenchymous tissue which alleviates oxygen limitation in their roots (Armstrong 1979). In some cases oxygen may leak out of submerged roots and oxidize toxic substances and nutrients in the rhizosphere and oxygenate marsh sediments (Howes et al. 1981, Armstrong et al. 1992, Howes and Teal 1994). For plants to facilitate others via soil aeration they must possess highly aerenchymous tissues; thus not just any species will do for a benefactor. Obviously, non-arenchymous species are not going to release much oxygen from their roots, but there is surprisingly large variation in oxygen release among aerenchymous wetland species. In general, oxygen release rates are affected by the redox potential of the soil, and different plant species may react differently to variation in redox potential. Wießner et al. (2002) compared oxygen release from the roots of several species and observed the highest rates were observed for Typha latifolia (1.4 mg/h per plant) followed by Phragmites australis (1.0 mg/h per plant), Juncus effusus (0.7 mg/h per plant), and Iris pseudacorus (0.3 mg/h per plant). They concluded that this “oxygen releasing behavior is a process dominating natural conditions within the rhizosphere” and that variation among species is an important factor to consider in the ecology of these systems and in phytoremediation. 5.2.1.4. Hydraulic lift Woody perennials can improve the water relations of understory plants through “hydraulic lift”, the movement of water from deep, moist soils to dry, surface soils at night when the stomata are closed and the lowest water potentials are in the upper soil layers (Richards and Caldwell 1987, Dawson 1993, Chapter 2.1). Dawson (1993) showed that all of 12 understory species examined derived some (3-60%) hydraulically lifted water from overstory Acer saccharum (sugar maple). The amount of hydraulically lifted water obtained by an understory species decreased markedly with increasing distance from the tree. In this case, as for many others, the positive relationship depends on the specific root architecture that permits a particular overstory species to hydraulically lift water and deposit it near the surface. The potential for species-specific facilitative effects based on differences in hydraulic lifting is well illustrated in a study by Carolyn Yoder and Robert Nowak (1999) of five Mojave Desert shrub species and a perennial grass. By measuring diurnal patterns of subcanopy soil moisture and conducting night lighting and day shading experiments they found evidence for substantial hydraulic lifting for all six species. However, hydraulic lift was not detected for all plants at a site at a given time. For example, at one sampling period in June 1995, lift was detected only beneath one Lycium pallidum shrub while at other
278
Species-Specific Positive Interactions
times hydraulic lift was detected beneath all species. The magnitude of changes in water potential also varied a great deal. Surface soil under Ambrosia dumosa increased from ≈-0.9 to ≈-0.8 MPa, but from -4.20 to -3.80 under Lycium pallidum. Ephedra nevadensis elicited diurnal cycles at 0.75 cm depth, but not at 0.35 cm. The depth effects of Yucca schidigera were the opposite. Ambrosia, Lycium, Larrea tridentata, and Achnatherum hymenoides showed effects at both depths. This variation may have been produced by phenological differences, variation in root architecture, or differences in soil conditions, but it points out the potential for species-specific effects of these plants on neighbors via hydraulic lift. Furthermore, Yoder and Nowak found that the CAM species studied, Yucca schidigera, created a pattern of surface soil water flux that was temporally opposite those of the other species. Soil water potential increased during the day while Yucca was not transpiring (CAM species transpire at night) and decreased at night when transpiration began. They noted that CAM species that lift during the day may enhance the water relations of surrounding species more than non-CAM plants, because C3 and C4 annuals photosynthesize and transpire at this time. Espeleta and colleagues (2004) found similar species-specific patterns in hydraulic lift in a study of Pinus palustris, three Quercus species and two perennial bunchgrasses in the sandhills of South Carolina. All but two of these species exhibited hydraulic lift but lifters varied substantially in the frequency of lifting throughout a growing season and the amplitude of the water lifted. Interestingly, the three Quercus species differed dramatically in their lifting capacity with Q. margaretta showing no lift and Q. incana and Q. laevis lifting large amounts of water frequently. They proposed that variation in the ability of these species to redistribute water from the deep soil to shallow soil may have strong effects on the overall water balance of sandhill plant communities. In a fascinating study, providing convincing evidence for the redistribution of soil water among individuals via hydraulic lift and speciesspecific variation in this function, Brooks et al. (2002) found that hydraulic redistribution may play an important role in northwestern coniferous forests. In a 20-year-old Pseudostuga menzeisii stand in late August, nocturnal hydraulic lift supplied 28% of the water in the upper 2 m of soil that was removed during the day by evaporation and transpiration. In an old-growth Pinus ponderosa stand approximately 35% of the daily water loss was replaced by hydraulic redistribution. Experimental application of deuterated water in plots was measured in the xylem water of ponderosa pine seedlings too far from the plots to be explained by movement of water through the soil – indicating facilitation through root redistribution. They argued that hydraulic redistribution enhances seedling survival.
Chapter 5
279
5.2.1.5. Disturbance It is evident that species may differ in the way that they moderate disturbance and by doing so ameliorate the effects of disturbance on neighbors. For example, only Carex nudata, the torrent sedge, can withstand spring flooding and provide safe sites for less tolerant species (Levine 1999, 2000). But to my knowledge no studies have explicitly examined species-specificity among potential benefactors in the context of disturbance. However, plant species certainly vary in their effects on disturbance regimes in ways that benefit other species, and it is reasonable to extrapolate from this to species-specific effects on neighbors. For example, Pandey and Rokad (1992) quantified sand dune development and stabilization behind different shrub species in the Thar Desert of India. They found that dune density and volume varied dramatically among species (Figure 5.15), and that these characteristics were related to the shape and leafiness of canopies. Other characteristics of the dunes (soil texture, bulk density, and nutrients) did not vary among shrub species. No mention was made of associated species, but many other studies have shown that stabilized dunes are rapidly colonized by other species (Hewett 1970, Carter 1991 Danin et al. 1998, Martínez 2003, Franks and Petersen 2002).
-1
Dune density Dune volume 20
10
10
0
0 Sal v oleo adora ides Ziz num iphus mu lair a
20
Cal o pro tropis cer a Cap dec paris idu a Eup thy horbia mif olia Lep pyr tadeni o te chn a ia May ema tenus rgin a ta Pro chil sopis ens is
Dune density (dunes ha )
30
3
Dune volume (m )
30
Figure 5.15. The volume and density of accumulated sand associated with individual shrub species in the Thar Desert of India. Drawn from data presented in Pandey and Rokad (1992) with permission from the Journal of Arid Environments.
280
Species-Specific Positive Interactions
5.2.2. Species-specific indirect effects 5.2.2.1. Consumers Some of the strongest positive effects are produced by shared defenses in which a palatable species is protected by its proximity to a well-defended species (see Chapter 3.1, Atsatt and O’Dowd 1976, McAuliffe 1984, 1986, Hay 1986). Repelling herbivores and seed predators requires specific morphological traits such as spines, tough tissues, or the possession of chemical defenses. Since not all potential benefactors in a community have these traits, most indirect facilitative interactions involving shared defense should be species-specific. On the other hand, some indirect interactions among consumers and their prey can appear to be species-specific on the surface, but after rigorous examination turn out to be quite general. A pair of experimental studies by Peter Hambäck and colleagues at the Swedish University of Agricultural Sciences and Umeå University in Sweden nicely illustrates how careful mechanistic studies of apparently obvious speciesspecific indirect facilitation can yield other results. In Chapter 3 I described their work on the relationship between Lythrum salicaria and Myrica gale, in which the former experienced less herbivory in thickets of the latter, a highly aromatic and apparently well-defended benefactor (Hambäck et al. 2000). Despite the apparent “shared defenses” nature of this relationship, Hambäck et al. (2003) conducted laboratory studies with insect specialist herbivores on Lythrum and found that they were not distracted by the odor of Myrica, suggesting that the aromatic defenses of Myrica were ineffective and could not therefore be shared by Lythrum. They then conducted experiments in which they tested the ability of three other shrubs and two artificially constructed thickets to deter specialist herbivory on Lythrum and found that all other shrubs and the fake shrubs provided protection from two specialist herbivores, Galerucella calmariensis and G. pusilla (Figure 5.16). Interestingly, the superior performance of Lythrum inside of shrub patches appeared to attract more of a third specialist insect, Nanophyes marmoratus. While suggesting that Lythrum benefits a great deal from the presence of neighbors, these results demonstrate that the originally described positive effect of Myrica, despite its strong scent, was associational and therefore not species-specific. Strong consumers also exist within the plant community. Interplant parasitism is a widespread phenomenon with over 3000 species of parasitic plants having been identified worldwide (Kuijit 1969). Most parasitic plants are highly host specific and thus have the potential to indirectly facilitate non-hosts in the community. In California salt marshes, Cuscuta salina mediates interactions by preferentially consuming the dominant Salicornia virginica, which indirectly facilitates several inferior competitors (Pennings and Callaway 1996). In one of four sites studied, Gibson and Watkinson (1992) found that the hemiparasite, Rhinathus minor, increased grassland plant diversity
Chapter 5
281 Northern
***
(B)
Southern
***
***
***
*
*
No. eggs per plant
100
10
1
0.1
Myrica
Artificial 1
Salix
Hippophae Artificial 2
Alnus
Figure 5.16. The number of Galerucella calmariensis/G. pusilla eggs (mean ± SE; ***, P < 0·001; **, P < 0·01; *, P < 0·05) on the host plant Lythrum salicaria when host plants were placed in either an open control area (open bar) or inside a thicket (solid bar). Thickets include Myrica gale, Alnus glutinosa, Hippophaë rhamnoides, Salix myrsinifolia-phylicifolia agg., and artificial thickets at Brännölandet 1) and Jungfruholm (2)]. Reprinted from Hambäck et al. (2003) with permission from Functional Ecology.
by suppressing a dominant competitor, and indirectly facilitating non-hosts. These indirect positive effects are not interchangeable among species. 5.2.2.2. Pollinators and dispersers As described in detail in Chapter 3.3, species with attractive flowers may attract pollinators for less attractive neighbors. For example, Thomson (1978) found that Hieracium florentinum received more pollinator visits when it was mixed with H. aurantiacum than when it was alone. Laverty and Plowright (1988) found that fruit and seed set of Podophyllum peltatum (mayapple) was enhanced when they were spatially associated with Pedicularis canadensis (lousewort). In later studies, Laverty (1992) found that Podophyllum, which produces no nectar, depends on infrequent visits from bumble bees that accidentally encounter them while collecting nectar from Pedicularis. This sort of facilitation has been rarely studied, but may be very common. Facilitation among co-flowering species, or benefits gained by an unattractive or nonreward producing species near a more attractive neighbor is likely to be very species-specific. The attractive neighbor must be able to appeal to pollinators that are also compatible with its unattractive neighbor. In northern Mexico, the wild chili, Capsicum annuum, is strongly skewed in its distribution to the understory of other shrub species (see Chapter 3, Tewksbury et al. 1998). However, Capsicum has much stronger associations with shrub species that have red fruits like those of Capsicum. Tewksbury and colleagues showed that Capsicum was dependent on its benefactors, primarily for
282
Species-Specific Positive Interactions
shade, but that the species-specificity of the relationship was due to dispersal of Capsicum fruits by birds to perch trees that had fruits similar in color to those of Capsicum. A similar example occurs for the rare and endangered Taxus baccata (yew) in the Mediterranean mountains of southern Spain. Garcia and Obeso (2000) found that the distribution of Taxus saplings was skewed towards the understory of fleshy fruited shrubs. Seedlings were also about 5 times more common under fleshy fruited shrubs than shrubs without fleshy fruits. As a conifer Taxus does not technically have fruits, but females do have fleshy red cones that look like fruits and are dispersed like fruits by birds. Because birds that eat fleshy fruits hang out in fleshy-fruited shrubs, Taxus tends to be deposited there. The species-specific indirect effect of being dispersed to particular shrubs did not stop there however. Taxus seedlings that were within shrubs were much less likely to be browsed by livestock than those in the open. These two studies provide insight into the complex interactions involving species-specific effects among plants determined by indirect interactions with dispersers (Figure 5.17). There are other species-specific indirect interactions that are driven by herbivores. Arid lands throughout the world are subject to heavy grazing. So it is for shrublands of southern Australia where rabbits and livestock have had major impacts for decades. Much of the vegetation is woodland dominated by Acacia papyrocarpa mixed with shrubs in the Chenopodiaceae family. Facelli and Temby (2002, see Figure 4.11, Chapter 4) experimented with the effects of two shrub species, Atriplex vesicaria (bladder saltbush) and Maireana sedifolia (pearly bluebush) on the biomass of the herbaceous species growing underneath them. When grazing was not excluded the biomass of the understory herbs was many times higher under the two shrubs than in the open and both shrub species had the same overall positive effect. When herbivores were excluded the positive effect remained, but it was not as strong. Most importantly, the total biomass of understory herbs and the biomass of Danthonia caespitosa beneath 1
Probability
0.8
0.6
0.4
Herbivory Seedling survival
0.2
0 0
20
40
60
Percentage of shelter
80
100
Figure 5.17. Probabilities of herbivore damage and seedling survival (from the first year to > than two-year) of Taxus baccata in southern Spain as a function of shelter by nurse shrubs. Reprinted from Garcia and Obeso (2003) with permission from Ecography.
Chapter 5
283
Maireana did not differ in or out of exclosures, whereas the exclosure effect of total understory biomass and Danthonia biomass under Atriplex was significant. This indicates that these two shrubs had species-specific capabilities to protect neighbors from herbivores, with Atriplex doing a relatively poor job compared to Maireana. Interestingly, strong overall species-specific differences between the shrubs were not apparent until exclosures were built. Inside exclosures Atriplex was much more facilitative in two of three comparisons than Maireana, a characteristic attributed to hydraulic lift of water by the root systems of Atriplex. The superior protection provided by Maireana against herbivores was attributed to this species’ low palatability. 5.2.2.3. Soil microbes Many plant species develop species-specific soil microbial communities in their rhizospheres (Chapter 3.6, Bever 1994, Bever et al. 1997, Johnson et al. 1997, Egerton-Warburton and Allen 2000, Stephan, 2000). Any changes elicited by these plant-soil microbe interactions have the potential to be species-specific. For example, Eom et al. (2000) found dramatic differences in the arbuscular mycorrhizal fungal communities among the rhizospheres of five different herbaceous species in tallgrass prairie. They followed this analysis with an experiment in which the grasses were grown in initially similar soil communities and allowed to interact with them over four months. Even when starting conditions were the same different species of herbs developed very different AM fungal species composition, species richness, and total spore numbers. Stephan et al. (2000) found that Trifolium repens had much stronger effects on the development of diverse communities of cultural soil bacteria than many other species. Because of this, T. repens was considered a “keystone species” in maintaining plant-microbial processes. Negative, density-dependent feedbacks commonly occur when the microbial communities that develop in the rhizosphere of a particular plant species have negative effects on their host, but not on other plant species (Van der Putten et al. 1993, Bever et al. 1996, 1997. As described in Chapter 3, indirect facilitation occurs when one species develops a soil biota more favorable for a neighbor than the soil biota established by the neighbor itself. Plant-soil microbe feedbacks can be highly species-specific. John Klironomos (2002) conducted an experiment on microbial feedbacks with five rare native species and five invasive species from grasslands in eastern Canada. He “trained” the soil microbial community by growing 10 different plant species in two sequential plantings in initially neutral soil that had been collected from a site without any of the 10 experimental species. After this initial training period he grew each of the ten species in “home” soil, that which they themselves had trained, and compared this growth to that in “foreign” soils, that trained by the other species. All of the native species experienced negative feedback with the soil microbial community (Figure 3.20, Chapter 3),
Species-Specific Positive Interactions
284
Relative plant abundance in the field
a process that should lead to indirect facilitation, and all of the invasive species experienced positive feedback, a process that could lead to conspecific facilitation and the exclusion of other species. Klironomos followed the soil feedback experiment with specific inoculations of isolates of pathogens/saprobes and arbuscular mycorrhizae. In this experiment rare natives showed strong negative feedback for the pathogen/saprobe filtrate when grown in soil with “a history of the same plant species”. The effects of filtrate with arbuscular mycorrhizae were consistently positive and did not differ among native and invasive species. Next, Klironomos compared the relative abundance of all 61 species in his grassland with the strength and direction of its feedback interactions with soil microbes. The relationship between abundance and feedback was striking (Figure 5.18). Most species experienced negative feedback (82%); however, for the relatively abundant species feedback interactions were consistently positive. The implications of these speciesspecific interactions are profound. First, they suggest that small scale plant soil microbial feedbacks create a shifting dynamic mosaic of species (see Chapter 6) that is crucial to the maintenance of biological diversity. The mechanisms driving these shifts are inherently facilitative because species with negative feedbacks are constantly creating safe sites for other species while the other species are constantly creating safe sites for them. Second, his results suggest that invaders may break up these facilitative mosaics. Because co-evolved pathogens for invasive plant species are absent or weak the negative microbial feedbacks that lead to indirect facilitation are missing, and positive feedbacks can establish among invaders and less host-specific mycorrhizal fungi. These positive feedbacks may eliminate the strong density-dependent controls on native species and allow invaders to run wild.
Feedback Figure 5.18. The relationship between relative species abundance in an old-field site and soil feedback response. P=0.0001. Numbers represent different plant species. Reprinted from Klironomos et al. (2002) with permission from Nature.
Chapter 5
285
Sarcodes sanguinea (snow plant) is a parasitic species in the Ericaceae without chlorophyll and does not physically attach to the roots or shoots of its hosts. Instead it is infected with species of ectomycorrhizal fungi that also infect associated conifers and provide conduits through which the parasite can acquire sustenance. In some parts of the Sierra Nevada Range of California the only associates of Sarcodes are the fungus species Rhizopogon ellenae and the tree Abies magnifica. Bidartondo et al. (2000) found that total mass of ectomycorrhizal fungal species associated with Abies was many times higher in soil cores near Sarcodes. Furthermore, there was a shift in fungal species dominance closer to Sarcodes. In cores collected over 500 cm from Sarcodes no Rhizopogon ellenae (the only species able to infect Sarcodes) were found. However, within the rootball of Abies trees, the abundance of R. elleanae was twice that of other ectomycorrhizal species. Although it is possible that Sarcodes preferentially colonized dense patches of relatively pure R. ellenae not found in the absence of the heteromycotroph, the results of Bidartondo and colleagues suggest that Sarcodes is able to stimulate the development of the fungal species that it relies on. Such interactions have important implications for species-specificity in plant microbial interactions. With Elmar Veenendaal of the Harry Oppenheimer Okavango Research Centre of the University of Botswana, I investigated the potential for soil microbial communities to drive the differences described above (see Figure 5.10 above) for understory communities beneath Colophospermum mopane (mopane) and Acacia erioloba in northern Botswana. The two tree species possess different root symbionts. We grew the perennial grass species, Urochloa pannicoides, in three different shade treatments, in fresh soil collected from the open or beneath Colophospermum or Acacia, and in heatsterilized soil from the three microhabitats. We chose Urochloa because it was present in all environments, but most abundant under tree canopies, and because it underwent a distinct shift from similar abundances under the two tree species at the beginning of the rainy season to a strong preference for Colophospermum at the end of the rainy season. Shade was the most important effect (see Figure 5.11 above), but soil source was also significant, with Urochloa growing taller in soils from under Colophospermum than in soils from Acacia or the open (Fsoil source=11.74, df-2,117, P<0.001). This effect, however, could simply be due to more nutrient enriched soil under Colophospermum. However, soil sterilization was also highly significant (Fsoil sterilization=41.03, df=1,117, P<0.001), as was a significant soil source by sterilization interaction (F=8.47, df=2,117, P=0.001). Sterilization of soils eliminated the highly positive effect of Colophospermum soil on Urochloa growth, indicating that this effect may have been due to the soil microbial community. Differences in the understory soil microbes beneath Colophospermum and Acacia appear to play an important role in speciesspecific and indirect beneficial effects on an understory grass.
286
Species-Specific Positive Interactions
Other studies have shown that trees accumulate soil pathogens that suppress conspecific seedlings (Packer and Clay 2000), but to my knowledge no other study has directly linked tree canopy effects on soil microbes to the general facilitative effect of the tree on other species (but see Reinhart and Callaway 2004). As described in Chapter 3, fungal intermediaries may also mediate species-specific facilitative interactions among plants. Grime et al. (1987) found that labeled 14C was transferred from Festuca ovina to many other plant species (including Centaurea nigra) in artificial microcosms that shared a common mycorrhizal network, but not to others that did not share the network. Mutual infection led to decreased biomass of the dominant Festuca and increased biomass of otherwise competitively inferior species, and experimental microcosms that were infected with mycorrhizae were more diverse than those that were not infected. Marler et al. (1999a) tested the effects of other Festuca and Centaurea species, F. idahoensis and C. maculosa, and found that mycorrhizae mediated strong positive effects of Festuca on Centaurea. When Centaurea was grown with large Festuca in the presence of mycorrhizae, they were 66% larger than in the absence of mycorrhizae. Mycorrhizae have also been found to move nutrients among plant species (Walter et al. 1996). Simard et al. (1997) used reciprocal isotope labeling to document bi-directional transfer of carbon among forest trees in the northwest United States. The rate of transfer varied in different shade treatments, but not all species involved in the study participated equally in the interaction. Species-specificity in fungi-mediated positive interactions among plants may be due to different fungal communities on different plant species or different abilities of plants to utilize the same fungal symbionts; however, the broad range of species-specific effects of mycorrhizae on plants make it is unlikely that the interacting species in these systems are very interchangeable. However, we still know very little about complex interactions among plants and soil microbes.
5 .2.3. Species-specificity due to similar positive effects, but different negative effects As described in detail in Chapter 4, facilitative effects rarely occur in the absence of competition. Therefore, different benefactor species may have the same positive effects, but quite different negative effects. For example, Suzán et al. (1996) suggested that Olneya’s superior facilitative ability, relative to other similar species, may be due to its phreatophytic life history and deeply distributed root architecture, thus reducing niche overlap (e.g. Cody 1986) and accentuating its positive mechanisms. The degree of root overlap between Quercus douglasii (blue oaks) and understory annuals has been shown to determine either a negative or positive effect of the overstory oak in savanna and woodlands of central California (Callaway et al. 1991, see Chapter 4). We found that blue oaks added considerable
Chapter 5
287
amounts of nutrients to the soil beneath their canopies and that soil and litter bioassays demonstrated strong facilitative effects of these components on the growth of a dominant understory grass, Bromus diandrus. In the field, however, the expression of this facilitative mechanism was determined by plasticity in the root architecture of individual trees. In experiments designed to quantify the development of lateral roots (a process crucial to understanding subspeciesspecificity in Q. douglasii effects), I found that Quercus douglasii seedlings increased lateral root weight and number when taproots were denied access to deeper water but seedlings decreased total root mass (Callaway 1990). The bottom line is that the plastic root architecture of Q. douglasii appears to play a major role in determining how the competitive effects of some individuals overwhelmed facilitative effects, leading to highly specific net effects. The poor performance of Larrea tridentata as a nurse plant (Hutto et al. 1986, McAuliffe 1988, Figure 5.1 of this chapter) may be due to the strong negative effects this species has on perennial neighbors (Fonteyn and Mahall 1981). Mahall and Callaway (1991, 1992) found that Larrea substantially inhibited the root elongation rates of Ambrosia dumosa, and that these negative effects were reduced by the addition of small amounts of activated carbon, a strong adsorbent to organic molecules (Cheremisinoff and Ellerbusch 1978). Thus Larrea canopies may have the potential to facilitate some shade-requiring species (see Yeaton 1978, Casper 1996), but some prospective beneficiaries appear to be eliminated by root allelopathy and competition. Muller (1953) documented strong positive associations between Ambrosia dumosa and many species of desert annuals; however, Encelia farinosa shrubs in the same area did not harbor any annual species. He attributed this difference to the inhibitory effects of leachates from Encelia leaves. One of the best examples of how differences in competitive effects may generate species-specific facilitation is a study conducted by Jennifer Dunne and Tom Parker in the chaparral of central California (1999). They observed that mature Pseudotsuga menziesii individuals were scattered throughout patches of chaparral, but primarily in association within different species of shrubs in the genus Arctostaphylos. They planted two cohorts of P. menziesii seeds at three different sites under two Arctostaphylos species and under Adenostoma fasiculatum, another dominant chaparral shrub. Survival of P. menziesii was much higher beneath Arctostaphylos glandulosa than under Adenostoma fasiculatum. They examined mortality due to insect herbivory, fungi, and disturbance and could find no evidence that these factors were important determinants of the survival differences. Instead, survival data and spatial patterns appeared to correspond with summer drought-caused mortality during the first year after germination. Because of this, Dunne and Parker used thermocouple psychrometers to measure soil water potentials at 7 and 20 cm below the soil surface in Adenostoma and Arctostaphylos stands. There were no apparent differences in soil moisture in the different stands at the end of the winter rainy season, but as the summer drought progressed soil beneath
288
Species-Specific Positive Interactions
Adenostoma became much drier than soil beneath Arctostyphylos glandulosa. Although site effects remain an alternative hypothesis to shrub effects, the dramatic differences in seedling survival and soil moisture occurred in stands with a common microclimate, soil type, and topography, and which occurred as little as 5 m apart. In addition to the possibility that microsites may have differed, they proposed that 1) soil moisture may be better retained in the thicker letter of Arctostyphylos glandulosa, 2) Adenostoma may be a superior competitor, 3) Arctostyphylos glandulosa may be a hydraulic lifter, or 4) longterm biotic effects of shrubs on soils may have differed. In similar hot dry climates some species may ameliorate the harshness of the environment more effectively that others. In conjunction with reforestation efforts in mountain forests in the Mediterranean basin, Castro et al. (2002) planted Pinus sylvestris and P. nigra either in the open, under various species of “spiny shrubs” and under Salvia oxydon. Salvia ameliorated the xeric conditions more than the spiny shrubs, and provided better habitat for the survival of pine seedlings. It was not clear why Salvia maintained more mesic conditions than the other shrubs, but the most apparent difference was canopy density. The relationship described above between beneficiary quality and soil moisture for Arctostyphylos and Pseudotsuga is correlational, and other mechanisms might cause the species-specific relationship described by Dunne and Parker. In the same system, but in an intergeneric comparison of Arctostaphylos and Adenostoma, Horton et al. (1999) reported that Pseudotsuga establishes only in Arctostaphylos. They found no significant differences between Arctostaphylos and Adenostoma in allelopathic effects, subcanopy light or temperature, or soil nitrate and ammonium. However, Pseudotsuga seedlings that had been transplanted into an Arctostaphylos patch were infected with 17 species of mycorrhizal fungi that colonized both Pseudotsuga and Arctostaphylos. Fifty-six of 66 Pseudotsuga seedlings were colonized by fungal species that also colonized Arctostaphylos within the same soil core. Horton and colleagues argued that Adenostoma patches did not offer adequate ectomycorrhizal inoculum for Pseudotsuga establishment. Their results provide evidence that particular ectomycorrhizal fungi associated with different shrub species contribute to the species-specific relationship between Pseudotsuga seedling establishment and shrubs.
5 .2.4. Species-specific interactions and life histories The species-specific effects of benefactors on beneficiaries depend on the extent to which resources and stress are altered in particular environments (Holmgren et al. 1997, see Chapter 2.3, Figure 2.2). However, the effects of these changes depend on the life history of the benefactors including patterns of biomass allocation, photosynthetic pathways, timing of water use, and phenollogy and lifespan. Considering this kind of life history variation in studies is
Chapter 5
289
probably crucial for integrating particular studies into community theory. For example, if ephemeral species are rarely facilitated then to base theory on studies of facilitation and ephemeral species might be misleading. In a very nice comparison of species with different life history characteristics, Tewksbury and Lloyd (2001) found that Olneya tesota canopies had stronger facilitative effects on understory perennials than on ephemerals the Sonoran Desert. They hypothesized that this was because ephemerals avoided drought-stress because they are adapted for very rapid development, high photosynthetic rates, and early reproduction. Consequently, ephemerals grow primarily during a time when water is relatively available and the long-term stress so indicative of the Sonoran Desert is not a factor. In contrast, perennials must tolerate the worst stress periods that occur in an area and may be more sensitive to differences in temperature and water availability, particularly during early development. Hoffman (1996) evaluated the effect of canopy cover on 12 species of trees and shrubs in the cerrado savanna of Brazil by sowing seeds into sites with three different densities of mature woody cover. Of these 12 species three had life histories that restricted them to relatively mesic forest conditions whereas eight had life histories restricting them to more xeric savanna. Overall, establishment of test species was greater under the canopies of trees indicating a broad facilitative effect, but there was species-specificity that correlated with life history. All species with “forest life histories” were strongly facilitated by canopy trees, whereas fewer species with “savanna life histories” required nurse plants and the facilitative effect was not as strong. Early successional development of vegetation in many of the temperate forests on the South Island of New Zealand typically proceeds from mosses and lichens to herbs and grasses to colonization by the nitrogen-fixing shrub (Carmichaelia odorata) (Bellingham et al. 2001). They examined the relationships between the development of Carmichaelia stands over time and soil characteristics and vegetation in four successional stages. Then they conducted experiments designed to measure different facilitative and inhibitory effects of soils from the successional gradient on three different tree species that varied in their development with increasing development of Carmichaelia stands. Over the chronosequence, there was a significant development of soil organic horizons and a large increase in soil nitrogen, as well as strong attenuation of light penetrating the canopy. They found that the effects of shade were not species-specific, and were always inhibitory for all three experimental species. But in contrast they found that Carmichaelia litter had much stronger positive effects on the shoot mass and foliar phosphorus concentrations of the trees Griselinia littoralis and Weinmannia racemosa than Metrosideros umbellata. Furthermore, only Griselinia and Weinmannia responded favorably to nitrogen-rich soils collected from old Carmichaelia stands. These species-specific responses corresponded with the increasing abundance of Griselinia and Weinmannia in late successional forests. Metrosideros peaked in abundance in young stands of Carmichaelia. These
290
Species-Specific Positive Interactions
species-specific responses of tree seedlings to the effects of Carmichaelia on soil nutrients may help to explain species composition during the development of forest communities after disturbance.
5.2.5. Communication We know little about signaling and communication among plants, less about how communication may have facilitative effects (see Chapter 2.10), and even less about species-specific communication and facilitation. However, the mechanisms by which plants may communicate; primarily chemicals and light wavelength transmission and reflection, differ among species and have strong potential to drive species-specific positive interactions. Furthermore, from studies of chemical interactions among the roots of parasitic plants and their hosts (Bouwmeester et al. 2003), and among species experiencing herbivore attack (Karban et al. 2000) we know that chemical signaling among plants can be amazingly species-specific and precise. Although peculiar, this potential is a rich and unexplored topic in community ecology. Plant species differ in the way that they transmit and reflect different light wavelengths, red:far red ratios for example, and in their responses to different wavelengths. As noted in Chapter 2.10, declines in the ratio of red:farred wavelengths of light that occur in the shade of plants can alter the growth of neighboring plants (Morgan and Smith 1979). To my knowledge the only study that has made an effort to examine the effects of species-specific wavelength emissions was that by Marcuvitz and Turkington (2000). They designed experiments using pots to eliminate any possibility of root interactions and to determine whether “the mere presence” of three different species of neighboring grasses, signaled via alterations of reflected red:far red ratios, could influence the growth and morphology of Trifolium repens. As they put it, this could “send a signal of impending competition”. Depending on the time of day, there were significant differences in both the quantity and quality of light penetrating the canopies of the different grass species. At midday, when red:far red ratios would be affected only by the way the grass canopies reflected far red wavelengths to Trifolium, there were no differences in red:far red ratios among the different species and the control. During the afternoon and at dusk though, when both transmission and reflectance contributed to red:far red ratios, ratios were decreased by all three grass species in comparison to the control. In the afternoon, red:far red ratios were reduced significantly more by Dactylis glomerata than Holcus lanatus or Lolium perenne, and at dusk ratios were reduced the most by Holcus. One factor in the differences among species appeared to be the heavy pubescence on Holcus plants which increases the reflection of red light off of its leaves (Thompson and Harper 1988). Dactylis canopies provided unique signals by lowering red:far red ratios while at the same time allowing more total light to reach Trifolium. Despite the effects of the grasses on red:far red ratios, in the absence of shade, there were no effects
Chapter 5
291
of different red:far red ratios on Trifolium growth. New leaves appeared to develop in the direction of increased far red wavelengths which may have strongly reduced the negative effects of neighboring plants. As a package, Marcuvitz and Turkington’s work demonstrated intriguing species-specific effects on light quality, and suggested how differences in light quality might elicit responses from neighbors, but no evidence for the effects of the signal. Even seeds may be able to transmit and receive signals. A number of authors have reported density-dependent seed germination responses in which higher densities of conspecific seeds inhibit germination (Linhart 1976, Murray 1998) or in which high seed densities accelerate germination (Miller 1987, Bergleson and Perry 1989, Dyer et al. 2000). Goldberg et al. (2001) demonstrated negative density dependence in seed germination using whole natural communities of desert annual plants. Mechanistic hypotheses for such signaling among seeds have included the detection of CO2 emitted from germinating neighbors and the release and detection of chemicals released from the seed coat. If the primary signals governing density-dependent responses among seeds are chemical then species-specific responses are possible. Whether or not species-specific signals are facilitative depends on whether or not the signal from another species actually benefits the recipient either by stimulating it to wait for better times, or stimulating it to hustle in what will become a competitive environment. Species-specific stimulation processes have been demonstrated, but to my knowledge the link between stimulation or retardation and ultimate benefit or detriment awaits investigation. An excellent example of species-specific stimulation is found in experiments carried out by Andrew Dyer and colleagues at the University of California at Davis (2000). They compared mean time to emergence for the native bunchgrass species, Nassella pulchra, when seeds were germinated alone, in a low density (five) of different neighbor seeds, and in a high density (from 15 to 50 depending on neighbor seed mass) of different neighbor seeds. Overall the presence of seeds of other species increased the germination rate (decreased the mean time to germination) of Nassella pulchra (Figure 5.19). They did not investigate the individual pairwise comparisons due to low statistical power, but their results strongly suggest that some species accelerate Nassella germination more than others. Interestingly, although most other studies on seed effects have shown negative, or suppressive effects (as did most of Dyer et al.’s intraspecific tests), there were no negative effect of neighbor seeds on Nassella. They concluded that their results supported the hypothesis that propagules “should vary with specific information obtained from the immediate vicinity of the propagule”.
292
Species-Specific Positive Interactions Low density
Time to emergence (hours)
104 102 100 98 96 94
Control At
Ab
Af
Bd
Bh Hm Tc
Neighbor species
High density
Time to emergence (hours)
104 102 100 98 96 94
Control At
Ab
Af
Bd
Bh Hm Tc
Neighbor species
Figure 5.19. Time to emergence of Nassella pulchra seeds planted without interspecific neighbors (open circles), at low densities of interspecific neighbors (a), and at high densities of interspecific neighbors (b). Error bars are 1 SE. At = Aegilops triuncialis, Ab = Avena barbata, Af = Avena fatua, Bd = Bromus diandrus, Bh = Bromus hordeaceus, Hm = Hordeum murinum, Tc = Taeniatherum caput-medusae. Reprinted from Dyer et al. (2000) with permission from Ecology Letters.
Lortie and Turkington (2002) explored the consequences of accelerated germination in the context of species-specific effects of seeds and seedlings on germination from natural Negev desert seed banks in two pot experiments. The first experiment tested the effect of adding seeds from either Erodium laciniatum or Erucaria pinnata, two large dominant annual species not present in the original seed bank, to pots sown with a natural density of the natural seed bank community from the desert soil. In the second experiment the same natural seed bank community was used, but seedlings of Erodium, Erucaria, or the smaller annual Trifolium tomentosum were established before adding seed bank. In both experiments they found evidence for species-specific increases in germination rates. In “dry” treatments, Erodium and Erucaria seeds positively affected germination from the seed bank only when they were planted together. When pots were kept relatively wet the addition of Erodium seeds resulted in significantly higher germination from the seed bank than any other treatment. Interestingly, the germination of Erodium was enhanced by the presence of the seed bank. The strongest species-specific effects, however, were seen when
Chapter 5
293
seedlings, rather than seeds, were established before germination of the seed bank. Early in the growth period, the density and biomass (2x) of annual species were higher with Erodium seedlings than with either Erucaria or Trifolium, or when the seed bank was germinated alone. Erodium promoted the germination of seeds of other species and also appeared to nurse them in later lifestages. The finding that Erodium had this positive effect, while other annuals did not, suggests a species-specific facilitative effect produced by some kind of chemical signaling. However, the ultimate facilitative effect of one plant stimulating the germination of another is uncertain. But in general, getting a head start is a good thing for annual plants because those that germinate first grow larger and complete their life cycle sooner (Miller 1987, Bergleson and Perry 1989). The fact that Lortie and Turkington found strong facilitation among relatively similar annual species in a greenhouse is remarkable, and indicative of unusually intense processes. This is because greenhouse conditions often ameliorate the harsh natural conditions that reveal positive mechanisms, leaving competition as the dominant interaction.
5 .2.6. Implications for community ecology Although some facilitative relationships appear to be due to simple nonspecies-specific changes in the biophysical environment, many others appear to be species-specific. Most positive interactions among plants also appear to be commensal, with only one species benefiting from the presence of the other. However, positive effects may be reciprocal (Pugnaire et al. 1996a) and some facilitative processes such as resource sharing via mycorrhizae and root grafts, and associational defenses have the potential to be mutualistic. If beneficiaries opportunistically take advantage of favorable changes caused by benefactors, coevolution is unlikely, but if strong reciprocal species-specific interactions exist, they have the potential to promote coevolution within plant communities. Species-specific positive interactions also have important consequences for community theory. Highly species-specific positive interactions within plant communities are strong evidence for interdependence in plant communities.
CHAPTER 6
POSITIVE INTERACTIONS AND COMMUNITY ORGANIZATION
Facilitative processes are highly relevant to several conceptual models in ecology. For example, Bruno et al. (2003) pointed out how facilitation provided new perspectives on the niche (Figure 6.1), the relative importance of species interactions along productivity gradients, diversity-community attribute relationships, and disturbance-diversity relationships. More recently, Lortie et al. (2004) argued convincingly that ubiquitous positive interactions among plants demonstrate the inadequacy of individualistic theory as a general foundation for plant community ecology (also see Callaway 1997b). In Chapter 4, I discussed gradient-interaction relationships in detail and in the Introduction I touched on facilitation and individualistic theory. In this chapter I explore the importance of positive interactions for niche theory, the role of diversity in community function, stability, productivity, and evolution, and how positive interactions make the case against the concept of individualistic plant communities.
herbivory
indirect competitive interactions
indirect competitive interactions
abiotic conditions
resource enhancement
protection from herbivores
competition
FUNDAMENTAL NICHE
REALIZED NICHE
abiotic amelioration
Figure 6.1. Diagram illustrating the process by which facilitation can expand the realized niche. Redrawn from Bruno et al. (2003) with permission from Trends in Ecology and Evolution.
295
296
Positive Interactions and Community Organization
6.1. POSITIVE INTERACTIONS AND THE EXPANSION OF NICHE SPACE Implicit throughout this book is the idea that many plant species increase the realized niches of other species. Typically it is assumed that the performance of a species along a set of relevant environmental variables is sufficient to explain its distribution or fundamental niche (Osmond et al. 1987, Shugart 1998) and competitive and consumer interactions have been incorporated into definitions of the realized niche. Accordingly, some biotic interactions such as herbivory and competition are now widely recognized as important driving forces shaping species distribution along gradients (Austin 1985, Louda et al. 1990). Discrepancies between realized and fundamental niches are virtually always attributed to resource competition (Ellenberg 1953, McIntosh 1967, Whittaker 1967, Austin and Smith 1989). However, van der Maarel et al. (1995) and Wilson et al. (1994) moved beyond competition and the realized niche and discussed “niche limitation” and “niche facilitation” based on spatial patterns and diversity. van der Maarel (1995) examined deviations from an expected variance in species richness in a grassland in order to differentiate between significantly left-skewed curves (‘niche limitation’), significantly right-skewed curves (‘niche facilitation’) and symmetrical curves, and found substantial evidence for niche facilitation. The experimental evidence described in this book clearly demonstrates that both positive and negative interactions must be incorporated into the concept of the niche. Competition and facilitation may act somewhat symmetrically at different margins of a species’ distribution with competition as a limiting factor at one extreme and facilitation an expanding factor at the other extreme.
6.2. POSITIVE INTERACTIONS AND THE ROLE OF DIVERSITY IN COMMUNITY FUNCTION
Understanding the effects of diversity on ecosystem functioning and emergent community properties has become one of ecology’s holy grails (Chapin et al. 2000, Loreau et al. 2001). A great deal of experimental research, based primarily on grassland communities, has found positive relationships between plant diversity and ecosystem attributes (Tilman et al. 1996, 1997, 2001, Naeem et al. 1996, Hooper and Vitousek 1997, Hector et al. 1999, Hooper et al. 2005). Clearly, direct and indirect positive effects of species on each other have the potential to be important mechanisms driving the relationship between community diversity and ecosystem productivity, stability, invasibility, and resource cycling. However, despite the ubiquitous occurrence of facilitation in natural communities positive interactions have not received a great deal of attention in the diversity-function literature. In fact, facilitation, as
Chapter 6
297
defined and described throughout this book, is often lumped together conceptually with “complementarity” in diversity-function theory. Complementarity is the idea that divergence in niche space among the species in a community can allow for greater total community utilization of resources. But complementarity is clearly not facilitation. Complementarity occurs when performance of a species is greater because neighbors do not substantially infringe on its resource requirements – reduced competition. In contrast, facilitation is when a species actually benefits from the presence of its neighbor. Separating these very different conceptual mechanisms may allow a better understanding of the role of diversity on community and ecosystem functioning. One of the central controversies in the diversity-function literature is whether diversity per se, versus the effects of particular species, enhances ecosystem function (Spehn et al. 2002). Direct facilitation, caused by mechanisms described in Chapters 2 and 3, is likely to have species-specific effects (Chapter 5). This is because direct facilitation is usually caused by particular traits, such as the ability to hydraulically lift water, mine nutrients from deep soils, bind mobile substrate, or tolerate high winds. However, direct facilitation can also be caused by non-specific processes such as wholecommunity boundary layer effects (see Mulder et al. 2000, below), and such effects may contribute to role of diversity on ecosystem functioning. Indirect facilitation involving consumers, if due to “shared resistance” (Chapter 3), is not likely to affect ecosystem function through diversity. This is because shared resistance is derived from particular defense traits of species. In contrast, indirect facilitation due to “associational resistance” is likely to drive diversityfunction relationships through the effects of diversity because associational resistance is derived simply through the complexity of community structure and chemical composition. Facilitation caused by indirect interactions among competitors (Chapter 3.7) may be the most overlooked process in the diversity-function debate. This is because indirect interactions among competitors are extremely difficult to measure, much less to connect experimentally to ecosystem function. Indirect interaction among competitors may be species-specific in some cases, and when so is likely to drive diversity-function relationships via the presence of certain species. However, the phenomenally complex indirect webs of interactions that may occur among competing plants (Miller 1994) are likely to affect ecosystem function through the diversity of a community. To my knowledge, the role of indirect facilitation among competitors has never been explored as a potential cause of diversity-function relationships. One of the best experimental analyses of a facilitative mechanism driving the effect of diversity on ecosystem function has been for stream invertebrates, taxa that have not received much attention in this book. Brad Cardinale and colleagues (2002) showed that increasing the species diversity of stream invertebrates induced facilitative interactions and led to non-additive
298
Positive Interactions and Community Organization
changes in resource consumption. This occurred because increasing diversity reduced the deceleration of water flow from upstream to downstream neighbors allowing more diverse communities to acquire a higher proportion of suspended resources than were captured by monocultures. This difference was based on fundamental differences in the thermodynamic properties of diverse communities, which raises the interesting possibility that similar processes may occur in terrestrial systems. In terrestrial systems “engineered” complexity and diversity can alter fluxes of energy and resources suggesting that species diversity may increase the possibility of positive interactions (Jones et al. 1994, 1997). As stated by Cardinale, “changes in species diversity may alter the probability of positive species interactions, resulting in disproportionally large changes in the functioning of ecosystems”. Strong facilitative interactions must increase community diversity, especially in cases in which beneficiaries have an obligatory or near-obligatory dependence on benefactors. However, whether or not increased diversity, driven by facilitation, feeds back to ecosystem function is less certain. As discussed in Chapter 4, many studies have found that the relative importance of facilitation increases with abiotic stress (Greenlee and Callaway 1996, Callaway 1998b, Choler et al. 2001, Pugnaire and Luque 2001, Callaway et al. 2002). Mulder et al. (2001) reasoned that if environmental stress creates conditions in which facilitation is important, then environmental stress may also create conditions in which plant species richness increases community productivity. In other words, “species that seem to be functionally redundant under constant conditions may add to community functioning under variable conditions” (Mulder et al 2001). They established experimental communities of bryophytes found growing near each other in New Zealand. These communities were assembled so that they contained 1, 2, 4, 8, 16, 24, or 32 species in trays that were grown under ideal, humid conditions for a year. After one year, 1 replicate of each mixture continued in the high moisture conditions, but other replicates were exposed to 5 days of much lower humidity and a 100-200% increase in light. Three months after the drought treatment they harvested the biomass in the trays. In the control trays, where humidity was high and light intensity was low, there was no relationship between species richness and community productivity (Figure 6.2). However, in the bryophyte communities exposed to short-term drought, the biomass of all species combined increased with the species richness of the community. Mulder et al. explored two different hypotheses to explain their results: 1) the insurance hypothesis – increased productivity in drought is due to the increased probability that the community contains species that are capable of dominating under the changed environmental conditions, 2) the positive interactions hypothesis – an increase in positive interactions in drought conditions among plants drives the relationship between diversity and productivity. This general mechanism is discussed at the beginning of this section. If the insurance hypothesis was to be supported, they
Chapter 6
299
expected to find three different principles in their results. First, the increase in biomass with species richness in drought conditions would have to be due to drought-resistant species showing proportionally greater increases in biomass at the expense of drought-sensitive species, resulting in greater variation in biomass between species in the drought treatment than in the control. However, they found that variation in biomass among species in drought conditions was not significantly different than variation in control conditions at all levels of diversity. Second, for the insurance hypothesis to be supported, increased biomass with diversity in drought conditions would have to be due to species that are the most resistant to drought also being those that show the greatest increase in biomass with diversity. Instead, they found that increased biomass under drought conditions was associated with increased survivorship for most species. Biomass per surviving plant did not change significantly with species richness in drought, but species demonstrating the greatest increases in biomass in the drought conditions were the least resistant to drought. Third, to support the insurance hypothesis the proportion of species contributing positively to increasing biomass should have remained constant or decreased with diversity. Instead, they found that for a given species pool, the number of species contributing positively to biomass increases also increased with diversity.
180
Control Drought Plot 1 Regr
Biomass (g/m2)
150 120 90 60 30 0 0
5
10
15
20
25
30
35
Number of species Figure 6.2. Total community biomass of bryophytes as a function of the number of species in the community. A) Mesic control conditions, F(1, 98) = 0.37, P = 0.60, B. Xeric drought conditions, F(1, 97) = 10.48, P = 0.002. Redrawn from Mulder et al. (2001) with permission from the Proceedings of the National Academy of Sciences.
300
Positive Interactions and Community Organization
Because none of the expectations for the insurance hypothesis were met, Mulder and colleagues argued that the alternative explanation, an increase in positive interactions among plant in stress conditions, was the reason for increased productivity with species diversity. These positive interactions could have occurred due to a general increase in humidity with plants with different architectures creating a more complex community boundary layer in diverse communities and trapping more transpiration. Tall bryophytes may have protected shorter species from photoinhibition, a decrease in photosynthetic capacity that often increases in drought conditions (Martin et al. 1985, Martin 1994). Regardless of the mechanism, the large increase in bryophyte survival, but not biomass per individual (for almost all of the species used in the experiment) indicates that physical conditions were ameliorated, rather than a reduction in competition intensity. This conclusion is supported further by the finding that the species most negatively affected by drought were the ones that showed the strongest positive response to diversity, and by the observation that increasing the number of different growth forms (rather than species) increased community biomass only in drought conditions. Also see Fenton and Bergeron (2006) for an example of facilitation among bryophytes. In Portugal, Maria Caldeira et al. (2001) constructed plots with different numbers of Mediterranean grasses and forbs (sown at 1, 8, and 14 species) to investigate mechanisms that might drive diversity-productivity relationships. After two years they measured plant cover, the number of species present, soil moisture in their plots, and stable carbon isotope ratios (δ13C) in the leaves of five different species with C3 physiology. They found that the total biomass and total cover in species-rich plots was significantly higher than in monocultures. This was interpreted by Caldeira and her colleagues as indicating either niche complementarity (different species morphologically or physiologically adapted to acquire resources from different parts of the environment) or facilitation driving increased productivity. However, fundamental findings in this study did not indicate niche complementarity. The niche complementarity hypothesis would have been supported if mixtures of species performed better as a community, but not as individuals. In other words, enhanced community performance would simply be due to the inclusion of species that were each better at exploiting resources in a particular dimension of space or time. Caldeira and her colleagues found that individuals performed better in mixtures. For four of the five individual species tested, leaf δ13C indicated significantly higher wateruse efficiencies in the species-rich mixtures than in monocultures (Figure 6.3). CO2 gradients were minimal, leaf temperatures varied less than 1oC, and their were no differences in leaf nitrogen concentration, suggesting that differences in δ13C were due to environmental modification within diverse communities and not differences in photosynthetic capacity among species. Increased plant richness and cover probably decreased surface evaporation losses, decreased leaf
Chapter 6
301
temperatures and the leaf to air vapor pressure gradient (see Kikvidze 1996), and perhaps increased dew accumulation. The fact that individual species had higher water-use efficiencies in diverse plots strongly indicates facilitation, not niche complementarity. Niche complementarity may have also occurred of course (suggested by lower soil water content in plots with mixtures), and neither mechanism was shown to be the cause of the diversity-productivity relationship - facilitation simply occurred in the same plots in which diversity enhanced productivity. However, this study, and that of Mulder et al. (2001) opens a window into the incorporation of facilitation into the issue of community diversity and ecosystem function. -27.40
Plantago lanceolata
-27.90 -28.40 -28.90 -29.40 -29.90 -30.40 -27.40
Holcus lanatus
-27.90 -28.40 -28.90 -29.40 -29.90 -30.40 -27.40
Dactylis glomerata
13
δ C (%0)
-27.90 -28.40 -28.90 -29.40 -29.90 -30.40
-27.40 -27.90
Trifolium subterraneum
-28.40 -28.90 -29.40 -29.90 -30.40 -27.40 -27.90
Ornithopus compressus
-28.40 -28.90 -29.40 -29.90 -30.40
Monocultures
Mixtures
Figure 6.3. Leaf carbon isotope ratio (δ13C) of five Mediterranean species grown in monocultures and species-rich mixtures (8 and 14 sown species). Error bars show 1 SE. Reprinted from Caldeira et al. (2001) with permission from Ecology Letters.
302
Positive Interactions and Community Organization
As discussed in Chapter 3, indirect mechanisms such as interspecific attraction of pollinators or dispersers, shared defenses against herbivores, or resistance to pathogens increase community diversity. Much like the direct facilitative mechanisms described by Mulder et al. (2001) and Caldeira et al. (2001), these indirect facilitative mechanisms can also drive the effects of community diversity on ecosystem function. Johannes Knops and colleagues (1999) built experimental communities with each plot assigned a diversity of either 1, 2, 4, 8, or 16 species chosen from a pool of 18 species (see Tilman 1996). Disease severity was quantified in the plots by assaying leaves of 20 plants per plots for four target species which occurred in 102 of the plots. All of these diseases are species-specific and had no alternate hosts among the other species used in the experiment. For each of the four target plant species, foliar disease was significantly negatively correlated with plant species richness. However, they found that disease severity was more strongly dependent on host plant density, not richness per se, suggesting that disease transmission was simply a function of the richer plots having lower densities of host plants. Although not a clear example of one species indirectly benefiting another, Knop et al.’s study suggests an approach that may eventually bear fruit for understanding indirect roles of facilitation in determining how diversity affects community properties. For example, examining plant diversity in the context of negative feedbacks between different species and the particular microbial communities that develop in their rhizospheres may be profitable. As described in Chapter 3, a number of studies have demonstrated negative feedbacks among soil biota and plants that occur when plant species accumulate host-specific pathogenic microbes (Van der Putten et al. 1993, Bever et al. 1996, 1997, Bever 2003, Callaway et al. 2004). Negative feedbacks should enhance community diversity by increasing species turnover rates. The turnover associated with these feedbacks may be much greater as the number of species increases, or a large number of species may retard the development of strong negative feedbacks in general. The facilitative mechanisms driving the effects of diversity in Mulder’s and Caldeira’s studies are not clear, but appear to be related to the effects of the community on water relations. In contrast, Spehn et al. (2002) investigated the effects of plant diversity on nitrogen accumulation in plants. They varied the number of plant species and functional groups (grasses, herbs and legumes) in experimental grassland communities across seven European experimental sites and two years after starting the experiment found that diversity and community composition had strong effects on ecosystem properties. Two years after sowing, nitrogen pools in Germany and Switzerland strongly increased in the presence of legumes, but less so at other sites (Figure 6.4). Over all sites, nitrogen concentrations in the tissues on non-legumes were very similar regardless of the presence of legumes, averaging 1.66±0.03% across all sites and diversity treatments. But legumes had a positive effect on nitrogen and also significantly increased above-ground plant biomass. The legume effect was the
Chapter 6
303
strongest at the German site, and there nitrogen fixed by legumes was transferred to grasses and forbs. There was a high degree of species-specificity in this process, and transfer depended on the particular legume species fixing N and the particular grass or forb taking it up. These results demonstrate that the facilitative effects, caused by nutrient addition (see Chapter 2), may contribute to the effects of diversity on ecosystem function. To my knowledge, no other studies have directly investigated the potential for facilitation to affect the relationship between community diversity and function. However, if the causal relationship between facilitation and productivity indicated by Spehn’s, Caldeira’s and Mulder’s research is a broad phenomenon, elucidating this relationship has the potential to profoundly alter the way we understand the role of positive interactions in community structure. The widespread idea that some species may be “redundant” in communities is not acceptable until so-called redundancy is shown in a realistically broad range of environmental conditions experienced by a community in nature. Ignoring the possibility that “redundant” species in non-stressful conditions may either elicit or respond to facilitative mechanisms in stressful conditions will result in an underestimation of the value of species diversity in plant communities. Sometimes the positive effects of species on diversity are only discernible at certain scales (also see section on scale below). Some studies of canopy effects in Mediterranean-climate shrub lands of Australia and South Africa indicate that overstory species reduce understory diversity (Cowling and Gxaba 1990, Keith and Bradstock 1994). However, studies in the mountain fynbos of South Africa suggest that the relationship between overstory cover of
Nitrogen concentration in above-ground biomass (%)
4
Sweden
Ireland
Great Britain
Germany
Switzerland
Greece
3 2 1 0 4
Portugal
3 2 1 0
0
500
1000
0
500
1000
0
500
1000
0
500
1000
1500
Above-ground biomass (g m-2) Figure 6.4. Effects of legumes on aboveground biomass (g m-2) and tissue nitrogen concentration (%). Closed symbols represent assemblages with legumes, open symbols assemblages without legumes. Reprinted from Spehn et al. (2002) with permission from Oikos.
304
Positive Interactions and Community Organization
proteas (species in the Proteaceae), and understory species are much more complex. Vlok and Yeaton (1999) found that when α-diversity in fynbos was measured using small quadrats (< 5 m2); overstory proteas decreased the number of understory species. However, when quadrats were increased to 100m2, higher species richness was associated with cover of proteas. They attributed this contradiction to the roles proteas play in maintaining community patchiness over spatial scales large enough to include protea patches and the matrix of vegetation surrounding protea patches, and over time-scales long enough to incorporate the effects of proteas on fire-generated community dynamics. At small scales, the competitive effects of proteas diminish understory diversity in part because they suppress other dominants, mostly species that can re-sprout rapidly after fire. This allows species otherwise suppressed by the re-sprouting dominants to persist, an indirect interaction (see Chapter 3). A crucial component of understanding the diversity-enhancing role of proteas at large spatial scales is repeated fire disturbance. Re-occurring fires do not allow proteas to expand their dominance across the landscape because they are obligate re-seeders after fire, requiring much more time than the re-sprouters to establish. Without fire, proteas may actually suppress diversity. With fire, proteas create small patches in which some species are protected from competitive exclusion by the resprouters which therefore increase landscape-scale diversity. Many ecologists have described a “humped-back” relationship between species richness and community biomass. This pattern has been explained by competitive exclusion of poor competitors at high community biomass (or productivity) and the exclusion of stress intolerant species at low community biomass. However, research on facilitation has shown a clear relationship between the effect of facilitation on community structure, community biomass, and the severity of the environment (see Chapter 4). Based on this observation Richard Michalet and colleagues (2006) explored whether biotic interactions might potentially shape both sides of the humped-back model for species richness commonly detected in plant communities. They proposed that facilitation promotes diversity at medium to high environmental severity levels, by expanding the realized niche of stress-intolerant competitive species into harsh physical conditions. However, they proposed that when environmental conditions become extremely severe the positive effects of the benefactors may decrease and thereby reduce diversity. The inclusion of facilitation into the classic humped back model of plant species diversity has the potential to generate a more complete framework for understanding the importance of plant interactions as drivers of community diversity.
Chapter 6
305
6.3. POSITIVE INTERACTIONS AND SPATIAL SCALE Very few studies of positive interactions have explicitly incorporated spatial scale into the design. Tewksbury and Lloyd (2001) used multivariate ordination in an imaginative way to examine the effects of the dominant Sonoran Desert tree, Olneya tesota at different scales. They used detrended correspondence analysis (DCA) to compare Olneya u nderstory to open matrix plant communities in 10 1-ha study sites on a 625 km long section of the Central Gulf Coast region of Sonora, Mexico. The key to understanding ordination analyses such as DCA is to know that their fundamental purpose is to create a picture in which samples (plots) are graphed near one another if the samples are similar in composition and far from one another if samples are different in composition. Five sites were located in mesic drainage channels of ephemeral watercourses and five were located on xeric upland bajadas. In each of the 10 sites they measured community composition in the understories of all trees in the 1-ha study site. This allowed them to measure the effect of Olneya at three scales: landscape, topographic (mesic versus xeric habitats), and local (at individual sites). At the landscape scale, they combined all plots from all sites into a single DCA to determine if Olneya plots were consistently different from plots in the open matrix across all xeric and mesic sites spanning the entire 625 km range of their study. At the scale of mesic and xeric habitats, they separated low water stress (mesic) sites from high water stress (xeric) sites and ordinated each in separate DCAs to determine if the effect of Olneya was consistent once the effect of water stress was removed. Finally, at the smallest scale (that of individual sites), they conducted separate DCA’s for each of the ten sites to determine the effect of Olneya canopies using only the species pool present within each site. The sources of variation in plant community structure (ordination scores in the DCA) were examined at each scale using ANOVA models. At the largest scale, the ANOVA model included water stress (xeric or mesic), the presence of Olneya, and the interaction. At the next smallest scale where they ordinated xeric and mesic sites separately, site and the presence of Olneya were important factors. At the site scale, they compared plant community structure between Olneya and control plots within each site using one-way ANOVA models. Conducting a hierarchical analysis in this fashion found that facilitation from Olneya canopies causes consistent and predictable changes in plantcommunity structure. In more detail, Olneya canopies had strong effects on species richness at all scales (Tewksbury and Lloyd, 2001), but Olneya canopies did not explain a significant amount of variation in plant community composition at the landscape scale along either DCA axis. (ANOVA for primary axis scores FOlneya=0.01, df=1,112, P>0.9; Figure 6.5). In contrast, water stress (mesic versus xeric sites) was highly significant for the scores on the primary axis (ANOVA for primary axis scores FOlneya=104.5, df=1,112, P<0.001), indicating
306
Positive Interactions and Community Organization
that the variation in total composition of plant communities at this scale was due more to differences in water availability among sites than the presence of Olneya. In other words, topography far out weighed the facilitative effects of tree canopies as a driver of community composition. However, when community composition was analyzed at the local scale, with paired xeric and mesic plots, Olneya had a large effect on plant community structure (Figure 6.6). The differences between composition of plant communities under Olneya canopies and plant communities in the open was significant at 6 of the 10 individual sites when separate ordinations were conducted for each site (Figure 6.7). Olneya plots were different than control plots in four of the five xeric sites, and only two of the five mesic sites. In other words, discerning the facilitative effect of Olneya required measurements at the appropriate scale.
300
Secondary axis
Olneya -- mesic sites Control -- mesic sites Olneya -- xeric sites
200
Control -- xeric sites Olneya -- overall Control -- overall
100
0 0
100
200
300
Primary axis Figure 6.5. Results of the Detrended Correspondence Analysis (DCA) ordination conducted across all plots (landscape-level ordination). Raw scores along axes 1 and 2 are shown for all plots (small symbols), and overall mean scores and 95% confidence intervals are shown for Olneya and control plots in mesic sites, xeric sites, and all sites combined (large symbols). Reprinted from Tewksbury and Lloyd (2001) with permission from Oecologia.
Chapter 6
307
Secondary axis
300
A: Xeric
250 200 150 100 50 0 0
Secondary axis
300
100
200
B: Mesic
300 Olneya Control
250 200 150 100 50 0 0
100
200
300
Primary axis Figure 6.6. Results of the DCA ordination conducted across all xeric plots (A) and all mesic plots (B). Raw scores along axes 1 and 2are shown for all Olneya plots (small dark-gray circles) and all control plots (small gray triangles). Overall mean score and 95% confidence intervals are shown for Olneya plots and control plots (large symbols). Reprinted from Tewksbury and Lloyd (2001) with permission from Oecologia.
The masking effect of mesic and xeric landscape features on canopy effects is not as evident in Quercus douglasii savanna in central California. Inspired by Tewksbury and Lloyd’s use of DCA, I analyzed two data sets of understory and open species collected from two locations, approximately 500 km apart. The first was located at Hasting Reservation in the coastal mountains of California near Carmel at 36.4o latitude. The second was at the Sedgwick Ranch near Santa Barbara, California at 34.5o latitude. DCA scores for understory communities at the two sites overlapped completely as did the scores for
308
Positive Interactions and Community Organization
open communities, indicating that the communities were very similar (Figure 6.8). However, DCA scores for the two understory communities were very distinct from the two open communities. These results, although much less extensive than those of Tewksbury and Lloyd, suggest that the effect of Q. douglasii canopies on understory community composition is stronger than large-scale landscape effects.
P = 0.75
120 100 80 60 40
100 80 60 40 20
125 100
M2
P = 0.63
75 P = 0.01
50
X3
100 80 60
P < 0.001 M3
40 P = 0.16 P = 0.008
160 140 120 100 80 60
X2
M1
P = 0.06 P = 0.91
100 80 60 40 20 0
P = 0.37
P = 0.02
X4
P = 0.03
20 200 150 100
M4
50 P = 0.009
P = 0.85
Secondary axis
40
P = 0.14
60
120 100 80 60 40 20
P = 0.18
80
X1
Mesic Sites
X5
P = 0.50
0 150 125 100
P = 0.63
100
P = 0.32
Xeric Sites
M5
75 P < 0.001
P = 0.39
50
0 75 150225300
0
75 150 225
Primary axis Figure 6.7. Mean DCA axis 1 and 2 scores for Olneya (dark gray triangles) and control (light gray circles) plots from ordinations conducted at each site. Error bars are 95% confidence intervals. P values are for difference between Olneya and control plots. Reprinted from Tewksbury and Lloyd (2001) with permission from Oecologia.
Chapter 6
309 350 300
DCA Y axis
250 200 150 100 50 0
0
50
100
150
200
250
300
350
DCA X axis Figure 6.8. Community composition plots of all herbaceous species under Quercus douglasii canopies (solid symbols) and in open grassland (open symbols) and at different sites in California. Circles are for the Hasting Natural History Reservation and triangles are for Sedgwick Ranch, UC Santa Barbara Preserve.
Another good example of how important spatial scale can be for identifying indirect facilitative interactions is that of the “magnet species” effect on pollinator attraction to neighboring plants. Johnson et al, (2003) correlated the visitation rate to flowers of the non-rewarding orchid, Anacamptis morio, which had been experimentally transplanted into different magnet communities of rewarding plants, with the density of sympatric nectar plants in 1-m2 and 100-m2 plots. At the level of whole meadows (>0.5–2 ha) visitation rate to flowers of A. morio was not correlated with the smallest patches of Geum rivale and Allium schoenoprasum magnets, but showed a significant positive relationship with density of these nectar plants in the smaller 100-m2 plots. In addition, visitation to flowers of A. morio was strongly and positively related to the density of A. schoenoprasum at the level of the meadow.
6.4. POSITIVE INTERACTIONS AND STABILITY IN PLANT COMMUNITIES
As demonstrated in detail in Chapters 2 and 3, many plant species recruit in environments favorably modified by other plant species. In many communities
310
Positive Interactions and Community Organization
this arrangement is predictable enough to suggest almost obligate dependency. In some cases positive effects are determined by species-specific benefactors (Chapter 5), making the case that facilitation may maintain particular components of the community and increase community stability. On the other hand, facilitation is often conditional or context-dependent (Chapter 4). In other words, the positive effects of one species on another may be very strong in certain abiotic conditions, but weak or even shift to competition in other abiotic conditions. In the case of beneficiaries with perennial life histories the net effect of benefactors may establish spatial association patterns, which can signal facilitation despite year-to-year variation in interaction strength (Greenlee and Callaway 1996). In the case of beneficiaries with annual life histories, however, it is possible to quantify annual variation in spatial patterns that suggest facilitation or competition and directly address questions of stability. Deserts provide highly variable environments over time because of extreme variation in precipitation, the primary limiting factor for plant growth and fitness. Tielbörger and Kadmon (1997) conducted a three-year study of the spatial relationships between annual plant species and shrubs in the Negev Desert of Israel with the intent of examining year-to-year variation in shrubannual interactions (see Chapter 4.2). They pointed out that the majority of the large number of studies on the positive effects of desert shrubs on annuals were based on one year’s measurements and that the general conclusion from these short-term studies was that “for annual plants, the shrubs produce a strong and relatively stable pattern of microsite differentiation” (see Schmida and Whittaker 1981). In contrast, Tielbörger and Kadmon found that the effects of shrubs were variable from year to year, and only one species, Launaea tenuliloba, showed a consistent and significant preference for shrub understories over three years that varied in precipitation amount and distribution. In two relatively wet years with a significant amount of precipitation in the fall and early winter ordination analyses demonstrated large differences between communities (emerging plants) under shrubs versus in openings. In an exceptionally dry year with precipitation in the late winter and spring the composition of the annual communities under shrubs and in the openings were much more similar. The community-scale differences seen for “emerging” plants were much less distinct than for “reproductive” plants, those that survived to produce seeds. For reproductive plants the largest compositional differences between shrub and open habitats were in the wettest year (168mm), but were much smaller in the year with intermediate precipitation (50mm). Shrub effects could not be tested in the drought year (35mm) because of the very low number of surviving annuals. Tielbörger and Kadmon argued that because their results demonstrated conditionality in the shrub-annual herb relationship, the positive effects of shrubs did not constitute a “stable source of niche differentiation for desert annual communities”. At the scale of some individual species this appeared to be true. However, only one
Chapter 6
311
“emerging” species and one “reproductive” species showed significant reversals in preference for shrubs or open spaces and 7 of 8 species with a significant microhabitat preference in at least one year retained that preference in the 168mm-year and the 50mm-year. One species, Ifloga spicata, which was abundant enough to analyze in the 35mm-precipitation year also retained its longterm preference for open spaces. Considering this degree of habitat consistency, shrubs in the Negev may not provide a completely predictable refuge for annuals, but for many species the relationship was quite stable. Tielbörger and Kadmon’s data also showed that even at the scale of communities, shrubs do not appear to stabilize community composition. Both annual and spatial variation in community composition of annuals was much higher under shrubs than in the openings. There are factors that should be considered when interpreting Tielbörger and Kadmon’s results. First, in the drought year in which communities under shrubs were shown to converge with those in openings extreme stress appeared to eliminate the number of reproductive species occurring in sample plots to 1 under shrubs and 3 in the openings. If extreme drought eliminated the ability of the vast majority of species simply to make it to maturity, then shrubs may be largely irrelevant during these years and comparisons of ordination analyses may be misleading. Second, sample sizes were quite small and shrubs were not identified to species, obscuring potential species-specific positive effects of benefactors that have been observed in other deserts (Went 1942, Muller 1953, Patten 1978). Even so, the dramatic year-toyear variation that occurs in these desert communities presents a strong case for the dangers of over-interpreting single-year studies, even for annual species, and questions the role of shrubs in deserts as highly stable, predictable microhabitats for annual species. In contrast to Tielbörger and Kadmon (1997) Andrew Wilby and Moshe Sachak (2004) of Ben-Gurion University found that shrubs in the Negev Desert provided substantial stability to annual communities. They tested the effects of granivores and shrubs on annual plant community dynamics and found that annual communities under shrubs were more species rich, and had higher plant density than in open spaces between the shrubs. Most importantly richness and diversity were temporally less variable over time (five years) under shrubs than in the open matrix. In fact, seven of the ten most common species had significantly lower coefficients of variation for abundance under shrubs than in the open, and no common species showed greater variation in abundance over time in the open. Species richness and abundance were also more resistant to drought under shrub understories compared with the open.
312
Positive Interactions and Community Organization
6.5. FACILITATION AND PRODUCTIVITY When overstory trees increase the productivity of the plants growing beneath them, the combination of tree productivity and herbaceous productivity is obviously much higher than in open areas without trees. For example, in Q. douglasii woodlands herbaceous productivity is over two times higher than in the open and total tree leaf and stem production of the oaks is estimated to be even greater than that of the herbs (R.M. Callaway, unpublished data). One of the most thorough investigations of facilitation on ecosystem productivity was conducted by Duncan Patten in the Sonoran Desert (Patten 1978). He harvested ephemerals in habitats created by trees and shrubs and in open spaces throughout two growing seasons and dried the harvested material for mass and caloric measurements. Mass measurements allowed computation of standard final biomass and repeated measurements of mass allowed computation of rates of productivity. Caloric measurements were compared to solar input to determine production efficiencies (percent solar energy used). Directly beneath the small tree, Cercidium microphyllum, seasonal rates of primary productivity (SPP) were 1.45 to 1.62 kg-1ha-1day-1 and under the shrub, Ambrosia dumosa, SPP reached 0.48 to 0.42 kg-1ha-1day-1 in a wet year (Figures 6.9 and 6.10). In the same year, SPP rates at the edge of Cercidium canopies were 0.76 and 0.37 kg-1ha-1day-1and in the open, 0.36 kg-1ha-1day-1. Although total productivity was much lower in a subsequent dry year, even stronger facilitation patterns were apparent for annuals under Cercidium (0.418 beneath to 0.045 in open).
Primary production (g-1m-2)
50
0
-50
-100
Cercidium Ambrosia Larrea Figure 6.9. Primary production of annual species beneath three perennial shrubs in the Sonoran Desert. Redrawn from Patten (1978) with permission from American Journal of Botany.
Chapter 6
313
Productivity under shrubs relative to open (%)
50 wet season (500 mm) dry season (100mm) 0
-50
-100
Cercidium Ambrosia Larrea Figure 6.10. Productivity under Sonoran Desert shrubs, relative to open productivity, in a dry season and a wet season. Redrawn from Patten et al. (1978) with permission from the American Journal of Botany.
Other interesting patterns were clear when productivity was converted to percent production efficiency (total cumulative solar energy entering a habitat divided by the caloric content of the annuals). The effect of Cercidium was even more striking when these calculations were made. In the open areas without perennial canopies production efficiency was 0.02% of solar energy converted to caloric energy. Under Cercidium this was over thirty times higher, 0.68%. Although an impressive example of how facilitation can affect productivity, we should also be impressed by how estimates of productivity (biomass) can differ from direct measurements, and how this difference can affect our understanding of the roles of species in communities. To my knowledge Patten’s study is the only one in which facilitative effects have been calculated as production (biomass), productivity (kg-1ha-1day-1), and production efficiency (chemical energy stored as a function of solar energy input). Quantified facilitative effects on productivity were 40% higher when calculated as biomass, 830% higher when calculated as productivity, and 3,300% higher when calculated as production efficiency because substantially less light energy was available under canopies (Figure 6.11). Production efficiency is quite derived and perhaps of less interest as a metric of interaction, but productivity as a metric is very important. Patten was measuring ephemeral species that
314
Positive Interactions and Community Organization
Proportional effect of Cercidium on productivity relative to open (%)
completed their life spans in weeks or months. Virtually all other similar studies reasonably assume that biomass of annual species at the end of the growing season is a reliable measure of annual productivity (Callaway et al. 1991, for example). Patten’s results, however, indicate that there is a much greater turnover of species or far more recruitment throughout the year in the understory of Cercidium than in the open – otherwise Cercidium’s effects on biomass would be much more similar to its effects on productivity. If biomass and productivity are this unconnected in other ecosystems we often may be substantially underestimating facilitative effects. The choice to use biomass or productivity as a metric for the effects of Ambrosia dumosa would appear to be even more important. The two-year average of biomass under Ambrosia indicated a 72% decrease relative to the open. However, comparing Patten’s calculations of seasonal productivity indicated a 20% increase under Ambrosia. 4000
3000
2000
1000
0
Biomass -1 -1 (kg ha )
Productivity -1 -1 -1 (kg ha day )
Production Efficiency (%)
Figure 6.11. Comparisons of different metrics of productivity determined by Patten et al. (1978) for the effects of Cercidium on desert annuals. Redrawn with permission from the American Journal of Botany.
While others have found Ambrosia to facilitate understory biomass (Holzapfel and Mahall 1999), these comparisons again suggest that biomass as a metric may underestimate facilitative effects. This can only be resolved by painstaking measurements and integrating careful demographic studies into community research. As discussed in Chapter 2, early successional Pinus albicaulis (whitebark pine) trees appear to facilitate Abies lasiocarpa (subalpine fir) near timberline under some environmental conditions in the Northern Rocky Mountains. In this context, Callaway et al. (2000) quantified the biomass
Chapter 6
315
allocation and productivity of harvested individual P. albicaulis trees and used these age-allocation relationships to estimate productivity in natural forests by applying regression models for harvested trees to all trees in each stand. Stands varied from 67 to 458 years in age. They hypothesized that that the increasing (and potentially facilitated) abundance of A. lasiocarpa during succession might maintain high rates of annual productivity as these forests aged. The estimated productivity of P. albicaulis in these forests increased for approximately 200-300 years and then slowly decreased over the next 200 years. In contrast, as stands shifted in dominance from pine to A. lasiocarpa with age, A. lasiocarpa appeared to maintain gradually increasing rates of whole-forest productivity until stands were approximately 400 years old. Although evidence for the facilitative affects of P. albicaulis and facilitated responses of A. lasiocarpa is only for high-elevation timberline forests, it suggests that facilitative processes might play an important role in maintaining high rates of productivity in some subalpine forests.
6.6. POSITIVE INTERACTIONS AND EXOTIC INVASION Although some facilitative interactions are quite species-specific (Chapter 5), many appear to be relatively general. If a species benefits from the shade of an oak canopy in Spain, there is a good chance that the same species will benefit from a different oak canopy in California. Furthermore, such general positive effects may tend to be less offset by competitive effects of the benefactor because there is some evidence that natives are surprisingly poor competitors with invaders. Freeman and Emlen (1995) inferred interactive strengths from spatial associations and performance of many species in a shrub-steppe community in western Utah and found that the aggressive invader, Bromus tectorum, did not appear to be negatively affected by any of the native species present; a pattern not true for any of the native species. In fact, native species “actually facilitated the reproduction of Bromus tectorum.” Callaway and Aschehoug (2000) found that Centaurea diffusa, a noxious weed in North America, had much stronger negative effects on grass species from North America than on closely related grass species from communities to which Centaurea is native. The stronger competitive effects of C. diffusa on North American natives were reflected in weaker effects of the natives on C. diffusa. Despite the growing body of research on how exotic and native plants interact, we do not know much about the importance of facilitation in exotic invasion, with a few exceptions. Lupinus arboreus (bush lupine) is native to California and derives 6070% of its nitrogen budget from nitrogen fixation (Bently and Johnson 1991). Lupinus arboreus may add as much as 185 kg/ha of nitrogen to subcanopy soils (Gadgil 1971, Palaniappan et al. 1979). Soil under Lupinus arboreus shrubs also has approximately twice the exchangeable ammonium concentration and 3-5 times
316
Positive Interactions and Community Organization
the exchangeable nitrate concentration as soils away from shrubs (Maron and Connors 1996). They found that the total aboveground biomass of herbaceous species was more than twice as much in patches where Lupinus had died as in lupine-free grassland. Additionally, the annual grass Bromus diandrus grew much larger in soil from Lupinus-patches than in soil from the open grassland. However, this facilitative effect was manifest primarily for non-native species such as B. diandrus, and nitrogen-rich patches associated with dead Lupinus shrubs were focal points for the facilitation and establishment of many exotic, weedy species. Dead Lupinus patches contained 47% fewer plant species overall and 57% fewer native species. Other experiments conducted by Pickart et al. (1998) showed that the removal of the Lupinus and/or the organic litter that collected under Lupinus did not decrease total or available soil nitrogen over a time span of four years, but highly altered the proportions of native and exotic species that occurred at their site. Experimental removal of the lupine plant and the duff layer increased the percent cover of native species ≈5x, and decreased the cover of exotic grasses from 35% to 0%. They also concluded that the nonnative species played a role in maintaining enriched soil nitrogen. Maron and Jefferies (1999) found that 57-70% of the net amount of nitrogen mineralized annually was taken up by the annual grasses that occupied dead lupine patches and returned to the soil when the annuals died. They estimated that the facilitative effect of the lupines on soil fertility would take 25 years to decline 50%, which demonstrates the longevity of some positive effects. Similar patterns were reported by Carino and Daehler (2002), who found that an inconspicuous annual legume, Chamaecrista nictitans, facilitated the invasion of Pennisetum setaceum (fountain grass) into native Heteropogon contortus grasslands in Hawaii. Over 60 years ago Went (1942) observed that the positive effect of shrub species on annuals tended to be more evident after benefactor mortality than before. Now, the ecology of the region where Went worked is changing, with facilitation of natives by shrubs replaced to a large degree by facilitation of exotic invaders by the same shrub species. Claus Holzapfel and Bruce Mahall (1999) found that native shrubs strongly facilitated exotic grasses and herbs. Considering that several of the invasive annuals they worked with occur at much lower abundances in their native land, it is possible that alien plants have altered natural balances among types of mechanisms that existed among native shrubs and herbs. Much like the patterns that emerged from Holzapfel and Mahall’s research, the positive effects of Quercus douglasii on understory productivity described in detail at the beginning of Chapter 4 was manifest almost exclusively through the facilitation of exotic Eurasian grasses such as Bromus diandrus (Callaway et al. 1991). Facilitative effects did not occur for the sole remaining native bunchgrass species, Nassella (Stipa) pulchra. In savannas, S. pulchra constituted less than 0.5% of the biomass beneath deep-rooted trees that demonstrated strong facilitative effects. In contrast, the biomass of S. pulchra under shallow-rooted trees, which have competitive effects on understory
Chapter 6
317
species, was 3.9% of the total herbaceous understory. In the open grassland Nassella did not occur in any sample plot. Cavieres et al. (2006) found that the nonnative Taraxacum officinale (dandelion) was facilitated at alpine communities of the central Chilean Andes by the cushion plant Azorella monantha. Seedling survival, net photosynthetic rates and stomatal conductance were all higher for seedlings growing within cushions than outside them, suggesting that the microclimatic modifications of the cushion facilitate the establishment and survival of T. officinale. Evan Siemann and Bill Rogers at Rice University observed that the first Sapium sebiferum (Chinese tallow) trees invading grasslands of southern Texas appeared to create high nitrogen and low light conditions that favored their offspring seedlings in competition with the native herbaceous vegetation (Siemann and Rogers 2003). They mimicked the effect of Sapium by manipulating nitrogen and light in field experiments. The growth of Sapium seedlings increased with nitrogen fertilization. Under shading the above-ground biomass of prairie vegetation did not change, prairie vegetation biomass decreased, and tree seedling growth increased. Sapium growth increased dramatically in the treatment with both nitrogen and shade. However, in a second experiment, Sapium growth increased in higher light suggesting that greater Sapium growth at low light levels in the first experiment was probably a consequence of decreased competitive interference from prairie vegetation (see indirect interactions among competitors, Chapter 3), rather than the preference of Sapium for low light. Although they did not manipulate the effects of mature trees on seedlings, their results suggest that intraspecific facilitation, perhaps the ‘invasional meltdown’ of Simberloff and von Holle (1999), are involved in Sapium invasions of grasslands. Chenopod shrublands in southern Australia are being invaded by the succulent Asclepiadaceae, Orbea variegata. Lenz and Facelli (2003) observed that Orbea often grows beneath native shrubs. They conducted a suite of experiments designed to ascertain whether or not facilitation was important to the invasion, and if so, what mechanisms were involved. Their results indicated that reduction in light and temperature under shrubs, rather than increased nutrients in subcanopy soils, are the primary facilitative mechanisms that benefit Orbea. Temperatures above 30oC, which are more likely to occur on the soil surface of open areas than under shrubs, inhibited seed germination. Seedling survival and the growth of established ramets were increased by shade. They concluded that the facilitative effect of shrubs was not obligatory, which is typical of facilitative relationships in general, but that Orbea gained substantially from the presence of the shrubs. Non-native invasive species can modify vegetation establishment after disturbance. Myrica faya, an invasive nitrogen-fixer, establishes after volcanic disturbance and strongly inhibits the establishment and growth of native species (Vitousek and Walker 1989, Walker and Vitousek 1991). In another volcanic system, Mt. Koma in Japan, native communities under non-native Larix
318
Positive Interactions and Community Organization
Seed set (no. seeds/no. florets)
kaempferi canopies have significantly greater richness and diversity than communities under the native Betula ermanii or in the open (Titus and Tzuyuzaki (2003). They hypothesized that Larix accelerates succession of a few smaller native species, but that overall succession will be deflected towards dominance by the introduced species. Heather Davis and colleagues (2004) at UC Davis documented strong Allee effects for invading Spartina alterniflora. Early invading, isolated S. alterniflora plants set about one-tenth the seed as plants growing in denser stands (Figure 6.12), but as the isolated recruits increased in density through clonal growth they produced more viable seeds.
0.5
a
Isolated Meadow
0.4
a
a
0.3
a
0.2
NE
0.1
Ø 0.0 0.6
Germination rate (no. germ.s/expected no. seeds)
(a)
(b)
a
0.5
a
0.4
a
0.3
a
NE
0.2 0.1
0 0.0 LI
PN
SO
SH
PX
Figure 6.12. Seed set (a) and germination rate (b) of Spartina alterniflora for isolated and meadow plants with at least some seed for each of the five sites. Sites with the same letter are not statistically different at p < 0.0. “NE” denotes no least-squares means were estimable. “0” indicates no seed was set, “Ø” indicates there were no seeds to test germination rate. Reprinted from Davis et al. (2004) with permission from the Journal of Ecology.
Chapter 6
319
6.7. FACILITATION AND CONSERVATION As noted in the introduction, despite the general paradigm of individualistic plant communities, conservationists assume a high degree of interdependence in communities when they argue for the preservation of natural systems and biological diversity (Freedman 1989 Erlich 1990, Erlich and Wilson 1991, Miller 1993, Noss 1994). Recently, there have been a number of studies that support that assumption. Acer opalus subsp. granatense is an endemic endangered tree existing in small populations with very few adults. The paucity of adults means that these populations are highly vulnerable to factors limiting recruitment. Lorena Gómez-Aparicio and colleagues at the University of Granada found that the tree is recruiting throughout its range in the Iberian Peninsula, but that microsites provided by shrubs play the crucial role for early establishment and long-term survival of this rare species (Gómez-Aparicio et al. 2005c). Padilla and Pugnaire (2006) argued that in extreme abiotic environments intentional planting of stress tolerant nurse species could play an important role in restoring highly degraded systems. This process is simple and inexpensive, simply “mimicking a natural process”. Mimicking a natural process appears to have a great deal of potential for restoring and re-establishing mangrove communities. Milbrandt and Tinsley (2006) found that mangrove seedlings were more common in patches of Batis maritima, a very salt tolerant species, than in mudflats without Batis. They also found that mortality of experimentally planted Avicennia germinans (black mangrove) seedlings was much lower in Batis patches than in open mudflats. They suggested that the primary mechanism for this facilitation was that Batis created a slight increase in elevation and argued that the process could be used in projects to protect or re-establish mangroves if sea-levels rise.
6.8. FACILITATION AND EVOLUTION IN PLANT COMMUNITIES Facilitative relationships among plants certainly do not imply evolutionary relationships. However, if plant communities are indeed individualistic, as argued by Gleason and virtually all current textbooks, then neither competitive nor facilitative interactions can be acting as selective forces – because if plants in a community are adapting to each other, communities can hardly be individualistic. With few exceptions (Aarssen et al. 1979, Turkington 1989, Menchaca and Connolly 1990, Turkington and Merhoff 1990), plant communities are not thought to consist of co-evolved species. One of the widest spread and common plant species in Mediterranean Europe, Thymus vulgaris, is composed of different eco- or chemotypes that
320
Positive Interactions and Community Organization
differ substantially in the biochemical composition of the essential oils produced in the leaves. In southern France, six different Thymus chemotypes can be distinguished based on the dominant monoterpene in the essential oil, which is either phenolic or non-phenolic (Ehlers and Thompson 2004). Bromus erectus is often spatially associated with Thymus, and this spatial association occurs for many different Thymus chemotypes. Ehlers and Thompson (2004) found that soils from within and away from Thymus patches in sites dominated by either phenolic or non-phenolic chemotypes affected the germination, growth and reproduction of B. erectus in ways suggesting that different populations of the grass were evolving to tolerate particular chemicals produced by Thymus chemotypes. Bromus erectus from non-phenolic Thymus patches performed significantly better on its home soil than on soil from a different non-phenolic or phenolic Thymus patches. This superior performance of matched local B. erectus to the same Thymus chemotypes was only observed for soil collected directly underneath Thymus plants and not on soil collected away from Thymus. These results suggest that B. erectus may be genetically adapting to soil modifications mediated by different Thymus chemotypes, and importantly this can only occur because the facilitative effects of the Thymus allows B. erectus into the chemically modified environment in the first place. The scenario described by Ehlers and Thompson (2004), to my knowledge, is unique in the ecological literature for plants, although there are roughly similar examples for microbes (Turner et al. 1996, Travisano and Rainey 2000). But it describes the most probable way that facilitation may drive evolutionary changes in beneficiaries. By “pulling” other species into an expanded niche (Figure 6.1), benefactors have the potential to expose beneficiaries to new abiotic or biotic environments to which the latter may adapt. This scenario is illustrated in Figure 6.13.
B
A C
Figure 6.13. Illustration of how facilitation might create novel evolutionary opportunities for beneficiaries (e.g. Ehlers and Thompson 2004). “A” represents the niche of a beneficiary species (e.g. Bromus erectus) in the absence of a benefactor species (e.g. Thymus vulgaris). “B” represents the niche of the benefactor species. “C” represents the realized niche of the beneficiary species in the presence of the benefactor, with shaded area representing the new environment to which the beneficiary can now evolve.
Chapter 6
321
An important consideration for the potential of facilitation-driven niche shift as selective agent is the fundamental issue of reproductive isolation. In other words, if the beneficiary represented in Figure 6.13 (A & C) reproduces primarily sexually, and individuals in habitat A can still mate with individuals in habitat C because the habitats are close to each other, then how can different genotypes emerge? The answer may be in recent evolutionary theory that postulates how evolutionary diversification into distinct phenotypes might occur among sympatric individuals of a single species (see a lucid discussion this by Day and Young 2004). They describe how such sympatric selection would provide an explanation for how selection favors the origin of different species, from a single intermixed population, and how these new species might coevolve. As described by Day and Young, disruptive selection can act on phenotypic variation within a species if the fundamental resources that a species relies on (the “resource base”) are more broadly available than the “specieswide spectrum” of resource use (see Abrams et al. 1993, Geritz et al. 1998). In this context selection favors diversification because of intense competition for the part of the resource base most utilized, making specialization on other parts of the resource base more profitable (Rosenzweig 1978). Clearly, facilitative scenarios such as described in Figure 6.13 provide a situation in which such diversification along a continuous, but broad, resource base exists. Recent research has also shown that facilitation is an ecological interaction can cause evolutionary diversification within the context of sympatric disruptive selection (Doebeli and Dieckmann 2000). Importantly, it is not only traits related to resource acquisition that are postulated to undergo evolutionary diversification in this context, but also unrelated life history and sexually selected traits (Day and Young 2004). Research on invasive exotic plants raises the possibility of coevolution within plant communities determined by interactions, though not necessarily facilitative, among plants. Erik Aschehoug and I compared the competitive effects of an invasive Eurasian forb, Centaurea diffusa (diffuse knapweed), on three bunchgrass species that co-exist with C. diffusa in Eurasia to the effects of C. diffusa on three bunchgrass species from North America with similar morphologies and sizes. Each of the North American species was closely related to one of the Eurasian grass species (Callaway and Aschehoug 2000). Centaurea diffusa had much stronger negative effects on North American species than it had on Eurasian species (Figure 6.14). Furthermore, none of the North American grass species (nor all species analyzed collectively) had a significant competitive effect on the biomass of C. diffusa, but Eurasian Koleria laerssenii, and all Eurasian species analyzed collectively, significantly reduced C. diffusa biomass. Centaurea diffusa had no effect on the amount of 32P acquired by Eurasian grass species, but significantly reduced 32P uptake of all North American species. Correspondingly, there were no competitive effects of North American grasses on the uptake of 32P by C. diffusa, but all Eurasian
322
Positive Interactions and Community Organization Koeleria 2
a 1
a
b
Grass biomass (g)
b b
c
0
Festuca 2
1
no competitor Centaurea, no carbon Centaurea, carbon
a
a
b b c
c
0
Psuedoroegneria-Agropyron 2
a
a
1
b c b
c
0
Montana
Caucasus
Figure 6.14. Total biomass of related Eurasian and North American bunchgrass species grown alone or with C. diffusa, and either with or without activated carbon in the soil. Error bars represent 1 SE. Means with different letters were significantly different in pair-wise comparisons. Reprinted from Callaway and Aschehoug (2000) with permission from Science.
species demonstrated strong negative effects on the amount of 32P acquired by C. diffusa. When activated carbon was added to ameliorate chemical effects there were contrasting effects on the interactions between C. diffusa and grass species from the different continents. When growing with C. diffusa, the biomass of two North American species, Festuca idahoensis and Pseudoroegneria spicata, increased significantly when activated carbon was added to the soil (Figure 6.14). However activated carbon substantially reduced the biomass of all Eurasian grass species growing with C. diffusa. Correspondingly, activated
Chapter 6
323
carbon gave C. diffusa a competitive disadvantage against North American grasses (Centaurea biomass was reduced) but a competitive advantage in the presence of Eurasian grasses (Centaurea biomass increased). Activated carbon was a strong disadvantage for Eurasian grasses competing for 32P with C. diffusa. In all cases 32P uptake by Eurasian grasses growing with C. diffusa decreased in the presence of activated carbon. The effects of activated carbon on 32P uptake by grasses corresponded with the effects of activated carbon on 32 P uptake by C. diffusa. Activated carbon enhanced uptake by C. diffusa in the presence of Eurasian grasses but reduced uptake in the presence of North American grasses. In sum, their results indicated that allelopathic chemicals released from C. diffusa were much more effective against plants that were naïve to these chemicals than against plants that had been exposed to them for long periods of time. There were also significant biogeographical differences for the total biomass and total resource uptake by both individuals combined within a pot. Pots with Eurasian grass species combined with C. diffusa produced 12% more total biomass and took up 63% more total phosphorus than pots with North American species planted with C. diffusa, suggesting that longterm association and coevolution among plant species may enhance community functions such as productivity and resource utilization. These results were supported by later studies comparing C. diffusa invasive success when seeded into microcosm communities consisting either of grass species from either the native or invaded region of C. diffusa. Centaurea diffuse performed far better when invading North American grass communities (Vivanco et al. 2004). Centaurea maculosa, a congener of C. diffusa, shows similar biogeographic differences in allelopathic effects. Centaurea maculosa roots exude (±)-catechin, an enantiomer with strong anti-microbial and anti-plant effects (Bais et al. 2003). When (±)-catechin was applied experimentally, the germination and growth of North American grasses was inhibited more than that of related Eurasian grasses. These results have been repeated in biogeographic field experiments (Thorpe et al. 2006b) and in other greenhouse experiments (W.M. He, Y. Feng, and R.M. Callaway, unpublished data). Centaurea maculosa invasions often cause high mortality in native populations, and there are few more powerful selective forces than mortality. Evidence has also been found for C. maculosa-drive selection in native North American plant communities (Callaway et al. 2005). Surviving individuals and their progeny from North American communities that had experienced extensive invasion by Centaurea maculosa had higher tolerances to the Eurasian invader than individuals from communities that did not experience invasion. One native species, Stipa occidentalis, demonstrated evidence for selection for resistance to (±)-catechin. Mealor et al. (2004) reported similar evidence for native species adapting to the invasive Acroptilon repens. If occurring, evolved tolerance may contribute to coexistence among natives and invaders, and selection driven by plant interactions suggest that natural plant
324
Positive Interactions and Community Organization
communities may be at least somewhat coevolved and therefore not fully individualistic. The results for these Centaurea species (also see Mallik and Pellissier 2000, Ehlers and Thompson 2004, and Prati and Bosdorf 2004) imply that natural biological communities may evolve in some way as functionally organized units (Goodnight 1990, Wilson 1997). Although not produced by facilitation, these studies are among the few that point to the importance of plant interactions for natural selection. Biogeographic variation in the allelopathic effects of Centaurea species may suggest some degree of coevolution among plant species, but there are not many studies that directly tie facilitative interactions to natural selection and evolution. In one of the few examples, Figeroa and colleagues (2003) examined invasion of the Odiel Marshes in the estuary of the Odiel and Tinto rivers in southwest Spain by a Sarcocornia species and found the invader to be a hybrid. The Sarcoconia hybrid appears to benefit from an unusual successional sequence in which the raised centers of Spartina martima patches are invaded by Sarcoconia perennis, a species common to lower marshes. However, once established in Spartina patches, S. perennis provides an opportunity for hybridization with S. fruticosa, a species common in the higher marshes. Hybrids only occur on Spartina patches with S. perennis. Figeroa et al. coined the establishment of the hybrid Sarcoconia “genetic facilitation”, and suggested that succession might be facilitated genetically through the establishment of conditions leading to hybridization rather than simply by the enhanced sediment accretion by earlier species. Sediment accretion prevents later successional marsh species from being submersed, and prevents anaerobic conditions in the low marsh that appear to exclude S. fruticosa. As described in detail in this book, plants have powerful modifying forces on their environments. This effect has been called “ecosystem engineering” by Clive Jones and colleagues although engineering is not limited to plants. Plants can favorably alter the availability of all fundamental resources, nutrients, water, and light; change the way that energy and materials are cycled, and profoundly alter the course of natural disturbance. All of these factors have important evolutionary implications in communities. K. Laland, F. Odling-Smee, and M. Feldman developed theoretical models in which ecosystem engineering, or as they call it “niche construction”, is responsible for selective feedback in evolution, whereby the niche construction of past generations can affect the evolution of their offspring (Laland et al. 1996, Odling-Smee et al. 1996 Laland et al. 1999). In other words, they argue that the effects of plants on patterns of light and shade, soil salinity, and nutrient cycling have the potential to alter future evolutionary trajectories. Their first model (Laland et al. 1996) assumed that the abundance of a key resource depended entirely on niche construction effects. With this constraint, they found that niche construction could cause evolutionary inertia or momentum, the fixation of otherwise deleterious alleles, establish unexpected stable polymorphisms, or
Chapter 6
325
eliminate expected stable polymorphisms. Clearly, the availability of resources is determined by facilitative environmental modification. In a second model both niche construction and independent abiotic processes were included (Laland et al. 1999). The results of the second model supported the conclusion of the first. Even with independent, abiotic effects on resource availability, the effects of environmental modification by an organism had the potential to override external source of genetic selection to “create new evolutionary trajectories and equilibria, generate and eliminate polymorphisms, and produce time lags in the response to selection as well as other unusual dynamics”. When the effects of environmental modification, or niche construction, opposed the external source of selection the former were particularly strong. The sort of facilitative effects described in this book almost always oppose the selective effects of the abiotic environment. Laland et al. also argue that their predictions fit patterns produced by other models. Incorporating the size or density of the niche constructing population, other demographic characteristics, and the effects of niche construction on other coevolving species (rather than just conspecifics) are steps that will greatly enhance the model’s predictive power. It has been exceptionally difficult to integrate ecosystem and community ecology with the evolutionary progress made in population ecology because the abiotic components of ecosystems do not evolve; however, models such as developed by Laland, Odling-Smee, and Feldman are crucial steps towards linking these fields and understanding the full evolutionary potential of facilitative interactions. Peter Attsat and Dennis O’Dowd (1976) in their landmark paper on plant “defense guilds” and herbivore-driven functional interdependence among plants argued that groups of “populations, races, closely related species, and unrelated but chemically similar species” form “gene conservation guilds”. In short, by forming guilds with other species plants may make it far more difficult for pathogens, parasites, and herbivores to win the genetic “arms race”. This “arms race” has made sustaining disease, herbivore, and pathogen-free agricultural monocultures exceptionally difficult without intense applications of pesticides and biological controls. In part this is due to the ease with which pathogens and herbivores can find their prey and multiply rapidly, but monoculture susceptibility is also due in part to substantial disadvantages for the consumed relative to the consumer in the evolutionary race between offense and defense. Genetic uniformity, whether within or among species provides unique opportunities for “evolutionary tracking” - the evolution of virulent pathogens and herbivores which can be rapidly selected for by monocultural hosts. Heterogeneity in host populations or communities decreases the exposure frequency of susceptible genotypes and can disrupt evolutionary tracking and specialization by pathogens and herbivores. If pathogens and herbivores cannot specialize they may be much less likely to become hyper virulent. To summarize, in heterogeneous mixtures of plants consumers may have to spend some of their time eating plant species that select for different consumer
326
Positive Interactions and Community Organization
characteristics than would other plant species. When this occurs the consumer cannot be maximally selected for either species and therefore does not become exceptionally virulent against either species. Most of the best evidence for these sorts of “gene conservation guilds” is from agriculture and focuses on different genotypes of crop species. However, there is a great deal of potential for gene conservation guilds to function as powerful facilitative mechanisms in natural plant communities and these guilds may contribute to a greater understanding of community diversity and function. If the sort of decoy-target effect on herbivores demonstrated in the potatopotato beetle system described in Attsatt and O’Dowd is controlled in any way by the relative abundance of neighbors; then the interactions may vary in function over time and space. For generalist herbivores the amount of an unpalatable species that is eaten depends somewhat on how abundant palatable species are in the same area. For example, when only toxic phenotypes of Lotus corniculatus are available the slug Agriolimax reticulatus consumes it. However, when nontoxic phenotypes are mixed with toxic phenotypes Agriolimax avoids the toxic phenotypes that it so readily eaten when there are no other choices. If one studied the relationship between Agriolimax and toxic phenotypes without considering the effect of the nontoxic phenotypes, selective pressures on these species would be completely misunderstood. Recently, Alfonso Valiente-Banuet of the Universidad Nacional Autonoma de Mexico and colleagues explored facilitative relationships among plant taxa that evolved during the drying and warming climate of the more recent Quaternary period and more ancient plant taxa that evolved during the wetter and warmer Tertiary (Valiente-Banuet et al. 2006). It was during the unusually dry Quaternary that most global deserts developed. The development of desert corresponded with the evolution of new species, but interestingly many mesic-adapted Tertiary species did not go extinct in the drier climate. Valiente-Banuet et al. compared the importance of facilitative interactions for recruitment between Tertiary and Quaternary species that exist today in Mediterranean climates around the world by analyzing biogeographic, paleobotanical, and ecological literature. Their results have profoundly important implications for the processes that sustain global biodiversity and for the nature of plant communities. They found that “modern” species, those that arose during the Quaternary, currently facilitate ancient Tertiary species. In other words, species that rapidly evolved to new stressful abiotic conditions appeared to be “pulling” other species that had not evolved to xeric conditions into modern communities by creating regeneration niches (subcanopy) that are similar to ancient conditions during the Tertiary. The reason these findings are so important, is that current plant communities certainly include mixtures of floristic elements that evolved in different geological eras, leading to the assumption that these taxa have been sorted independently of each other into extant assemblages. This assumption has been central to the idea that idea that communities are ‘‘merely a
Chapter 6
327
coincidence’’ (Gleason 1926). Valiente-Banuet et al.’s results indicate that “interdependent relationships among plants have played a central role in the preservation of the global biodiversity and provided a mechanism for stabilizing selection and the conservation of ecological traits over evolutionary time scales”. I explore the implications of this and other research on facilitation for community theory in the next section of this chapter.
6.9. REPLACING THE NOTION OF INDIVIDUALISTIC COMMUNITIES WITH THE “INTEGRATED COMMUNITY” Most ecologists today probably do not strongly identify with either “holistic” Clementsian or “individualistic” Gleasonian camps (see page 1 of this book). Most of us probably see plant communities as something in between. However, the presentation of communities as individualistic is common in general and specialized textbooks. More importantly, decades of retelling the tale of Gleasonian triumph and Clementsian superstition has deeply permeated the individualistic paradigm into many conceptual arenas of plant ecology (e.g. neutral theory, climate envelopes for future distributional predictions, and research emphases on resource competition). But as I have argued throughout this book, the individualistic paradigm is not tenable in the face of abundant evidence for facilitation, and other evidence for evolutionary relationships caused by interactions among plants (Callaway and Aschehoug 2000, Mallik and Pellisier 2000, Bais et al. 2003, Vivanco et al. 2004, Callaway and Ridenour 2004, Ehlers and Thompson 2004, Callaway et al. 2005) and in communities (Goodnight 1990, Inouye and Stinchcombe 2001, Whitham et al. 2003, Schweitzer et al. 2004). The classic and often artificial, historical dichotomy in viewpoints is one with lingering heuristic impacts. This is the perspective that plant communities are composed of individual species that share adaptations to particular abiotic conditions (Gleason 1926), versus the perspective that communities are highly codependent groups of species (Clements 1916). Clements compared the plant community, or association, to an organism, “able to essentially reproduce its component parts”, whereas Gleason argued that a plant community is “scarcely even a vegetational unit, but merely a coincidence”. These positions are a bit extreme, Gleason’s viewpoints are in fact more complex and hard to precisely pin down (Nicolson and McIntosh 2002), and broadly accepted definitions are hard to find. However, one of the most recent of these sounds a lot like Gleason himself: describing “vegetation as an assembly of individual plants belonging to different species distributed according to its own physiological requirements as constrained by competitive interactions.” (Moore 1990, see Nicolson and McIntosh 2002).
328
Positive Interactions and Community Organization A) Desert Transect 1
Transect 3
300
100 0
10
20
30
0
10
DCA Axis 1
B) Terrace
Transect 1
400
20
30
Transect 2
200 0 0
10
20
30
0
10
20
30
C) Riparian Transect 2
Transect 1
300
100 0
10
Presence
20
30
0
Quadrat sequence Abundance
10
20
Canopy
Figure 6.15. Results of DCA analyses showing spatial relationships between Prosopis velutina canopies and the DCA scores of quadrats in the understory. Flat dashed lines at the tops of the figures represent the presence of Prosopis canopies (“canopy”), solid lines below represent DCA scores based on understory species presence or absence (“presence”), and dashed lines below represent DCA scores based on understory species abundance. Reprinted from Schade et al. (2003) with permission from the Journal of Vegetation Science.
The arguments for the individualistic paradigm are multidimensional (Huntley 1991, Johnson and Mayeux 1992, Nicolson and McIntosh 2002), but descriptions of continuous distributions of species along environmental gradients have been a central component of the argument for individualistic communities since the idea was first articulated (Whittaker 1951). Most gradient analyses demonstrate the continuum of apparently randomly overlapping species that is used to argue for individualistic communities. However, most gradient analyses are not designed to quantify spatial relationships at a scale appropriate to detect positive associations, and some gradient analyses are at odds with the continuum. Schade et al. (2003) used Detrended Correspondence Analysis (DCA) ordinations to characterize the spatial relationship between understory whole-community composition with the presence of Prosopis velutina canopies in desert, river terrace, and riparian zones in the Sonoran Desert of Arizona. On transects placed through the
Chapter 6
329 Chaparral
Adenostoma cover ordination scores of understory community
ordination scores for annual species
Mojave Desert
Larrea tridentata Ambrosia dumosa
Quadrat sequence Figure 6.16. Stylized representation of graphical results presented in Schmida and Whittaker (1981) demonstrating strong nodality among species in plant communities.
vegetation, the presence of Prosopis on desert and terrace landforms was strongly associated with different DCA axis 1 scores for understory communities, whether based on understory species abundance or presence/absence (Figure 6.15). In other words, where Prosopis occurred on transects, quadrats contained different understory species than quadrats placed in open areas on the transects – as reflected in the differences in DCA scores. The distributions of species on the gradients quantified by Schade at al. were not continuous, but grouped into nodes, nodes that suggest some degree of interdependence of species in these communities. As described in Chapter 1, Robert Whittaker provided quantitative and articulate conceptual arguments for the individualistic theory. However, in a strikingly ironic development, Robert Whittaker posthumously published a paper in 1981 titled “Pattern and biological microsite effects in two shrub communities in southern California” (Schmida and Whittaker 1981). The central figures presented in this paper displayed virtually perfect correlations in
330
Positive Interactions and Community Organization
the ordinated composition of understory communities with the presence of either desert or chaparral shrubs – the nodal distribution of species along gradients indicative of holistic communities (Figure 6.16). It is hard to imagine tighter correlations among the distributions of plant species along gradients, yet these “strongly differentiated patterns” were referred to as microsites without borders! Despite the complete absence of a continuum in the results, there was no discussion of reconciling these findings with Whittaker’s paradigm building emphasis on the continuum (Whittaker 1951, 1952, 1956); or more importantly, rejecting the continuum as a universal descriptor of plant communities. Old paradigms die hard. Bruno et al. (2003) proposed that explicitly incorporating facilitation into ecological theory ‘‘...will fundamentally challenge some of our most cherished paradigms’’ and ‘‘...that current theory emphasizing competition or predation paints an incomplete, and in some cases misleading picture of our understanding of the structure and organization of ecological systems’’. Lortie et al. (2004) argued that one aspect of this incomplete picture was the debate on individualistic versus organismal communities. They stated that recent experimental efforts to understand the relative importance of positive or negative interactions allows ecologist “to explicitly reconsider what most ecologists appear to have done implicitly; our formal conceptual theory of the fundamental nature of communities.” They also proposed an alternative conceptual model for plant communities termed the ‘integrated community concept’. This model was based on the assertions that community composition is determined by (i) stochastic processes, (ii) the specific tolerances of species to the suite of local abiotic conditions, (iii) positive and negative direct and indirect interactions among plants, and (iv) direct interactions with other organisms. This was illustrated in a figure (Figure 6.17) adapted from (Krebs 2001) and a general concept of biological filters (Grime 1998, Laakso et al. 2001). Of course these and other processes may work together in ways not well-illustrated in this diagram, but the purpose of the diagram is to emphasize the integration of drivers in communities that have both individualistic and interdependent characteristics. This theme of integration has been a central focus throughout this book. If communities are determined by complex interactions among all the processes presented in Figure 6.17, communities will vary from collections of individual species to highly interdependent groups of species (also see Michelet et al. 1999). As stated by Lortie et al. (2004) “…communities (and even a single community) will encompass a range of different dependencies among species - or degrees of integration - determined by the relative importance, and variation in space and time, of each of the filters we proposed.” Lortie et al. (2004) suggested a graphic visualization for this range of different dependencies with a y-axis representing variation in interdependence among species and an x-axis representing a conceptual range of randomness and non-randomness (Figure 6.18). Our hope for this visualization was that it
Chapter 6
331
would reject the isolation of single processes as community drivers, such as just competition or just facilitation, improving the unproductive and simplistic debate over which type of interaction is more important in communities (see Shipley and Keddy 1987, Shouse 2003). This figure also represents an attempt to reconcile Whittaker’s classic continuum presented in virtually every ecological textbook with the highly nodal results presented above from Schmida and Whittaker (1981) and Schade et al. (2003). Lortie et al. (2004) argued “the real challenge is instead to determine when and where different processes are important, and to do this, we need to credit communities with more complexity, i.e. as a dynamic collection of species integrated to varying degrees (through competition and facilitation) inextricably linked to biotic and abiotic drivers.”
GLOBAL SPECIES POOL Chance biogeographical events (e.g. dispersal, presence of vectors, or distance to new environment).
Plant species capable of reaching an environment
Local environmental conditions
Mycorrhizae + microbes increase available nutrients
Species pool based upon physiological tolerances
Plant interactions
Competition
Herbivory
Facilitation
Direct interactions with other organisms Pollination
EXTANT PLANT COMMUNITY Figure 6.17. Processes or “filters” that structure a plant community. The Integrated Community concept proposes that all four processes can be important in determining a particular plant community but that the relative importance of each process varies among different places and time. Each process/filter is represented by a pair of horizontal lines and the corresponding description is in bold italics adjacent to the symbol (sub-sets of a process such as herbivory or competition are labeled in plain text). Solid arrows depict the movement of species through the filters, and hatched lines illustrate where each process might influence the plant community. Reprinted from Lortie et al. (2004) with permission from Oikos.
332
Positive Interactions and Community Organization
Degree of dependence
interdependence
x
x x
independence plant interactions non-random
abiotic conditions
stochastic processes random
Degree of randomness Figure 6.18. Visualization of how individuals within a plant community may function in either individualistic or interdependent ways. Different symbols represent individual species from different plant communities and each point is a value for a single species. The degree of interdependence for each species could be based on relative competitive intensity calculated from neighbor removal experiments, and randomness could be based on the relative importance of each of the three processes or filters that determine geographical ranges for each species. The ellipses represent possible groupings of species within a community. “X”s represent a centroid for each community. Strong species-specific facilitative interactions would be prominent in the upper left portion of the graph. Reprinted from Lortie et al. (2004) with permission from Oikos.
By formalizing the balance between independence and interdependence in our conceptual models of plant communities we may gain conceptual freedom to better understand the effect of invasions and climate change on natural communities. The Integrated Community concept may not include all of the processes that are important in communities, but the Integrated Community does not ignore the overwhelming evidence for positive interactions among plant species.
Chapter 6
333
6.10. CONCLUSIONS A complete understanding of positive interactions offers a different and more complete understanding of plant communities than that presented in the dominant individualistic paradigm. Community composition, biological diversity, and coexistence among plant species is not determined only by competitive and consumer interactions, but by powerful and ubiquitous facilitative effects. Because facilitative effects bring a substantial component of interdependence to plant community ecology and can be implicated in evolutionary divergence of plant species, plant communities cannot be understood by studying populations. Instead we must continue to develop new ways to integrate phenomenological, mechanistic, context-dependent, multispecies, and systems-based approaches. By doing this, advances in the next 50 years by community ecologists will certainly exceed the far-reaching progress achieved during the last 50 years.
REFERENCES
Aarrsen, L. W., and G. A. Epp. 1990. Neighbor manipulations in natural vegetation: a review. Journal of Vegetation Science 1:13-30. Aarssen, L. W., R. Turkington, and P. B. Cavers. 1979. Neighbour relationships in grass/legume communities. II. Temporal stability and community evolution. Canadian Journal of Botany 57:2695-2703. Abd El Rahman, A. A., and K. H. Batanouny. 1965a. The water output of the desert vegetation in the different microhabitats of Wadi Hoff. Journal of Ecology 53:139-145. Abd El Rahman, A. A., and K. H. Batanouny. 1965b. Transpiration of desert plants under different environmental conditions. Journal of Ecology 53:267-272. Abrams, P. A. 1987. On classifying interactions between populations. Oecologia 73:272-281. Adler, P. B., C. M. D’Antonio, and J. T. Tunison. 1998. Understory successsion following a dieback of Myrica faya in Hawai’i Volcanoes National Park. Pacific Science 52:69-78. Aggarwal, R. K., J. P. Gupta, S. K. Saxena, and K. D. Muthana. 1976. Studies on soil physico-chemical and ecological changes under five twelveyear old desert species of western Rajasthan. Indian Forester 102:863-872. Aggarwal, R. K., P. Kumar, and P. Raina. 1993. Nutrient availability from sandy soils underneath Prosopis cineraria (Linn. MacBride) compared to adjacent open site in an arid environment. Indian Forester 11:321-324. Agnew, A. D. Q. 1997. Switches, pulses and grazing in arid vegetation. Journal of Arid Environments 37:609-617. Agnew, A. D. Q., J. B. Wilson, and M. T. Sykes. 1993. A vegetation switch as the cause of a forest/mire ecotone in New Zealand. Journal of Vegetation Science 4:273-278. Aguiar, M. R., and O. E. Sala. 1994. Competition, facilitation, seed distribution and the origin of patches in a Patagonian steppe. Oikos 70:26-34. Aguiar, M. R., A. Soriano, and O. E. Sala. 1992. Competition and facilitation in the recruitment of seedlings in Patagonian steppe. Functional Ecology 6:66-70.
335
336
References
Aide, T. M., J. K. Zimmerman, and L. Herrera. 1995. Forest recovery in abandoned tropical pastures in Puerto Rico. Forest Ecology and Management 77:77-86. Aizen, M. A., and P. Feinsinger. 1994. Forest fragmentation, pollination and plant reproduction in a Chaco dry forest, Argentina. Ecology 75:330-351. Alados, C. L., P. Gotor, P. Ballester, D. Navas, J. M. Escos, T. Navarro, and B. Cabezudo. 2006. Association between competition and facilitation processes and vegetation spatial patterns in alpha steppes. Biological Journal of the Linnean Society 87:103-113. Alban, D. H. 1982. Effect of nutrient accumulation by aspen, spruce, and pine on soil properties. Soil Science Society of America Journal 46:853-861. Alftine, K. J., and G. P. Malanson. 2004. Directional positive feedback and pattern at an alpine tree line. Journal of Vegetation Science 15:3-12. Allee, W. C. 1931. Animal Aggregations: a Study in General Sociology. University of Chicago Press, Chicago, Illinois, USA. Allen, E. B., and M. F. Allen. 1990. The mediation of competition by soil fungi in successional and patchy environments. Pages 367-389 in J. B. Grace and D. Tilman, editors. Perspectives on Plant Competition. Academic Press, New York, USA. Amarasekare, P. 1998. Allee effects in metapopulation dynamics. American Naturalist 152:298-302. Amiotti, N. M., P. Zalba, L. F. Sanchez, and N. Peinemann. 2000. The impact of single trees on properties of loess-derived grassland soils in Argentina. Ecology 81:3283-3290. Anderson, L. J., M. S. Brumbaugh, and R. B. Jackson. 2001. Water and treeunderstory interactions: a natural experiment in a savanna with oak wilt. Ecology 82:33-49. Andow, D. A. 1991. Vegetational diversity and arthropod population response. Annual Review of Entomology 36:561-586. Anthelme, F., R. Michalet, and M. Saadouc. 2007. Positive associations involving the tussock grass Panicum turgidum Forssk. in the AïrTénéré Reserve, Niger. Journal of Arid Environments 68:348-362. Aphalo, P. J., and C. L. Ballaré. 1995. On the importance of informationacquiring systems in plant-plant interactions. Functional Ecology 9:5-14. Aplet, G. H., R. L. Loh, J. T. Tunison, and P. M. Vitousek. 2001. Experimental restoration of a dense M. faya stand. University of Hawaii.
References
337
Archer, S. 1990. Development and stability of grass/woody mosaics in a subtropical savanna parkland, Texas, U.S.A. Journal of Biogeography 17:453-463. Archer, S. 1995. Tree-grass dynamics in a Prosopis-thornscrub savanna parkland: reconstructing the past and predicting the future. Ecoscience 2:83-99. Archer, S., C. Scifres, and C. R. Bassham. 1988. Autogenic succession in a subtropical savanna: conversion of grassland to thorn woodland. Ecological Monographs 58:111-127. Archer, S. A. 1989. Have southern Texas savannas been converted to woodlands in recent history? American Naturalist 134:545-561. Armbrust, D. V., and J. D. Bilbro. 1997. Relating plant canopy characteristics to soil transport capacity by wind. Agronomy Journal 89:157-162. Armstrong, J., and W. Armstrong. 1990. Light-enhanced convective throughflow increases oxygentation in rhizomes and rhizosphere of Phragmites australis (Cav.) Trin ex Steud. New Phytologist 114:121-128. Armstrong, J., W. Armstrong, and P. M. Becket. 1992. Phragmites australis: Venturi- and humidity-induced pressure flows enhance rhizome aeration and rhizosphere oxidation. New Phytologist: 116:197-207. Armstrong, W. 1964. Oxygen diffusion from the roots of some British bog plants. Nature 204:801-802. Armstrong, W. 1979. Aeration in higher plants. Advances in Botanical Research 7:226-322. Arno, S. F., and R. P. Hammerly. 1990. Timberline: Mountain and Arctic Forest Frontiers. The Mountaineers, Seattle, Washington, USA. Arp, W. J. 1991. Effects of source-sink relations on photosynthetic acclimation to elevated CO2. Plant, Cell and Environment 14: 869-875. Arriaga, L., Y. Maya, S. Diaz, and J. Cancino. 1993. Association between nurse perennials and cacti in a heterogeneous dry forest in northwestern Mexico. Journal of Vegetation Science 4:349-356. Arroyo, M. T. K., L. A. Cavieres, A. Peñaloza, and M. A. Arroyo-Kalin. 2003. Positive associations between the cushion plant Azorella monantha (Apiaceae) and alpine plant species in the Chilean Patagonian Andes. Plant Ecology 169:12-129. Art, H. W., F. H. Borman, G. K. Voight, and G. M. Woodwell. 1974. Barrier island forest ecosystem: role of meteorologic nutrient inputs. Science 184:60-62. Atsatt, P. R., and D. O’Dowd. 1976. Plant defense guilds. Science 193:24-29. Austin, M. P. 1985. Continuum concept, ordination methods, and niche theory. Annual Review of Ecology and Systematics 16:39-61.
338
References
Azevedo, J., and D. L. Morgan. 1974. Fog precipitation in coastal California forests. Ecology 55:1135-1141. Badano, E. I., and L. A. Cavieres. 2006a. Impacts of ecosystem engineers on community attributes: effects of cushion plants at different elevations of the Chilean Andes. Diversity and Distributions 11:137-144. Badano, E. I., and L. A. Cavieres. 2006b. Ecosystem engineering across ecosystems: do engineer species sharing common features have generalized or idiosyncratic effects on species diversity? Journal of Biogeography 33:304-313. Bagnold, R. A. 1954. The Physics of Blown Sand and Desert Dunes. Methuen and Co., London, England. Bais, H. P., S. W. Park, T. L. Weir, R. M. Callaway, and J. M. Vivanco. 2004. How plants communicate using the underground information superhighway. Trends in Plant Science 9:26-32. Bais, H. P., R. Vepachedu, S. Gilroy, R. M. Callaway, and J. M. Vivanco. 2003. Allelopathy and exotic plants: from genes to invasion. Science 301:1377-1380. Baldwin, I. T., and J. C. Schultz. 1983. Rapid changes in tree leaf chemistry induced by damage: evidence for communication between plants. Science 221:277-279. Baldwin, I. T., L. Staszak-Kozinski, and R. Davidson. 1994. Up in smoke. I. Smoke-derived germination cues for a postfire annual, Nicotiana attenuata Torr. ex. Watson. Journal of Chemical Ecology 20: 2345-2347. Ball, M. C. 1994. The role of photoinhibition during tree seedling establishment at low temperatures. Pages 365-376 in N. R. Baker and J. R. Bowyer, editors. Photoinhibition of photosynthesis: from Molecular mechanisms to the field. Bioscientific Publishers, New York, New York, USA. Ball, M. C., J. J. G. Egerton, R. Leuning, R. B. Cunningham, and P. J. Dunne. 1997. Microclimate above grass adversely affects spring growth of seedling snow gum (Eucalyptus pauciflora). Plant, Cell and Environment 20:155-166. Ball, M. C., V. S. Hodges, and G. P. Laughlin. 1991. Cold-induced photoinhibition limits regeneration of snow gum at tree line. Functional Ecology 5:663-668. Ballare, C. L., R. A. Sanchez, A. L. Scopel, J. J. Casal, and C. M. Ghersa. 1987. Early detection of neighbour plants by phytochrome perception of spectral changes in reflected sunlight. Plant, Cell and Environment 10:551-557.
References
339
Ballare, C. L., A. L. Scopel, and R. A. Sanchez. 1990. Far-red radiation reflected from adjacent leaves: an early signal of competition in plant canopies. Science 247:329-332. Barbour, M. G., and A. F. Johnson. 1988. Beach and dune. Pages 296-322 in M. G. a. M. Barbour, J., editor. Terrestrial Vegetation of California. California Native Plant Society., Sacramento, California, USA. Bard, G. 1952. Secondary succession of the Piedmont of New Jersey. Ecological Monographs 22:195-215. Barko, J. W., and R. M. Smart. 1986. Sediment-related mechanisms of growth limitation in submersed macrophytes. Ecology 67:161-175. Barnes, P. W., and S. Archer. 1996. Influence of an overstorey tree (Prosopis glandulosa) on associated shrubs in a savanna parkland: implications for dynamics. Oecologia 105:493-500. Barnes, P. W., and S. Archer. 1999. Tree-shrub interactions in a subtropical savanna parkland: competition or facilitation? Journal of Vegetation Science 10:525-536. Barney, M. A. 1973. Vegetation changes following fire in the pinyon-juniper type of west-central Utah. Journal of Range Management 27:91-96. Barret, L. I. 1931. Influence of forest litter on the germination and early survival of chestnut oak, Quercus montana, Willd. Ecology 12: 476-484. Barros-Henriques, R. P., and J. D. Hay. 1992. Nutrient content and the structure of a plant community on a tropical beach-dune system in Brazil. Acta Oecologia 13:101-117. Barth, R. C. 1980. Influence of pinyon pine trees on soil chemical and physical properties. Soil Science Society of America Journal 44:112-114. Barth, R. C., and J. O. Klemmedson. 1978. Shrub-induced spatial patterns of dry matter, nitrogen, and organic carbon. Soil Science Society of America 42:804-809. Bartolome, J. W., P. C. Muick, and M. P. McClaran. 1987. Natural Regeneration of Californian hardwood. Pacific Southwest Forest and Range Experiment Station PSW-100, United States Forest Service, Berkeley. Bates, J. W. 1988. The effect of shoot spacing on the growth and branch development of the moss Rhytidiadelphus triquetrus. New Phytologist 109:499-504. Baumeister, D., and R. M. Callaway. 2006. Facilitative effects of Pinus flexilis during succession: a hierarchy of mechanisms benefits other plant species. Ecology 87:1816-1930. Beattie, A. J. 1976. Plant dispersion, pollination and gene flow in Viola. Oecologia 25:291-300.
340
References
Beatty, S. W. 1984. Influence of microtopography and canopy species on spatial patterns of forest understory plants. Ecology 65:1406-1419. Bedford, B. L., D. R. Bouldin, and B. D. Beliveau. 1991. Net oxygen and carbon-dioxide balances in solutions bathing roots of wetland plants. Journal of Ecology 79:943-959. Bellingham, P. J., L. R. Walker, and D. A. Wardle. 2001. Differential facilitation by a nitrogen-fixing shrub during primary succession influences relative performance of canopy tree species. Journal of Ecology 89:861-875. Belsky, A. J. 1994. Influences of trees on savanna productivity: tests of shade, nutrients, and tree-grass competition. Ecology 75: 922-932. Belsky, A. J., R. G. Amundson, J. M. Duxbury, S. J. Riha, A. R. Ali, and S. M. Mwonga. 1989. The effects of trees on their physical, chemical and biological environments in a semi-arid savanna in Kenya. Jounrnal of Applied Ecology 26: 1005-1024. Bendix, M., T. Tornberg, and H. Brix. 1994. Internal gas transport in Typha latifolia L. and Typha angustifolia L. 1. humidity-induced pressurization and convective throughflow. Aquatic Botany 41:41-65. Bennett, B. C. 1986. Patchiness, diversity, and abundance relationships of vascular epiphytes. Selbyana 9:70-75. Bently, B. L., and N. D. Johnson. 1991. Plants as food for herbivores: the roles of nitrogen fixation and carbon dioxide enrichment. Pages 257-272 in P. W. Price, T. M. Lewinsohn, G. W. Fernandez, and W. W. Benson, editors. Plant animal interactions: evolutionary ecology in tropical and temperate regions. Wiley, New York, New York, USA. Benzing, D. H. 1980. Bromeliad epiphytism. Pages 211-264 in The Biology of the Bromeliads. Mad River Press, Eureka, California, USA. Benzing, D. H. 1981. Bark surfaces and the origin and maintenance of diversity among angiosperm epiphytes: a hypothesis. Selbyana 5:248-255. Benzing, D. H., and A. Renfrow. 1974. The nutritional status of Encyclia tempense and Tillandsia circinata on Taxodium ascendens and the availability of nutrients to epiphytes on this host in south Florida. Bulletin of the Torrey Botanical Club 101:191-197. Berendse, F., and R. Aerts. 1984. Competition between Erica tetralix L. and Molina aerulea (L.) Moench as affected by the availability of nutrients. Acta Oecologia 5:3-14. Bergelson, J., and R. Perry. 1989. Inter-specific competition between seeds: relative planting date and density affect seedlings emergence. Ecology 70:1639-1644. Bernhard-Reversat, F. 1982. Biogeochemical cycle of nitrogen in semi-arid savanna. Oikos 38:321-332.
References
341
Bertiller, M. B., J. O. Ares, P. Graff, and R. Baldi. 2000. Sex-related spatial patterns of Poa ligularis in relation to shrub patch occurrence in northern Patagonia. Journal of Vegetation Science 11:9-14. Bertness, M. D. 1988. Peat accumulation and the success of marsh plants. Ecology 69:703-713. Bertness, M. D. 1991a. Interspecific interactions among high marsh perennials in a New England salt marsh. Ecology 72:125-137. Bertness, M. D. 1991b. Zonation of Spartina patens and Spartina alterniflora in a New England salt marsh. Ecology 72:138-148. Bertness, M. D. 1998. Searching for the role of positive interactions in plant communities. Trends in Ecology and Evolution 13:133-134. Bertness, M. D., and R. M. Callaway. 1994. Positive interactions in communities. Trends in Ecology and Evolution 9:191-193. Bertness, M. D., and P. J. Ewanchuk. 2002. Latitudinal and climate-driven variation in the strength and nature of biological interactions in New England salt marshes. Oecologia 132:392-401. Bertness, M. D., and S. D. Hacker. 1994. Physical stress and positive associations among marsh plants. American Naturalist 144:363-372. Bertness, M. D., and S. W. Shumway. 1993. Competition and facilitation in marsh plants. American Naturalist 142:718-724. Bertness, M. D., and S. M. Yeh. 1994. Cooperative and competitive interactions in the recruitment of marsh elders. Ecology 75:2416-2429. Bever, J. D. 1994. Feedback between plants and their soil communities in an old field community. Ecology 75:1965-1977. Bever, J. D. 2003. Soil community feedback and the coexistence of competitors: conceptual frameworks and empirical tests. New Phytologist 167:465-473. Bever, J. D., J. B. Morton, J. Antonovics, and P. A. Schultz. 1996. Hostdependent sporulation and species diversity of arbuscular mycorrhizal fungi in a mown grassland. Journal of Ecology 84:71-82. Bever, J. D., K. M. Westover, and J. Antonovics. 1997. Incorporating the soil community into plant population dynamics: the utility of the feedback approach. Journal of Ecology 85:561-573. Bidartondo, M. I., A. M. Kretzer, E. M. Pine, and T. D. Bruns. 2000. High root concentration and uneven ectomycorrhizal diversity near Sarcodes sanguinea (Ericaceae): a cheater that stimulates its victims? American Journal of Botany 87:1783-1788. Billings, W. D. 1950. Vegetation and plant griowth as affected by chemically altered rocks in the western Great Basin. Ecology 31:62-74. Billings, W. D. 1969. Vegetational pattern near alpine timberline as affected by fire-snowdrift interactions. Vegetatio 19:192-207.
342
References
Billings, W. D., and W. B. Drew. 1938. Bark factors affecting the distribution of corticolous bryophytic communities. American Midland Naturalist 20:302-330. Billings, W. D., and A. F. Mark. 1957. Factors involved in the persistence of montane tree-less balds. Ecology 38:140-142. Bjorkman, J. 1960. Monotropa hypopitys L. - an epiparasite on tree roots. Physiologia Plantarum 13:308-327. Blackburn, W. H., and P. T. Teller. 1970. Pinyon and juniper invasion in black sagebrush communities in east-central Nevada. Ecology 51:841-848. Bleher, B., and K. Bohning-Gaese. 2001. Consequences of frugivore diversity for seed dispersal, seedling establishment and the spatial pattern of seedlings and trees. Oecologia 129:385-394. Blennow, K., A. R. G. Lang, P. Dunne, and M. C. Ball. 1998. Cold-induced photoinibition and growth of snow gum (Eucalyptus pauciflora) under differing temperature and radiation regimes in fragmented forests. Plant, Cell and Environment 21:407-416. Bliss, L. C. 1971. Arctic and alpine plant life cycles. Annual Review of Ecology and Systematics 2:405-438. Blundon, D., J. MacIssac, and R. T. Dale. 1983. Nucleation during primary succession in the Canadian Rockies. Canadian Journal of Botany 71:1093-1096. Boettcher, S. E., and P. J. Kalisz. 1990. Single-tree influence on soil properties in the mountains of eastern Kentucky. Ecology 71:1365-1372. Bonan, G. B. 1999. Frost followed the plow: impacts of deforestation on the climate of the United States. Ecological Applications 9:1305-1315. Bond, B. J., B. T. Farnsworth, R. A. Coulombe, and W. E. Winner. 1999. Foliage physiology and biochemistry in response to light gradients in conifers with varying shade tolerance. Oecologia 120:183-192. Bormann, F. H. 1966. The structure, function, and ecological significance of root grafts in Pinus strobus L. Ecological Monographs 36:1-26. Bouwmeester, H. J., R. Matusova, S. Zhongkui, and M. H. Beale. 2003. Secondary metabolite signaling in host–parasitic plant interactions. Current Opinion in Plant Biology 6:358-364. Bowman, W. D., J. C. Schardt, and S. K. Schmidt. 1996. Symbiotic N-fixation in alpine tundra: ecosystem input and variation in fixation rates among communities. Oecologia 108:345-350. Bratton, S. P. 1975. A comparison of the beta diversity functions of the overstory and herbaceous understory of a deciduous forest. Bulletin of the Torrey Botanical Club 102:55-60.
References
343
Bray, J. R. 1955. The savanna vegetation of Wisconsin and an application of the concepts order and complexity to the field of ecology. Dissertation. University of Wisconsin, Madison. Breshears, D. D., J. W. Nyan, C. E. Heil, and B. P. Wilcox. 1998. Effects of woody plants on microclimate in a semiarid woodland: soil temperature and evaporation in canopy and intercanopy patches. International Journal of Plant Sciences 159:1010-1017. Bressolier, C., and Y. F. Thomas. 1977. Studies on wind and plant interactions on French Atlantic coastal dunes. Journal of Sedimentary Petrology 47:331-338. Brewbaker, J. L., and B. H. Kwack. 1963. The essential role of calcium ion in pollen germination and pollen tube growth. American Journal of Botany 50:859-865. Brewbaker, J. L., and S. K. Majumder. 1961. Cultural studies of the pollen population effect and the self-incompatibility inhibition. American Journal Botany 48:457-464. Brewer, J. S., J. M. Levine, and M. D. Bertness. 1997. Effects of biomass removal and elevation on species richness in a New England salt marsh. Oikos 80:333-341. Brink, R. A. 1924. The physiology of pollen IV. Chemotropism: effects on growth of grouping grains; formation and function of callose plugs; summary and conclusions. American Journal of Botany 11:417-436. Briones, M. J. I., and P. Ineson. 1996. Decomposition of Eucalyptus leaves in litter mixtures. Soil Biology Biochemistry 28:1381-1388. Briones, O., C. Montana, and E. Ezcurra. 1998. Competition intensity as a function of resource availability in a semiarid system. Oecologia 116:365-372. Brooke, R. C., E. B. Peterson, and V. J. Krajina. 1970. The subalpine Mountain Hemlock zone. Ecology of Western North America 2:148-349. Brooker, R., Z. Kikvidze, F. I. Pugnaire, R. M. Callaway, P. Choler, C. Lortie, and M. Michalet. 2005. The importance of importance. Oikos 109: 63-70. Brooker, R. W. 2006. Plant-plant interactions and environmental change. New Phytologist 171:271-284. Brooker, R. W., and T. V. Callaghan. 1998. The balance between positive and negative plant interactions and its relationship to environmental gradients: a model. Oikos 81:196-207. Brooker, R. W., D. Scott, S. C. F. Palmer, and E. Swaine. 2006. Transient facilitative effects of heather on Scots pine along a grazing disturbance gradient in Scottish moorland. Journal of Ecology 94:637-645.
344
References
Brooks, J. R., F. C. Meinzer, R. Coulombe, and J. Gregg. 2002. Hydraulic redistribution of soil water during summer drought in two contrasting Pacific Northwest coniferous forests. Tree Physiology 22:1107-1117. Brown, B. J. and J. J. Ewel. 1987. Herbivory in complex and simple tropical successional ecosystems. Ecology 68:108-116. Brown, J. H., and A. Kodric-Brown. 1979. Convergence, competition, and mimicry in a temperate community of hummingbird-pollinated flowers. Ecology 60:1022-1035. Brown, J. R., and S. Archer. 1987. Woody plant seed dispersal and gap formation in a North American subtropical woodland: the role of domestic herbivores. Vegetatio 73:73-80. Brown, J. R., and S. Archer. 1989. Woody plant invasion of grasslands: establishment of honey mesquite (Prosopis glandulosa var. glandulosa) on sites differing in herbaceous biomass and grazing history. Oecologia 80:19-26. Bruin, J., M. Dicke, and M. W. Sabellis. 1992. Plants are better protected against spider mites after exposure to volatiles from infested conspecifics. Experientia 48:525-529. Bruno, J. F. 2000. Facilitation of cobble beach communities through habitat modification by Spartina alterniflora. Ecology 81:1179-1192. Bruno, J. F., and C. W. Kennedy. 2000. Patch-size dependent habitat modification and facilitation on New England cobble beaches by Spartina alterniflora. Oecologia 122:98-108. Bruno, J. F., J. J. Stachowitcz, and M. E. Bertness. 2003. Inclusion of facilitaiton into general ecological theory. Trends in Ecology and Evolution 18:119-125. Buckley, D. S., T. L. Sharik, and J. G. Isebrands. 1998. Regeneration of northern red oak: positive and negative effects of competitor removal. Ecology 79:65-78. Bullock, S. H. 1991. Herbivory and the demography of the chaparral shrub Ceanothus greggii (Rhamnaceae). Madrono 38:63-72. Bush, J. K., and O. W. Van Auken. 1986. Changes in nitrogen, carbon, and other surface soil properties during secondary successsion. Soil Science Society of America 50:1597-1601. Buss, L. W., and J. B. C. Jackson. 1979. Competitive networks: nontransitive competitive relationships in cryptic coral reef communities. American Naturalist 113:223-234. Byers, D. L. 1995. Pollen quantity and quality as explanation for low seed set in small populations exemplified by Eupatorium (Asteraceae). American Journal of Botany 82:1000-1006.
References
345
Byers, J. E., K. Cuddington, C. G. Jones, T. S. Talley, A. Hastings, J. G. Lambrinos, J. A. Crooks, and W. G. Wilson. 2006. Using ecosystem engineers to restore ecological systems. Trends in Ecology and Evolution 21:493-500. Caccia, F. D., and C. L. Ballare. 1998. Effects of tree cover, understory vegetation, and litter on regeneration of Douglas-fir (Pseudotsuga menziesii) in southwestern Argentina. Canadian Journal of Forest Research 28:683-692. Caldeira, M. C., R. J. Ryel, J. H. Lawton, and J. S. Pereira. 2001. Mechanisms of positive biodiversity-production relationships: insights provided by 13 C analysis in experimental mediterranean grassland plots. Ecology Letters 4:439-443. Caldwell, M. M. 1990. Water parasitism stemming from hydraulic lift: a quantitative test in the field. Israel Journal of Botany 39:395-402. Caldwell, M. M., T. E. Dawson, and J. H. Richards. 1998. Hydraulic lift: consequences of water efflux from the roots of plants. Oecologia 113:151-161. Caldwell, M. M., D. M. Eissenstat, J. H. Richards, and M. F. Allen. 1985. Competition for phosphorus: Differential uptake for dual isotopelabeled soil interspaces between shrub and grass. Science 229:384-386. Caldwell, M. M., and J. H. Richards. 1989. Hydraulic lift: water efflux from upper roots improves effectiveness of water uptake from deep roots. Oecologia 79:1-5. Callaghan, T. V. 1987. Plant population processes in arctic and boreal regions. Pages 58-68 in M. Sonesson, editor. Research in Arctic Life and Earth Sciences: Present Knowledge and Future Perspectives. Ecological Bulletins, London, UK. Callaghan, T. V., and U. Emanuelsson. 1985. Population structure process of tundra plants and vegetation. Pages 399-439 in J. White, editor. The Population Structure of Vegetation. Junk, Dordrecht. Callaway, R. M. 1990. Effects of soil water distribution on the lateral root development of three species of California oaks. American Journal of Botany 77:1469-1475. Callaway, R.M. 1990. Effects of Quercus douglasii on grassland productivity and nutrient cycling in central California. PhD Dissertation, University of California, Santa Barbara, California, USA. Callaway, R. M. 1992. Effect of shrubs on recruitment of Quercus douglasii and Quercus lobata in California. Ecology 73:2118-2128. Callaway, R. M. 1994. Facilitative and interfering effects of Arthocnemum subterminale on winter annuals. Ecology 75:681-686. Callaway, R. M. 1995. Positive interactions among plants. Botanical Review 61:306-349.
346
References
Callaway, R. M. 1997. Positive interactions in plant communities and the individualistic-contiuum concept. Oecologia 112:143-149. Callaway, R. M. 1998a. Are positive interactions species-specific? Oikos 82: 202-207. Callaway, R. M. 1998b. Competition and facilitation on elevation gradients in subalpine forests of the northern Rocky Mountains, USA. Oikos 82:561-573. Callaway, R. M., and E. T. Aschehoug. 2000. Invasive plants versus their new and old neighbors: a mechanism for exotic invasion. Science 290:521-523. Callaway, R. M., R. W. Brooker, C. P, Z. Kikvidze, C. J. Lortie, R. Michalet, L. Paolini, F. I. Pugnaire, B. J. Cook, E. T. Aschehoug, C. Armas, and B. Newingham. 2002. Positive interactions among alpine plants increase with stress: a global experiment. Nature 417:844-848. Callaway, R. M., and S. C. Sabraw. 1994. Effects of variable precipitation on the structure and diversity of a winter annual plant community in a central California marsh. Journal of Vegetation Science 5:433-438. Callaway, R. M., K. O. Reinhart, T. S.C., and S. C. Pennings. 2001b. Effects of epiphytic lichens on host preference of the vascular epiphyte Tillandsia usneoides. Oikos 94:433-441. Callaway, R. M., and C. M. D’Antonio. 1991. Shrub facilitation of coast live oak establishment in central California. Madrono 38:158-169. Callaway, R. M., and F. W. Davis. 1993. Vegetation dynamics, fire, and the physical environment in coastal central California. Ecology 74:1567-1578. Callaway, R. M., and F. W. Davis. 1998. Recruitment of Quercus agrifolia in central California: the importance of shrub-dominated patches. Journal of Vegetation Science 9:647-656. Callaway, R. M., E. H. DeLucia, D. Moore, R. Nowak, and W. H. Schlesinger. 1996. Competition and facilitation: contrasting effects of Artemisia tridentata on desert vs. montane pines. Ecology 77:2130-2141. Callaway, R. M., D. Kikodze, M. Chiboshvili, and L. Khetsuriani. 2005. Unpalatable plants protect neighbors from grazing and increase plant community diversity. Ecology 86:1856-1862. Callaway, R. M., D. Kikodze, and Z. Kikvidze. 2000. Facilitation by unpalatable weeds may conserve plant diversity in overgrazed meadows in the Caucasus Mountains. Oikos 89:275-282. Callaway, R. M., and L. King. 1996. Temperature-driven variation in substrate oxygenation and the balance of competition and facilitation. Ecology 77:1189-1195.
References
347
Callaway, R. M., B. E. Mahall, C. Wicks, J. Pankey, and C. A. Zabinski. 2003. Soil fungi and the effects of an invasive forb on native versus naturalized grasses: neighbor identity matters. Ecology 84:129-135. Callaway, R. M., and N. M. Nadkarni. 1991. Seasonal patterns of nutrient deposition in a Quercus douglasii woodland in central California. Plant and Soil 137:209-222. Callaway, R. M., N. M. Nadkarni, and B. E. Mahall. 1991. Facilitation and interference of Quercus douglasii on understory productivity in central California. Ecology 72:1484-1499. Callaway, R. M., B. Newingham, C. A. Zabinski, and B. E. Mahall. 2001a. Compensatory growth and competitive ability of and invasive weed are enhanced by soil fungi and neighbors. Ecology Letters 4:1-5. Callaway, R. M., and S. C. Pennings. 2000. Facilitation may buffer competitive effects: indirect and diffuse interactions among salt marsh plants. American Naturalist 156:416-424. Callaway, R. M., S. C. Pennings, and C. L. Richards. 2003. Phenotypic plasticity and interactions among plants. Ecology 84:1115-1128. Callaway, R. M., K. O. Reinhart, G. W. Moore, D. J. Moore, and S. C. Pennings. 2002. Epiphyte host preferences and host traits: mechanisms for species-specific interactions. Oecologia 132:221-230. Callaway, R. M., and W. M. Ridenour. 2004. Novel weapons: a biochemically based hypothesis for invasive success and the evolution of increased competitive ability. Frontiers in Ecology and the Environment 2:436-433. Callaway, R. M., A. Sala, and R. Keane. 2000. Succession may maintain high leaf area:sapwood ratios and productivity in old subalpine forests. Ecosystems 3:254-268. Callaway, R. M., G. C. Thelen, A. Rodriguez, and W. E. Holben. 2004. Soil biota and exotic invasion. Nature 427:731-733. Callaway, R. M., and L. R. Walker. 1997a. Competition and facilitation: a synthetic approach to interactions in plant communities. Ecology 78:1958-1965. Camarero, J. J., and E. Gutiérrez. 1999. Structure and recent recruitment at alpine forest-pasture ecotones in the Spanish central Pyrenees. Ecoscience 6:451-464. Campbell, B. M., T. Lynam, and J. C. Hatton. 1990. Small-scall patterning in the recruitment of forest species during succession in tropical dry forest, Mozambique. Vegetatio 87:51-57. Canfield, R. H. 1948. Perennial grass composition as an indicator of condition of southwestern mixed grass ranges. Ecology 29:190-204. Canham, C. D., K. D. Coates, P. Bartemucci, and P. Quaglia. 1999. Measurement and modeling of spatially explicit variation in light
348
References
transmission through interior cedar–hemlock forests of British Columbia. Canadian Journal of Forest Research 29:1775-1783. Canham, C. D., A. C. Finzi, S. Pacala, and D. H. Burbank. 1994. Causes and consequences of resource heterogeneity in forests: interspecific variation in light transmission by canopy trees. Canadian Journal of Forest Research 24:337-349. Cantero, J. J., M. Partel, and M. Zobel. 1999. Is species richness dependent on the neighboring stands? An analysis of the community patterns in mountain grasslands of central Argentina. Oikos 87:346-354. Cappuccino, N. 2004. Allee effect in an invasive alien plant, pale swallowwort Vincetoxicum rossicum (Asclepiadaceae). Oikos 106:3-8. Cardinale, B. J., M. A. Palmer, and S. L. Collins. 2002. Species diversity enhances ecosystem functioning through interspecific facilitation. Nature 415:426-429. Carey, E. V., M. J. Marler, and R. M. Callaway. 2004. Mycorrhizae transfer carbon from a native grass to an invasive weed: evidence from stable isotopes and physiology. Plant Ecology 172:133-141. Carino, D. A., and C. C. Daehler. 2002. Can inconspicuous legumes facilitate alien grass invasions? Partridge peas and fountain grass in Hawai’i. Ecography 25:33-41. Carlsson, B. A., and T. V. Callaghan. 1991. Positive plant interactions in tundra vegetation and the importance of shelter. Journal of Ecology 79:973-983. Carlyle, J. C., and D. C. Malcolm. 1986. Nitrogen availability beneath pure spruce and mixed larch and spruce stands growing on deep peat. I. Net mineralization measured by field and laboratory incubations. Plant and Soil 93:95-113. Carrillo-Garcia, A., Y. Bashan, and G. J. Bethlenfalvay. 2000a. Resourceisland soils and the survival of the giant cactus, cardon, of Baja California Sur. Plant and Soil 218:207-214. Carrillo-García, A., Y. Bashan, E. D. Rivera, and G. J. Bethlenfalvay. 2000b. Effects of Resource-Island Soils, Competition, and Inoculation with Azospirillum on Survival and Growth of Pachycereus pringlei, the Giant Cactus of the Sonoran Desert. Restoration Ecology 8:65-73. Carrillo-Garcia, A., J. L. L. de la Luz, Y. Bashan, and G. J. Bethlenfalvay. 1999. Nurse plants, mycorrhizae, and plant establishment in a disturbed area of the Sonoran Desert. Restoration Ecology 4:321-335. Carter, A. J., and T. G. O’Connor. 1991. A two-phase mosaic in a savanna grassland. Journal of Vegetation Science 2:231-236. Carter, R. W. G. 1991. Coastal Environments. Academic Press, London, UK.
References
349
Casal, J. J., V. A. Deregibus, and R. A. Sanchez. 1985. Variations in tiller dynamics and morphology in Lolium multiflorum Lam. vegetative and reproductive plants as affected by difference in red/far-red irradiation. Annals of Botany 56:553-559. Casal, J. J., R. A. Sanchez, and V. A. Deregibus. 1987. The effect of light quality on shoot extension growth in three species of grasses. Annals of Botany 59:1-7. Case, T. J. 1991. Invasion resistance, species build-up and community collapse in metapopulation models with interspecies competition. Biological Journal of the Linnean Society 42:239-266. Casper, B. B. 1996. Demographic consequences of drought in the herbaceous perennial Cryptantha flava: effects of density, associations with shrubs, and plant size. Oecologia 106:144-152. Castellanos, E. M., M. E. Figueroa, and A. J. Davy. 1994. Nucleation and facilitation in saltmarsh succession: interactions between Spartina maritima and Arthrocnemum perenne. Journal of Ecology 82:239-248. Castro, J., R. Zamora, J.A. Hodar and J. Gomez. 2004. Seedling establishment of a boreal tree species (Pinus sylvestris) at its southernmost distribution limit: consequences of being in a marginal Mediterranean habitat. Journal of Ecology 92:266-277. Castro, J., R. Zamora, J.A. Hodar and J.M. Gomez. 2002. Use of shrubs as nurse plants: a new technique for reforestation in Mediterranean mountains. Restoration Ecology 10:297-305. Castro, J., R. Zamora, J. A. Hódar, and J. M. Gomez. 2004. Benefits of using shrubs as nurse plants for reforestation in Mediterranean mountains: a 4-year study. Restoration Ecology 12:352-358. Cavieres, L. A., M. T. K. Arroyo, A. Penaloza, M. Molina-Montenegro, and T. C. 2002. Nurse effect of Bolax gummifera cushion plants in the alpine vegetation of the Chilean Patagonian Andes. Journal of Vegetation Science 13:547–554. Cavieres, L. A., E. I. Badano, A. Sierra-Almeida, S. Gómez-Gonzalez, and M. A. Molina-Montenegro. 2006. Positive interactions between alpine plant species and the nurse cushion plant Laretia acaulis do not increase with elevation in the Andes of central Chile. New Phytologist 169:59-69. Cavieres, L. A., A. Penaloza, C. Papic, and M. Tambutti. 1998. Nurse effect of Laretia acaulis (Umbelliferae) in the high Andes of central Chile. Revista Chilena de Historia Natural 71:331-341 (in Spanish). Cavieres, L. A., C. L. Quiroz, M. A. Molina-Montenegro, A. A. Munoz, and A. Pauchard. 2006. Nurse effect of the native cushion plant Azorella monantha on the invasive non-native Taraxacum officinale in the
350
References
high-Andes of central Chile. Perspectives in Plant Ecology, Evolution and Systematics 7:217-226. Cerdá, A. 1997. The effect of patchy distribution of Stipa tenacissima L. on runoff and erosion. Journal of Arid Environments 36:37-51. Challinor, D. 1968. Alteration of surface soil by four tree species. Ecology 49:286-290. Chapin, F. S. I., and G. R. Shaver. 1985. Individualistic growth response of tundra plant species to environmental manipulations in the field. Ecology 66:564-576. Chapin, F. S. I., L. R. Walker, C. L. Fastie, and L. C. Sharman. 1994. Mechanisms of primary succession following deglaciation at Glacier Bay. Ecological Monographs 64:149-175. Chapin, F. S. I., E. S. Zavaleta, V. T. Eviner, R. L. Naylor, P. M. Vitousek, H. L. Reynolds, D. U. Hooper, S. Lavorel, O. E. Sala, S. E. Hobbie, M. C. Mack, and S. Diaz. 2000. Consequences of changing biodiversity. Nature 405:234-242. Chapman, A. S., and A. R. O. Chapman. 1999. Effects of cordgrass on saltmarsh fucoids: reduced desiccation and light availability, but no changes in biomass. Journal Experimental Marine Biology 238:69-91. Chapman, K., J. B. Whittaker, and O. W. Heal. 1988. Metabolic and faunal activity in litters of three mixtures compared with pure stands. Agriculture, Ecosystems and Environment 24:33-40. Chapman, V. J. 1974. Salt marshes and salt deserts of the world. Pages 3-19 in R. J. Reimold and W. H. Queen, editors. Ecology of Halophytes. Academic Press, New York, USA. Charley, J. L., and N. E. West. 1975. Plant-induced soil chemical patterns in some shrub-dominated semi-desert ecosystems of Utah. Journal of Ecology 63:945-964. Chen, J., and J. M. Stark. 2000. Plant species effects and carbon and nitrogen cycling in a sagebrush-crested wheatgrass soil. Soil Biology and Biochemistry 32:47-57. Chiariello, N. R., J. C. Hickman, and H. Mooney. 1982. Endomycorrhizal role for interspecific transfer of phosphorus in a community of annual plants. Science 217:941-943. Chock, J. S., and A. C. Mathieson. 1976. Ecological studies of the salt marsh ecad scorpiodies (Hornemann) Hauck of Ascophyllum nodosum (L.) Le Jolis. Journal of Experimental Marine Biology 23:171-190. Choler, P., R. Michalet, and R. M. Callaway. 2001. Facilitation and competition on gradients in alpine plant communities: revisiting the “individualistic” hypothesis. Ecology 82:3295-3308.
References
351
Christian, S., D. Béguin, A. Buttler, and H. Müller-Schärer. 2005. Safe sites for tree regeneration in wooded pastures: A case of associational resistance? Journal of Vegetation Science 16:209-214. Christie, D. A., and J. J. Armesto. 2003. Regeneration microsites and tree species coexistence in temperate rain forests of Chiloé Island, Chile. Journal of Ecology 91:776-784. Cleavitt, N. 2004. Comparative ecology of a lowland and a subalpine species of Mnium in the northern Rocky Mountains. Plant Ecology 174:205-216. Clements, F. E. 1916. Plant Succession. Carnegie Institute Washington Publications 242. Cody, M. L. 1993. Do cholla cacti (Opuntia spp. subgenus Cylindropuntia) use or need nurse plants in the Mojave Desert? Journal of Arid Environments 24:139-154. Compton, R. H. 1929. The vegetation of the Karoo. Journal of the Botanical Society of South Africa 15:13-21. Connell, J. H. 1983. On the prevalance and relative importance of interspecific competition: evidence from field experiments. American Naturalist 122:661-696. Connell, J. H., and R. O. Slatyer. 1977. Mechanisms of succession in natural communities and their role in community stabilty and organization. American Naturalist 111:1119-1144. Connor, J. M., and B. L. Willoughby. 1997. Effects of blue oak canopy on annual forage production. Pages 321-326 in USDA Forest Service General Technical Report PSW-GTR-160, Berkeley, California, USA. Corak, S. J., D. G. Blevins, and S. G. Pallardy. 1987. Water transfer in an alfalfa/maize association. Plant Physiology 84:582-586. Coult, D. A., and K. B. Vallance. 1958. Observations on the gaseous exchanges which take place between Menyanthes trifotiata L. and its environment. Journal of Experimental Botany 9:384-402. Couteron, P., A. Mahamane, P. Ouedraogo, and J. Seghieri. 2000. Differences between banded thickets (tiger bush) at two sites in West Africa. Journal of Vegetation Science 11:321-328. Cowles, H. C. 1899. The ecological relations of the vegetation on the sand dunes of Lake Michigan. Botanical Gazette 27:95-117. Cowling, R. M., and T. Gxaba. 1990. Effects of a fynbos overstory shrub on community structure: implications for the maintenance of community-wide species richness. South African Journal of Ecology 1:1-7. Cross, A. F., and W. H. Schlesinger. 1999. Plant regulation of soil nutrient distribution in the northern Chihuahuan Desert. Plant Ecology 145:11-25.
352
References
Crozier, C. R., and R. E. J. Boerner. 1986. Stemflow induced soil nutrient heterogeneity in a mixed mesophytic forest. Bartonia 52:1-8. Cruzan, M. B. 1986. Pollen tube distributions in Nicotiana glauca: evidence for density dependent growth. American Journal Botany 73:902-907. Cullings, K. W., T. Szaro, and T. D. Bruns. 1996. Evolution of extreme specialization within a lineage of ectomycorrhizal epiparasites. Nature 379:63-67. Cunliffe, R. N., M. L. Jarman, E. J. Moll, and R. I. Yeaton. 1990. Competitive interactions between the perennial shrub Leipoldtia constricta and an annual forb, Gorteria diffusa. South African Journal of Botany 56: 34-37. Curtis, J. T. 1959. The Vegetation of Wisconsin. The University of Wisconsin Press, Madison. Dacey, J. W. H. 1981. Pressurized ventilation in the water lily. Ecology 62:1137-1147. Dacey, J. W. H., and B. L. Howes. 1984. Water uptake by roots controls water table movement and sediment oxidation in short Spartina marsh. Science 224:487-489. Dakora, F. D., and D. A. Phillips. 2002. Root exudates as mediators of mineral acquisition in low-nutrient environments. Plant and Soil 245:35-47. Dale, M. R. T., E. A. John, and D. J. Blundon. 1991. Contact sampling for the detection of interspecific association: a comparison in two vegetation types. Journal of Ecology 79:781-792. Dancette, C., and J. F. Poulain. 1969. Influence of Acacia albida on pedoclimatic factors and crop yields. African Soils 14:143-184. Daniels, J. D., and R. O. Lawton. 1991. Habitat and host preference of Ficus crassiuscula, a neotropical strangling fig of the lower-montane rainforest. Journal of Ecology 79:129-141. Danin, A., S. Rae, M. Barbour, N. Juravcic, P. Connors, and E. Uhlinger. 1998. Early primary succession on dunes at Bodega Head, California. Madrono 45:101-109. Darwin, C. 1858. The Origin of Species by Means of Natural Selection. Murrary, London, UK. Davidson, D. W. 1980. Some consequences of diffuse competition in a desert ant community. American Naturalist 116: 92-105. Davis, H. G., C. M. Taylor, J. C. Civille, and D. R. Strong. 2004. An Allee effect at the front of a plant invasion: Spartina in a Pacific estuary. Journal of Ecology 92:321-327. Davis, M. A., K. J. Wrage, P. B. Reich, M. G. Tjoelker, T. Schaeffer, and C. Muermann. 1999. Survival, growth, and photosynthesis of tree
References
353
seedlings competing with herbaceous vegetation along a water-light nitrogen gradient. Plant Ecology 145:341-350. Davis, M. H., and S. R. Simmons. 1994. Far-red light reflected from neighbouring vegetation promotes shoot elongation and accelerates flowering in spring barley plants. Plant, Cell and Environment 17:829-836. Dawson, T. E. 1993. Hydraulic lift and water use by plants: implications for water balance, performance and plant-plant interactions. Oecologia 95:565-574. Dawson, T. E. 1996. Determining water use by trees and forests from isotopic, energy balance, and transpiration analyses: the roles of tree size and hydraulic lift. Tree Physiology 16:263-272. Dawson, T. E. 1998. Fog in the California redwood forest: ecosystem inputs and use by plants. Oecologia 117:476-485. Dawson, T. E. and J. R. Ehleringer. 1998. Plants isotopes, and water use: a catchment-scale perspective. Pages 211-260 in C. Kendall and J. J. McDonnell, editors. Isotope Tracers in Catchment Hydrology. Elsevier Science, New York, USA. Day, T. and K. A. Young. 2004. Competitive and facilitative evolutionary diversification. BioScience 54:101-109. Day, T. A. and R. G. Wright. 1989. Positive plant spatial association with Eriogonum ovalifolium in primary succession on cinder cones: seedtrapping nurse plants. Vegetatio 80:37-45. De Jong, T. J., and P. G. L. Klinkhamer. 1988. Seedling establishment of the biennials Cirsium vulgare and Cynoglossum officinale in a sand-dune area: the importance of water for differential survival and growth. Journal of Ecology 76:393-402. de la Cruz, A., C. T. Hackney, and N. Bhardwaj. 1989. Temporal and spatial patterns of redox potential (Eh) in three tidal marsh communities. Wetlands 9:181-190. de Viana, M. L., S. Sühring, and B. F. J. Manly. 2001. Application of randomization methods to study the association of Trichocereus pasacana (Cactaceae) with potential nurse plants. Plant Ecology 156:193-197. DeAngelis, D. L., W. M. Post, and C. C. Travis. 1986. Positive Feedback in Natural Systems. Springer, Berlin, Germany. Deckmyn, G., E. Cayenberghs, and R. Ceulemans. 2001. UV-B and PAR in single and mixed canopies grown under different UV-B exclusions in the field. Plant Ecology 154:125-133. Dejean, A., I. Olmstead, and R. R. Snelling. 1995. Tree-epiphyte-ant relationships in the low innundated forest of the Sian Ka’an Biosphere Reserve, Quintana Roo, Mexico. Biotropica 27:57-70.
354
References
del Moral, R., and L. C. Bliss. 1993. Mechanisms of primary succession: insights resulting from the eruption of Mount St. Helens. Advances in Ecological Research 24:1-66. del Moral, R., and R. M. Wood. 1993. Early primary succession on the volcano Mount St. Helens. Journal of Vegetation Science 4:223-234. DeLucia, E. H., W. H. Schlesinger, and W. D. Billings. 1988. Water relations and the maintenance of Sierran conifers on hydrothermally altered rock. Ecology 69:303-311. DeLucia, E. H., W. H. Schlesinger, and W. D. Billings. 1989. Edaphic limitations to growth and photosynthesis in Sierran and Great Basin vegetation. Oecologia 78:184-190. DeLucia, E. H., and W. K. Smith. 1987. Air and soil temperature limitations on photosynthesis in Englemann spruce during summer. Canadian Journal of Forest Research 17:527-533. Dennis, W. M., and W. T. Bateson. 1974. The floating log and stump communities in the Santee Swamp of South Carolina. Castanea 39:166-170. Dickie, I. A., S. A. Schnitzer, P. B. Reich, and S. E. Hobbie. 2005. Spatially disjunct effects of co-occurring competition and facilitation. Ecology Letters 8:1191-1200. Diekman, M., and U. Falkengren-Grerup. 1998. A new species index for forest vascular plants: development of functional indices based on mineralization rates of various forms of soil nitrogen. Journal of Ecology 86:269-283. Dijkstra, F. 2003. Calcium mineralization in the forest floor and surface soil beneath different tree species in the northeastern U.S. Forest Ecology and Management 175:185-194. Dijkstra, F. A., C. Geibe, S. Holmstro, U. S. Lundstro, and N. van Breemen. 2001. The effect of organic acids on base cation leaching from the forest floor under six North American tree species. European Journal of Soil Science 52:205-214. Dijkstra, F. A., and M. M. Smits. 2002. Tree species effects on calcium cycling: the role of calcium uptake in deep soils. Ecosystems 5:385-398. Dijkstra, F. A., N. van Breemen, A. Jongmans, G. R. Davies, and G. E. Likens. 2003. Calcium weathering in forested soils and the effect of different tree species. Biogeochemistry 62:253-275. Doak, D. F., D. Bigger, E. Harding-Smith, M. A. Marvier, R. O’Malley, and D. Thomson. 1998. The statistical inevitability of stability-diversity relationships in community ecology. American Naturalist 154:314-329. Dodds, W. K. 1997. Interspecific interactions: constructing a general neutral model for interaction type. Oikos 78:377-383.
References
355
Dodds, W. K., and G. M. Henebry. 1996. The effect of density dependence on community structure. Ecological Modeling 93:33-42. Dolch, R. a. T. T. 2000. Defoliation of alders (Alnus glutinosa) affects herbivory by leaf beetles on undamaged neighbours. Oecologia 125:504-511. Donovan, L. A., and J. H. Richards. 2000. Juvenile shrubs show differences in stress tolerance, but no competition or facilitation, along a stress gradient. Journal of Ecology 88:1-16. Dormann, C. F., and R. W. Brooker. 2002. Facilitation and competition in the high Arctic: the importance of the experimental approach. Acta Oecologia 23:297-301. Doutt, R. L., and J. Nakata. 1973. The Rubus leafhopper and its egg parasitoid: an endemic biotic system useful in grape-pest management. Environmental Entomology 2:381-386. Drivas, E. P., and R. L. Everett. 1988. Water relations characteristic of competing singleleaf pinyon seedlings and sagebrush nurse plants. Forest Ecology and Management 23:27-37. Dudley, T. L., and N. B. Grimm. 1994. Modification of macrophyte resistance to disturbance by an exotic grass, and implications for desert stream succession. Verhandlungen International Verein Limnologie 25: 1456-1460. Dunne, J. A., and V. T. Parker. 1999. Species-mediated soil moisture availability and patchy establishment of Pseudotsuga menziesii in chaparral. Oecologia 119:36-45. Dunwiddie, P. W. 1977. Recent tree invasion of subalpine meadows in the Wind River Mountains, Wyoming. Arctic and Alpine Research 9:393-399. Dupraz, C., V. Simorte, M. Dauzat, G. Bertoni, A. Bernadac, and P. Masson. 1999. Growth and nitrogen status of young walnuts as affected by intercropped legumes in a Mediterranean climate. Agroforestry Systems 43:71-80. Dyer, A. R., A. Fenech, and R. K.J. 2000. Accelerated seedling emergence in interspecific competitive neighborhoods. Ecology Letters 3:523-529. Dzwonko, Z., and S. Loster. 1997. Effects of dominant trees and anthropogenic disturbances on species richness and floristic composition of secondary communities in southern Poland. Journal of Applied Ecology 34:861-870. Ebersohn, J. P., and P. Lucas. 1965. Trees and soil nutrients in south west Queensland. Queensland Journal of Agricultural and Animal Science 22:431-436.
356
References
Eccles, N. S., K. J. Esler, and R. M. Cowling. 1999. Spatial pattern analysis in Namaqualand desert plant communities: evidence for general positive interactions. Plant Ecology 142:71-85. Eckstein, R. L. 2005. Differential effects of interspecific interactions and water availability on survival, growth and fecundity of three congeneric grassland herbs. New Phytologist 166:525-536. Eckstein, R. L., and T. W. Donath. 2005. Interactions between litter and water availability affect seedling emergence in four familial pairs of floodplain species. Journal of Ecology 93:807-816. Egerton, J. J. G., J. C. G. Banks, A. Gibson, R. B. Cunningham, and M. C. Ball. 2000. Facilitation of seedling establishment: reduction in irradiance enhances winter growth of Eucalyptus pauciflora. Ecology 81:1437-1449. Egerton, J. J. G., and S. D. Wilson. 1993. Plant competition over winter in alpine shrubland and grassland, Snowy Mountains, Australia. Arctic and Alpine Research 2:124-129. Eggler, W. A. 1941. Primary succession on volcanic deposits in southern Idaho. Ecological Monographs 3:278-298. Ehlers, B. K., and J. Thompson. 2004. Do co-occurring plant species adapt to one another? The response of Bromus erectus to the presence of different Thymus vulgaris chemotypes. Oecologia 141:511-515. Ehrenfeld, J. G. 1990. Dynamics and processes of Barrier Island vegetation. Aquatic Science 2:437-480. El-Bana, M. I., I. Nijs, and F. Kockelbergh. 2002. Microenvironmental and vegetational heterogeneity induced by phytogenic nebkhas in an arid coastal ecosystem. Plant and Soil 247:283-293. Ellison, L., and W. R. Houston. 1958. Production of herbaceous vegetation in openings and under canopies of western aspen. Ecology 39:337-345. Emerman, S. H., and T. E. Dawson. 1996. Hydraulic lift and its influence on the water content of the rhizosphere: an example from sugar maple: Acer saccharum. Oecologia 108:273-278. Engelaar, W. M. H. G., J. C. Symens, H. J. Laanbrock, and C. W. P. M. Blom. 1995. Preservation of nitrifying capacity and nitrate availability in waterlogged soils by radial oxygen loss from roots of wetland plants. Biology and Fertility of Soils 20:243-248. Enright, N. J., and B. B. Lamont. 1989. Seed banks, fire season, safe sites and seedling recruitment in five co-occurring Banksia species. Journal of Ecology 77:1111-1122. Eom, A.-H., D. C. Hartnett, and G. W. T. Wilson. 2000. Host plant species effects on arbuscular mycorrhizal fungal communities in tallgrass prairie. Oecologia 122:435-444.
References
357
Erlich, P. R. 1990. Habitats in crisis: why should we care about the loss of species? Forest Ecology and Management 35:5-11. Erlich, P. R., and E. O. Wilson. 1991. Biodiversity studies: science and policy. Science 253:758-762. Escudero, A., B. Garcia, J. M. Gomez, and E. Luis. 1985. Nutrient cycling in Quercus rotundifolia and Q. pyrenaica ecosystems of Spain. Oecologia Plantarum 6:73-86. Espeleta, J. F., J. B. West, and L. A. Donovan. 2004. Species-specific patterns of hydraulic lift in co-occurring adult trees and grasses in a sandhill community. Oecologia 138:341-349. Espigares, T., A. López-Pintor, and J. M. R. Benayas. 2004. Is the interaction between Retama sphaerocarpa and its understorey herbaceous vegetation always reciprocally positive? Competition-facilitation shift during Retama establishment. Acta Oecologia 26:121-128. Espinar, J. L., L. V. García, P. García Murillo, and J. Toja. 2002. Submerged macrophyte zonation in a Mediterranean salt marsh: a facilitation effect from established helophytes? Journal of Vegetation Science 13:831-840. Everett, R. L., S. Koniak, and J. Budy. 1986. Pinyon seedling distribution among soil surface microsites. Research Paper INT-363, United States Department of Agriculture, Forest Service, Intermountain Research Station, Ogden, Utah, USA. Facelli, J., and A. M. Temby. 2002. Multiple effects of shrubs control the distribution and performance of annual plants in arid lands of South Australia. Austral Ecology 27:422-432. Facelli, J. M., and D. J. Brock. 2000. Patch dynamics in arid lands: localized effects of Acacia papyrocarrpa on soils and vegetation of open woodlands of South Australia. Ecography 23:479-491. Facelli, J. M., and S. T. A. Pickett. 1991. Indirect effects of litter on woody seedlings subject to herb competition. Oikos 62:129-138. Falik, O., P. Reides, M. Gersani, and A. Novoplansky. 2005. Root navigation by self inhibition. Plant, Cell and Environment 28:562-569. Fan, T. W.-M., A. N. Lane, J. Pedler, D. Crowley, and R. M. Higashi. 1997. Comprehensive analysis of organic ligands in whole root exudates using nuclear magnetic resonance and gas chromatography–mass spectrometry. Analytical Biochemistry 251:57-68. Farmer, E. E., and C. A. Ryan. 1990. Interplant communication: airborne methyl jasmonate induces synthesis of proteinase inhibitors in plant leaves. Proc. Natl. Acad. Sci. USA 87:7713-7716. Farmer, E. E., H. Weber, and S. Vollenweider. 1998. Fatty acid signaling in Arabidopsis. Planta 206:167-174.
358
References
Feinsinger, P., G. K. Murray, S. Kinsman, and W. H. Busby. 1986. Floral neighborhood and pollination success in four hummingbird-pollinated cloud forest plant species. Ecology 67:449-464. Feldman, T. S., W. F. Morris, and W. G. Wilson. 2004. When can two plant species facilitate each other’s pollination? Oikos 105:197-207. Fensham, R. J., and D. W. Butler. 2004. Spatial pattern of dry rainforest colonizing unburnt Eucalyptus savanna. Austral Ecology 29:121-128. Fenton, N. J., and Y. Bergeron. 2006. Facilitative succession in a boreal bryophyte community driven by changes in available moisture and light. Journal of Vegetation Science 17:65-76. Ferguson, B. G. 2001. Post-agricultural tropical forest succession: patterns, processes and implications for conservation and restoration. Dissertation. University of Michigan, Ann Arbor, Michigan, USA. Ferland, C., and L. Rochefort. 1997. Restoration techniques for Sphagnumdominated peatlands. Canadian Journal of Botany 75:1110-1118. Figeroa, M. E., J. M. Castillo, S. Redondo, T. Luque, E. M. Castellanos, F. J. Nieva, C. J. Luque, A. E. Rubio-Casal, and A. J. Davy. 2003. Facilitated invasion by hybridization of Sarcocornia species in a saltmarsh succession. Journal of Ecology 91:616-626. Fike, J., and W. A. Niering. 1999. Four decades of old field vegetation development and the role of Celastrus orbiculatus in the northeastern United States. Journal of Vegetation Science 10:483-492. Finzi, A. C., C. D. Canham, and N. Van Breemen. 1998b. Canopy tree-soil interactions within temperate forests: species effects on pH and cations. Ecological Applications 8:447-454. Finzi, A. C., N. Van Breemen, and C. D. Canham. 1998a. Canopy tree-soil interactions within temperate forests: species effects on soil carbon and nitrogen. Ecological Applications 8:440-446. Fischer, M., and D. Matthies. 1998a. Effects of population size on performance in the rare plant Gentianella germanica. Journal of Ecology 86:195-204. Fischer, M., and D. Matthies. 1998b. Experimental demography of the rare Gentianella germanica: seed bank formation and microsite effects on seedling establishment. Ecography 21:269-278. Fischer, M., M. van Kluenen, and B. Schmid. 2000. Genetic allee effects on performance, plasticity, and genetic stability in a clonal plant. Ecology Letters 3:520-539. Fitter, A. H., J. D. Graves, N. K. Watkins, D. Robinson, and C. Scrimgeour. 1998. Carbon transfer between plants and its control in networks of arbuscular mycorrhizas. Funtional Ecology 12:406-412.
References
359
Flores-Martinez, A., E. Ezcurra, and S. Sanchez. 1994. Effect of Neobuxbaumia tetetzo on growth and fecundity of its nurse plant Mimosa luisiana. Journal of Vegetation Science 6:73-78. Flores, J., and E. Jurado. 2003. Are nurse-protégé interactions more common among plants from arid environments? Journal of Vegetation Science 14:911-916. Fogel, B. N., C. M. Crain, and M. D. Bertness. 2004. Community level engineering effects of Triglochin maritima (seaside arrowgrass) in a salt marsh in northern New England, USA. Journal of Ecology 92:589-600. Fonteyn, P. J., and B. E. Mahall. 1981. An experimental analysis of structure in a desert plant community. Journal of Ecology 69:883-896. Foster, B. L. 2002. Competition, facilitation, and the distribution of Schizachryium scoparium along a topographic-productivity gradient. Ecoscience 9:355-363. Fowler, N. 1986. The role of competition in plant communities in semiarid regions. Annual Review of Ecology and Systematics 17:89-110. Francis, R., and D. J. Read. 1984. Direct transfer of carbon between plants connected by vesicular-arbuscular mycorrhizal mycelium. Nature 307:53-56. Franco-Pizana, J., T. E. Fulbright, and D. T. Gardiner. 1995. Spatial relations between shrubs and Prosopis glandulosa canopies. Journal of Vegetation Science 6:73-78. Franco-Pizana, J. G., T. E. Fulbright, D. T. Gardiner, and A. R. Tipton. 1996. Shrub emergence and seedling growth in microenvironment created by Prosopis glandulosa. Journal of Vegetation Science 7:257-264. Franco, A. C., and P. S. Nobel. 1989. Effect of nurse plants on the microhabitat and growth of cacti. Journal of Ecology 77:870-886. Franklin, J. F., W. H. Moir, G. W. Douglas, and C. Wiberg. 1971. Invasion of subalpine meadows by trees in the Cascade Range, Washington and Oregon. Arctic and Alpine Research 3:215-224. Franks, S. J., and C. J. Peterson. 2002. Burial disturbance leads to facilitation among coastal dune plants. Plant Ecology 168:13-21. Freedman, B. 1989. Environmental Ecology. Academic Press, New York, USA. Freeman, D. C., and J. M. Emlen. 1995. Assessment of interspecific interactions in plant communities: an illustration from the cold desert saltbush grasslands of North America. Journal of Arid Environments 31:179-198. Frei, J. K., and C. H. Dodson. 1972. The chemical effect of certain bark substrates on the germination and early growth of epiphytic orchids. Bulletin of the Torrey Botanical Club 99:301-307.
360
References
Freiberg, M. 2001. The influence of epiphyte cover on branch temperature in a tropical tree. Plant Ecology 153:241-250. Frelich, L. E., R. R. Calcote, M. B. Davis, and J. Pastor. 1993. Patch formation and maintenance in an old-growth hemlock-hardwood forest. Ecology 74:513-527. Frelich, L. E., S. Sugita, P. B. Reich, M. B. Davis, and S. K. Friedman. 1998. Neighbourhood effects in forests: implications for within-stand patch structure. Journal of Ecology 86:149-161. Fuentes, E. R., R. D. Otaiza, M. C. Alliende, A. Hoffman, and A. Poiani. 1984. Shrub clumps of the Chilean matorral vegetation: structure and possible maintenance mechanisms. Oecologia 62:405-411. Fulbright, T. E., J. O. Kuti, and A. R. Tipton. 1997. Effects of nurse-plant canopy light intensity on shrub seedling growth. Journal of Range Management 50:607-610. Futuyma, D. J. 1979. Evolutionary Biology. Sinauer, Sundlerland, Massachusetts, USA. Gadgil, R. L. 1971. The nutritional role of Lupinus arboreus in coastal sand dune forestry: 3. nitrogen distribution in the ecosystem before planting. Plant and Soil 35:113-126. Galen, C., and T. Gregory. 1989. Interspecific pollen transfer as a mechanism for competition: consequences of foreign pollen contamination for seed set in the alpine wildflower, Polemonium viscosum. Oecologia 81:120-123. Garcia-Moya, E., and C. M. McKell. 1970. Contribution of shrubs to the nitrogen economy of a desert-wash plant community. Ecology 51: 81-88. García, D., and J. R. Obeso. 2003. Facilitation by herbivore-mediated nurse plants in a threatened tree, Taxus baccata: local effects and landscape level consistency. Ecography 26:739-748. Garner, W., and Y. Steinberger. 1989. A proposed mechanism for the formation of ‘Fertile Islands’ in the desert ecosystem. Journal of Arid Environments 16:257-262. Garth, R. E. 1964. The ecology of Spanish moss (Tillandsia usneoides): its growth and distribution. Ecology 45:470-481. Gass, L., and P. W. Barnes. 1998. Microclimate and understory structure of live oak (Quercus fusiformis) clusters in central Texas, USA. Southwestern Naturalist 43:183-194. Gauslaa, Y., M. Ohlson, and J. Rolstad. 1998. Fine-scale distribution of the epiphytic lichen Usnea longissima on two even-aged neighboring Picea abies trees. Journal of Vegetation Science 9:95-102.
References
361
Gehring, C. A., and T. G. Whitham. 1991. Herbivore-driven mycorrhizal mutualism in insect-susceptible pinyon pine. Nature 353:556-557. Geiger, R. 1965. The Climate Near the Ground. Harvard University Press, Cambridge, Massachusetts, USA. Gentry, A. H., and C. Dodson. 1987. Contribution of nontrees to species richness of a tropical rain forest. Biotropica 19:149-156. Georgiadis, N. J. 1989. Microhabitat variation in an African savanna: effects of woody cover and herbivores in Kenya. Journal of Tropical Ecology 5:95-108. Gerard, V. A. 1999. Positive interactions between cordgrass, Spartina alterniflora, and the brown alga, Ascophyllum nodosum ecad scorpioides, in a mid-Atlantic coast salt marsh. Journal Experimental Marine Biology 239:157-164. Gerdol, R., L. Brancaleoni, M. Menghini, and R. Marchesini. 2000. Response of dwarf shrubs to neighbor removal and nutrient addition and their influence on community structure in a subalpine heath. Journal of Ecology 88:256-266. Germino, M. J., and W. K. Smith. 1999. Sky exposure, crown architecture, and low-temperature photoinhibition in conifer seedlings at alpine treeline. Plant, Cell and Environment 22:407-415. Germino, M. J., and W. K. Smith. 2000. Differences in microsite, plant form, and low-temperature photoinhibition in alpine plants. Arctic, Antarctic, and Alpine Research 32:388-396. Germino, M. J., W. K. Smith, and C. Resor. 2002. Conifer seedling distribution and survival in an alpine-treeline ecotone. Plant Ecology 162:157-168. Gersani, M., J. S. Brown, E. E. O’Brien, G. M. Maina, and Z. Abramsky. 2001. Tragedy of the commons as a result of root competition. Journal of Ecology 89:660-669. Gersper, P. L., and N. Holowaychuk. 1970. Some effects of stemflow from forest canopy trees on chemical properties of soils. Ecology 52: 691-702. Ghazoul, J., K. A. Liston, and T. J. B. Boyle. 1998. Disturbance-induced density-dependent set in Shorea siamensis (Dipterocarpaceae), a tropical forest tree. Journal of Ecology 86:462-473. Gibson, C. C., and A. R. Watkinson. 1992. The role of the hemiparasitic annual Rhinanthus minor in determining grassland community structure. Oecologia 89:62-68. Gigon, A. 1994. Positive interaktionen bei pflanzen in TrespenHalbtrockenrasen. Verhandlungen der Gesellschaft fur Okologie 23:1-6.
362
References
Gigon, A., and P. Ryser. 1986. Positive Interaktionen zwischen Pflanzenarten. Verhoff. Geobot. Inst. ETH 87:372-387. Gill, D. S., and P. L. Marks. 1991. Tree and shrub colonization of old fields in central New York. Ecological Monographs 61:183-205. Gleason, H. A. 1926. The individualistic concept of the plant association. Bulletin of the Torrey Botanical Club 53:7-27. Godinez-Alvarez, H., A. Valiente-Banuet, and L. V. Banuet. 1998. Biotic interactions and the population dynamics of the long-lived columnar cactus, Neobuxbaumia tetetzo in the Tehuacan Valley, Mexico. Canadian Journal Botany 77:203-208. Gold, W. G., and L. C. Bliss. 1995. Water limitations and plant community development in a polar desert. Ecology 76:1558-1568. Goldberg, D. E., and A. M. Barton. 1992. Patterns and consequences of interspecific competition in natural communities: a review of field experiments with plants. American Naturalist 139:771-801. Goldberg, D. E., R. Turkington, L. Olsvig-Whittaker, and A. R. Dyer. 2001. Density dependence in an annual plant community: variation among life history stages. Ecology 71:423-446. Goldman, C. R. 1961. The contribution of alder trees (Alnus tenuifolia) to the primary productivity of Castle Lake, California. Ecology 42:282-288. Gómez-Aparicio, L., J. M. Gómez, and R. Zamora. 2005a. Microhabitats shift rank in suitability for seedling establishment depending on habitat type and climate. Journal of Ecology 93:1194-1202. Gómez-Aparicio, L., J. M. Gómez, R. Zamora, and J. L. Boettinger. 2005d. Canopy vs. soil effects of shrubs facilitating tree seedlings in Mediterranean montane ecosystems. Journal of Vegetation Science 16:191-198. Gómez-Aparicio, L., F. Valladares, and R. Zamora. 2006. Differential light responses of Mediterranean tree saplings: linking ecophysiology with regeneration niche in four co-occurring species. Tree Physiology 26:947-958. Gómez-Aparicio, L., F. Valladares, R. Zamora, and J. L. Quero. 2005b. Response of tree seedlings to the abiotic heterogeneity generated by nurse shrubs: an experimental approach at different scales. Ecography 28:757-768. Gómez-Aparicio, L., R. Zamora, H. M. Gómez, J. A. Hódar, J. Castro, and E. Baraza. 2004. Applying plant facilitation to forest restoration: a meta-analysis of the use of shrubs as nurse plants. Ecological Applications 14:1128-1138. Gómez-Aparicio, L., R. Zamora, and J. M. Gómez. 2005c. The regeneration status of the endangered Acer opalus subsp. granatense throughout its
References
363
geographical distribution in the Iberian Peninsula. Biological Conservation 121:195-206. Gómez, J. M. 2005. Long-term effects of ungulates on performance, abundance and spatial distribution of two montane herbs. Ecological Monographs 75:231-258. Goodnight, C. J. 1990. Experimental studies of community evolution. 1. The response to selection at the community level. Evolution 44:1625-1636. Grace, J. 1989. Tree lines. Philosophical Transactions Royal Society London B 324:233-245. Graves, J. D., N. K. Watkins, A. H. Fitter, D. Robinson, and C. Scrimgeour. 1997. Interspecific transfer of carbon between plants linked by a common mycorrhizal network. Plant and Soil 192:153-159. Greenlee, J. T., and R. M. Callaway. 1996. Abiotic stress and the relative importance of interference and facilitation in montane bunchgrass communities in western Montana. American Naturalist 148:386-396. Griggs, R. F. 1956. Competition and succession on a Rocky Mountain fellfield. Ecology 37:8-20. Grime, J. P. 1977. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. American Naturalist 111:1169-1194. Grime, J. P. 1998. Benefits of plant diversity to ecosystems: immediate, filter and founder effects. Journal of Ecology 86:902-910. Grime, J. P., J. M. L. Mackey, S. H. Hillier, and D. J. Read. 1987. Floristic diversity in a model system using experimental microcosms. Nature 328:420-422. Groom, M. J. 1998. Allee effects limit population viability of an annual plant. American Naturalist 151:487-496. Grosse, W., H. B. Buchel, and H. Tiebel. 1991. Pressurized ventilation in wetland plants. Aquatic Botany 39:89-98. Grosvernier, P., Y. Matthey, and A. Buttler. 1995. Microclimate and physical properties of peat: new clues to the understanding of bog restoration processes. Pages 435-450 in B. D. Wheeler, S. C. Shaw, W. J. Fojt, and R. A. Robertson, editors. Restoration of Temperate Wetlands. John Wiley & Sons, Chichester, UK. Grover, H. D., and H. B. Musick. 1990. Shrubland encroachment in southern New Mexico, USA: an analysis of desertification processes in the American southwest. Climate Change 17:305-330. Guevara, S., J. Meave, P. Moreno-Casasola, and J. Laborde. 1992. Floristic composition and structure of vegetation under isolated trees in neotropical pastures. Journal of Vegetation Science 3:655-664. Guevara, S., E. Purata, and E. Van der Maarel. 1986. The role of remnant forest trees in tropical forest succession. Vegetatio 66:77-84.
364
References
Guo, Q., P. W. Rundel, and D. W. Goodall. 1998. Horizontal and vertical distribution of desert seed banks: patterns, causes, and implications. Journal of Arid Environments 38:465-478. Gutierrez, J. R., P. L. Meserve, L. C. Contreras, H. Vasquez, and F. M. Jaksic. 1993. Spatial distribution of soil nutrients and ephemeral plants underneath and outside the canopy of Porlieria chilensis shrubs (Zygophyllaceae) in arid coastal Chile. Oecologia 95:347-352. Haase, P., F. Pugnaire, S. C. Clark, and L. D. Incoll. 1996. Spatial patterns in a two-tiered semi-arid shrubland in southeastern Spain. Journal of Vegetative Science 7:527-534. Habeck, J. R. 1969. A gradient analysis of a timberline zone at Logan Pass, Glacier Park, Montana. Northwest Science 43:65-73. Hacker, S. D., and M. D. Bertness. 1999. Experimental evidence for factors maintaining plant species diversity in a New England salt marsh. Ecology 80:2064-2073. Hacker, S. D., and S. D. Bertness. 1995. Morphological and physiological consequences of a positive plant interaction. Ecology 76:2165-2175. Hadley, E. B., and L. C. Bliss. 1964. Energy relationships of alpine plants on Mt Washington, New Hampshire. Ecological Monographs 34:331-357. Hadley, J. L., and W. K. Smith. 1983. Influence of wind exposure on needle desiccation and mortality for timberline conifers in Wyoming, USA. Arctic and Alpine Research 15:127-135. Hadley, J. L., and W. K. Smith. 1986. Wind effects on needles of subalpine conifers: seasonal influence on mortality. Ecology 67:12-19. Hadley, J. L., and W. K. Smith. 1989. Wind erosion of leaf surface wax in alpine timberline conifers. Arctic and Alpine Research 21:392-398. Hale, M. E. 1952. Vertical distribution of cryptogams in a virgin forest in Wisconsin. Ecology 33:398-406. Hale, M. E. 1965. Vertical distribution of cryptogams in a red maple swamp in Connecticut. The Bryologist 68:193-197. Halvorson, W. L., and D. T. Patten. 1975. Productivity and flowering of winter ephemerals in relation to Sonoran Desert shrubs. American Midlands Naturalist 93:311-319. Hambäck, P. A., J. Agren, and L. Ericson. 2000. Associational resistance: insect damage to purple loosestrife reduced in thickets of sweet gale. Ecology 81:1784-1794. Hambäck, P. A., and A. P. Beckerman. 2003. Herbivory and plant resource competition: a mechanistic review of two interacting interactions. Oikos 101:26-37. Hambäck, P. A., J. Pettersson, and L. Ericson. 2003. Are associational refuges species-specific? Functional Ecology 17:87-93.
References
365
Harley, C. D. G., and M. D. Bertness. 1996. Structural interdependence: an ecological consequence of morphological responses to crowding in marsh plants. Functional Ecology 10:654-661. Harmon, M. E., and J. F. Franklin. 1989. Tree seedlings on logs in Picea– Tsuga forests of Oregon and Washington. Ecology 70:48-59. Harris, L. G., A. W. Ebeling, D. R. Laur, and R. J. Rowley. 1984. Community recovery after storm damage: a case of facilitation in primary succession. Science 224:1336-1338. Harrison, J. S., and P. A. Werner. 1984. Colonization by oak seedlings into a heterogeneous successional habitat. Canadian Journal of Botany 62:559-563. Hart, G. E., and D. R. Parent. 1974. Chemistry of throughfall under Douglas fir and Rocky Mountain juniper. American Midland Naturalist 92:191-201. Hartley, C. 1918. Stem lesions caused by excessive heat. Journal of Agricultural Research 14:595-604. Hasselquist, N., M. J. Germino, T. McGonigle, and W. K. Smith. 2005. Variability of Cenococcum colonization and its ecophysiological significance for young conifers at alpine-treeline. New Phytologist 165:867-873. Hastwell, G. T., and J. M. Facelli. 2003. Differing effects of shade-induced facilitation on growth and survival during the establishment of a chenopod shrub. Journal of Ecology 91:941-950. Hawksworth, F. G., and D. Wiens. 1996. Dwarf mistletoes: biology, pathology, and systematics. USDA Forest Service Agriculture Handbook 709. 410 p. Haworth, K., and G. R. McPherson. 1995. Effects of Quercus emoryi trees on precipitation distribution and microclimate in a semi-arid savanna. Journal of Arid Environments 31:153-170. Hay, M. E. 1986. Associational plant defenses and the maintenance of species diversity: turning competitors into accomplices. American Naturalist 128:617-641. Hazlett, D. L., and G. R. Hoffman. 1975. Plant species distributional patterns in Artemisia tridentata- and Artemisia cana-dominated vegetation in western North Dakota. Botanical Gazette 136:72-77. Heap, A. J., and E. I. Newman. 1980. An evaluation of techniques for measuring vesicular-arbuscular mycorrhizas. New Phytologist 85:169-171. Hector, A. et. al. 1999. Plant diversity and productivity experiments in Eurpoean grasslands. Science 286:1123-1127. Hector, A., E. Bazeley-White, M. Loreau, S. Otway, and B. Schmid. 2002. Overyielding in grassland communities: testing the sampling effect
366
References
hypothesis with replicated biodiversity experiments. Ecology Letters 5:502-511. Heilbronn, T. D., and W. H. Walton. 1984. Plant colonization of actively sorted stone stripes in the subantarctic. Arctic and Alpine Research 16:161-172. Heinrich, B. 1979. Resource heterogeneity and patterns of movement in foraging bumblebees. Oecologia 40:235-245. Hellmers, H., M. K. Genthe, and F. Ronco. 1970. Temperature affects growth and development of Englemann spruce. Forest Science 16:447-452. Henriquez, J. M., and C. H. Lusk. 2005. Facilitation of Nothofagus antarctica (Fagaceae) seedlings by the prostrate shrub Empetrum rubrum (Empetraceae) on glacial moraines in Patagonia. Austral Ecology 30:877-882. Herrera, C. M. 1984. Seed dispersal and fitness determinants in wild rose: combined effects of hawthorn, birds, mice, and browsing ungulates. Oecologia 63:36-393. Hertling, U. M., and R. A. Lubke. 1999. Indigenous and Ammophila arenaria-dominated dune vegetation on the South African Cape Coast. Applied Vegetation Science 2:157-168. Heschel, M. S., and K. N. Paige. 1995. Inbreeding depression, environmental stress, and population size variation in scarlet gilia (Ipomopsis aggregata). Conservation Biology 9:126-133. Hetrick, B. A. D., G. T. Wilson, and D. C. Hartnett. 1989. Relationship between mycorrhizal dependence and competitive ability of two tallgrass prairie grasses. Canadian Journal of Botany 67:2608-2615. Hetrick, B. A. D., G. W. T. Wilson, and T. C. Todd. 1990. Differential responses of C3 and C4 plants to mycorrhizal symbiosis, phosphorus fertilization, and soil microorganisms. Canadian Journal of Botany 68:461-467. Hewett, D. G. 1970. The colonization of sand dunes after stablization with marran grass (Ammophila arenaria). Journal of Ecology 58:653-658. Hicks, D. J. 1980. Intrastand distribution patterns of southern Appalachian cove forest herbaceous species. American Midlands Naturalist 104:209-223. Hicks, K. L., and J. O. Tahvanainen. 1974. Niche differentiation by Cruciferfeeding flea beetles (Coleoptera: Chrysomelidae). American Midland Naturalist 91:406-423. Hietz, P., and O. Briones. 1998. Correlation between water relations and within-canopy distribution of epiphytic ferns in a Mexican cloud forest. Oecologia 114:305-316.
References
367
Hill, M. O. 1979. DECORANA - A FORTRAN program for detrended correspondence analysis. Cornell University, Ithaca, New York, USA. Hillier, S. H. 1990. Gaps, seed banks, and plant species diversity in calcareous grasslands. Pages 57-66 in D. W. H. Walton and D. A. Wells, editors. Calcareous Grasslands: Ecology and Management. Bluntisham Books, Huntingdon, UK. Hils, M., and J. Vankat. 1982. Species removals from a first-year old-field plant community. Ecology 63:705-711. Hirrel, M. C., and J. W. Gerdemann. 1979. Enhanced carbon transfer between onions infected with a vesicular-arbuscular mycorrhizal fungus. New Phytologist 83. Hjältén, J., K. Danell, and P. Lundberg. 1993. Herbivore avoidance by association: vole and hare utilization of woody plants. Oikos 68: 125-131. Hjältén, J., and P. W. Price. 1997. Can plants gain protection from herbivory herbivory by association with unpalatable neighbors? A field experiment in a willow-sawfly system. Oikos 78:317-322. Hobbie, S. E. 1992. Effects of plant species on nutrient cycling. Trends in Ecology and Evolution 7:336-339. Hobbie, S. E. 1996. Temperature and plant species control over litter decomposition in Alaskan tundra. Ecological Monographs 66: 503-522. Hoffman, W. A. 1996. The effects of fire and cover on seedling establishment in a neotropical savanna. Journal of Ecology 84:383-393. Hoffman, W. A. 2000. Post-establishment seedlings success in the Brazilian Cerrado: a comparison of savanna and forest species. Biotropica 32:62-69. Høgh-Jensen, H., and J. K. Schjoerring. 1997. Interactions between white clover and ryegrass under contrasting nitrogen availability: N2 fixation, N fertilizer recovery, N transfer and water use efficiency. Plant and Soil 197:187-199. Holbrook, N. M., and F. E. Putz. 1989. The influence of neighbors on tree form: effects of lateral shade and prevention of sway on the allometry of Liquidambar styraciflua (sweetgum). American Journal of Botany 76:40-49. Holland, V. L. 1973. A study of soil and vegetation under Quercus douglasii compared to open grassland. Dissertation. University of California, Berkeley, California, USA. Holland, V. L. 1980b. Effect of blue oak on rangeland forage production in central California. Pages 314-318 in T. R. Plumb, editor. USDA General Technical Report, Pacific Southwest Station-44, No. 319. Berkeley, California, USA.
368
References
Holland, V. L., and J. Morton. 1980a. Effect of blue oak on nutritional quality of rangeland forage in central California. Pages 319-322 in T. R. Plumb, editor. USDA General Technical Report, Pacific Southwest Station-44, No. 319, Berkeley, California, USA. Holm, S. O. 1994. Pollination density - effects on pollen germination and pollen tube growth in Betula pubescens Ehrh. in northern Sweden. New Phytologist 126:541-547. Holmes, R. D., and K. Jepson-Innes. 1989. A neighborhood analysis of herbivory in Bouteloua gracilis. Ecology 70:971-976. Holmgren, M. 2000. Combined effects of shade and drought on tulip poplar seedlings: trade-off in tolerance or facilitation? Oikos 90:67-78. Holmgren, M., M. Scheffer, and M. A. Huston. 1997. The interplay of facilitation and competition in plant communities. Ecology 78: 1966-1975. Holmgren, M., A. M. Segura, and E. R. Fuentes. 2000. Limiting mechanisms in the regeneration of the Chilean matorral. Plant Ecology 147:49-57. Holt, R. D. 1977. Predation, apparent competition, and the structure of prey communities. Theoretical Population Biology 12:197-229. Holt, R. D., and J. H. Lawton. 1994. The ecological consequences of shared natural enemies. Annual Review of Ecology and Systematics 25: 495-520. Holtmeier, F., and G. Broll. 1992. The influence of tree islands and microtopography on pedoecological conditions in the forest-alpine tundra ecotone on Niwot Ridge, Colorado Front Range, U.S.A. Arctic and Alpine Research 24:216-228. Holzapfel, C., and B. E. Mahall. 1999. Bi-directional facilitation and interference between shrubs and associated annuals in the Mojave Desert. Ecology 80:1747-1761. Holzapfel, C., K. Tielbörger, H. A. Paragb, J. Kigel, and M. Sternberga. 2006. Annual plant-shrub interactions along an aridity gradient. Basic and Applied Ecology 7:268-279. Hooper, D. U., F. S. Chapin, J. J. Ewel, A. Hector, P. Inchausti, S. Lavorel, J. H. Lawton, D. Lodge, M. Loreau, S. Naeem, B. Schmid, H. Setäl, A. J. Symstad, J. Vandermeer, and D. A. Wardle. 2005. Effects of biodiversity on ecosystem functioning: a consensus of current knowledge and needs for future research. Ecological Monographs 75:3-35. Hooper, D. U., and P. M. Vitousek. 1997. The effects of plant composition and diversity on ecosystem processes. Science 277:1302-1305. Horn, H. S. 1975. Markovian properties of forest succession. Pages 196-211 in M. L. Cody and J. M. Diamond, editors. Ecology and Evolution of
References
369
Communities. Harvard University Press, Cambridge, Massachusetts, USA. Horton, T. R., T. Bruns, and V. T. Parker. 1999. Ectomycorrhizal fungi in Arctostaphylos patches contribute to the establishment of Pseudotsuga menziesii. Canadian Journal of Botany 77:93-102. Horton, J. L., and S. C. Hart. 1998. Hydraulic lift: a potentially important ecosystem process. Trends in Ecology and Evolution 13:232-235. Houle, G. 1996. No evidence of interspecific interactions between plants in the first stage of succession on coastal dunes in subarctic Quebec, Canada. Canadian Journal of Botany 75:902-915. Houle, G., and D. L. Phillips. 1989. Seed availability and biotic interactions in granite outcrop plant communities. Ecology 70:1307-1316. Howes, B. L., J. W. H. Dacey, and D. D. Goehringer. 1986. Factors controlling the growth form of Spartina alterniflora: feedbacks between aboveground production, sediment oxidation, nitrogen and salinity. Journal of Ecology 74:881-898. Howes, B. L., R. W. Howarth, J. M. Teal, and I. Valiela. 1981. Oxidationreduction potentials in a salt marsh: spatial patterns and interactions with primary production. Limnology and Oceanography 26:350-360. Hsiao, T. S., and F. G. 1968. The role of secondary plant substances in the food specificity of the Colorado potato beetle. Annals Entomological Society America 61:485-492. Huang, X., M. J. Singer, and R.A. Dahlgren. 1997. Oak tree and grazing impacts on soil properties and nutrients in a California oak woodland. Biogeochemistry 39:45-64. Hubbell, S. P. 2001. The Unified Neutral Theory of Biodiversity and Biogeography. Princeton University Press, Princeton, New Jersey USA. Huckle, J. M., J. A. Potter, and R. H. Harris. 2000. Influence of environmental factors on the growth and interactions between salt marsh plants: effects of salinity, sediment and waterlogging. Journal of Ecology 88:492-505. Huffaker, C. B., and C. E. Kennett. 1959. A ten year study of vegetational changes associated with biological control of Klamath weed. Journal of Rangeland Management 12:69-82. Huisman, J., and H. Olff. 1998. Competition and facilitation in multispecies plant-herbivore systems of productive environments. Ecology Letters 1:25-29. Hull, J. C., and C. H. Muller. 1976. Responses of California annual grassland species to variations in moisture and fertiliaztion. Journal of Range Management 29:49-52.
370
References
Humberto, S., G. P. Nabhan, and D. T. Patten. 1996. The importance of Olneya tesota as a nurse plant in the Sonoran Desert. Journal of Vegetation Science 7:635-644. Hunter, A. F., and L. W. Aarssen. 1988. Plants helping plants. Bioscience 38:34-40. Huntley, B. 1991. How plants respond to climate change: migration rates, individualism and the consequences for the plant communities. Journal of Botany 67:15-22. Huntley, L. B., D. Doley, D.J. Yayes and A. Boonsaner. 1997. Water balance of an Australian subtropical rainforest at altitude: the ecological and physiological significance of intercepted cloud and fog. Australian Journal Botany 45:311-329. Huston, M. A., and T. Smith. 1987. Plant succession: life history and competition. American Naturalist 130:168-198. Hutto, R. L., J. R. McAuliffe, and L. Hogan. 1986. Distributional associates of the saguaro (Carnegiea gigantea). Southwestern Naturalist 31: 469-476. Ibáñez, I., and E. W. Schupp. 2001. Positive and negative interactions between environmental conditions affecting Cercocarpus ledifolius seedling survival. Oecologia 129:543-550. Ibáñez, I., and E. W. Schupp. 2002. Effects of litter, soil surface conditions, and microhabitat on Cercocarpus ledifolius Nutt. seedling emergence and establishment. Journal of Arid Environments 52:209-221. Ikeda, H., and K. Okutomi. 1992. Effects of species interactions on community organization along a trampling gradient. Journal of Vegetation Science 3:217-222. Ineson, P., and K. B. McTiernan. 1992. Decomposition of foliar litter mixtures: a microcosm experiment. Pages 703-706 in A. Teller, P. Mathy, and J. N. R. Jeffers, editors. Responses of Forest Ecosystems to Environmental Changes. Elsevier Applied Science, Barking, UK. Ingerpuu, N., J. Liira, and M. Pärtel. 2005. Vascular plants facilitated bryophytes in a grassland experiment. Plant Ecology 180:69-75. Ingwerson, J. B. 1985. Fog drip, water yield, and timber harvesting in the Bull Run municipal watershed, Oregon. Water Resources Bulletin 21: 469-473. Inouye, B., and S. J.R. 2001. Relationships between ecological interaction modifications and diffuse coevolution: similarities, differences, and causal links. Oikos 95:353-360. Ishikawa, C. M., and C. S. Bledsoe. 2000. Seasonal and diurnal patterns of soil water potential in the rhizosphere of blue oaks: evidence for hydraulic lift. Oecologia 125:459-465.
References
371
Izhaki, I., P. B. Walton, and U. Safriel. 1991. Seed shadows generated by frugivorous birds in an eastern mediterranean scrub. Journal of Ecology 79:575-590. Jackson, J., and A. J. Ash. 1998. Tree-grass relationships in open eucalypt woodlands of northeastern Australia - influence of trees on pasture production, forage quality and species diversity. Agroforestry Systems 40:159-176. Jackson, J., and A. J. Ash. 2001. The role of trees in enhancing soil nutrient availability for native perennial grasses in open Eucalypt woodlands of north-east Queensland. Australian Journal of Agricultural Research 52:377-386. Jackson, J. T., and O. W. Van Auken. 1997. Seedling survival, growth and mortality of Juniperus ashei (Cupressaceae) in the Edwards Plateau region of central Texas. Texas Journal of Science 49:267-278. Jackson, L. E., R. B. Strauss, M. K. Firestone, and J. W. Bartolome. 1990. Influences of tree canopies on grassland productivity and nitrogen dynamics in deciduous oak savanna. Agricultural Ecosystems and Environments 32:89-105. Jackson, R. B., and M. M. Caldwell. 1993. Geostatistical patterns of soil heterogeneity around individual perennial plants. Journal of Ecology 81:683-692. Jacquez, G. M., and D. T. Patten. 1996. Chesnya nubigena on a Himalayan glacial moraine: a case of facilitation in primary succession. Mountain Research and Development 16:265-273. Jaksic, F. M., and E. R. Fuentes. 1980. Why are native herbs in the Chilean matorral more abundant beneath bushes: microclimate or grazing? Journal of Ecology 68:665-669. Jennerston, O. 1988. Pollination in Dianthus deltoides (Caryophyllaceae): effects of habitat fragmentation on visitation and seed set. Conservation Biology 2:359-356. Jernakoff, P., and J. Nielsen. 1988. Plant-animal associations in two species of seagrasses in western Australia. Aquatic Botany 60:359-376. Jespersen, D. N., B. K. Sorrell, and H. Brix. 1998. Growth and root oxygen release by Typha latifolia and its effects on sediment methanogenesis. Aquatic Botany 61:165-180. Joffre, R., and S. Rambal. 1988. Soil water improvement by trees in the rangelands of southern Spain. Oecologia Plantarum 9:405-422. Joffre, R., and S. Rambal. 1993. How tree cover influences the water balance of Mediterranean rangelands. Ecology 74:570-582. Johansson, R. 1974. Ecology of vascular epiphytes in West African rain forest. Acta Phytogeographica Suecica 59:1-29.
372
References
Johnson, D. M., M. J. Germino, and W. K. Smith. 2004. Abiotic factors limiting photosynthesis in Abies lasiocarpa and Picea engelmannii seedlings below and above alpine timberline. Tree Physiology 24: 377-387. Johnson, H. B., and H. S. Mayeux. 1992. Viewpoint: a view on species additions and deletions and the balance of nature. Journal of Range Management 45:322-333. Johnson, N. C., J. H. Graham, and F. A. Smith. 1997. Functioning of mycorrhizal associations along the mutualism-parasitism continuum. New Phytologist 135:575-585. Johnson, S. D., C. I. Peter, L. A. Nilsson, and J. Agren. 2003. Pollination success in a deceptive orchid is enhanced by co-occurring rewarding magnet plants. Ecology 84:2919-2927. Johnson, W., C. M. McKell, R. A. Evans, and L. J. Berry. 1959. Yield and quality of annual range forage following 2, 4-D application on blue oak trees. Journal of Range Management 12:18-20. Johnson, W. B., C. E. Sasser, and J. G. Gosselink. 1985. Succession of vegetation in an evolving river delta, Atchafalaya Bay, Louisiana. Journal of Ecology 73:973-986. Jolliffe, P. A. 1997. Are mixed populations of plant species more productive than pure stands? Oikos 80:1-8. Jonasson, S. 1992. Plant responses to fertilization and species removal in tundra related to community structure and clonality. Oikos 63: 420-429. Jones, C. G., J. H. Lawton, and M. Shachak. 1994. Organisms as ecosystem engineers. Oikos 69:373-386. Jones, C. G., J. H. Lawton, and M. Shachak. 1997. Positive and negative effects of organisms as physical ecosystem engineers. Ecology 78:1946-1957. Jones, M. B. 1963. Yield, percent nitrogen, and total nitrogen uptake of various California annual grassland species fertilized with increasing rates of nitrogen. Agronomy Journal 55:254-257. Jordan, D. N., and W. K. Smith. 1994. Energy balance analysis of night-time leaf temperatures and frost formation in a subalpine environment. Agricultural and Forest Meteorology 77:359-372. Jordan, D. N., and W. K. Smith. 1995. Radiation frost and the relationship between sky exposure and leaf size. Oecologia 103:43-48. Jordan, R. A., and J. M. Hartman. 1995. Safe sites and the regeneration of Clethra alnifolia L. (Clethraceae) in wetland forests of central New Jersey. American Midland Naturalist 133:112-123.
References
373
Jordano, P., and E. W. Schupp. 2000. Seed disperser effectiveness: the quantity component and patterns of seed rain for Prunus mahaleb. Ecological Monographs 70:591-615. Joy, D. A., and D. R. Young. 2002. Promotion of mid-successional seedling recruitment and establishment by Juniperus virginiana in a coastal environment. Plant Ecology 160:125-135. Jumpponen, A., K. Mattson, J. M. Trappe, and R. Ohtonen. 1998. Effects of established willows on primary succession on Lyman Glacier forefront, North Cascade Range, Washington, U.S.A.: evidence for simultaneous canopy inhibition and soil facilitation. Arctic and Alpine Research 30:31-39. Jurena, P. N., and O. W. Van Auken. 1998. Woody plant recruitment under canopies of two Acacias in a southwestern Texas grassland. Southwestern Naturalist 43:195-203. Karban, R., I. T. Baldwin, K. J. Baxter, G. Laue, and G. W. Felton. 2000. Communication between plants: induced resistance in wild tobacco plants following clipping of neighboring sagebrush. Oecologia 125:66-71. Karlson, R. H., and J. B. C. Jackson. 1981. Competitive networks and community structure: a simulation study. Ecology 62:670-678. Kay, B. L. 1987. Long-term effects of blue oak removal on forage production, forage quality, soil and oak regeneration. Pages 351-357 in T. R. Plumb and N. H. Pillsbury, editors. USDA, General Technical Report, PSW-100, Berkeley, California, USA. Keeley, J. E. 1988. Population variation in root grafting and a hypothesis. Oecologia 52:364-366. Keeley, J. E. 1991. Seed germination and life history syndromes in the California chaparral. Botanical Review 57:81-116. Keeley, J. E. 1992. Recruitment of seedlings and vegetative sprouts in unburned chaparral. Ecology 73:1194-1208. Keeley, J. E., and C. J. Fotheringham. 1997. Trace gas emissions in smokeinduced seed germination. Science 276:1248-1251. Keeley, J. E., and C. J. Fotheringham. 1998a. Mechanism of smoke-induced seed germination in a post-fire germination annual. Journal of Ecology 86:27-36. Keeley, J. E., and C. J. Fotheringham. 1998b. Smoke-induced seed germination in Californian chaparral. Ecology 79:2320-2336. Keeley, S. C., and A. W. Johnson. 1977. A comparison of the pattern of herb and shrub growth in comparable sites in Chile and California. American Midland Naturalist 97:120-132.
374
References
Keith, D. A., and R. A. Bradstock. 1994. Fire and competition in Australian heath: a conceptual model and field investigations. Journal of Vegetation Science 5:347-354. Kellman, M. 1979. Soil enrichment by neotropical savanna trees. Journal of Ecology 67:565-577. Kellman, M. 1985. Forest seedling establishment in Neotropical savannas: transplant experiments with Xylopia frutescens and Calophyllum brasiliense. Journal of Biogeography 12:373-379. Kellman, M. 1989. Mineral nutrient dynamics during savanna-forest transformation in Central America. Pages 85-103 in J. Proctor, editor. Mineral Nutrients in Tropical Forest and Savanna Ecosystems. Blackwell Scientific Publications, Oxford, UK. Kellman, M., and A. Carty. 1986. Magnitude of nutrient influxes from atmospheric sources to a Central American Pinus caribaea woodland. Journal of Applied Ecology 23:211-226. Kellman, M., and M. Kading. 1992. Facilitation of tree seedling establishment in a sand dune succession. Journal of Vegetation Science 3:679-688. Kellman, M., and K. Miyanishi. 1982. Forest seedling establishment in Neotropical savannas: observations and experiments in the Mountain Pine Ridge savanna, Belize. Journal of Biogeography 9:193-206. Kennard, D. G., and B. H. Walker. 1973. 1973. Relationships between tree canopy cover and Panicum maximum in the vicinity of Fort Victory. Rhodesian Journal of Agricultural Research 11:145-153. Kennedy, P. G., A. D. Izzo, and T. D. Bruns. 2003. There is high potential for the formation of common mycorrhizal networks between understorey and canopy trees in a mixed evergreen forest. Journal of Ecology 91:1071-1080. Kernan, C., and N. Fowler. 1995. Differential substrate use by epiphytes in Corcovado National Park, Costa Rica: a source of guild structure. Journal of Vegetation Science 83:65-73. Kery, M., D. Matthies, and H. Spillman. 2000. Reduced fecundity and offspring performance in small populations of the declining grassland plants Primula veris and Gentiana lutea. Journal of Ecology 88:17-30. Kikvidze, Z. 1996. Neighbour interaction and stability in subalpine meadow communities. Journal of Vegetation Science 7:41-44. Kikvidze, Z., L. Khetsuriani, D. Kikodze, and R. M. Callaway. 2006. Seasonal shifts in competition and facilitation in subalpine plant communities of the central Caucasus. Journal of Vegetation Science 17:77-82. Kikvidze, Z., F.I. Pugnaire, R.W. Brooker, P. Choler, C.J. Lortie, R. Michalet and R.M. Callaway. 2005. Linking patterns and processes in alpine plant communities: a global study. Ecology 86:1395-1400.
References
375
Kitzberger, T., D. F. Steinaker, and T. T. Veblen. 2000. Establishment of Austrocedrus chilensis in Patagonian forest-steppe ecotones: facilitation and climatic variability. Ecology 81:1914-1924. Klanderud, K. 2005. Climate change effects on species interactions in an alpine plant community. Journal of Ecology 93:127-137. Klanderud, K., and Ø. Totland. 2004. Habitat dependent nurse effects of the dwarf-shrub Dryas octopetala on alpine and arctic plant community structure. Ecoscience 11:410-420. Klanderud, K., and Ø. Totland. 2005. The relative importance of neighbours and abiotic environmental conditions for population dynamic parameters of two alpine plant species. Journal of Ecology 93:493-501. Kleb, H. R., and S. D. Wilson. 1997. Vegetation effects on soil resource heterogeneity in prairie and forest. American Naturalist 150:283-298. Klemmedson, J. O., and A. R. Tiedemann. 1986. Long-term effects of mesquite removal on soil characteristics: II. nutrient availability. Soil Science Society America Journal 50:476-480. Klinger, L. F. 1990. Global patterns in community succession: bryophytes and forest decline. Memoirs of the Torrey Botanical Club 24:1-50. Klironomos, J. 2002. Feedback with soil biota contributes to plant rarity and invasiveness in communities. Nature 417:67-70. Knops, J. M. H., D. Tilman, N. M. Haddad, S. Naeem, C. E. Mitchell, J. Haarstad, M. E. Ritchie, K. M. Howe, P. B. Reich, E. Siemann, and J. Groth. 1999. Effects of plant species richness on invasion dynamics, disease outbreaks, insect abundances, and diversity. Ecology Letters 2:286-293. Ko, L. J., and P. B. Reich. 1993. Oak tree effects on soil and herbaceous vegetation of savannas and pastures in Wisconsin. American Midland Naturalist 130:31-42. Kobe, R. K., S. W. Palcala, J. A. J. Silander, and C. D. Canham. 1995. Juvenile tree survivorship as a function of shade tolerance. Ecological Applications 5:527-532. Kochian, L. V., O. A. Hoekenga, and M. A. Pineros. 2004. How do crop plants tolerate acid soils? Mechanisms of aluminum tolerance and phosphorus efficiency. Annual Review of Plant Biology 55:459-493. Korner, C. 1998. A re-assessment of high elevation treeline positions and their explanation. Oecologia 115:445-459. Korner, C. 1999. Alpine plants: stressed or adapted? Pages 297-311 in M. C. Press, J. C. Scholes, and M. G. Barker, editors. Physiological Plant Ecology. Blackwell Science Ltd., Oxford, UK. Kosiba, P., and J. Sarosiek. 1993. Growth as a function of density of population of some selected species of mosses. Cryptogamic Botany 3:117-122.
376
References
Kost, C., and M. Heil. 2006. Herbivore-induced plant volatiles induce an indirect defence in neighbouring plants. Journal of Ecology 94: 619-628. Krannitz, P. G., and M. M. Caldwell. 1995. Root growth responses of three Great basin perennials to intraspecific and interspecific contact with other roots. Flora 190:161-167. Krause, G. H. 1994. Photoinhibition induced by low temperatures. Pages 331-348 in N. R. Baker and J. R. Bowyer, editors. Photoinhibition of Photosynthesis: Molecular Mechanisms to the Field. Bios Scientific Publishers, Oxford, UK. Krebs, C. J. 2001. Ecology. The experimental analysis of distribution and abundance. Benjamin Cummings, New York, USA. Kruess, A., and T. Tscharntke. 2000. Species richness and parasitism in a fragmented landscape: experiments and field studies with insects on Vicia sepium. Oecologia 122:129-137. Kunstler, G., J. Chadoeuf, E. K. Klein, T. Curt, M. Bouchaud, and J. Lepart. In press. Tree colonization of sub-Mediterranean grasslands: Effects of dispersal limitation and shrub facilitation. Canadian Journal of Forest Research. Kunstler, G., C. Thomas, M. Bouchaud, and J. Lepart. 2006. Indirect facilitation and competition in tree species colonization of subMediterranean grasslands. Journal of Vegetation Science 17:379-388. Laakso, J., V. Kaitala, and E. Ranta. 2001. How does environmental variation translate into biological processes? Oikos 92:119-122. Laland, K. N., F. J. Odling-Smee, and M. W. Feldman. 1996. On the evolutionary consequences of niche construction. Journal Evolutionary Biology 9:293-316. Laland, K. N., F. J. Odling-Smee, and M. W. Feldman. 1999. Evolutionary consequences of niche construction and their implications for ecology. Proceedings of the National Academy of Sciences 96: 10242-10247. Lamont, B. B., P. G. L. Klinkhamer, and E. T. F. Witkowski. 1993. Population fragmentation may reduce fertility to zero in Banksia goodi - a demonstration of the Allee effect. Oecologia 94:446-450. Langley, C. M. 1996. Search images: selective attention to specific visual features of prey. Journal of Experimental Psychology: Animal Behavior 22:152-163. Larcher, W. 1995. Physiological Plant Ecology, 3rd edition. Springer-Verlag, New York, New York, USA. Laverty, T. M. 1992. Plant interactions for pollinator visits: a test of the magnet species effect. Oecologia 89:502-508.
References
377
Laverty, T. M., and R. C. Plowright. 1988. Fruit and seed set in mayapple (Podophyllum peltatum): influence of intraspecific factors and local enhancement near Pedicularis canadensis. Canadian Journal of Botany 66:173-178. Lawlor, L. R. 1979. Direct and indirect effects of n-species competition. Oecologia 45:355-364. Lawrence, D. B., R. E. Schoenike, A. Quispel, and G. E. Bonds. 1967. The role of Dryas drummondii in vegetation development following ice recession at Glacier Bay, Alaska, with special reference to its nitrogen fixation by root nodules. Journal of Ecology 55:793-813. Leake, J. R. 1994. The biology of myco-heterotrophic (‘saprophytic’) plants. New Phytologist 127:171-216. Lee, J., R. S. Oliveira, T. E. Dawson, and I. Fung. 2005. Root functioning modifies seasonal climate. PNAS 102:17576-17581. Lee, T. D., P. B. Reich, and M. G. Tjoelker. 2003. Legume presence increases photosynthesis and N concentrations of co-occurring non-fixers but does not modulate their responsiveness to carbon dioxide enrichment. Oecologia 137:22-31. Lenz, T. I., and J. M. Facelli. 2003. Shade facilitates an invasive stem succulent in a chenopod shrubland in South Australia. Austral Ecology 28:480-490. Leonard, O. A. 1956. Effect on blue oak (Quercus douglasii) of 2, 4-D and 2, 4, 5-T concentrates applied to cut trunks. Journal of Range Management 9:15-19. Lesica, P., and R. K. Antibus. 1991. Canopy soils and epiphyte richness. National Geographic Research & Exploration 7:156-165. Levin, S. A. 1970. Community equilibria and stability, and an extension of the competitive exclusion principle. American Naturalist 104:413-423. Levine, J. M. 1999. Indirect facilitation: evidence and predictions from a riparian community. Ecology 80:1762-1769. Levine, J. M. 2000. Complex interactions in a streamside plant community. Ecology 81:3431-3444. Levine, J. M., S. D. Hacker, H. D. G. Christopher, and M. D. Bertness. 1998. Nitrogen effects on an interaction chain in a salt marsh community. Oecologia 117:266-272. Levine, S. H. 1976. Competitive interactions in ecosystems. American Naturalist 110:903-910. Levitt, J. 1972. Responses of plants to environmental stresses. Academic Press, New York, New York, USA. Li, L., C. Tang, Z. Rengel, and F. Zhang. 2003. Chickpea facilitates phosphorus uptake by intercropped wheat from an organic phosphorus source. Plant and Soil 248:297-303.
378
References
Li, X., and S. D. Wilson. 1998. Facilitation among woody plants establishing in an old field. Ecology 79:2694-2705. Lichter, J. 2000. Colonization constraints during primary succession on coastal Lake Michigan sand dunes. Journal of Ecology 88:825-839. Linhart, Y. B. 1976. Density-dependent seed germination strategy in colonizing versus non-colonizing species. Journal of Ecology 64:375-380. Lippmaa, T. 1939. The unistratal concept of plant communities (the unions). American Midlands Naturalist 21:111-145. Lloret, F., J. Peñuelas, and M. Estiarte. 2005. Effects of vegetation canopy and climate on seedling establishment in Mediterranean shrubland. Journal of Vegetation Science 16:67-76. Lodhi, M. A. K. 1977. The influence and comparison of individual forest trees on soil properties and possible inhibition of nitrification due to intact vegetation. American Journal of Botany 64:260-264. Longpre, M. H., Y. Bergeron, D. Pare, and M. Beland. 1994. Effect of companion species on the growth of jack pine (Pinus banksiana). Canadian Journal of Forest Research 24:1846-1853. Lookingbill, T. R., and M. A. Zavala. 2000. Spatial pattern of Quercus ilex and Quercus pubescens recruitment in Pinus halepensis dominated woodlands. Journal of Vegetation Science 11:607-612. López-Pintor, A., T. Espigares, and J. M. Rey Benayas. 2003. Spatial segregation of plant species caused by Retama sphaerocarpa influence in a Mediterranean pasture: a perspective from the soil seed bank. Plant Ecology 167:107-116. Loreau, M., and A. Hector. 2001. Partitioning selection and complementarity in biodiversity experiments. Nature 412:72-76. Loreau, M., S. Naeem, P. Inchausti, J. Bengtsson, J. P. Grime, A. Hector, D. U. Hooper, M. A. Huston, D. Raffaelli, B. Schmid, D. Tilman, and D. A. Wardle. 2001. Biodiversity and ecosystem functioning: current knowledge and future challenges. Science 294:804-808. Lortie, C. J., R. W. Brooker, P. Choler, Z. Kikvidze, R. Michalet, F. I. Pugnaire, and R. M. Callaway. 2004. Rethinking plant community theory. Oikos 107:433-438. Lortie, C. J., R. W. Brooker, Z. Kikvidze, and R. M. Callaway. 2004. The value of stress and limitation in an imperfect world. Journal of Vegetation Science 15:577-580. Lortie, C. J., and R. M. Callaway. 2006. Meta-analysis and premature rejection of the stress-gradient hypothesis: a reply to Maestre et al. Journal of Ecology 94:7-16.
References
379
Lortie, C. J., E. Ellis, A. Novoplansky, and R. Turkington. 2005. Implications of spatial pattern and local density on community-level interactions. Oikos 109:495-502. Lortie, C. J., D. T. Ganey, and B. P. Kotler. 2000. The effects of gerbil foraging on the natural seedbank and consequences on the annual plant community. Oikos 90:39-407. Lortie, C. J., and R. Turkington. 2002. The facilitative effects of seeds and seedlings on emergence from the seed bank of a desert annual plant community. Ecoscience 9:106-111. Lovett, G. M., W. A. Reiners, and R. K. Olson. 1982. Cloud droplet deposition in subalpine balsam fir forests: hydrological and chemical inputs. Science 218:1303-1304. Lovett, G. M., and G. M. Rueth. 1999. Soil nitrogen transformations in beech and maple stands along a nitrogen deposition gradient. Ecological Applications 9:1330-1344. Ludwig, F., H. de Kroon, F. Berendse, and H. H. T. Prins. 2004. The influence of savanna trees on nutrient, water and light availability and the understorey vegetation. Plant Ecology 170:93-105. Macior, L. W. 1971. Co-evolution of plants and animals: systemic insights from plant-insect interactions. Taxon 20:17-28. Madgwick, H. A. I., and J. D. Ovington. 1959. The chemical composition of precipitation in adjacent forest and open plots. Forestry (Oxford):14-22. Maestre, F. T. 2002. La restauración de la cubierta vegetal en zonas semiáridas en función del patrón espacial de factores bióticos y abióticos. PhD. Thesis. Fundación Biblioteca Virtual Miguel de Cervantes, Alicante, Spain. Maestre, F. T., S. Bautista, and J. Cortina. 2003b. Positive, negative, and net effects in grass-shrub interactions in Mediterranean semiarid grasslands. Ecology 84:3186-3197. Maestre, F. T., S. Bautista, and J. Cortina. 2003a. Positive, negative and net effects in grass-shrub interactions in semiarid Mediterranean steppes. Ecology 84:3186-3197. Maestre, F. T., S. Bautista, J. Cortina, and J. Bellot. 2001. Potential of using facilitation by grasses to establish shrubs on a semiarid degraded steppe. Ecological Applications 11:1641-1655. Maestre, F. T., and J. Cortina. 2004. Do positive interactions increase with abiotic stress? A test from a semi-arid steppe. Proceedings of the Royal Society of London B Supplement 271:S231-333. Maestre, F. T., J. Cortina, and S. Bautista. 2004. Mechanisms underlying the interaction between Pinus halepensis and the native late-
380
References
successional shrub Pistacia lentiscus in a semi-arid plantation. Ecography 27:776-786. Maestre, F. T., F. Valladares, and J. F. Reynolds. 2005. Is the change of plantplant interactions with abiotic stress predictable? A meta-analysis of field results in arid environments. Journal of Ecology 93:748-757. Magee, T. K., and J. A. Antos. 1992. Tree invasion into a mountain-top meadow in the Oregon Coast Range, USA. Journal of Vegetation Science 3:485-494. Magistad, O. C., and J. F. Breazeale. 1929. Plant and soil water relations at and below the wilting percentage. Arizona Agricultural Experiment Station Technical Bulletin 25. Maguire, D. A., and R. T. T. Forman. 1983. Herb cover effects on tree seedling patterns in a mature hemlock-hardwood forest. Ecology 64:1367-1380. Mahall, B. E., and R. M. Callaway. 1992. Root communication mechanisms and intracommunity distributions of two Mojave Desert shrubs. Ecology 73:2145-2151. Mahall, B. E., and R. M. Callaway. 1996. Geographic and genetic contributions to root communication in Ambrosia dumosa. American Journal of Botany 83:93-88. Mahall, B. E., V. T. Parker, and P. J. Fonteyn. 1981. Growth and photosynthetic responses of Avena fatua L. and Bromus diandrus Roth. and their ecological significance in California savannas. Photosynthetica 15:5-15. Maher, E. L., M. J. Germino, and N. J. Hasselquist. 2005. Interactive effects of tree and herb cover on survivorship, physiology, and microclimate of conifer seedlings at the alpine tree-line ecotone. Canadian Journal of Forest Research 35:567-574. Maillette, L. 1988. Apparent commensalism among three Vaccinium species on a climatic gradient. Journal of Ecology 76:877-888. Mallik, A. U., and F. Pellissier. 2000. Effects of Vaccinium myrtillus on spruce regeneration: testing the notion of coevolutionary significance of allelopathy. Journal of Chemical Ecology 26:2197-2209. Man, R., and V. J. Lieffers. 1998. Seasonal photosynthetic responses to light and temperature in white spruce (Picea glauca) seedlings planted under an aspen (Populus tremuloides) canopy and in the open. Tree Physiology 17:437-444. Maranga, E. K. 1984. Influence of Acacia tortilis trees on the distribution of Panicum maximum and Digitaria macroblephara in south central Kenya. M.S. Thesis. Texas A & M University, College Station, Texas, USA.
References
381
Maranon, T., and J. W. Bartolome. 1993. Reciprocal transplants of herbaceous communities between Quercus agrifolia woodland and adjacent grassland. Journal of Ecology 81:673-682. Marchand, P. J. 1972. Wind and the winter-exposed plant. Rhodora 74: 528-531. Marchand, P. J., and B. F. Chabot. 1978. Winter water-relations of treeline plant species on Mt. Washington, New Hampshire. Arctic and Alpine Research 17:1-5. Marcuvitz, S., and R. Turkington. 2000. Differential effects of light quality provided by different grass neighbours, on the growth and morphology of Trifolium repens L. (white clover). Oecologia 125:293-300. Marler, M. J., C. A. Zabinski, and R. M. Callaway. 1999a. Mycorrhizae indirectly enhance competitive effects of an invasive forb on a native bunchgrass. Ecology 80:1180-1186. Marler, M. J., C. A. Zabinski, T. Wojtowicz, and R. M. Callaway. 1999b. Mycorrhizae and fine root dynamics of Centaurea maculosa and native bunchgrasses in western Montana. Northwest Science 73:217-224. Maron, J. L., and P. G. Connors. 1996. A native nitrogen-fixing shrub facilitates weed invasion. Oecologia 105:302-312. Maron, J. L., and R. L. Jefferies. 1999. Bush lupine mortality, altered resource availability, and alternative vegetation states. Ecology 80:443-454. Marr, J. W. 1977. The development and movement of tree islands near the upper limit of tree growth in the southern Rocky Mountains. Ecology 58:1159-1164. Martens, S. N., D. D. Breshears, C. W. Meyer, and F. J. Barnes. 1997. Scales of above-ground and below-ground competition in a semi-arid woodland detected from spatial pattern. Journal of Vegetation Science 8:655-644. Martin, C. E. 1994. Physiological ecology of the Bromeliaceae. Botanical Review 60:2-82. Martin, C. E., M. K.W., C. A. Eades, and A. F. Pitzer. 1985. Morphological and physiological responses to irradiance in the CAM epiphyte Tillandsia usneoides L. (Bromeliaceae). Botanical Gazette 146: 489-494. Martin, C. E., and J. N. Siedow. 1981. Crassulacean Acid Metabolism in the epiphyte Tillandsia usneoides L. (Spanish moss). Plant Physiology 68:335-339. Martin, T. E. 2007. Climate correlates of 20 years of trophic changes in a high elevation riparian system. Ecology 88:366-372. Martínez, M. L. 2003. Facilitation of seedling establishment by an endemic shrub in tropical coastal sand dunes. Plant Ecology 168:333-345.
382
References
Martinez, M. L., O. Perez-Maqueo, and V. M. Vásquez. 2004. Facilitative interactions on coastal dunes in response to seasonal weather fluctuations and benefactor size. Ecoscience 11:390-398. Martínez, M. L., G. Vázquez, and S. Sánchez-Colón. 2001. Spatial and temporal dynamics during primary succession on tropical coastal sand dunes. Journal of Vegetation Science 12:361-372. Martinsen, G. D., E. M. Driebe, and T. M. Whitham. 1998. Indirect interactions mediated by changing plant chemistry: beaver browsing benefits beetles. Ecology 79:192-200. Maschinski, J., and T. G. Whitham. 1989. The continuum of plant responses to herbivory: the influence of plant association, nutrient availability, and timing. American Naturalist 134:1-19. Matthies, D. 2003. Positive and negative interactions among individuals of a root hemiparasite. Plant Biology 5:79-84. Matthies, D., B. Schmid, and P. Schmid-Hempel. 1995. The importance of population processes for the maintenance of biological diversity. Gaia 4:199-209. McAuliffe, J. R. 1984a. Sahuaro-nurse tree associations in the Sonoran Desert: competitive effects of the sahuaros. Oecologia 64:319-321. McAuliffe, J. R. 1984b. Prey refugia and the distributions of two Sonoran Desert cacti. Oecologia 65:82-85. McAuliffe, J. R. 1986. Herbivore-limited establishment of a Sonoran Desert tree: Cercidium microphyllum. Ecology 67:276-280. McAuliffe, J. R. 1988. Markovian dynamics of simple and complex desert plant communities. American Naturalist 131:459-490. McAuliffe, J. R. 1990. Paloverdes, pocket mice, and bruchid beetles: interrelationships of seeds, dispersers, and seed predators. Southwestern Naturalist 35:329-337. McAuliffe, J. R. 1991. Demographic shifts and plant succession along a late Holocene soil chronosequence in the Sonoran Desert of Baja California. Journal of Arid Environments 20:165-178. McClaran, M. P., and J. W. Bartolome. 1989. Effect of Quercus douglasii (Fagaceae) on herbaceous understory along a rainfall gradient. Madrono 36:141-153. McConn, M., R. A. Creelman, E. Bell, J. E. Mullet, and J. Browse. 1997. Jasmonate is essential for insect defense in Arabidopsis. Proceedings of the National Academy of Science 94:5473-5477. McCune, B., and J. A. Antos. 1981. Correlations between forest layers in the Swan Valley, Montana. Ecology 62:1196-1204.
References
383
McCune, B., and M. J. Medford. 1997. PC-ORD. Multivariate analysis of ecological data, Version 3.0. MjM Software Design, Gleneden Beach, Oregon, USA. McDrew, M. C. 1983. Plant injury and adaptation to oxygen deficiency in the root environment. Plant and Soil 75:179-199. McLeod, K. W., and P. G. Murphy. 1977. Establishment of Ptelea trifoliata on Lake Michigan sand dunes. American Midlands Naturalist 97:350362. McNaughton, S. J. 1978. Serengeti ungulates: feeding selectivity influences the effectiveness of plant defense guilds. Science 199:806-807. McNaughton, S. J. 1979. Grazing as an optimization process: grass-ungulate relationships in the Serengeti. American Naturalist 124:863-886. McNaughton, S. J. 1983. Serengeti grassland ecology: the role of composite environmental factors and contingency in community organization. Ecological Monographs 53:291-320. McPherson, G. R., H. A. Wright, and D. B. Wester. 1988. Patterns of shrub invasion in semiarid Texas grasslands. American Midlands Naturalist 120:391-397. Mealor, B. A., A. L. Hild, and N. L. Shaw. 2004. Native plant community composition and genetic diversity associated with long-term weed invasions. Western North American Naturalist 64:503-513. Meiners, S. J., and D. L. Gorchov. 1998. Effects of distance to Juniperus virginiana on the establishment of Fraxinus and Acer seedlings in old fields. American Midland Naturalist 139:353-364. Menchaca, L., and J. Connolly. 1990. Species interactions in white cloverryegrass mixtures. Journal of Ecology 78:223-232. Menges, E. S. 1991. Seed germination percentage increases with population size in a fragmented prairie species. Conservation Biology 5:158-164. Menges, E. S., and R. W. Dolan. 1998. Demographic viability of populations of Silene regia in midwestern prairies: relationships with fire management, genetic variation, geographic location, population size and isolation. Journal of Ecology 86:63-78. Messier, C., S. Parent, and Y. Bergeron. 1998. Effects of overstory and understory vegetation on the understory light environment in mixed boreal forests. Journal of Vegetation Science 9:511-520. Michalet, M., R. W. Brooker, L. A. Cavieres, Z. Kikvidze, C. J. Lortie, F. I. Pugnaire, A. Valiente-Banuet, and R. M. Callaway. 2006. Do biotic interactions shape both sides of the humped-back model of species richness in plant communities? Ecology Letters 9:767-773. Michalet, R., C. Rolland, D. Joud, D. Gafta, and R. M. Callaway. 2002. Spatial associations between canopy and understory species increase
384
References
along a rainshadow gradient in the Alps. Journal of Vegetation Science 165:145-160. Micheli, F., K. L. Cottingham, J. Bascompte, O. N. Bjørnstad, G. L. Eckert, J.M. Fischer, T. H. Keitt, B. E. Kendall, J. L. Klug and J. A. Rusak. 1999. The dual nature of communities. Oikos 85:161-169. Migenis, L. E., and J. D. Ackerman. 1993. Orchid-porophyte relationships in a forest watershed in Puerto Rico. Journal of Tropical Ecology 9: 231-240. Milbrandt, E. C., and M. N. Tinsley. 2006. The role of saltwort (Batis maritima L.) in regeneration of degraded mangrove forests. Hydrobiologia 568:369-377. Milchunas, D. G., and I. Noy-Meir. 2002. Grazing refuges, external avoidance of herbivory and plant diversity. Oikos 99:113-130. Miller, E. A., and C. B. Halpern. 1998. Effects of environment and grazing disturbance on tree establishment in meadows of the central Cascade Range, Oregon, USA. Journal of Vegetation Sciences 9:265-282. Miller, G. T. 1993. Environmental Science. Wadsworth, Belmont, California, USA. Miller, T. E. 1987. Effects of time on survival and growth in an early old-field community. Oecologia 72:272-278. Miller, T. E. 1994. Direct and indirect species interactions in an early oldfield plant community. American Naturalist 143:1007-1025. Miller, T. E., and P. A. Werner. 1987. Competitive effects and responses between plant species in a first-year old-field community. Ecology 68:1201-1210. Minnich, R. A. 1984. Snow drifting and timberline dynamics on Mt. San Gorgonio, California, USA. Arctic and Alpine Research 16:395-412. Miriti, M. N. 2006. Ontogenetic shift from facilitation to competition in a desert shrub. Journal of Ecology 94:973-979. Mitchell, G., and W. Arthur. 1998. Population interactions in primary succession: an example of contramensalism involving rockcolonizing bryophytes. Lindbergia 23:81-85. Moeller, D. A. 2004. Facilitative interactions among plants via shared pollinators. Ecology 85:3289-3301. Moen, J. 1993. Positive versus negative interactions in a high alpine block field: germination of Oxyria digyna seeds in a Ranunculus glacialis community. Arctic and Alpine Research 25:201-206. Monk, C. D., and F. C. Gabrielson. 1985. Effects of shade, litter and root competition on old-field vegetation in South Carolina. Bulletin Torrey Botanical Club 112:383-392.
References
385
Monteith, L. G. 1960. Influence of plants other than the food plants of their host on host-finding by tachinid parasites. The Canadian Entomologist 92:641-652. Mooney, H. A., S. L. Gulmon, P. W. Rundel, and J. Ehleringer. 1980. Further observations on the water relations of Prosopis tamarugo of the northern Atacama Desert. Oecologia 44:177-180. Moora, M., and M. Zobel. 1996. Effect of arbuscular mycorrhiza on inter- and intraspecific competition of two grassland species. Oecologia 108: 79-84. Moore, P. D. 1990. Vegetation’s place in history. Nature 347:710. Morgan, D. C., and H. Smith. 1979. A systematic relationship between phytochrome-controlled development and species habitat for plants grown in simulated natural radiation. Planta 142:187-193. Moro, M. J., F. I. Pugnaire, P. Hasse, and J. Puigdefabregas. 1997. Mechanisms of interaction between Retama sphaerocarpa and its understory layer in a semi-arid environment. Ecography 20:175-184. Morris, W. F., and D. M. Wood. 1989. The role of lupine in succession on Mount St. Helens: facilitation or inhibition? Ecology 70:697-703. Morrow, P. A., D. W. Tonkyn, and R. J. Goldburg. 1989. Patch colonization by Trirhabda canadensis (Coleoptera: Chrysomelidae): effects of plant species composition and wind. Oecologia 81:43-50. Mott, J. J., and A. J. McComb. 1974. Patterns in annual vegetation and soil microrelief in an arid region of western Australia. Journal of Ecology 62:115-126. Mulder, C. P. H., D. D. Uliassi, and D. F. Doak. 2001. Physical stress and diversity-productivity relationships: the role of positive interactions. Proceedings of the National Academy of Sciences 98:6704-6708. Muller, C. H. 1953. The association of desert annuals with shrubs. American Journal of Botany 40:53-60. Muñoz, A. A., P. Chacón, F. Pérez, E. S. Barnert, and J. J. Armesto. 2003. Diversity and host tree preferences of vascular epiphytes and vines in a temperate rainforest in southern Chile. Australian Journal of Botany 51:381-391. Murphy, A. H., and L. J. Berry. 1973. Range pasture benefits through tree removal. California Agriculture 27:8-10. Murphy, A. H., and B. Crampton. 1964. Quality and yield of forage as affected by chemical removal of blue oak (Quercus douglasii). Journal of Range Management 17:142-144. Murphy, S. D., and L. W. Aarssen. 1995. In vitro allelopathic effects of pollen from three Hieracium species (Asteraceae) and pollen transfer to sympatric Fabaceae. American Journal of Botany 82:37-45.
386
References
Murray, B. R. 1998. Density-dependent seed germination and the role of leachate. Australian Journal of Ecology 23:411-418. Muslin, E. H., and P. H. Homann. 1992. Light as a hazard for the desiccation resistant “resurrection” fern Polypodium polypodioides L. Plant, Cell and Environment 15:81-89. Nabhan, G. P. 1997. Why chiles are hot. Natural History 6:26-29. Nadkarni, N. M. 1986. The nutritional effects of epiphytes on host trees with special reference to alteration of precipitation chemistry. Selbyana 9:44-51. Naeem, S., K. Hakansson, J. H. Lawton, M. J. Crawley, and L. J. Thompson. 1996. Biodiversity and plant productivity in a model assemblage of plant species. Oikos 76:259-265. Naeem, S., L. J. Thompson, S. P. Lawler, J. H. Lawton, and R. M. Woodfin. 1994. Declining biodiversity can alter the performance of ecosystems. Nature 368:734-737. Nara, K. 2006a. Pioneer dwarf willow may facilitate tree succession by providing late colonizers with compatible ectomycorrhizal fungi in a primary successional volcanic desert. New Phytologist 171:187-198. Nara, K. 2006b. Ectomycorrhizal networks and seedling establishment during early primary succession. New Phytologist 169:169-176. Nara, K., and T. Hogetsu. 2004. Ectomycorrhizal fungi on established shrubs facilitate subsequent seedling establishment of successional plant species. Ecology 85:1700-1707. Newman, E. I., C. L. N. Devoy, N. J. Easen, and K. J. Fowles. 1994. Plant species that can be linked by VA mycorrhizal fungi. New Phytologist 126:691-693. Nicholson, M., and R. P. McIntosh. 2002. H. A. Gleason and the individualistic hypothesis revisited. Bulletin of the Ecological Society of America 83:133-142. Niering, W. A., R. H. Whittaker, and C. H. Lowe. 1963. The saguaro: a population in relation to environment. Science 142:15-23. Nobel, P. S. 1980a. Morphology, nurse plants, and minimum apical temperatures for young Carnegia gigantea. Botanical Gazette 144:188-191. Nobel, P. S. 1980b. Morphology, surface temperatures, and northern limits of columnar cacti in the Sonoran Desert. Ecology 61:1-7. Nobel, P. S. 1982. Low temperature tolerance and cold hardening of cacti. Ecology 63:1650-1656. Nobel, P. S. 1984a. High temperature responses of North American cacti. Ecology 65:643-651.
References
387
Nobel, P. S. 1984b. Extreme temperatures and thermal tolerances for seedlings of desert succulents. Oecologia 62:310-317. Noble, D. L., and R. R. Alexander. 1977. Environmental factors affecting natural regeneration of Englemann spruce in the central Rocky Mountains. Forest Science 23:420-429. Noble, I. R. 1980. Interactions between tussock grass (Poa spp.) and Eucalyptus pauciflora seedlings near treeline in south-eastern Australia. Oecologia 45:350-353. Noss, R. F., and A. Y. Cooperrider. 1994. Saving Nature’s Legacy: Protecting and Restoring Biodiversity. Island Press, Washington, D.C., USA. Nunez, C. L., M. A. Aizen, and C. Ezcurra. 1999. Species associations and nurse plant effect in patches of high-Andean vegetation. Journal of Vegetation Science 10:357-364. Nunez, M., and D. M. J. S. Bowman. 1986. Nocturnal cooling in a high altitude stand of Eucalyptus delegatensis as related to stand density. Australian Journal of Forest Research 16:185-197. O’Connell, L. M., and M. O. Johnston. 1998. Male and female pollination success in a deceptive orchid, a selective study. Ecology 79:1246-1260. O’Connor, T. G. 1996. Hierarchical control over seedling recruitment of the bunch-grass Themeda triandra in a semi-arid savanna. Journal of Applied Ecology 33:1095-1106. Odling-Smee, F. J., K. N. Laland, and M. W. Feldman. 1996. Niche construction. American Naturalist 147:641-648. Oesterheld, M., and M. Oyarzábal. 2004. Grass-to-grass protection from grazing in a semi-arid steppe. Facilitation, competition, and mass effect. Oikos 107:576-582. Olesen, J. M., and S. K. Jain. 1994. Fragmented plant populations and their lost interactions. Pages 417-426 in V. Loeschcke, J. Tomiuk, and S. K. Jain, editors. Conservation Genetics. Birkhauser, Basel, Switzerland. Olff, H., J. Huisman, and B. F. van Tooren. 1993. Species dynamics and nutrient accumulation during early primary succession in coastal sand dunes. Journal of Ecology 81:693-706. Olff, H., F. W. M. Vera, J. Bokdam, E. S. Bakker, J. M. Gleichman, K. de Maeyer, and R. Smit. 1999. Shifting mosaics in grazed woodlands driven by the alternation of facilitation and competition. Plant Biology 1:127-137. Oliveira, R. S., T. E. Dawson, S. S. O. Burgess, and D. C. Nepstad. 2005. Hydraulic redistribution in three Amazonian trees. Oecologia 145:354-363. Olofsson, J. 2004. Positive and negative plant-plant interactions in two contrasting arctic-alpine plant communities. Arctic, Antarctic, and Alpine Research 36:464–467.
388
References
Olofsson, J., J. Moen, and L. Oksanen. 1999. On the balance between positive and negative plant interactions in harsh environments. Oikos 86: 539-543. Onipchenko, V. G., and M. S. Blinnikov. 1994. Experimental investigation of alpine plant communities in the northwestern Caucasus. Veroffentlichungen des Geobotanishen, Zurich 115. Oostermeijer, J. G. B., M. W. van Eijck, and J. C. M. den Nijs. 1994. Offspring fitness in relation to population size and genetic variation in the rare perennial plant species Gentiana pneumonanthe (Gentianaceae). Oecologia 97:287-296. Ovalle, C., and J. Avendano. 1987. Interactions de la strate ligneuse avec la strate herbacee dans les formations d’Acacia caven (Mil.) Hook. et Arn. au Chili. I. Influence de l’arbe et la phenologie de la strate herbacee. Acta Oecologia 8:385-404. Ovington, J. D. 1955. Studies of the development of woodland conditions under different trees. Journal of Ecology 43:1-25. Owens, M. K., R. B. Wallace, and S. R. Archer. 1995. Landscape and microsite influences on shrub recruitment in a disturbed semi-arid Quercus-Juniperus woodland. Oikos 74:493-502. Pabst, R. J., and T. A. Spies. 1997. Distribution of herbs and shrubs in relation to landform and canopy cover in riparian forests of Coastal Oregon. Canadian Journal of Botany 76:298-315. Packer, A., and K. Clay. 2000. Soil pathogens and spatial patterns of seedling mortality in a temperate tree. Nature 404:278-281. Padilla, F. M., and F. I. Pugnaire. 2006. The role of nurse plants in the restoration of degraded environments. Frontiers in Ecology and the Environment 4:196-202. Page, J. P., and R. Michalet. 2003. A test of the indirect facilitation model in a temperate hardwood forest of the northern French Alps. Journal of Ecology 91:932-940. Pahlsson, L. 1974. Influence of vegetation on microclimate and soil moisture on a Scanian hill. Oikos 25:176-186. Palaniappan, V. M., R. H. Marrs, and A. D. Bradshaw. 1979. The effect of Lupinus arboreus on the nitrogen status of china clay wastes. Journal of Applied Ecology 16:825-830. Pandey, A. N., and M. V. Rokad. 1992. Sand dune stabilisation: an investigation in the Thar Desert of India. Journal of Arid Environments 22:287-292. Pandey, C. B., A. K. Singh, and D. K. Sharma. 2000. Soil properties under Acacia nilotica trees in a traditional agroforestry system in central India. Agroforestry Systems 49:53-61.
References
389
Parker, C. 1988. Environmental relationships and vegetation associates of columnar cacti in the northern Sonoran Desert. Vegetation 78:125-140. Parker, C. 1989. Nurse plant relationships of columnar cacti in Arizona. Physiographic Geography 10:322-355. Parker, M. A. 1982. Association with mature plants protects seedlings from predation in an arid grassland shrub, Gutierrezia microcephala. Oecologia 53:276-280. Parker, V. T., and C. H. Muller. 1982. Vegetational and environmental changes beneath isolated live oak trees (Quercus agrifolia) in a California annual grassland. American Midland Naturalist 107:69-81. Parsons, A. N., J. M. Welker, P. A. Wookey, M. C. Press, T. V. Callaghan, and J. A. Lee. 1994. Growth responses of four sub-Arctic dwarf shrubs to simulated environmental change. Journal of Ecology 82:307-318. Pasonen, H. L., and M. Kapyla. 1998. Pollen-pollen interactions in Betula pendula in vitro. New Phytologist 138:481-487. Patten, D. T. 1978. Productivity and production efficiency of an upper Sonoran Desert ephemeral community. American Journal of Botany 65:891-895. Peek, M. S., and I. N. Forseth. 2003. Microhabitat dependent responses to resource pulses in the aridland perennial, Cryptantha flava. Journal of Ecology 91:457-466. Pelletier, B., J. W. Fyles, and P. Dutilleul. 1999. Tree species control and spatial structure of forest floor properties in a mixed-species stand. Ecoscience 6:202-214. Penfound, W. T., and F. G. Deiler. 1947. On the ecology of Spanish moss. Ecology 28:455-458. Pennings, S., and R. M. Callaway. 1992. Salt marsh plant zonation: the importance and intensity of competition and physical factors. Ecology 73:681-690. Pennings, S. C., and M. D. Bertness. 2001. Salt marsh communities. Pages 289-316 in M. D. Bertness, S. D. Gaines, and M. E. Hay, editors. Marine Community Ecology. Sinauer Associates, Inc., Sunderland, Massachusetts, USA. Pennings, S. C., and R. M. Callaway. 1996. Impact of a native parasitic plant on salt marsh vegetation structure and dynamics. Ecology 77: 1410-1419. Pennings, S. C., E. R. Selig, L. T. Houser, and M. D. Bertness. 2003. Geographic variation in positive and negative interactions among salt marsh plants. Ecology 84:1527-1538.
390
References
Petanidou, T., H. C. M. den Nijs, and A. C. Ellis-Adam. 1993. Comparative pollination ecology of two rare Dutch Gentiana species, in relation to population size. Acta Horticulturae 288:308-312. Petit, S., and C. R. Dickson. 2005. Grass tree (Xanthorrhoea semiplana, Lliliaceae) facilitation of the endangered pink lipped spider orchid (Caladenia syn. Arachnorchis behrii, Orchidaceae) varies in south Australia. Australian Journal of Botany 53:455-464. Petranka, J. W., and J. K. McPherson. 1979. The role of Rhus copallina in the dynamics of the forest-prairie ecotone in north-central Oklahoma. Ecology 60:956-965. Pfister, C. A., and M. E. Hay. 1988. Associational plant refuges: convergent patterns in marine and terrestrial communities result from differing mechanisms. Oecologia 77:118-129. Phillips, F. J. 1909. A study of pinyon pine. Botanical Gazette 48:216-223. Phillips, F. J. 1910. The dissemination of junipers by birds. Forest Quarterly 8:60-73. Pickart, A. J., L. M. Miller, and T. E. Duebendorfer. 1998. Yellow bush lupine invasion in northern California coastal dunes. I. Ecological impacts and manual restoration techniques. Restoration Ecology 6: 59-68. Pickett, S. T. A., S. L. Collins, and J. J. Armesto. 1987. Models, mechanisms, and pathways of succession. Botanical Review 53:335-371. Pickett, S. T. A., and P. S. White. 1985. The Ecology of Natural Disturbance and Patch Dynamics. Academic Press, Orlando, Florida, USA. Plaisted, K. C., and N. J. MacIntosh. 1995. Visual search images for cryptic stimuli in pigeons: implications for the search images and search rate hypotheses. Journal of Experimental Psychology: Animal Behavior Processes 39:335-354. Pontecorvo, G., and M. Bokhari. 1975. Hedge-like habit of Juniperus excelsa at high altitude on the southern Zagros Mountains in Iran. Proceedings of Royal Society of London B 188:507-508. Poole, R. W., and B. J. Rathcke. 1979. Regularity, randomness, and aggregation in flower phenologies. Science 203:470-471. Prati, D., and O. Bosdorf. 2004. Allelopathic inhibition of germination by Alliaria petiolata (Brassicaceae). American Journal of Botany 91:285-288. Proffitt, C. E., R. L. Chiasson, A. B. Owens, K. R. Edwards, and S. E. Travis. 2005. Spartina alterniflora genotype influences facilitation and suppression of high marsh species colonizing an early successional salt marsh. Journal of Ecology 93:404-416.
References
391
Pugnaire, F., C. Armas, and F. Valladares. 2004. Soil as a mediator in plantplant interactions in a semi-arid community. Journal of Vegetation Science 15:85-92. Pugnaire, F. I., P. Haase, J. Puigdefabregas, M. Cueto, L. D. Incoll, and S. C. Clark. 1996b. Facilitation and succession under the canopy of Retama sphaerocarpa (L.) Boiss. in a semi-arid environment in South-east Spain. Oikos 76:455-464. Pugnaire, F. I., P. Hasse, and J. Puidefabregas. 1996a. Facilitation between higher plant species in a semiarid environment. Ecology 77:1420-1426. Pugnaire, F. I., and R. Lazaro. 2000. Seed bank and understorey species composition in a semi-arid environment: the effect of shrub age and rainfall. Annals of Botany 86:807-813. Pugnaire, F. I., and M. T. Luque. 2001. Changes in plant interactions along a gradient of environmental stress. Oikos 93:42-49. Pusenius, J., and R. S. Ostfeld. 2002. Mammalian predator scent, vegetation cover and tree seedling predation by meadow voles. Ecography 25:481-487. Pysek, P., and J. Liska. 1991. Colonization of Sibbaldia tetrandra cushions on alpine scree in the Pamiro-Alai Mountains, central Asia. Arctic and Alpine Research 23:263-272. Quinos, P. M., P. Insausti, and A. Soriano. 1998. Facilitative effect of Lotus tenuis on Paspalum dilatatum in a lowland grassland of Argentina. Oecologia 114:427-431. Rabinowitz, D., J. Rapp, V. Sork, B. Rathcke, G. Reese, and J. Weaver. 1981. Phenologcial properties of wind- and insect-pollinated prairie plants. Ecology 62:49-56. Radevanoki, S. A., and G. E. Wickens. 1967. The ecology of Acacia albida on mantle soils in Zalinger, Jebel Marra Sudan. Journal of Applied Ecology 4:569-579. Raffaele, E., and T. T. Veblen. 1998. Facilitation by nurse shrubs of resprouting behavior in a post-fire shrubland in northen Patagonia, Argentina. Journal of Vegetation Science 9:693-698. Ramirez, J. M., P. J. Rey, J. M. Alcantara, and A. M. Sanchez-Lafuente. 2006. Altitude and woody cover control recruitment of Helleborus foetidus in a Mediterranean mountain area. Oikos 29:375-384. Ratliff, R. D., D. A. Duncan, and S. E. Westfall. 1991. California oakwoodland overstory species affect herbage understory: Management implications. Journal of Rangeland Management 44:306-310. Rausher, M. D. 1981. The effect of native vegetation on the susceptibility of Aristolochia reticulata (Aristolochiaceae) to herbivore attack. Ecology 62:1187-1195.
392
References
Read, D. J., R. Francis, and R. D. Finlay. 1985. Mycorrhizal mycelia and nutrient cycling in plant communities. Pages 193-217 in A. H. Fitter, editor. Ecological Interactions in Soil. Blackwell Scientific, Oxford, UK. Reader, R. J. 1975. Competitive relationships of some bog ericads for major insect pollinators. Canadian Journal of Botany 53:1300-1305. Reader, R. J., S. D. Wilson, J. W. Belcher, I. Wisheu, P. A. Keddy, D. Tilman, E. C. Morris, J. B. Grace, J. B. McGraw, H. Olff, R. Turkington, E. Klein, Y. Leung, B. Shipley, R. van Hulst, M. E. Johansson, C. Nilsson, J. Gurevitch, K. Grigulis, and B. E. Beisner. 1994. Plant competition in relation to neighbor biomass: an intercontinental study with Poa pratensis. Ecology 75:1753-1760. Rebele, F. 2000. Competition and coexistence of rhizomatous perennial plants along a nutrient gradient. Plant Ecology 147:77-94. Rebertus, A. J., B. R. Burns, and T. T. Veblen. 1991. Stand dynamics of Pinus flexilis-dominated subalpine forest in the Colorado Front Range. Journal of Vegetation Science 2:445-458. Reinhart, K. O., and R. M. Callaway. 2004. Soil biota facilitate exotic Acer invasions in Europe and North America. Ecological Applications 14:1737-1745. Reinhart, K. O., J. Gurnee, R. Tirado, and R. M. Callaway. 2006. Invasion through quantitative effects: intense shade as a driver of invasive success and native decline. Ecological Applications 16:1821-1831. Rey, P. J., and J. M. Alcántara. 2000. Recruitment dynamics of a fleshyfruited plant (Olea europea): connecting patterns of seed dispersal to seedling establishment. Journal of Ecology 88:622-633. Reyes-Olivas, A., E. García-Moya, and L. López-Mata. 2002. Cacti-shrub interactions in the coastal desert of northern Sinaloa, Mexico. Journal of Arid Environments 52:431-445. Reynolds, J. F., R. A. Virginia, P. R. Kemp, A. G. De Soyza, and D. C. Tremmel. 1999. Impact of drought on desert shrubs: effects of seasonality and degree of resource island development. Ecological Monographs 69:69-106. Rheinhardt, R. D. 1992. Disparate distribution patterns between canopy and subcanopy life-forms in two temperate North-American forests. Vegetatio 103:67-77. Rhoades, D. F. 1983. Responses of alder and willow to attack by tent caterpillars and webworms: evidence for pheromonal sensitivity of insects. Pages 55-68 in P. A. Hedin, editor. Plant Resistance to Insects. American Chemical Society, Washington, D.C, USA.
References
393
Rice, K. J., and E. S. Nagy. 2000. Oak canopy effects on the distribution patterns of two annual grasses: the role of competition and soil nutrients. American Journal of Botany 87:1699-1706. Richards, J. H., and M. M. Caldwell. 1987. Hydraulic lift: substantial nocturnal water transport between soil layers by Artemisia tridentata roots. Oecologia 73:486-498. Richardson, D. M., N. Allsopp, C. M. D’Antonio, S. J. Milton, and M. Rejmanek. 2000. Plant invasions - the role of mutualisms. Biological Review 75:65-93. Rigg, L. S., N. J. Enright, G. L. W. Perry, and B. P. Miller. 2002. The role of cloud combing and shading by isolated trees in the succession from maquis to rain forest in New Caledonia. Biotropica 34:199-210. Riginos, C., S. J. Milton, and T. Wiegand. 2005. Context-dependent interactions between adult shrubs and seedlings in a semi-arid shrubland. Journal of Vegetation Science 16:331-340. Robinson, D., and A. Fitter. 1999. The magnitude and control of carbon transfer between plants linked by a common mycorrhizal network. Journal of Experimental Botany 50:9-13. Robinson, M. D. 2004. Growth and abundance of desert annuals in an arid woodland in Oman. Plant Ecology 174:137-145. Roll, J., R. J. Mitchell, R. J. Cabin, and D. L. Marshall. 1997. Reproductive success increases with local density of conspecifics in a desert mustard (Lesquerella fendleri). Conservation Biology 11:738-746. Ronco, F. 1970. Influence of high light intensity on survival of planted Englemann spruce. Forest Science 16:331-339. Root, R. B. 1972. The influence of vegetational diversity on the population ecology of a specialized herbivore, Phyllotreta cruciferae (Coleoptera: Chrysomelidae). Oecologia 10:321-346. Root, R. B. 1973. Organization of a plant-arthropod association in simple and diverse habitats: the fauna of collards (Brassica oleracea). Ecological Monographs 43:95-124. Rosenzweig, M. L. 1978. Competitive speciation. Biological Journal of the Linnean Society 10:275-289. Rostagno, C. M., H. F. del Valle, and L. Videla. 1991. The influence of shrubs on some chemical and physical properties of an aridic soil in north-eastern Patagonia, Argentina. Journal of Arid Environments 20:179-188. Rousset, O., and J. Lepart. 1999. Shrub facilitation of Quercus humilis regeneration in succession on calcareous grasslands. Journal of Vegetation Science 10:493-502.
394
References
Rousset, O., and J. Lepart. 2000. Positive and negative interactions at different life stages of a colonizing species (Quercus humilis). Journal of Ecology 88:401-412. Rousset, O., and J. Lepart. 2002. Neighbourhood effects on the risk of an unpalatable plant being grazed. Plant Ecology 165:197-206. Rubio-Casal, A. E., J. M. Castillo, C. J. Luque, and M. E. Figueroa. 2001. Nucleation and facilitation in salt pans in Mediterranean salt marshes. Journal of Vegetation Science 12:761-770. Rudgers, J. A., and J. L. Maron. 2003. Facilitation between coastal dune shrubs: a non-nitrogen fixing shrub facilitates establishment of a nitrogen-fixer. Oikos 102:75-84. Rumbaugh, M. D., D. A. Johnson, and G. A. Van Epps. 1982. Forage yield and quality in a Great Basin shrub, grass, and legume pasture experiment. Journal of Range Management 35:604-609. Russell, F. L., and N. L. Fowler. 2004. Effects of white-tailed deer on the population dynamics of acorns, seedlings and small saplings of Quercus buckleyi. Plant Ecology 173:59-72. Ryan, P., E. Delhaize and D.L. Jones. 2001. Function and mechanism of organic anion exudation from plant roots. Annual Review of Plant Physiology and Plant Molecular Biology 52:527-560. Rykiel, E. J., Jr., and T. L. Cook. 1986. Hardwood-red cedar clusters in the post oak savannas of Texas. The Southwestern Naturalist 31:73-78. Ryser, P. 1993. Influences of neighboring plants on seedling establishment in limestone grassland. Journal of Vegetation Science 4:195-202. Sabelis, M. W., J. A., and M. R. Hunt. 2001. The enemy of my enemy is my ally. Science 291:2104-2105. Salonen, V. 1987. Relationship between seed rain and establishment of vegetation in two areas abandoned after peat harvesting. Holarctic Ecology 10:171-174. Salonen, V. 1991. Effects of artificial plant cover on plant colonization of a bare peat surface. Journal of Vegetation Science 3:109-112. Sans, F. X., J. Escarre, V. Gorse, and J. Lepart. 1998. Persistence of Picris hieracioides populations in old fields: an example of facilitation. Oikos 83:283-292. Sans, F. X., J. Escarré, J. Lepart, and F. Hopkins. 2002. Positive vs. negative interactions in Picris hieracioides L., a mid-successional species of Mediterranean secondary succession. Plant Ecology: 171:109-122. Sarig, S., G. Barness, and Y. Steinberger. 1994. Annual plant growth and soil characteristics under desert halophyte canopy. Acta Oecologica 15:521-527.
References
395
Scanlan, J. C., and S. Archer. 1991. Simulated dynamics of succession in a North American subtropical Prosopis savanna. Journal of Vegetation Science 2:625-634. Schade, J. D., R. Sponseller, S. L. Collins, and A. Stiles. 2003. The influence of Prosopis canopies on understorey vegetation: effects of landscape position. Journal of Vegetation Science 14:743-750. Schat, H. 1984. A comparative ecophysiological study on the effects of waterlogging and submergence on dune slack plants: growth, survival and mineral nutrition in sand culture experiments. Oecologia 62:279-286. Schat, H., and K. Van Beckhoven. 1991. Water as a stress factor in the coastal dune system. Pages 76-89 in J. Rozema and J. A. C. Verkleij, editors. Ecological Responses to Environmental Stresses. Kluwer Academic Publishers, Amsterdam, The Netherlands. Schemenauer, R. S., H. Fuenzalida and P. Cereceda. 1988. A neglected water resource: the camanchaca of South America. Bulletin American Meteorological Society 69:138-147. Schemske, D. W. 1981. Floral convergence and pollinator sharing in two beepollinated tropical herbs. Ecology 62:946-954. Schenck, J., B. E. Mahall, and R. M. Callaway. 1999. Spatial segregation of roots. Advances in Ecology 28:145-180. Schenk, H. J., C. Holzapfel, H. J.G., and B. E. Mahall. 2003. Spatial ecology of a small desert shrub on adjacent geological substrates. Journal of Ecology 91:383-395. Schenk, H. J. and B. E. Mahall. 2002. Positive and negative plant interactions contribute to a north-south-patterned association between two desert shrub species. Oecologia 132:402-410. Schiffers, K., and K. Tielbörger. 2006. Ontogenetic shifts in interactions among annual plants. Journal of Ecology 94:336-341. Schimel, J. P., R. G. Cates, and R. Ruess. 1998. The role of balsam popular secondary chemicals in controlling soil nutrient dynamics through succession in the Alaskan taiga. Biogeochemistry 42:221-234. Schlesinger, W. H., and P. L. Marks. 1977. Mineral cycling and the niche of Spanish moss, Tillandsia usneoides L. American Journal of Botany 64:1254-1262. Schlesinger, W. H., and W. A. Reiners. 1974. Deposition of water and cations on artificial foliar collectors in fir krummholtz of New England mountains. Ecology 55:378-386. Schmida, A., and R. H. Whittaker. 1981. Pattern and biological microsite effects in two shrub communities in southern California. Ecology 62:234-251.
396
References
Schoener, T. W. 1980. Ecological Interactions. Pages 226-230 in S. P. Parker, editor. McGraw-Hill Encyclopedia of Environmental Science. McGraw-Hill, New York, USA. Schoener, T. W. 1983. Field experiments on interspecific competition. American Naturalist 122:240-285. Scholes, R. J. 1990. The influence of soil fertility on the ecology of southern African dry savannas. Journal of Biogeography 17:415-419. Scholes, R. J., and S. R. Archer. 1997. Tree-grass interactions in savannas. Annual Review of Ecology and Systematics 28:517-544. Schupp, E. W. 1995. Seed-seedling conflicts, habitat choice, and patterns of plant recruitment. American Journal of Botany 82:399-409. Schuster, W. S., and R. J. Hutnick. 1987. Community development on 35-year-old planted minespoil banks in Pennsylvania. Reclamation and Revegetation Research 6:109-120. Schweitzer, J. A., J. K. Bailey, B. J. Rehill, G. D. Martinsen, S. C. Hart, R. L. Lindroth, P. Keim, and T. G. Whitham. 2004. Genetically based trait in a dominant tree affects ecosystem processes. Ecology Letters 7:127-134. Seastedt, T. R. 1984. The role of microarthropods in decomposition and mineralization processes. Annual Review Entomology 29:25-46. Seiwa, K. 1998. Advantages of early germination for growth and survival of seedlings of Acer mono under different overstorey phenologies in deciduous broad-leaved forests. Journal of Ecology 86:219-228. Shevtsova, A., A. Ojala, N. S., M. Vieno, and E. Haukoija. 1995. Growth and reproduction of dwarf shrubs in a subarctic plant community: annual variation and above-ground interactions with neighbors. Journal of Ecology 83:263-275. Shipley, B., and P. A. Keddy. 1987. The individualistic and community-unit concepts and falsifiable hypotheses. Vegetatio 69:47-55. Shirley, H. L. 1945. Reproduction of upland conifers in the Lake States as affected by root competition and light. American Midland Naturalist 33:537-612. Shore, J. S. 1993. Pollination genetics of the common milkweed, Asclepias syriaca L. Heredity 70:101-108. Shreve, F. 1910. The rate of establishment of the giant cactus. Plant World 13:235-241. Shreve, F. 1917. The establishment of desert perennials. Journal of Ecology 5:210-216. Shreve, F. 1931. Physical conditions in sun and shade. Ecology 12:96-104.
References
397
Shumway, S. W. 2000. Facilitative interactions between a sand dune shrub and species growing beneath the shrub canopy. Oecologia 124:138-148. Shumway, S. W., and M. D. Bertness. 1992. Salt stress limitation of seedling recruitment in a salt marsh plant community. Oecologia 92:490-497. Shumway, S. W., and M. D. Bertness. 1994. Patch size effects on marsh plant secondary succession mechanisms. Ecology 75:564-568. Siemann, E., and W. E. Rogers. 2003. Changes in light and nitrogen availability under pioneer trees may indirectly facilitate tree invasions of grasslands. Journal of Ecology 91:923-931. Sih, A., and M.-S. Baltus. 1987. Patch size, pollinator behavior, and pollinator limitation in catnip. Ecology 68:1679-1690. Silander, J. A., and J. Antonovics. 1982. Analysis of interspecific interactions in a coastal plant community - a perturbation approach. Nature 298:557-560. Silander, J. A. J. 1978. Density-dependent control of reproductive success in Cassia biflora. Biotropica 10:292-296. Silber, A., and M. Raviv. 1996. Effects on chemical surface properties of tuff by growing rose plants. Plant and Soil 186:353-360. Silvertown, J., and J. B. Bastow. 1994. Community structure in a desert community. Ecology 75:409-417. Simard, S. W., D. S. Perry, M. D. Jones, D. D. Myrold, D. M. Durall, and R. Molina. 1997. Net transfer of carbon between ectomycorrhizal tree species in the field. Nature 388:579-582. Simberloff, D., and B. Von Holle. 1999. Positive interactions of nonindigenous species: invasional meltdown? Biological Invasions 1:21-32. Singer, M. C. 1971. Evolution of food-plant preference in the butterfly Euphydryas editha. Evolution 25:383-389. Singh, K. S., and P. Lal. 1969. Effect of Khejari (Prosopis spicigera Linn) and Babool (Acacia arabica) trees on soil fertility and profile characteristics. Annals of Arid Zones 8:33-36. Skorupa, J. P., and J. M. Kasenene. 1982. Tropical forest management: can rates of natural treefalls help us? Oryx 18:96-101. Slocum, M. 2001. How tree species differ as recruitment foci in a tropical pasture. Ecology 82:2547-2559. Slocum, M. G., and C. C. Horvitz. 2000. Seed arrival under different genera of trees in a neotropical pasture. Plant Ecology 149:51-62. Smethurst, P. J., N. D. Turvery, and P. M. Attiwill. 1986. Effects of Lupinus spp. on soil nutrient availability and the growth of Pinus radiata D. Don seedlings on a sandy podzol in Victoria, Australia. Plant and Soil 95:183-190.
398
References
Smit, C., J. Den Ouden, and H. Müller-Schärer. 2006. Unpalatable plants facilitate tree sapling survival in wooded pastures. Journal of Ecology 43:305-312. Smith, H., J. J. Casal, and G. M. Jackson. 1990. Reflection signals and the perception by phytochrome of the proximity of neighboring vegetation. Plant, Cell and Environment 13:73-78. Smith, S. D., B. Didden-Zoppy, and N. P.S. 1984. High-temperature responses of North American cacti. Ecology 65:643-651. Smith, S. D., D. T. Patten, and R. K. Monson. 1987. Effects of artificially imposed shade on a Sonoran Desert ecosystem: microclimate and vegetation. Journal of Arid Environments 13:65-82. Smith, T. M., and P. S. Goodman. 1987. Successional dynamics in an Acacia nilotica-Euclea divinorum savanna in southern Africa. Journal of Ecology 75:603-610. Smith, T. M., and M. A. Huston. 1989. A theory of the spatial and temporal dynamics of plant communities. Vegetatio 83:49-69. Smith, W. K., M. J. Germino, T. E. Hancock, and D. M. Johnson. 2003. Another perspective on altitudinal limits of alpine timberlines. Tree Physiology 23:1101-1112. Solomon, B. P. 1981. Response of a host-specific herbivore to resource density, relative abundance, and phenology. Ecology 62:1205-1214. Soriano, A., and O. E. Sala. 1986. Emergence and survival of Bromus setifolius seedlings in different microsites of a Patagonian arid steppe. Israel Journal Botany 35:91-100. Spehn, E. M., M. Scherer-Lorenzen, M. B. Schmid, A. Hector, M. C. Caldeira, P. G. Dimitrakopoulos, J. A. Finn, A. Jumpponen, G. O’Donnovan, J. S. Pereira, E.-D. Schulze, A. Y. Troumbis, and C. Körner. 2002. The role of legumes as a component of biodiversity in a cross-European study of grassland biomass nitrogen. Oikos 98:205-218. Stachowicz, J. J. 2001. Mutualism, facilitation, and the structure of ecological communities. Bioscience 51:235-246. Stadler, B., and B. Michalzik. 1998. Linking aphid honeydew, throughfall, and forest floor solution chemistry of Norway spruce. Ecology Letters 1:13-16. Stahelin, R. 1943. Factors influencing the natural restocking of high altitude burns by coniferous trees in the central Rocky Mountains. Ecology 24:19-30. Steenberg, W. F., and C. H. Lowe. 1969. Critical factors during the first year of life of the saguaro (Cereus giganteus) at Saguaro National Monument. Ecology 50:825-834.
References
399
Steenberg, W. F., and C. H. Lowe. 1976. Ecology of the saguaro. I. The role of freezing weather in a warm-desert plant population. Sympoium Series Number 1 Number 1, National Park Service, Washington, D.C., USA. Steenberg, W. F., and C. H. Lowe. 1977. Ecology of the saguaro. II. Reproduction, germination, establishment, growth, and survival of the young plant. National Park Service, Washington, D.C., USA. Stephan, A., and A. H. Meyer. 2000. Plant diversity affects culturable soil bacteria in experimental grassland communities. Journal of Ecology 88:988-998. Sthultz, C. M., C. A. Gehring, and T. G. Whitham. 2006. Shifts from competition to facilitation between a foundation tree and a pioneer shrub across spatial and temporal scales in a semiarid woodland. New Phytologist 173:135-145. Stiling, P., A. M. Rossi, and M. V. Cattell. 2003. Associational resistance mediated by natural enemies. Ecological Entomology 28:587-592. Stock, W. D., T. S. Dlamini, and R. M. Cowling. 1999. Plant induced fertile islands as possible indicators of desertification in a succulent desert ecosystem in northern Namaqualand, South Africa. Plant Ecology 142:161-167. Stone, L., and A. Roberts. 1991. Conditions for a species to gain an advantage from the presence of competitors. Ecology 72:1964-1972. Strauss, S. Y. 1991. Indirect effects in community ecology: their definition, study and importance. Trends in Ecology and Evolution 6:206-210. Stuart, T. S. 1968. Revival of respiration and photosynthesis in dried leaves of Polypodium polypodioides. Planta 83:185-206. Suding, K. N., and D. E. Goldberg. 1999. Variation in the effects of vegetation and litter on recruitment across productivity gradients. Journal of Ecology 87:436-449. Suzan, H., G. P. Nabhan, and D. P. Patten. 1994. Nurse plant and floral biology of a rare night-blooming cereus, Peniocereus striatus (Brandengee) F. Buxbaum. Conservation Biology 8:4461-4470. Svenning, J. C., and F. Skov. 2002. Mesoscale distribution of understorey plants in managed temperate forest (Kalø, Denmark): the importance of environment and dispersal. Plant Ecology: 173:169-185. Sydes, C., and J. P. Grime. 1981. Effects of tree litter on herbaceous vegetation in deciduous woodland. II. An experimental investigation. Journal of Ecology 69:249-262. Symstad, A. J., D. Tilman, J. Willson, and J. M. H. Knops. 1998. Species loss and ecosystem functioning: effects of species identity and community composition. Oikos 81:389-397.
400
References
Tahvanainen, J. O. 1972. Phenology and microhabitat selection of some flea beetles (Coleoptera: Chrysomelidae) on wild and cultivated crucifers in Cental New York. Entomologica Scandinavia 3:120-138. Tahvanainen, J. O., and R. B. Root. 1972. The influence of vegetational diversity on the population ecology of a specialized herbivore, Phyllotreta cruciferae (Coleoptera: Chrysomelidae). Oecologia 10:321-346. Takahashi, K. 1997. Regeneration and coexistence of two subalpine conifer species in relation to dwarf bamboo in the understory. Journal of Vegetation Science 8:529-536. Talley, S. M., R. O. Lawton, and W. N. Setzer. 1996a. Host preferences of Rhus radicans (Anacardiaceae) in a southern deciduous hardwood forest. Ecology 77:1271-1276. Talley, S. M., W. N. Setzer, and B. R. Jackes. 1996b. Host associations of two adventitious-root-climbing vines in a North Queensland tropical rainforest. Biotropica 28:356-366. Taylor, D. L., and T. D. Bruns. 1997. Independent, specialized invasions of ectomycorrhizal mutualism by two nonphotosynthetic orchids. Proceedings of the National Academy of Science 94:4510-4515. ter Steege, H., and J. H. C. Cornelissen. 1989. Distribution and ecology of vascular epiphytes in lowland rainforest of Guyana. Biotropica 21:331-339. Tewksbury, J. J., and J. D. Lloyd. 2001. Positive interactions under nurseplants: spatial scale, stress gradients and benefactor size. Oecologia 127:425-434. Tewksbury, J. J., G. P. Nabhan, D. Norman, H. Suzan, J. Tuxhill, and J. Donovan. 1998. In situ conservation of wild chiles and their biotic associates. Conservation Biology 13:1-10. Tewksbury, J. T., and C. A. Petrovich. 1994. The influences of ironwood as a habitat modifier species: a case study on the Sonoran Desert coast of the Sea of Cortez. Pages 29-54 in G. P. Nabhan and J. L. Carr, editors. Ironwood: an Ecological and Cultural Keystone of the Sonoran Desert. Conservation International, Washington, D.C., USA. Theodose, T. A., and W. D. Bowman. 1997. The influence of interspecific competition on the distribution of an alpine graminoid: evidence for the importance of plant competition in an extreme environment. Oikos 79:63-74. Thomas, B. D., and W. D. Bowman. 1998. Influence of N2-fixing Trifolium on plant species composition and biomass production in alpine tundra. Oecologia 115:26-31. Thomas, C. D. 1986. Butterfly larvae reduce host plant survival in vicinity of alternative host species. Oecologia 70:113-117.
References
401
Thomas, F. I. M., C. D. Cornelisen, and J. M. Zande. 2000. Effects of water velocity and canopy morphology on ammonium uptake by seagrass communities. Ecology 81:2704-2713. Thomas, H. 1984. Effects of drought on growth and competitive ability of perennial ryegrass and white clover. Journal of Applied Ecology 21:591-602. Thomas, S. C., C. B. Halpern, D. A. Falk, D. A. Liguori, and K. A. Austin. 1999. Plant diversity in managed forests: understory response to thinning and fertilization. Ecological Applications 9:864-879. Thompson, J. N. 2005. The Geographic Mosaic of Coevolution. University of Chicago Press, Chicago, Illinois, USA. Thompson, L., and J. L. Harper. 1988. The effect of grasses on the quality of transmitted radiation and its influence on the growth of white clover (Trifolium repens). Oecologia 75:343-347. Thomson, J. D. 1978. Effects of stand composition on insect visitation in twospecies mixtures of Hieracium. American Midland Naturalist 100:431-440. Thomson, J. D. 1981. Spatial and temporal components of resource assessment by flower-feeding insects. Journal of Animal Ecology 50:49-59. Thomson, J. D. 1982. Patterns of visitation by animal pollinators. Oikos 39:241-250. Thomson, J. D. 1989. Germination schedules of pollen grains: implications for pollen selection. Evolution 43:220-223. Thorpe, A. S. 2006. Biochemical effects of C. maculosa on soil nutrient cycles and plant communities. The University of Montana, PhD Dissertation, Missoula, Montana, USA. Thorpe, A. S., V. Archer, and T. H. DeLuca. 2006. The invasive forb, Centaurea maculosa, increases phosphorus availability in Montana grasslands. Applied Soil Ecology 32:118-122. Tiedemann, A. R., and J. O. Klemmedson. 1973. Effects of mesquite on physical and chemical properties of the soil. Journal of Range Management 26:27-29. Tiedemann, A. R., and J. O. Klemmedson. 1977. Effect of mesquite trees on vegetation and soils in the desert grasslands. Journal of Range Management 30:361-367. Tiedemann, A. R., and J. O. Klemmedson. 1986. Long-term effects of mesquite removal on soil characteristics: I. nutrients and bulk density. Soil Science Society of America Journal 50:472-475. Tielbörger, K., and R. Kadmon. 1995. The effect of shrubs on the emergence, survival and fecundity of four coexisting annual species in a sandy desert ecosystem. Ecoscience 2:141-147.
402
References
Tielbörger, K., and R. Kadmon. 1997. Relationships between shrubs and annual communities in a sandy desert ecosystem: a three year study. Plant Ecology 130:191-201. Tielbörger, K., and R. Kadmon. 2000. Temporal environmental variation tips the balance between facilitation and interference in desert plants. Ecology 81:1544-1553. Tilman, D. 1985. The resource-ratio hypothesis for plant succession. American Naturalist 125:827-852. Tilman, D. 1988. Plant Strategies and the Dynamics and Structure of Plant Communities. Princeton University Press, Princeton, New Jersey, USA. Tilman, D., J. Knops, D. Wedin, P. Reich, M. Ritchie, and E. Siemann. 1997. The influence of functional diversity and composition on ecosystem processes. Science 277:1300-1302. Tilman, D., P. B. Reich, J. Knops, D. Wedin, T. Mielke, and C. Lehman. 2001. Diversity and productivity in a long-term grassland experiment. Science 294:843-845. Tilman, D., D. Wedin, and J. Knops. 1996. Productivity and sustainability influenced by biodiversity in grassland ecosystems. Nature 379:718-720. Tilman, D., and D. A. Wedin. 1991a. Plant traits and resource reduction for five grasses growing on a nitrogen gradient. Ecology 72:685-700. Tilman, D., and D. A. Wedin. 1991b. Dynamics of nitrogen competition between successional grasses. Ecology 72:1038-1049. Tirado, R., and F. Pugnaire. 2003. Shrub spatial aggregation and consequences for reproductive success. Oecologia 136:296-301. Titus, J. H., and S. Tsuyuzaki. 2003. Influence of a non-native invasive tree on primary succession at Mt. Koma, Hokkaido, Japan. Plant Ecology 169:307-315. Toft, C., and D. Elliott-Fisk. 2002. Patterns of vegetation along a spatiotemporal gradient on shoreline strands of a desert basin lake. Plant Ecology 158:21-39. Toh, I., M. Gillespie, and D. Lamb. 1999. The role of isolated trees in facilitating tree seedling recruitment at a degraded sub-tropical rainforest site. Restoration Ecology 7:228-237. Totland, Ø., and J. Esaete. 2002. Effects of willow canopies on plant species performance in a low-alpine community. Plant Ecology 161:157-166. Toumey, J. W., and E. J. Neethling. 1924. Insolation as a factor in the natural regeneration of certain conifers. Yale University School of Forestry Bulletin 11:1-63.
References
403
Tranquillini, W. 1980. Winter desiccation as the cause for alpine timberline. Pages 263-267 in U. Benecke and M. R. Davis, editors. Mountain Environments and Subalpine Tree Growth. Forest Research Institute, New Zealand Forest Service, Wellington, New Zealand. Travis, J. M. J., R. W. Brooker, and C. Dytham. 2004. The interplay of positive and negative species interactions across an environmental gradient: insights from an individual based simulation model. Proceedings Royal Society London B (Suppl.: Biology Letters). Travisano, M., and P. B. Rainey. 2000. Studies of adaptive radiation using model microbial systems. American Naturalist 156:S35–S44. Trenbath, B. R. 1974. Biomass productivity of mixtures. Advances in Agronomy 26:177-210. Tuittila, E.-S., H. Rita, H. Vasander, and J. Laine. 2000. Vegetation patterns around Eriophorum vaginatum L. tussocks in a cut-away peatland in southern Finland. Canadian Journal of Botany 78:47-58. Tupas, G. L., and P. E. Sajise. 1977. The role of savanna trees in plant succession: Ecological conditions associated with tree clumps. Philippine Journal of Biology 6:229-244. Turkington, R. 1989. The growth, distribution and neighbour relationships of Trifolium repens in a permanent pasture. V. The coevolution of competitors. Journal of Ecology 77:717-733. Turkington, R., and P. A. Jolliffe. 1996. Interference in Trifolium repensLolium perene mixtures: short- and long-term relationships. Journal of Ecology 84:563-571. Turkington, R., and L. A. Mehrhoff. 1990. The role of competition in structuring pasture communities. Pages 308-366 in J. B. Grace and D. Tilman, editors. Perspectives on Plant Competition. Academic Press, New York, USA. Turner, P. E., V. Souza, and R. E. Lenski. 1996. Tests of ecological mechanisms promoting the stable coexistence of two bacterial genotypes. Ecology 77:2119-2129. Turner, R. M., S. M. Alcorn, G. Olin, and J. A. Booth. 1966. The influence of shade, soil, and water on saguaro seedling establishment. Botanical Gazette 127:95-102. Turner, R. M., S. M. Alcorn, and G. Olin. 1969. Mortality of transplanted saguaro seedlings. Ecology 50:835-844. Turner, T. 1983. Facilitation as a successional mechanism in a rocky intertidal community. American Naturalist 121:729-737. Twolan-Strutt, L., and P. A. Keddy. 1996. Above- and below-ground competition intensity in two contrasting wetland plant communities. Ecology 77:259-270.
404
References
Ugolini, F. C., and D. H. Mann. 1979. Biopedological origin of peatlands in southeast Alaska. Nature 281:366-368. Valdeyron, G., B. Dommee, and P. Vernet. 1977. Self-fertilization in malefertile plants of a gynodioecious species: Thymus vulgaris L. Heredity 62:17-26. Vale, T. R. 1981. Tree invasion of montane meadows in Oregon. American Midland Naturalist 105:61-69. Valiente-Banuet, A., A. Bolongaro, O. Briones, E. Ezcurra, M. Rosas, H. Nunez, G. Barnhard, and E. Vasquez. 1991. Spatial relationships between cacti and nurse shrubs in a semi-arid environment in central Mexico. Journal of Vegetation Science 2:15-20. Valiente-Banuet, A., A. V. Rumebe, M. Verdú, and R. M. Callaway. 2006. Modern Quaternary plant lineages promote diversity through facilitation of ancient Tertiary lineages. Procedings of the National Academy of Sciences 103:16812-16817. Valiente-Banuet, A., and E. Ezcurra. 1991. Shade as a cause of the association between the cactus Neobuxbaumia tetetzo and the nurse plant Mimosa luisana in the Tehuacan Valley, Mexico. Journal of Ecology 79: 961-971. Van Auken, O. W., and J. K. Bush. 1985. Secondary succession on terraces of the San Antonio River. Bulletin of the Torrey Botanical Club 112:158-166. Van Auken, O. W., E. M. Gese, and K. Connors. 1985. Fertilization response of early and late successional species: Acacia smallii and Celtis laevigata. Botanical Gazette 146:564-569. van Breeman, N. 1995. How Sphagnum bogs down other plants. Trends in Ecology and Evolution 10:270-275. Van Breemen, N., and A. C. Finzi. 1998. Plant-soil interactions: ecological aspects and evolutionary implications. Biogeochemistry 42:1-19. van de Koppel, J., A. H. Altieri, B. R. Silliman, J. F. Bruno, and M. D. Bertness. 2006. Scale-dependent interactions and community structure on cobble beaches. Ecology Letters 9:45-50. van der Heijden, M. G. A., J. N. Klironomos, M. Ursic, P. Moutoglis, R. Streitwolf-Engel, T. Boller, A. Wiemken, and I. R. Sanders. 1998. Mycorrhizal fungal diversity determines plant biodiversity, ecosystem variability and productivity. Nature 396:69-72. van der Maarel, E., V. Noest, and M. W. Palmer. 1995. Variation in species richness on small grassland quadrats: niche structure or small-scale plant mobility? Journal of Vegetation Science 6:741-752.
References
405
Van der Putten, W. H., C. Van Dijk, and B. A. M. Peters. 1993. Plant-specific soil-borne diseases contribute to succession in foredune vegetation. Nature 53:362-364. Van Treuren, R., R. Bulsma, N. J. Ouborg, and W. Van Delden. 1993. The effects of population size and plant density on outcrossing rates in locally endangered Salvia pratensis. Evolution 47:1094-1104. Vandermeer, J. 1980. Saguaros and nurse trees: a new hypothesis to account for population fluctuations. Southwest Naturalist 25:357-360. Vandermeer, J. 1990. Indirect and diffuse interactions: complicated cycles in population embedded in a large community. Journal of Theoretical Biology 142:429-442. Vasek, F. C., and L. J. Lund. 1980. Soil characteristics associated with a primary plant succession on a Mojave Desert dry lake. Ecology 61:1013-1028. Vazquez, G., P. Moreno-Casasola, and O. Barrera. 1998. Interaction between algae and seed germination in tropical dune slack species: a facilitation process. Aquatic Botany 60:409-416. Veblen, T. T., D. H. Ashton, F. M. Schlegel, and A. T. Veblen. 1977. Distribution and dominance of species in the understory of a mixed evergreen-deciduous Nothofagus forest in south-central Chile. Journal of Ecology 65:815-830. Veblen, T. T., A. T. Veblen, and F. M. Schlegel. 1979. Understory patterns in mixed evergreen-deciduous Nothofagus forests in Chile. Journal of Ecology 67:809-823. Verdú, M., and P. García-Fayos. 1996. Nuleation processes in a Mediterranean bird-dispersed plant. Functional Ecology 10:275-280. Verdú, M., and P. García-Fayos. 2003. Frugivorous birds mediate sex-biased facilitation in a dioecious nurse plant. Journal of Vegetation Science 14:35-42. Verdú, M., P. Villar-Salvador, and P. García-Fayos. 2004. Gender effects on the post-facilitation performance of two dioecious Juniperus species. Functional Ecology 18:87-93. Vetaas, O. R. 1992. Micro-site effects of trees and shrubs in dry savannas. Journal of Vegetation Science 3:337-334. Vieira, I. C. G., C. Uhl, and D. Nepstad. 1994. The role of the shrub Cordia multispicata Cham. as a ‘succession facilitator’ in an abandoned pasture, Paragominas, Amazonia. Vegetatio 115:91-99. Villalba, R., and T. T. Veblen. 1997. Regional patterns of tree population age structures in northern Patagonia: climatic and disturbance influences. Journal of Ecology 85:113-124. Virginia, R. A. 1986. Soil development under legume tree canopies. Forest Ecology and Management 16:69-79.
406
References
Virginia, R. A., and W. M. Jarrell. 1983. Soil properties in a mesquitedominated Sonoran desert ecosystem. Soil Science Society of America Journal 47:138-144. Vitousek, P. M., and L. R. Walker. 1989. Biological invasion by Myrica faya in Hawaii: plant demography, nitrogen fixation, and ecosystem effects. Ecological Monographs 59:247-265. Vitousek, P. M., L. R. Walker, L. D. Whiteaker, D. Mueller-Dombois, and P. A. Matson. 1987. Biological invasion by Myrica faya alters ecosystem development in Hawaii. Science 238:802-804. Vivanco, J. M., H. P. Bais, F. R. Stermitz, G. C. Thelen, and R. M. Callaway. 2004. Biogeographical variation in community response to root allelochemistry: novel weapons and exotic invasion. Ecology Letters 7:285-292. Vlok, J. H. J., and R. I. Yeaton. 1999. The effect of overstorey proteas on plant species richness in South African mountain fynbos. Diversity and Distributions 5:213-222. Vogelmann, H. W., T. Siccama, D. Leedy and D.C. Ovitt. 1968. Precipitation from fog moisture in the Green Mountains of Vermont. Journal of Ecology 49:1205-1207. Wahl, M., and M. E. Hay. 1995. Associational resistance and shared doom: effects of epibiosis on herbivory. Oecologia 102:329-340. Walker, L., and R. del Moral. 2003. Primary Succession: Ecosystem Assembly on Barren Landscapes. Cambridge University Press, Cambridge, UK. Walker, L. R. 1999. Ecosystems of Disturbed Ground. Elsevier, Amsterdam, The Netherlands. Walker, L. R., and F. S. I. Chapin. 1986. Physiological controls over seedling growth in primary succession on an Alaskan floodplain. Ecology 67:1508-1523. Walker, L. R., and F. S. I. Chapin. 1987. Interactions among processes controlling successional change. Oikos 50:131-135. Walker, L. R., and E. A. Powell. 1999. Regeneration of the Mauna Kea silversword Argyroxiphium sandwicense (Asteraceae) in Hawaii. Biological Conservation 89:61-70. Walker, L. R., D. B. Thompson, and F. H. Landau. 2001. Experimental manipulations of fertile islands and nurse plant effects in the Mojave Desert, USA. Western North American Naturalist 61:25-35. Walker, L. R., and P. M. Vitousek. 1991. An invader alters germination and growth of a native dominant tree in Hawai’i. Ecology 72:1449-1455.
References
407
Wallace, A., and E. M. Romney. 1980. The role of pioneer species in revegetation of disturbed desert areas. Great Basin Natural Memoirs 4:31-33. Walter, L. E., D. C. Hartnett, B. A. D. Hetrick, and A. P. Schwab. 1996. Interspecific nutrient transfer in a tallgrass prairie plant community. American Journal of Botany 83:180-184. Wan, C., R. E. Sosebee, and B. L. McMichael. 1993. Does hydraulic lift exist in shallow-rooted species? A quantitative examination with a halfshrub Gutierrezia sarothrae. Plant and Soil 153:11-17. Wardle, D. A., K. I. Bonner, G. M. Barker, G. W. Yeates, K. S. Nicholson, R. D. Bardgett, R. N. Watson, and A. Ghani. 1999. Plant removals in perennial grassland: vegetation dynamics, decomposers, soil biodiversity, and ecosystem properties. Ecological Monographs 69:535-568. Wardle, D. A., M. Nilsson, C. Gallet, and Z. O. Zackisson. 1998. An ecosystem-level perspective of allelopathy. Biological Review 73:305-319. Wardle, P. 1968. Englemann spruce (Picea englemannii Engle.) at its upper limits on the Front Range, Colorado. Ecology 49:483-495. Wasser, N. M. 1978. Interspecific pollen transfer and competition between co-occurring plant species. Oecologia 36:223-236. Watkins, N. K., A. H. Fitter, J. D. Graves, and D. Robinson. 1996. Carbon transfer between C3 and C4 plants linked by a common mycorrhizal network, quantified using stable carbon isotopes. Soil Biology and Biogeochemistry 28:471-477. Webb, L. J., J. G. Tracey, and K. P. Haydock. 1967. A factor toxic to seedlings of the same species associated with living roots of the nongregarious subtropical rainforest tree Grevillea robusta. Journal of Applied Ecology 4:13-25. Wedin, D. A., and D. Tilman. 1990. Species effects on nitrogen cycling: a test with perennial grasses. Oecologia 84:433-441. Weiner, P. W. 1985. Size hierarchies in experimental populations of annual plants. Ecology 66:743-752. Weir, T. L., H. P. Bais, V. J. Stull, R. M. Callaway, G. C. Thelen, W. M. Ridenour, S. Bhamidi, F. R. Stermitz, and J. M. Vivanco. 2006. Oxalate contributes to the resistance of Gaillardia grandiflora and Lupinus sericeus to a phytotoxin produced by Centaurea maculosa. Planta 223:785-795. Weisberg, P. J., and W. L. Baker. 1995. Spatial variation in tree regeneration in the forest-tundra ecotone, Rocky Mountain National Park, Colorado. Canadian Journal of Forest Research 25:1326-1339.
408
References
Welden, C. W., and W. L. Slauson. 1986. The intensity of competition versus its importance: an overlooked distinction and some implications. Quarterly Review of Biology 61:23-44. Welden, C. W., W. L. Slauson, and R. T. Ward. 1990. Spatial pattern and interference in pinon-juniper woodlands of northwest Colorado. Great Basin Naturalist 50:313-319. Weltzin, J. F., and M. B. Coughenhour. 1990. Savanna tree influence on understorey vegetation and soil nutrients in northwestern Kenya. Journal of Vegetation Science 1:325-332. Weltzin, J. F., and G. R. McPherson. 1999. Facilitation of conspecific seedling recruitment and shifts in temperate savanna ecotones. Ecological Monographs 69:513-534. Went, F. W. 1940. Soziologie der epiphyten eines tropischen Urwaldes. Annales Jardin Botanique Buitenzorg 50:1-98. Went, F. W. 1942. The dependence of certain annual plants on shrubs in southern California deserts. Bulletin of the Torrey Botanical Club 69:101-114. Werner, P. A., and A. L. Harbeck. 1982. The pattern of tree seeding establishment relative to staghorn sumac cover in Michigan old fields. American Midlands Naturalist 108:124-132. Westover, K. M., A. C. Kennedy, and S. E. Kelley. 1997. Patterns of rhizosphere microbial community structure associated with co-occurring plant species. Journal of Ecology 85:863-873. Whisler, S. J., and A. A. Snow. 1992. Potential for the loss of selfincompatibility in pollen-limited populations of mayapple (Podophyllum peltatum). American Journal of Botany 79:1273-1285. White, J. A., and T. G. Whitham. 2000. Associational susceptibility of cottonwood to a box elder herbivore. Ecology 81:1795-1803. Whitfield, J. 2002. Neutrality versus the niche. Nature 417:480-481. Whitham, T. G., W. P. Young, G. D. Martinsen, C. A. Gehring, J. A. Schweitzer, S. M. Shuster, G. M. Wimp, D. G. Fischer, J. K. Bailey, R. L. Lindroth, S. Woolbright, and C. R. Kuske. 2003. Community and ecosystem genetics: a consequence of the extended phenotype. Ecology 84:559-573. Whittaker, R. H. 1951. A criticism of the plant association and climatic climax concepts. Northwest Sciences 25:17-31. Whittaker, R. H. 1953. A consideration of climax theory: the climax as population and pattern. Ecological Monographs 23:41-78. Whittaker, R. H. 1956. Vegetation of the Great Smoky Mountains. Ecological Monographs 23:1-80.
References
409
Whittaker, R. H., L. E. Gilbert, and J. H. Connell. 1979a. Analysis of twophase pattern in a mesquite grassland, Texas. Journal of Ecology 67:935-952. Whittingham, J., and D. J. Read. 1982. Vesicular-arbuscular mycorrhiza in natural vegetation systems III. Nutrient transfer between plants with mycorrhizal interconnections. New Phytologist 90:277-284. Widen, B. 1993. Demographic and genetic effects on reproduction as related to population size in a rare, perennial herb Senecio integrifolius. Biological Journal of the Linnean Society 50:179-195. Wied, A., and C. Galen. 1998. Plant parental care: conspecific nurse effects in Frasera speciosa and Cirsium scopulorum. Ecology 79:1657-1668. Wießner, A., P. Kuschk, M. Kästner, and U. Stottmeister. 2002. Abilities of helophyte species to release oxygen into rhizospheres with varying redox conditions in laboratory-scale hydroponic systems. International Journal of Phytoremediation 4:1-15. Wilby, A., and M. Sachak. 2004. Shrubs, granivores and annual plant community stability in an arid ecosystem. Oikos 106:209-216. Williams, K., M. M. Caldwell, and J. H. Richards. 1993. The influence of shade and clouds on water potential: the buffered behavior of hydraulic lift. Plant and Soil 157:83-95. Williams, S. L. 1990. Experimental studies of Caribbean seagrass bed development. Ecological Monographs 60:449-469. Wilson, D. S. 1992. Complex interactions in metacommunities, with implications for biodiversity and higher levels of selection. Ecology 73:1984-2000. Wilson, D. S. 1997. Biological communities as functionally organized units. Ecology 78:2018-2024. Wilson, J. B. 1989. Root competition between three upland grasses. Functional Ecology 3:447-451. Wilson, J. B., and A. D. Q. Agnew. 1992. Positive-feedback switches in plant communities. Advances in Ecological Research 23:263-336. Wilson, J. B., R. K. Peet, and M. T. Sykes. 1994. What constitutes evidence of community structure? A reply to van der Maarel, Noest & Palmer. Journal of Vegetation Science 6:753-758. Wilson, K. A., and A. H. Fitter. 1984. The role of phosphorus in vegetational differentitation in a small valley mire. Journal of Ecology 72:463-473. Wilson, S. D. 1993. Competition and resource availability in heath and grassland in the Snowy Mountains of Australia. Journal of Ecology 81:445-451.
410
References
Wilson, S. D., and P. A. Keddy. 1986. Measuring diffuse competition along an environmental gradient: results from a shoreline plant community. American Naturalist 127:862-869. Wilson, S. D., and D. Tilman. 1995. Competitive responses of eight old-field plant species in four environments. Ecology 76:1169-1180. Wipf, S., C. Rixen, and C. P. H. Mulder. 2006. Advanced snowmelt causes shift towards positive neighbour interactions in a subarctic tundra community. Global Change Biology 12:1-11. Wolff, K., B. Friso, and J. M. M. van Damme. 1988. Outcrossing rates and male sterility in natural populations of Plantago coronopus. Theoretical and Applied Genetics 76:190-196. Wood, D. M., and R. del Moral. 1987. Mechanisms of early primary succession in subalpine habitats on Mount St. Helens. Ecology 68:780-790. Wood, D. M., and W. F. Morris. 1990. Ecological constraints to seedling establishment on the Pumice Plain, Mount St. Helens, Washington. American Journal of Botany 77:1411-1418. Woods, K. D., and R. H. Whittaker. 1981. Canopy-understory interaction and the internal dynamics of mature hardwood and hemlock-hardwood forests. Pages 305-323 in D. C. West, H. H. Shugart, and D. B. Botkin, editors. Forest Succession: Concepts and Applications. Springer-Verlag, New York, New York, USA. Wooton, J. T. 1994. The nature and consequences of indirect effects in ecological communities. Annual Review of Ecology and Systematics 25:443-466. Wright, A. J. 1981. The analysis of yield-density relationships in binary mixtures using inverse polynomials. Journal of Agricultural Science Cambridge 96:561-567. Yarranton, G. A., and R. G. Morrison. 1974. Spatial dynamics of a primary succession. Journal of Ecology 62:417-428. Yavitt, J. B., and E. L. J. Smith. 1988. Spatial patterns of mesquite and associated herbaceous species in an Arizona desert grassland. American Midland Naturalist 109:89-93. Yeaton, R. I. 1978. A cyclical relationship between Larrea tridentata and Opuntia leptocaulis in the northern Chihuahuan Desert. Journal of Ecology 66:651-656. Yeaton, R. I. 1988. The structure and function of the Namib Dune grasslands: characteristics of the environmental gradients and species distributions. Journal of Ecology 76:744-758. Yeaton, R. I. 1990. The structure and function of Namib Dune grasslands: species interactions. Journal of Arid Environments 18:343-349.
References
411
Yeaton, R. I., and K. J. Elser. 1990. The dynamics of a succulent karoo vegetation. Vegetatio 88:103-113. Yeaton, R. I., A. R. Manzanares, G. V. Castillo, and S. Vielegas. 1987. Tree succession in the subalpine forest of the neo-volcanic range, southcentral Mexico. Southwestern Naturalist 32:335-345. Yeaton, R. I., and A. Romero-Manzanares. 1986. Organization of vegetation mosaics in the Acacia schaffneri-Opuntia streptacantha association, southern Chihuahuan Desert, Mexico. Journal of Ecology 65:586-595. Yoder, C. K., and R. S. Nowak. 1999. Hydraulic lift among native plant species in the Mojave Desert. Plant and Soil 215:93-102. Young, D. R. 1983. Comparison of intraspecific variations in the reproduction and photosynthesis of an understory herb, Arnica cordifolia. American Journal of Botany 70:728-734. Zabinski, C. A., L. Quinn, and R. M. Callaway. 2002. Phosphorus uptake, not carbon transfer, explains arbuscular mycorrhizal enhancement of Centaurea maculosa in the presence of native grasses. Functional Ecology 16:758-765. Zahawi, R. A., and C. K. Augspurger. 1999. Early plant succession in abandoned pastures in Ecuador. Biotropica 31:540-552. Zamfir, M., and D. E. Goldberg. 2000. The effect of initial density on interactions between bryophytes at individual and community levels. Journal of Ecology 88:243-255. Zhang, F., and L. Li. 2003. Using competitive and facilitative interactions in intercropping systems enhances crop productivity and nutrient-use efficiency. Plant and Soil 248:305-312. Zincke, P. J. 1962. The pattern of influence of individual forest trees on soil properties. Ecology 43:130-133. Zou, C. B., P. W. Barnes, S. Archer, and C. R. McMurtry. 2005. Soil moisture redistribution as a mechanism of facilitation in savanna tree-shrub clusters. Oecologia 145:32-40. Zvereva, E. L., and M. V. Kozlov. 2004. Facilitative effects of top-canopy plants on four dwarf shrub species in habitats severely disturbed by pollution. Journal of Ecology 92:288-296.
INDEX
conspecifics, 12, 52, 82, 106, 129, 134, 140, 142, 144, 145, 146, 148, 154, 156, 161, 163, 170, 211, 267, 325 continuum, 222, 328, 330, 331 cushion plant, 42, 317
algae, 52, 106, 176 Allee, 108, 109, 110, 140, 318 allelopathic, 42, 84, 85, 86, 139, 154, 168, 175, 181, 265, 272, 288, 323, 324 allelopathy, 2, 144, 253, 265, 287 alpine, 8, 16, 40, 53, 73, 75, 88, 89, 90, 91, 92, 108, 156, 195, 202, 205, 207, 210, 211, 218, 236, 317 Alpine Pals, 207 AM fungi, 155, 156, 157, 158 annual communities, 110, 214, 247, 310, 311 annual plants, 62, 190, 191, 213, 226, 228, 229, 292, 293, 310 annuals, 48, 74, 106, 110, 172, 175, 190, 191, 192, 214, 215, 225, 226, 228, 229, 230, 247, 251, 260, 278, 287, 291, 292, 310, 311, 312, 313, 316 arctic, 41, 66, 90, 92, 194, 205, 208, 219 associational resistance, 119, 129, 130, 133, 136, 297
deciduous forest, 27, 85, 168, 171, 264 decoys, 138 desert, 7, 8, 28, 29, 30, 60, 72, 104, 106, 113, 140, 213, 215, 216, 220, 225, 226, 227, 228, 230, 242, 243, 247, 256, 257, 259, 260, 262, 276, 289, 291, 292, 310, 311, 313, 329, 330 diffuse competition, 201, 202 disease, 302, 325 dispersal, 3, 112, 113, 121, 151, 152, 153, 154, 282 disturbance, 11, 12, 23, 55, 88, 104, 106, 107, 108, 117, 143, 176, 194, 208, 230, 236, 251, 279, 287, 290, 295, 304, 317, 324 diversity, 5, 9, 10, 11, 12, 15, 18, 90, 99, 109, 116, 126, 130, 144, 145, 151, 160, 176, 177, 185, 202, 207, 210, 220, 222, 235, 237, 255, 256, 280, 284, 295–304, 311, 318, 319, 326, 333 drought, 9, 12, 22, 23, 27, 42, 46, 48, 50, 66, 89, 108, 123, 124, 126, 153, 154, 188, 210, 215, 216, 230, 231, 287, 289, 299, 300, 310, 311 dunes, 7, 37, 52, 104, 230, 261, 279
bryophyte, 9, 210, 262, 263, 298, 300 chaparral, 136, 287, 330 chelator, 85, 86, 272 cobble beach communities, 104 cold temperatures, 33, 35 colonization, 42, 49, 74, 155, 156, 233, 234, 241, 289 commensalism, 2 communication, 21, 113, 114, 116, 127, 129, 146, 290 complementarity, 211, 297, 301 conifers, 24, 26, 35, 54, 93, 156, 256, 262, 285 conservation, 3, 10, 319, 326, 327
ecosystem engineering, 21, 44, 103, 324 ecosystem function, 9, 276, 296–298, 302, 303 ecotone, 43, 81, 89, 103, 123, 169 413
414
Index
epiphyte, 49, 50, 83, 84, 99, 100, 176, 264, 273 evolution, 2, 135, 293, 319, 324, 325, 326
mutualism, 2, 159 mycorrhizae, 10, 56, 118, 154, 155, 157, 160, 284, 286, 293
fertility, 29, 59, 60, 63–67, 75, 80, 143, 230, 243, 256, 316 fire, 51, 114, 143, 304 fog, 22, 23, 195, 261
niche, 4, 9, 11, 16, 135, 176, 211, 286, 295–297, 300, 304, 310, 320, 324, 325 niche construction, 324, 325 nitrogen, 9, 12, 25, 29, 56, 58, 59, 61–64, 66, 68, 70, 72, 73, 75, 76, 78, 80, 82, 84, 85, 88, 97, 98, 107, 113, 131, 151, 163, 181, 185, 191, 192, 222, 236, 242, 250, 261, 271, 273, 274, 275, 276, 289, 300–302, 315–317 nitrogen fixers, 29, 276 nurse plant, 17, 18, 21, 25, 27, 29, 35, 79, 125, 236, 240, 257, 259, 260, 287 nutrient deposition, 57, 67, 78
grassland, 7, 12, 34, 38, 39, 41, 46, 47, 52, 58, 61, 63, 64, 68, 72, 73, 80, 81, 89, 91, 122, 141, 152, 169, 170, 175, 181, 182, 184, 193, 194, 222, 236, 240, 248–250, 260, 270, 276, 280, 283, 296, 302, 309, 316, 317 herbivory, 4, 18, 117, 119, 122–124, 127, 128, 130–132, 134–136, 139, 153, 176, 192, 193, 246, 251–253, 280, 287, 296, 331 heteromycotrophs, 158, 159 hydraulic lift, 18–22, 24, 45, 191, 236, 241, 277, 278, 283 indirect facilitation, 18, 118, 127, 135, 139, 146, 155, 159, 161, 171, 172, 280, 283, 284, 297 indirect interactions, 1, 10, 30, 57, 106, 114, 117, 118, 121, 130, 136, 145, 151, 164, 169, 174, 177, 280, 282, 297, 317, 330 individualistic, 1, 3, 5, 10, 118, 295, 319, 324, 327, 328, 329, 330, 332, 333 integrated community concept, 330 invasive plants, 10 leaf leachates, 99 leaf temperature, 24, 30, 40 lichen, 84, 90, 176 litter, 25, 28, 47, 48, 53, 57, 58, 63, 65, 69, 70, 71, 78, 81, 82, 100, 137, 169, 175, 181, 236, 244, 272, 275, 287, 289, 316 litterfall, 21, 56, 57, 66, 80, 275 magnet species, 147, 149, 309 meta-analysis, 196, 198, 200, 213
oaks, 38, 42, 46, 47, 50, 51, 175, 179, 248, 250, 264, 286, 312 old fields, 161, 169 oxygen, 94, 97, 98, 99, 101, 277 parasites, 113, 118, 121, 122, 325 parasitic plants, 15, 118, 158, 280, 290 parental care, 108 pathogens, 101, 109, 154, 160, 284, 286, 302, 325 peatlands, 48 photoinhibition, 33–36, 50, 263, 300 pollen, 11, 108, 113, 140–142, 144–146, 148–150 pollinator, 11, 108, 110, 117, 140, 143–145, 252, 281, 309 prairie, 42, 43, 64, 74, 110, 143, 168, 219, 223, 283, 317 predation, 2, 27, 38, 110, 123, 130, 131, 143, 152, 246, 253, 330 production efficiency, 313 productivity, 5, 9, 15, 16, 21, 39, 45, 58–64, 70, 72, 75, 78, 96, 99, 116, 139, 143, 179, 180, 182–184, 192–198, 201, 205, 210, 212, 214, 215, 217, 219, 223, 227–230, 232, 236–238, 248, 256, 260, 272, 295, 296, 298, 300, 301, 304, 305, 312–315, 323
Index rainforest, 21, 23, 77 recruitment, 32, 33, 38, 54, 70, 75, 81, 88, 89, 101, 109, 122, 127, 153, 169, 171, 231, 236, 240, 241, 244, 260, 268, 314, 319, 326 redox potentials, 235 regeneration niches, 326 resource islands, 62, 64–66, 79 riparian, 82, 106, 121, 185, 270, 271, 328 root architecture, 181, 183, 255, 277, 287 root competition, 16, 76, 82, 111, 115, 184, 186, 189, 190, 215, 223, 224 saguaro, 2, 27–30, 71, 111, 240, 257–260 salinity, 11, 55, 64, 95, 163, 189, 211, 212, 233, 235, 236, 324 salt marshes, 38, 55, 94, 101, 171, 188, 211, 213, 252, 280 savanna, 7, 30, 45, 46, 59, 62, 67, 77, 80, 180, 182, 194, 236, 246, 268, 286, 289, 307 seed bank, 11, 110–112, 152, 190, 228, 292, 293 seed shadows, 110, 150 seed-seedling conflict, 246 shade, 4, 12, 16–18, 23–25, 27–38, 40–44, 46, 48, 55, 56, 78–82, 89, 93, 114, 115, 120, 122, 123, 127, 134, 135, 153, 158, 168–170, 175, 186, 187, 192, 215, 220, 223, 224, 231, 239, 246, 248, 250, 252, 255, 256, 259, 261, 267–269, 271, 272, 282, 285, 286, 287, 289, 290, 315, 317, 324 shared defense, 118, 120, 121, 133, 280 shared doom, 136 shrubland, 34, 91, 260, 276 signaling, 113, 116, 290, 291, 293 soil biota, 1, 162, 163, 258, 269, 283, 302 soil bulk density, 47, 241 soil compaction, 47, 108, 154 soil enrichment, 58, 64 soil fungi, 154, 157, 160 soil microbes, 117, 118, 160, 162, 164, 283, 284, 286 soil moisture, 21–25, 30–32, 38, 42, 45–48, 50–54, 64, 72, 76, 78, 88,
415 108, 127, 151, 153, 154, 163, 168, 184–186, 189, 265, 277, 288, 300 soil oxygenation, 94, 97, 99 species-specific, 5, 9, 11, 12, 28, 29, 44, 83–85, 176, 229, 255, 256, 258, 260, 262, 263, 265, 267–269, 272, 274, 275, 277–285, 288, 290–293, 297, 303, 310, 311, 315, 332 steppe, 21, 30, 31, 54, 111, 126, 169, 244, 276, 315 stress, 4, 11, 16, 18, 23, 25, 28, 37, 49, 52, 55, 66, 89, 90, 93, 103, 183, 192–196, 198–202, 205, 207, 209–217, 219–221, 223, 225, 227–229, 231–237, 241, 246, 250, 253, 254, 289, 298, 300, 305, 311, 319 stress gradient hypothesis, 198, 201, 213 stress gradients, 196, 219 subalpine meadow, 40 substrate, 17, 44, 47, 52, 65, 66, 75, 85, 99–106, 137, 185, 233, 251, 272, 297 succession, 12, 23, 74, 82, 100, 102, 150, 154, 232, 235, 267, 314, 318, 324 temperature, 1, 16, 24, 25, 29, 30, 33, 35, 36, 41, 42, 45, 46, 52, 66, 72, 78, 83, 88, 89, 102, 115, 123, 151, 196, 205, 210, 211, 219, 220, 236, 258, 273, 288, 317 throughfall, 16, 21, 45, 57, 58, 65, 66, 77, 80, 83, 84, 139, 181, 255, 264, 273–275 timberline, 35, 54, 89, 91, 93, 232, 314 trampling, 108 transpiration, 20, 25, 36, 38, 45, 46, 88, 278, 300 tree islands, 33, 54, 88 tropical, 21, 52, 77, 84, 99, 103, 243, 244, 261, 264 tundra, 53, 66, 76, 88, 90, 272 water-use-efficiency, 41, 262 wetland, 94, 102, 277 wind, 4, 11, 42, 48, 53, 88–93, 104, 108, 110, 111, 151, 152, 154, 169, 205, 207, 208, 236 woodlands, 18, 38, 45, 47, 59, 62, 66, 80, 100, 135, 175, 222, 240, 268, 286, 312