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!
(λ) ml–1
50
25
107
10
n = 332
LL 106
8
20
6
15
4 2
10
0 0
5
1
2
3
105 104 103 102 101 100
Particles,
(λ) ml–1
0
0 25
n = 260
10
20
MY
8
107 106 105 104 103 102 101 100
6
15
4
10
2 0
5
0
1
2
Particles,
(λ) ml–1
Mean number of species, <S > (λ) m–2
156
3
0 0
5
10 15 20 Mean body-length, λ (mm)
25
30
Figure 8.4 Species- and particle-size frequency distribution in three stream ecosystems (Seebach (SB), Llwch (LL) and Mynach (MY)). Mean number of species per 1 m2 given over 300 sizeclasses in increments of 100 mm for the total size range 30 mm. The figure insets show the most abundant species over 300 size classes in increments of 10 mm for body sizes 3 mm. Mean particle density per 1 ml over 600 size classes in increments of 5 mm is shown superimposed in the inset as open circles (SB), triangles (LL) and squares (MY). n is the number of species in each stream community.
emphasizes the distinctness for multifractality of the BSDs in all streams (Fig. 8.6a). This result strongly contrasts to spectra, which would collapse into a narrow range of or even into a single value for a strictly monofractal, homogeneous size distribution with a single scaling exponent. In the three streams, the multifractal spectra were characterized by similar entropy dimensions, (1) ¼ D1 with values close to 1, which predicts that the irregularities of the species-frequency distribution are even across body sizes (Figs. 8.4 & 8.6a; Table 8.3). The fraction of small-sized particles from interstitial sediments may serve as a potential food and habitat source for the benthic communities. The particle-size distribution displayed a decline in the mean number of particles with increasing particle size (inset Fig. 8.4). Scaling irregularities increased with particle size and were particularly pronounced in the SB (inset Fig. 8.4). The interstitial particle
BODY SIZE AND SCALE INVARIANCE
Table 8.3 Summary of parameters obtained by multifractal analysis of body- and particle-size distributions in the streams Seebach (SB), Llwch (LL) and Mynach (MY). a(1), a(0) and a(1) are the Lipschitz-Ho¨lder exponents with 1 SE for q ¼ 1, q ¼ 0 and q ¼ 1 respectively. r2 are the coefficients of variation of the linear regression of log partition function on log size for each a(q). Stream
SB LL MY
Body-size distribution ( 1)
r2
(0)
r2
(1)
r2
1.039 0.078 1.131 0.024 1.075 0.011
0.998 0.999 0.999
1.012 0.027 1.044 0.003 1.030 0.001
0.997 0.999 0.999
0.882 0.002 0.949 0.001 0.968 0.001
0.998 0.999 0.999
Stream
SB LL MY
Particle-size distribution 2
( 1)
r
1.389 0.096 1.244 0.047 1.289 0.033
0.995 0.996 0.999
40
(0)
r2
(1)
r2
1.138 0.004 1.087 0.002 1.098 0.002
0.999 0.999 0.999
0.792 0.003 0.842 0.003 0.767 0.002
0.996 0.996 0.997
SB
30 20 10 0
Log partition function, log µq(λ)
–10 –20 40 30
LL
q = –10
20 10 0 –10 –20
q = 10
40 MY
30 20 10 0 –10 –20
–1.0 –0.8 –0.6 –0.4 –0.2 0.0 0.2 0.4 0.6 0.8 Log body-size, λ (mm)
Figure 8.5 Log partition function, log q(l), as a function of log body-size (l) for q-moments in the range between 10 and 10 at 0.5 lag increments for the streams SB, LL and MY. Linear regressions are superimposed for each q-moment (for all r2 0.95; P < 0.001). Significant linear relationships are similar for particle-size distributions not shown here.
157
Multifractal spectrum, f (α)
P. E. SCHMID AND J. M. SCHMID-ARAYA
(a)
(b)
1.0
1.0
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
SB LL MY
0.0 0.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Lipschitz-Hölder exponent, α
Figure 8.6 Multifractal spectra for the (a) species- and (b) particle-size distributions in the streams SB, LL and MY.
distribution displayed multifractal properties (over all 10 q 10; r2 > 0.99), with wider f()-spectra than those for the species-size distributions (Fig. 8.6b). The values of the entropy dimension, (1) ¼ D1, differed from 1 (Table 8.3), which implies that particle-size irregularities are unevenly distributed and concentrate in several subintervals, which was particularly evident for the SB and MY (Figs. 8.4 & 8.6b). The various scaling domains of interstitial particles are similar to those from different soil PSDs (Posadas et al., 2001). To test the similarity of the multifractal patterns between organisms and particles the Re´nyi exponents, q, of the BSD were plotted against the q of the PSD for each stream (Fig. 8.7). All stream ecosystems were characterized by two linear sections of the q relationships, where the larger section defined trends for q < 0 and the shorter, steeper section of the curve defined trends for q > 0. The well-defined difference between the two curve sections strengthens the finding of multifractal patterns governing both BSD and PSDs in streams. Particularly striking is the similarity between BSD and PSDs scaling exponents in the first-order LL as shown by the closeness of the curve to a slope of unity (Fig. 8.7). Moreover, the relationship between the Re´nyi exponents of the BSD 10
Rényi exponent of body-size distribution, τq
158
5 0 –5 –10 –15
SB LL MY
–20 –20 –15 –10 –5 0 5 Rényi exponent of particle-size distribution, τq
Figure 8.7 The Re´nyi exponents, (q), of the body-size versus (q) of particle-size distribution in the streams SB, LL and MY. The dotted line gives a slope of unity.
BODY SIZE AND SCALE INVARIANCE
and PSD are similar between the Oberer Seebach (SB) and Afon Mynach (MY) for larger species and particle sizes. A distinct departure occurred between the two streams for smaller, more common species and particles. Further evidence for a close link between habitat- and body-size is given by recent experiments conducted in a Japanese mountain stream. This study revealed that body sizes of invertebrates tended to decline with decreasing crevice size, thus, strongly underlining the effect of fractal habitat complexity on body-size distribution (Taniguchi & Tokeshi, 2004; see also Townsend & Thompson, this volume).
Multifractal species-area relationships The species-area relationship (SAR) is one of the most widely studied patterns in community ecology and is often expressed as a single power-law curve. However, the SAR should not be characterized by a single curve that assigns one species richness value to each sampled area, as different samples of equal area differ in species number. Therefore, the SAR should be given as the relationship between area and its mean number of species, <S>. <S> was calculated for non-overlapping but equally shaped sampling areas taken over the sampling period. In order to estimate the mean number of species in an increasing area of size l ¼ k2 (where k ¼ 1, 2, 3, . . . ! l maximum area), 106 Monte-Carlo randomizations without replications of the data matrix were conducted for each k-set. By randomly pooling samples of the observed data sets taken at arbitrary sampling positions and/or occasions, it was possible to obtain samples incorporating invertebrate species over different developmental stages, body-size ranges and densities. The species-area curves were obtained by averaging species richness of those randomly pooled sample sites for each area of size l. Some authors have suggested that the SAR following a power-law curve holds at all spatial scales and implies self-similarity in the distribution and abundance of species (e.g. Harte et al., 1999). A problem with this monofractal assumption is that scale invariance of SARs is not maintained over all spatial scales, but that there are different scaling domains underlying SARs over different spatial scales. Thus, irregularities in the species-abundance distribution may be statistically the same across areas resulting in a spectrum of fractal subsets. To test for different scaling domains underlying the SARs of the stream communities, the method of Re´nyi’s generalized dimension (Dq) was used for the speciesabundance data. This method follows Eqs. (8.7) to (8.11) outlined in Fractal properties of size-structured communities (see above), where in the current context, pi is defined as the relative abundance of species i in an area l. The relationships between mean species richness and sampled area is displayed as a power-law function for each of the three streams (Fig. 8.8; Table 8.4). The mean values for the parameter z (Table 8.4) differed significantly between stream communities (ANCOVA: F2,41 ¼ 54.82, P < 0.001) due to a higher z-value in the LL. No differences in slopes were found between the SB and MY
159
P. E. SCHMID AND J. M. SCHMID-ARAYA
Table 8.4 Regression parameters of the power-law relationship between the mean number of species and sampled areas of size l ¼ k2 in the streams Seebach (SB), Llwch (LL) and Mynach (MY). c gives the intercept, z the slope, r2 is the variance explained by the correlation of area with mean number of species per area, F is the value of the variance ratio test, df are the degrees of freedom, P-values are given for the F-test. Bootstrap confidence limits (95%) are given in parentheses for c and z values. Stream
c (95% CL)
z (95% CL)
r2
SB
134.50 (107.90, 205.12) 34.36 (28.64, 56.89) 92.05 (87.30, 96.16)
0.233 (0.140, 0.282) 0.448 (0.338, 0.498) 0.200 (0.190, 0.214)
0.903
LL MY
Mean number of species, <S > (λ)
160
F
df
P
139.46
1,15
<0.001
0.968
395.06
1,13
<0.001
0.992
1694.93
1,13
<0.001
1000
100
10
SB LL MY 1 1
10 100 Area, λ (m–2)
1000
Figure 8.8 Species-area relationship (SAR) of all benthic invertebrate species for three geographically separate stream ecosystem, the SB, LL and MY. The plot shows the linear regression of log mean number of species on log area, in l ¼ k2 intervals (see details in text). The regression equations are given in Table 8.4.
community (ANCOVA: F1,28 ¼ 2.49, P ¼ 0.126). A higher value of z for the SAR of the LL is attributed to a rapid increase in species richness with increasing area, with a ten-fold increase in species number over a spatial scale encompassing an order of magnitude (Fig. 8.8). In contrast, over the same spatial scale species richness increased four-fold in the SB but only two-fold in the MY. Moreover, a wider range of the scale-dependent coefficient c for the SB and LL invertebrate communities implies higher sample-to-sample variations in species richness at small spatial scales compared to the MY. Empirical observations suggested that the parameter z of the SAR is about 1/4 for many ecosystems (Rosenzweig, 1995). The z-values of the SAR differed from a 1/4 assumption for the LL (t ¼ 13.45; df ¼ 13, P < 0.001) and MY (t ¼ 9.46, df ¼ 13, P < 0.001) but were close to 1/4 for the SB community (t ¼ 0.49, df ¼ 15, P ¼ 0.628). The results indicate that z-values, and therefore SARs, do not display a common scale-invariant behaviour across stream systems.
BODY SIZE AND SCALE INVARIANCE
80 60 40 20 0
SB
log Partition function, log μ q(λ)
–4 –8 1
10
100
80 60 40 20 0
1000 LL
–4 –8 1
10
100
100 80 60 40 20 0
1000 MY
–4 –8 1
10 100 Area, λ (m2)
1000
Figure 8.9 Log partition function, log q (l), as a function of log area (l) for q-moments in the range between 10 and 10 at 0.5 lag increments given for the streams Seebach (SB), Llwch (LL) and Mynach (LL). For each stream the upper panel gives the partition function for q < 0 and the lower panel for q > 0. Linear regressions are superimposed for each q-moment (all q < 1 and q > 2 with r2 > 0.95; P < 0.001).
Rényi’s generalised dimension, Dq
All streams are characterized by a class of multifractal distributions, as the test criterion of linearity between the log-partition function and log area is fulfilled, particularly for all q-moment <2 (Fig. 8.9). The spectra of the generalized Re´nyi dimensions for q-moments are a monotonically decreasing function for all q-moments < 2 in each stream community (Fig. 8.10). With q-moments >2, Dq increases slightly in all ecosystems (Fig. 8.10) while the linearity criterion 1.0
SB LL MY
0.8 0.6 0.4 0.2 0.0 –10
–5
0
Moment, q
5
10
Figure 8.10 Re´nyi’s generalized dimensions (Dq) versus q-moments for the SAR of all benthic invertebrate species in three streams, the Seebach (SB), Llwch (LL) and Mynach (MY). Parameters are given in Table 8.5.
161
162
P. E. SCHMID AND J. M. SCHMID-ARAYA
Table 8.5 Summary of parameters obtained by multifractal analysis of species-abundance distributions across different areas in the streams Seebach (SB), Llwch (LL) and Mynach (MY). Dq are Re´nyi’s generalized dimensions given with 1SE; D0 is the capacity dimension, which equals to the slope of the species-area relationship; D1 is the entropy dimension; D2 is the correlation dimension; r2 are the coefficient of variation of the relationship between log partition function and log area for each Dq. Stream
SB LL MY
Species abundance across areas 2
D0
r
D1
r2
D2
r2
0.233 0.025 0.448 0.031 0.200 0.024
0.959 0.973 0.968
0.126 0.001 0.241 0.003 0.109 0.002
0.833 0.901 0.921
0.105 0.013 0.116 0.021 0.080 0.009
0.868 0.894 0.925
between the log-partition function and log area remains roughly fulfilled (Fig. 8.9; r2 > 0.95; P < 0.001). Excluding slight methodological differences, these results resemble those obtained for the species-abundance distribution of a tropical forest in Panama by Borda-de-A´gua et al. (2002), where Dq was not defined for q > 1. This more general class of multifractal distributions, for which not all q-moments exist, also indicates certain scale-related limitations in the application of Re´nyi’s generalized dimension for species-abundance distributions. Nevertheless, if a single fractal dimension does characterize the SAR then values of D0, D1 and D2 would be all equal. This equality would only be attained if the densities are equally represented among species. The capacity dimension, D0, is equivalent to the value for the single fractal dimension and to the exponent z of the SAR, which differs between the three streams (Table 8.5). Comparatively low values of the capacity dimensions indicate that irregularities in the abundance distribution are more concentrated in samples from smaller areas, emphasizing small-scale patchiness in streams. An entropy dimension, D1, with a value close to 1 signifies an evenly spread distribution of irregularities in species-abundance patterns across different scales. However, D1 values varied among the three stream communities, ranging from 0.11 in the MY, 0.13 in the SB, to 0.24 in the LL (Table 8.5). Therefore, higher values of D1 in the LL implied a greater evenness in the species-abundance distribution contrasting with the other stream communities. Similarly, the low values of the correlation dimension (D2) between communities illustrated uncorrelated irregularities of speciesabundance patterns within the same spatial scale range. Following Borda-de-A´gua et al. (2002), results of the generalized dimension strengthen the positive relationship between the density and spatial range of species. Considering the extremes of the distribution (e.g. q ! 10; q ! þ10), that respectively correspond to the rarer but larger and the abundant but
BODY SIZE AND SCALE INVARIANCE
smaller species, the density of the larger species declines faster with area than the abundance of smaller species. The reason for this observation is that rare species have a smaller geographic range than abundant, mostly small-sized, organisms (Schmid et al., 2002). Small aquatic invertebrates such as testate amoebae or benthic rotifers have a much wider geographic distribution than larger species, and free-living microbial eukaryotes are probably sufficiently abundant to have a global distribution (sensu Finlay, 2002; see also Finlay & Esteban, this volume).
Conclusions During the last few decades, evidence for scale invariance has appeared in several fields other than physics, and ecology is no exception. All these rather recent findings on power-law relationships and scaling in ecology, ranging from species-biomass invariance to body-size mediated patterns of community structure, are suggestive of an underlying scale-free pattern of complex systems. Biomass invariance across sites and stream systems may reflect the relative balance between resource supply and use, mediated and maintained through size-dependent resource partitioning processes of communities (Schmid et al., unpublished). The fit of species-abundance patterns to random resourceassembly, but not to neutral models, emphasizes the link between resource dependence, body size and species composition, irrespective of trophic group. This also stresses that differences between species in response to variations in resource supply and environmental heterogeneity question the core assumption of demographically identical species under Hubbell’s (2001) neutral concept. It is also evident that the dispersal probability of individuals plays a limited role in shaping patterns of observed benthic community structure at local scales. Moreover, food in the form of biofilm-enriched particles is not a limiting factor in most temperate stream ecosystems (Schmid et al., 2000), while habitat complexity has a profound influence on species composition and size structure (e.g. Jeffries, 1993; Schmid et al., 2002; Taniguchi & Tokeshi, 2004). In benthic stream systems, resource complexity is the product of physical and biological processes acting on the shape and size distribution of organic particles (Schmid et al., unpublished). The results presented here substantiate the view that the dynamics of the interstitial habitat and the various scaling domains of particles, which serve both as food and habitat, influence the size distribution of invertebrates and therefore, diversity and species composition in stream communities. Fractals are certainly the simplest method we have for quantifying a measure across a range of scales. As such, they provide the best null model against which to judge the real behaviour of natural patterns across different spatial scales, just as stochasticity is the null model against which to test spatial patterns at a single scale. Moreover, measures on aggregates such as (a) species-abundance distributions which influence SARs, (b) population densities underlying BSDs, or (c) particle densities underlying PSDs, give rise to a very different type of scaling
163
164
P. E. SCHMID AND J. M. SCHMID-ARAYA
known as multifractal. Thus, through the examination of multifractal measures, subtle differences in spatial and size structure can be explored. Fine-scale patterns embedded within coarser patterns apparently reflect ecological processes operating across a wide size spectrum and spatial scales, such as the resource utilization among invertebrate species in streams. The fact that multifractals can mirror compound phenomena may prove increasingly useful in the characterization, modelling and understanding of complex phenomena in ecology.
Acknowledgements This work has been supported by grants from NERC (NER/A/S/2001/00566), the Royal Society, and in parts from the Austrian Science Fund FWF: P15597-B03. We thank Dr Luı´s Borda-de-A´gua and an anonymous reviewer for valuable comments and suggestions on an early draft of this chapter.
References Ankersmid, B. V. (2002). Particle size of certified standards using the time-of-transition method. Ankersmid Application Note, 1.1, 1. Bittelli, M., Campbell, G. S. & Flury, M. (1999). Characterization of particle-size distribution in soils with a fragmentation model. Soil Science Society America Journal, 63, 782–788. Borda-de-A´qua, L., Hubbell, S. P. & McAllister, M. (2002). Species-area curves, diversity indices, and species abundance distributions: a multifractal analysis. The American Naturalist, 159, 138–155. Brown, J. H. (1995). Macroecology. Chicago: University of Chicago Press. Brown, J. H. & West, G. B. (2000). Scaling in Biology. Santa Fe Institute Studies in the Science of Complexity. Oxford: Oxford University Press. Brown, J. H., Gupta, V. K., Li, B.-L. et al. (2002). The fractal nature of nature: power laws, ecological complexity and biodiversity. Philosophical Transactions of the Royal Society London B, 357, 619–626. Brown, J. H., Gillooly, J. F., Allan, A. P., Savage, V. M. & West, G. B. (2004). Towards a metabolic theory of ecology. Ecology, 85, 1771–1789.
Calder, W. A. III (1996). Size, Function, and Life History. Mineola, New York: Dover Publications, Inc. Chhabra, A. & Jensen, R. V. (1989). Direct determination of the f(a) singularity spectrum. Physical Review Letters, 62, 1327–1330. De Bartolo, S. D., Gabriele, S. & Gaudio, R. (2000). Multifractal behaviour of river networks. Hydrology and Earth System Sciences, 4, 105–112. Dodds, P. S., Rothman, D. H. & Weitz, J. S. (2001). Re-examination of the ‘3/4 law’ of metabolism. Journal of Theoretical Biology, 209, 9–27. Dornelas, M., Connolly, S. R. & Hughes, T. P. (2006). Coral reef diversity refutes the neutral theory of biodiversity. Nature, 440, 80–82. Drake, J. B. & Weishampel, J. F. (2000). Multifractal analysis of canopy height measures in a longleaf pine savanna. Forest Ecology and Management, 128, 121–127. Enquist, B. J. & Niklas, K. J. (2001). Invariant sclaing relations across tree-dominated communities. Nature, 410, 655–660. Etienne, R. S. (2005). A new sampling formula for neutral biodiversity. Ecology Letters, 8, 253–260. Evertsz, C. J. G. & Mandelbrot, B. B. (1992). Multifractal measures. In Chaos and Fractals. New Frontiers of Science, ed. H. Peitgen,
BODY SIZE AND SCALE INVARIANCE
H. Ju¨rgens and D. Saupe. New York: Springer Verlag, pp. 921–953. Feder, H. (1988). Fractals. New York: Plenum Press. Fesl, C. (2002). Niche-oriented speciesabundance models: different approaches of their application to larval chironomid (Diptera) assemblages in a large river. Journal of Animal Ecology, 71, 1085–1094. Finlay, B. J. (2002). Global dispersal of free-living microbial eukaryote species. Science, 296, 1061–1063. Gisiger, T. (2001). Scale invariance in biology: coincidence or footprint of a universal mechanism? Biological Review, 76, 161–209. Halsey, T. C., Jensen, M. H., Kadanoff, L. P., Procaccia, I. & Shraiman, B. I. (1986). Fractal measures and their singularities: the characterization of strange sets. Physical Review A, 33, 1141–1151. Harte, J., Kinzig, A. & Green, J. (1999). Self-similarity in the distribution and abundance of species. Science, 284, 334–336. Hentschel, H. & Procaccia, I. (1983). The infinite number of generalized dimensions of fractal and strange attractors. Physica D, 8, 435–444. Hubbell, S. P. (2001). A Unified Neutral Theory of Biodiversity and Biogeography. Princeton, NJ: Princeton University Press. Jeffries, M. (1993). Invertebrate colonization of artificial pond weeds of differing fractal dimension. Oikos, 67, 142–148. Johnson, G. D., Tempelman, A. & Patil, G. P. (1995). Fractal based methods in ecology: a review for analysis at multiple spatial scales. Coenosis, 10, 123–131. Keylock, C. J. (2005). Simpson diversity and the Shannon-Wiener index as special cases of a generalized entropy. Oikos, 101, 205–207. Kolmogorov, A. N. (1959). Entropy per unit time as a metric invariant of automorphisms. Mathemathical Review, 21, 2035. Kravchenko, A. N., Boast, C. W. & Bullock, D. G. (1999). Multifractal analysis of soil spatial variability. Agronomy Journal, 91, 1033–1041.
Kropp, J., von Bloh, W., Block, A., Klenke, Th. & Schellnhuber, H.-J. (1994). Characteristic multifractal element distributions in recent bioactive marine sediments. In Fractals and Dynamic Systems in Geosciences, ed. J. H. Kruhl. Berlin: Springer, pp. 369–375. Kunin, W. E. (1998). Extrapolating species abundance across spatial scales. Science, 281, 1513–1515. Lennon, J. J., Kunin, W. E. & Hartley, S. (2002). Fractal species distributions do not produce power-law species-area relationships. Oikos, 97, 378–386. Mandelbrot, B. B. (1974). Intermittent turbulence in self similar cascades: divergence of high moments and dimension of the carrier. Journal of Fluid Mechanics, 62, 331–358. Mandelbrot, B. B. (1989). Multifractal measures, especially for the geophysicist. Pure Applied Geophysics, 131, 5–42. Manrubia, S. C. & Sole´, R. V. (1996). Selforganized criticality in rainforest dynamics. Chaos, Solutions and Fractals, 7, 523–541. Margalef, R. (1996). Information and uncertainty in living systems, a view from ecology. BioSystems, 38, 141–146. Margalef, R. (1997). Our Biosphere, ed. O. Kinne. Oldendorf/Luhe, Germany: Ecology Institute. Marquet, P. A., Quin˜ones, R. A., Abades S. et al. (2005). Scaling and power-laws in ecological systems. The Journal of Experimental Biology, 208, 1749–1769. McGill, B. J. (2003). A test of the unified neutral theory of biodiversity. Nature, 422, 881–884. Milne, B. T. (1998). Motivation and beliefs of complex system approaches in ecology. Ecosystems, 1, 449–456. Nee, S., Read, A. F., Greenwood, J. J. D. & Harvey, P. H. (1991). The relationship between abundance and body size in British birds. Nature, 351, 312–313. Pascual, M., Ascioti, F. A. & Caswell, H. (1995). Intermittency in the plankton: a multifractal analysis of zooplankton biomass
165
166
P. E. SCHMID AND J. M. SCHMID-ARAYA
variability. Journal of Plankton Research, 17, 1209–1232. Peters, R. H. (1983). The Ecological Implications of Body Size. Cambridge: Cambridge University Press. Posadas, A. N. D., Gime´nez, D., Bittelli, M., Vaz, C. M. P. & Flury, M. (2001). Multifractal characterization of soil particle-size distributions. Soil Science Society America Journal, 65, 1361–1367. Re´nyi, A. (1955). On a new axiomatic theory of probability. Acta Mathematica Hungarica, 6, 285–335. Rodriguez-Iturbe, I. & Rinaldo, A. (2001). Fractal River Basins. Chance and SelfOrganisation. Cambridge, UK: Cambridge University Press. Rosenzweig, M. L. (1995). Species Diversity in Space and Time. Cambridge, UK: Cambridge University Press. Schmid, P. E. (1997). Stochasticity in resource utilization by a larval Chironomidae (Diptera) community in the bed sediments of a gravel stream. In Groundwater/ Surfacewater Ecotones: Biological and Hydrological Interactions and Management Options, ed. J. Gilbert, J. Mathieu & F. Fournier. Cambridge: Cambridge University Press, pp. 21–29. Schmid, P. E. (2000). The fractal properties of habitat and patch structure in benthic ecosystems. Advances in Ecological Research, 30, 339–401. Schmid, P. E. & Schmid-Araya, J. M. (1997). Predation on meiobenthic assemblages: resource use of a tanypod guild (Chironomidae, Diptera) in a gravel stream. Freshwater Biology, 38, 67–91. Schmid, P. E. & Schmid-Araya, J. M. (2002). Trophic relationships in temporary and permanent freshwater meiofauna. In Freshwater Meiofauna: Biology and Ecology, ed. S. D. Rundle, A. L. Robertson and J. M. Schmid-Araya. Leiden, The
Netherlands: Backhuys Publisher, pp. 295–319. Schmid, P. E., Tokeshi, M. & Schmid-Araya, J. M. (2000). Relation between population density and body size in stream communities. Science, 289, 1557–1560. Schmid, P. E., Tokeshi, M. & Schmid-Araya, J. M. (2002). Scaling in stream communities. Proceedings of the Royal Society London B, 269, 2587–2594. Schmitt, F. G. & Seuront, L. (2001). Multifractal random walk in copepod behavior. Physica A, 301, 375–396. Schroeder, M. (1991). Fractals, Chaos, Power Laws. Minutes from an Infinite Paradise. New York: W. H. Freeman and Company. Sugihara, G. (1980). Minimal community structure: an explanation of species abundance patterns. The American Naturalist, 116, 770–787. Stanley, H. E., Amaral, L. A. N., Gopikrishnan, P. et al. (2000). Scale invariance and universality: organising principles in complex systems. Physica A, 281, 60–68. Taniguchi, H. & Tokeshi, M. (2004). Effects of habitat complexity on benthic assemblages in a variable environment. Freshwater Biology, 49, 1164–1178. Tokeshi, M (1993). Species abundance patterns and community structure. Advances in Ecological Research, 24, 111–186. Tokeshi, M. (1996). Power fraction: a new explanation of relative abundance patterns in species-rich assemblages. Oikos, 75, 543–550. Tokeshi, M. (1999). Species Coexistence. Ecological and Evolutionary Perspectives. Oxford: Blackwell Science. Turcotte, D. L. (1986). Fractals and fragmentation. Journal of Geophysical Research, 91, 1921–1926. West, G. B., Brown, J. H. & Enquist, B. J. (1999). The fourth dimension of life: fractal geometry and allometric scaling of organisms. Science, 284, 1677–1679.
CHAPTER NINE
Body size and biogeography B . J . FINLAY Natural Environment Research Council, UK
G . F . ESTEBAN Natural Environment Research Council, UK
Introduction June 9th, having received early in the morning some rain-water in a dish . . . and exposed it to the air about the third story of my house . . . I did not think I should then perceive any living creatures therein; yet viewing it attentively, I did, with admiration, observe a thousand of them in one drop of water, which were the smallest sort that I had seen hitherto. (From a letter written in 1676 by Antonie van Leeuwenhoek, who had a passion for designing and building ‘magnifying glasses’.)
Leeuwenhoek was almost certainly the first person to see protozoa and other microfauna, and the first to record their huge population sizes. He could not explain how the microbes got into the ‘dish of rainwater’, and this rather disappointing level of understanding has not changed much in 300 years. The debate is rather polarized. On one hand are those who draw attention to the possibility that a significant proportion of free-living microbial species may be geographically restricted (examples include Mann & Droop, 1996; Foissner, 1999). On the other hand are those who recognize that this is at odds with the alternative hypothesis that neutral dispersal of small organisms is driven by extraordinarily large numbers of the individuals themselves. Although the probability of an individual microbe being transported over great distance is vanishingly small, a multitude of interacting processes and events, both common and rare (hurricanes, transport in wet fur and feathers, etc.), in combination with huge population sizes, are expected to drive large-scale dispersal across all spatial scales. A fitting analogy is the purchase of lottery tickets – buying many tickets increases the probability of winning. Similarly, if a speciespopulation is big enough, some individuals will, for purely statistical reasons, be transported over great distances. Over time, therefore, the dispersal of microbes may be essentially random, with the rate and scale of dispersal determined mainly by global population size (Finlay, Monaghan & Maberly, 2002). This view is supported by the discovery of ‘signatures’ of randomness (Fig. 9.1) in the spatial distribution of soil protozoan species, and by the observation that local and global abundances of a wide range Body Size: The Structure and Function of Aquatic Ecosystems, eds. Alan G. Hildrew, David G. Raffaelli and Ronni Edmonds-Brown. Published by Cambridge University Press. # British Ecological Society 2007.
168
B. J. FINLAY AND G. F. ESTEBAN
Figure 9.1 Random distribution of soil protozoa – the spatial distributions of rare and abundant species of testate amoebae in a grassland soil. Each line is the regression through six data points for each species. For clarity only the six most abundant species and the six rarest species are included. The bold line represents the equality of variance and mean. The rare species may simply have been transported randomly to the site but unable to produce a population in this particular soil habitat type. Further details are in Finlay et al. (2001).
Figure 9.2 The local and global abundances of individuals in soil protozoan species are significantly correlated (p < 0.001). This analysis is based on frequency of detection (a surrogate measure of abundance per species) of 95 ciliated protozoan species in 150 soil samples from a onehectare upland grassland in Scotland, and of the same species in 606 soil samples collected worldwide. Adapted from Finlay et al. (2001).
of microbial species are correlated (Fig. 9.2). Unlike most macroscopic organisms, whose dispersal and spatial distribution are determined by historical factors such as continental drift, or physical barriers such as mountain ranges, these probably have little influence on the geographical distribution of microbes. The microbial species that thrive in a particular habitat are probably the result of habitat properties alone and, with their capacity for ubiquitous dispersal, we can assume that they thrive wherever their preferred habitat type is realized on Earth.
BODY SIZE AND BIOGEOGRAPHY
Table 9.1 The two provinces of biodiversity – adapted from Finlay, Esteban and Fenchel (2004).
Absolute abundance Rates of dispersal Rates of allopatric speciation Rates of species extinction Relative number of endemics Global number of morphospecies Proportion of global species pool found locally Cryptic persistence of species Persistence of specific morphotypes over geological time Large-scale distribution determined by historical factors e.g. continental drift
Macro-organisms
Micro-organisms
Low Low High High High High Low Variable Low
High High Low Low Low/None Low High High High
High
Low
One of the more awkward and contentious problems for supporters of the ‘restricted distribution’ of microbial species has been the difficulty in establishing the existence of microbial ‘endemics’. This problem may be insurmountable. We can be confident in the knowledge that a particular large mammal species does not exist outside of its ‘endemic range’, but it is probably impossible to demonstrate that a microbial species does not exist elsewhere in the biosphere. Moreover, if the geographical ranges of microbial ‘endemics’ cannot be established, the term ‘endemic microbial species’ is probably meaningless. The fundamental characteristics of biodiversity at the microbial and macroscopic levels (Table 9.1) are markedly different – indeed they appear to represent two distinctly different ‘provinces’ of biodiversity. That most macroscopic organisms have biogeographies is well established, but the case for ubiquitous neutral dispersal and cosmopolitan distribution driven by population abundance in free-living organisms <1 mm has not yet received unanimous support.
The absolute abundance of microbial species-populations Various independent studies confirm that the abundance of individuals within morphospecies is inversely related to body size, and that microbial species therefore contain extremely large numbers of individuals (see for example Damuth 1981; Schmid, Tokeshi & Schmid-Araya, 2000; Finlay, 2002). This is readily confirmed with empirical evidence from any small ecosystem.
169
170
B. J. FINLAY AND G. F. ESTEBAN
Figure 9.3 Body size and areal abundance of free-living organisms, adapted from Finlay (2002), where more detailed information is available. In the microbial eukaryotes, each data point represents a species. The ‘mammals’ data are included only to establish the existence of a consistent pattern stretching over 12 orders of magnitude in abundance.
A one-hectare freshwater pond will support something in the region of 1018 bacteria, 1016 protists and 1011 small animals (Finlay & Maberly, 2000). Most of the organisms that Leeuwenhoek observed were protozoa, microalgae, and a diversity of mixotrophic protists that mostly lie in the size range 2–200 mm. Flagellates are very small (mostly 2–5 mm, and almost all <20 mm). Most amoebae are 5–50 mm, and most ciliates are 15–200 mm. Thus the protozoa, together with meiofauna such as rotifers, tardigrades and gastrotrichs (almost all <1 mm), have body sizes roughly 100 times bigger than the bacteria on which many of them feed. Compared with macroscopic animals, protozoa are extremely abundant: one gram of soil typically contains around 15 000 naked amoebae (Finlay et al., 2000), and every millilitre of fresh- or seawater on the planet supports 102 to 106 heterotrophic flagellates (Berninger, Finlay & Kuuppo-Leinikki, 1991). The absolute abundance of organisms in any aquatic habitat is typically astronomical (Finlay, 2002) – indeed the areal abundance of individuals is more than, say, ten orders of magnitude greater than that of average-sized mammal species (Fig. 9.3). At the global scale, the number of individuals of a typical protozoan species is virtually incalculable. One consequence of great abundance is that the probability of dispersal is significantly elevated. An individual protozoon may not survive transport through groundwaters, or attached to a duck’s foot, or carried across the Atlantic in a harmattan dust cloud. However, population sizes with astronomical dimensions must increase the probability that geographical barriers will be crossed in the course of time, leaving few if any absolute barriers to (free-living) microbial dispersal.
BODY SIZE AND BIOGEOGRAPHY
The ubiquitous dispersal of microbes initiates a train of patterns and phenomena, such as low rates of allopatric speciation, species extinction and endemism, as described below.
The cosmopolitan–biogeography transition Over the past 200 years or so, luminaries such as Darwin, Ehrenberg and Beijerinck formed an opinion based largely on anecdotal evidence that small species had cosmopolitan distribution and large species did not. Until quite recently, the body-size range of the transition zone remained unknown, although Lawton (1998) surmised that it might be in the region of 1 mm. The largest (and least abundant) macroscopic organisms may not be cosmopolitan, but the smallest (and most abundant), ranging from bacteria (Glo¨ckner et al., 2000; Hagstro¨m, Pinhassi & Zweifel, 2000; Massana, DeLong & Pedro´s-Alio´, 2000) to protists (Fenchel, Esteban & Finlay, 1997; Finlay & Clarke, 1999a, b; Darling et al., 2000; Lee & Patterson, 2000; Finlay et al., 2001; Fenchel & Finlay, 2004), tardigrades (Morgan & King, 1976), nematodes (Abebe & Coomans, 1995) and rotifers (Shiel & Green, 1996) do in many cases appear to be cosmopolitan. This suggests that the global biota can be divided into two broad groups – small species each consisting of a single, cosmopolitan metapopulation, and larger species composed of one or more regionally restricted metapopulations. There must be a crossover body-size range – a cosmopolitan-biogeography transition – but neither the steepness of this transition zone, nor the organism size range in which it occurs, was known. Finlay and Fenchel (2004) obtained extensive inventories of free-living eukaryote species from two contrasting aquatic habitats – a brackish, two-hectare, estuary-like habitat (Niva8 Bay, north of Copenhagen) that is permanently covered by the waters of the Øresund, and a eutrophic freshwater pond in the English Lake District, Priest Pot (Finlay & Maberly, 2000). The pond consists of one hectare of open water, surrounded by roughly one hectare of floating bog. By restricting taxonomic coverage to aquatic eukaryotes, it was possible to apply a single species concept (the morphospecies) to all biota. Finlay, Fenchel and their collaborators searched all accessible resources for geographic records worldwide of the species recorded in Priest Pot and Niva8 Bay, but the level of undersampling of meiofauna and protists created difficulties. Much published information is available for the northern hemisphere, but some taxonomic groups (e.g. marine ciliates, rotifers and gastrotrichs) have hardly been studied in the southern hemisphere. Almost all geographical records for southern hemisphere chrysomonads, heliozoans and heterotrophic flagellate, for example, can be attributed to the efforts of four or five workers in the last three decades. The authors obtained extensive inventories of naked amoebae in both Priest Pot and Niva8 Bay, but there has been only very limited sampling and species-level identification of naked amoebae outside of the Holarctic.
171
172
B. J. FINLAY AND G. F. ESTEBAN
Figure 9.4 The proportion of eukaryote species at Priest Pot that are cosmopolitan, arranged in logarithmic size classes. The principal taxa contributing to the upper shoulder of the curve are protozoa, gastrotrichs, rotifers, ostracods, nematodes and cladocerans. The main taxa in the lower part of the curve are triclads, molluscs, aquatic insects and leeches. Adapted from Finlay and Fenchel (2004).
Of the >2000 eukaryote taxa recorded, those identified to species level with sufficient associated geographical information were allocated to size classes, and the percentage of cosmopolitan species calculated for each size class. The curve that emerged for Priest Pot (Fig. 9.4) appeared fairly smooth, apart from the low value datapoint marking the smallest size class (arrow), which was almost certainly a consequence of undersampling worldwide. Most species in that size class were chrysomonads, and species-level identification requires electron microscopy and other specialized techniques. Nevertheless, 43 of the 50 species of the flagellated chrysomonad Paraphysomonas known worldwide were recorded from Priest Pot.
Testing the theory of cosmopolitan distribution Cryptic protist diversity exceeds the diversity of ‘active’ species, and the habitat selects Ubiquitous dispersal of free-living microbial eukaryotes implies that a large species richness will exist wherever life is possible. This will rarely be obvious, however, because most microbial species are usually encysted or in some other cryptic, inactive state, waiting for the arrival of conditions that support population growth. However, with patience and experimental manipulation (e.g. Finlay, Esteban & Fenchel, 1996a; Fenchel et al., 1997) it is usually possible to coax these cryptic species to excyst and grow so that they become detectable. The limited information available indicates that the ‘seedbank’ of cryptic species richness can be very diverse. In order to investigate this, we examined
BODY SIZE AND BIOGEOGRAPHY
Figure 9.5 The ubiquitous ‘seedbank’. A small sample of sediment from a freshwater lake was ‘manipulated’ (see text and Finlay et al., 1996a; Fenchel et al., 1997) to produce a wide variety of niches for ciliate species – e.g. a range of microbial food sources, redox gradients, changes in light and temperature. About 20 ‘trophic’ species were identified at the beginning of the experiment, and this number increased to more than 130 species over a period of three months.
approximately one gram of superficial lake sediment, and recorded 20 active ciliate species at the time of sampling. In response to experimental treatments (e.g. addition of a variety of microbial food sources, establishment of redox gradients, and changes in light and temperature regimes) over a period of three months, the species total rose to 137, at which time it was still increasing (Fig. 9.5). More specific evidence that the ‘habitat selects’ is provided by the discovery of typical marine ciliate morphospecies living in inland evaporation salt pans in Central Spain at 300 km from the nearest coastal marine environment (Esteban & Finlay, 2004). These salt pans are fed from saline underground water originating in Triassic marine sediments. Water flows to the surface and into permanent evaporation pools, where the salt concentration varies from brackish to characteristic seawater, and up to hypersaline. Conversely, when samples from a hypersaline coastal lagoon in Southern Spain were progressively diluted with freshwater, a large number of typically freshwater protozoan species appeared (Esteban & Finlay, 2003). It is surprising that simple
173
174
B. J. FINLAY AND G. F. ESTEBAN
experiments such as these are not carried out more frequently, especially when the rewards (for example, discovering the true dimensions of microbial eukaryote diversity in a very small sample) are so easily gained. Dispersal of free-living protists is essentially random Any contender for a randomly distributed free-living species will have small body size, high absolute abundance, be easily dispersed and have minimal or nonexistent social or parasitic behaviour. Probably all soil protozoa fit the bill. They are adapted for life in soil, they form protective cysts (Finlay & Fenchel, 2001), there is evidence they spend much of their time as cysts, and the cysts remain viable for long periods of time (e.g. 49 years, Goodey, 1915). Thus, encysted soil protozoa could fit the criteria for random distribution. To test this hypothesis, we examined the possibility that the rarer testate amoebae species in a one-hectare grassland site in the southern uplands of Scotland could be randomly distributed (for a review see Esteban et al., 2006). We compared the spatial distributions of six rare species and six abundant species, obtaining a mean and variance for each species across the 25 plots in which the grassland site was divided, and repeating this on six sampling occasions (Finlay et al., 2001). We found that the rare and abundant species did appear to have different spatial distributions. Like most animal species the abundant testate species had aggregated distributions (Fig. 9.1), whereas the distributions of rare species were close to random (logvariance was closely proportional to the log-mean, the intercept was small, and we concluded that the distributions of these species were effectively random). Unlike the abundant species, with sustained population growth at this site, the rare species were probably transported randomly to the site but unable to initiate population growth in this particular habitat type. Cosmopolitan distribution implies the existence of similar species inventories in similar habitats irrespective of geographical distance This could be tested by investigating the ciliated protozoa living in a habitat that is separated from Northern Europe by geographical barriers and great distance – the sediment of a Holocene volcanic crater-lake in Australia (Finlay et al., 1999; Esteban et al., 2000). Of the 85 ciliate species recorded, all were already known from northern Europe by the year 1935 except one that was described in 1985. As the crater is only 2 km from the Southern Ocean, seaspray renders the water brackish (2–5%), and eight of the species recorded were typical of brackish waters. One of them (Tracheloraphis caudata) was a typical marine interstitial ciliate. This species is fragile and it does not form protective cysts – so, it was unclear how it may have reached the crater-lake sediment. The same can also be said for the anaerobic ciliate Plagiopyla frontata, a marine ciliate that contains unmistakable endosymbiotic methanogenic bacteria in both Northern Europe and Australia.
BODY SIZE AND BIOGEOGRAPHY
Endemic species Endemic protist species – if they exist – would seriously challenge the theory of ubiquitous dispersal. Until recently, the large and distinctive ciliate Loxodes rex had been found only in freshwater in tropical Africa (Dragesco, 1970), where it was considered to be an example of an endemic ciliate. However, 30 years later, it was discovered thriving in a pond in Thailand (Esteban et al., 2001), where it joined an expanding collection of large ciliate species previously believed to be endemics – a status they all subsequently lost with additional sampling effort. Neobursaridium gigas was originally recorded from Argentina (Balech, 1941), and subsequently in Uganda (Nilsson, 1962). The very large (2–3 mm) Avelia martinicensis from the West Indies (Nouzare`de, 1975) was found in Brazil and, later, in Saudi Arabia. Condylostoma reichi from Israel (Wilbert & Kahan, 1981), was found in tropical Africa (Dragesco & Dragesco-Kerne´is, 1986). These few examples show that it is difficult to find protist species whose status as ‘endemics’ stands the test of time. Disproof of ubiquitous dispersal might have been provided by the protist community described by Gajewskaja (1933) from Lake Baikal – the world’s oldest and deepest lake. But the dozen or so new ‘forms’ and new species have all been found in other parts of the world. Loxodes rex would, because of its very large size (2 mm long), fragility, and lack of cyst formation, have been a prime candidate for an endemic ciliate, but it is not and it seems to thrive in tropical freshwater habitats on different continents (Esteban et al., 2001). In contrast, the unusual desmids of Tasmania (Tyler, 1996) indeed appear to be endemic. A possible explanation, however, is that unusual habitats (supporting unusual, even endemic, semi-aquatic plants) provide local microhabitats for particular, globally rare, micro-organisms. Thus, the unusual desmids may still disperse over a large area, but remain rare or undiscovered outside their place of discovery. It is generally agreed that the composition of diatom communities is dominated by cosmopolitan species with high dispersal ability (e.g. Soinen, Paavola & Muotka, 2004), but the relative importance of (a) physical barriers that may limit species dispersal, and (b) the extent to which diatoms have limited access to suitable habitat, was unknown. Recent analyses (Potapova & Charles, 2002) suggest that the availability of suitable habitat is the principal determinant of diatom community structure in rivers. Further, if the presence of a diatom in a habitat is limited primarily by the quality of the local environment rather than by physical barriers to dispersal, the diatom can potentially be found wherever the same habitat type occurs. In a quantitative approach using canonical correspondence analysis, Potapova and Charles (2002) showed that habitat factors consistently explained more variation than physical barriers. At the national (USA) scale, more than two-thirds of explicable variation in diatom species composition could be attributed to habitat factors, with less than one-third attributable to the role of physical barriers and other spatial factors.
175
176
B. J. FINLAY AND G. F. ESTEBAN
The list of the most frequently occurring diatom species consists solely of taxa with cosmopolitan distribution. This is presumably because common species are adapted to the most common habitat, which in the case of the United States and Europe is a nutrient-enriched, moderately alkaline river of the temperate climatic zone. Abundance of this habitat type in one continent increases the probability of dispersal to other parts of the world – with (a) the consequence that locally abundant diatom species tend to be globally abundant (Finlay et al., 2002), and (b), the corollary, that rare habitats such as oligotrophic waters, support much lower dispersal. The thesis that locally abundant or rare species tend to be globally abundant or rare is a recurring theme within this chapter and in macroecology more generally (Bell, 2001). See, for example, exactly the same phenomenon revealed within the soil protozoa (Fig. 9.2). Local:global species ratios Within any taxonomic group of small organisms, a large proportion of the global species richness is found in a local area, and the local:global species ratio for that group tends to be very high (Finlay & Clarke, 1999a, b; Hillebrand & Azovsky, 2001). For example, the percentage of the global species pool recorded from Priest Pot is >15% for ciliates, >50% for heliozoa and >80% for chrysomonads (all recorded species were actually recorded from roughly one tenth of a cubic centimetre of sediment). It follows that, if cosmopolitan distribution is the product of high absolute abundance, populations of the larger, non-cosmopolitan organisms will have low local:global ratios. This appears to be the case – the percentage of the global species pool of aquatic insects recorded in Priest Pot is <0.001, but with decreasing mean body size, local species richness represents an increasing fraction of global species richness, and it approaches 100% in the case of protozoan species with body sizes 0.004 mm. Finlay and Fenchel (2004) found that a systematic relationship does exist. Even in the case of two small aquatic ecosystems – one freshwater, the other marine – the local:global species ratio increases with decreasing organism size, incorporating the combined freshwater and marine species in a single relationship (Fig. 9.6). Undersampling The problem of undersampling persists because sampling regimes are often inadequate for detecting protists and other microfauna. This is especially so for the rarer species. The frequently cited argument of undersampling – ‘this species must be an endemic because it is morphologically so distinctive that if it existed elsewhere, it would have been recorded by now’ – might be convincing for large mammal species but, with respect to the protists, it evades the crucial question. It can confidently be demonstrated that a large mammal species does not exist outside of its endemic range, but how are we to demonstrate that a particular protist species does not exist elsewhere on the planet?
BODY SIZE AND BIOGEOGRAPHY
Figure 9.6 Local species richness calculated as a ratio of the known or ‘best estimate’ global species richness for the same taxonomic group (adapted from Finlay & Fenchel, 2004). Moving from taxa with the largest to the smallest mean size, the symbols represent, (a) Priest Pot (filled symbols) – name of taxon followed in parenthesis by organism size (mm), and number of species in taxonomic group: fish (220, 5); Odonata (43, 15); leeches (20, 10); Chironomidae (14, 12); Mollusca (9.8, 71); Tricladida (6, 3); Ostracoda (0.804, 9); Nematoda (0.8, 71); Cladocera (0.73, 31); Rotifera (0.236, 22); gastrotrichs (0.164, 6); testate amoebae (0.097, 39); ciliates (0.076, 274); euglenids (0.048, 21); Synura (0.047, 6); Korotnevella (0.045, 6); naked amoebae (0.037, 64), Cochliopodium (0.03, 13); heliozoans (scale-bearing; 0.0271, 57); diatoms (0.0234, 93); Mallomonas (0.016, 24); heterotrophic flagellates (0.0158, 86), Spiniferomonas (0.006, 8); Paraphysomonas (0.005, 40); Cyathobodo (0.005, 7); Luffisphaera (0.004, 15). (b) Niva8 Bay (open symbols) – fish (165, 9); polychaetes (45, 9); Mollusca (30, 11); Crustacea (13, 44); Nematoda (1.3, 55); Turbellaria (1.2, 34); Ostracoda (0.557, 15); gastrotrichs (0.29, 4); ciliates (0.11, 201); dinoflagellates (0.0446, 103); Prasinophyta (0.0337, 25); naked amoebae (0.0242, 42); flagellates (0.0077, 75). The least squares regression through the combined data set is highly significant (p < 0.001).
The ‘problem’ is readily illustrated with the following simple example (Finlay & Fenchel, 2004). Consider the aforementioned proposition ‘if it did exist elsewhere, it would have been found already’. On the basis of results from previous sampling campaigns, we might predict that a certain ‘rare’ protist species lives in the oxygen-depleted hypolimnion of a productive freshwater pond. Priest Pot, in the UK, is such a pond. It has an area of one hectare, and a 1.75 m-deep hypolimnion holding roughly 6300 tonnes of water. Assume that the protist species we are looking for exists in the pond as a population of one million individuals. This figure may appear to have little to do with the idea of rarity, but each individual would, on average, be surrounded by 6.3 litres of water, giving it a volume
177
178
B. J. FINLAY AND G. F. ESTEBAN
fraction in the region 10 11 –10 13 if we consider, for example, the size range that includes most ciliate species. But ‘rarity’ is a relative term that is strongly influenced by organism size. A million perch in the same pond would never be considered ‘rare’, but a protist population of a million individuals would probably never be detected – even after years of sampling. For most of the relatively rare protist species (those that are orders of magnitude scarcer than the abundant species; Finlay, Esteban & Fenchel, 1998), undersampling is inevitable using typical sampling procedures. For that reason alone, ‘endemic’ species are not found elsewhere (see also Fenchel & Finlay, 2003). However (see above), a growing body of evidence indicates that selective-enrichment culture techniques can reveal the rare or cryptic diversity of protists (Fenchel et al., 1997), such as freshwater ciliates in hypersaline ponds (Esteban & Finlay, 2003), but the use of such methods is not common practice in the search for species that may have escaped from their ‘endemic’ ranges. Figure 9.7 illustrates this point. Species concepts Is the resolution of morphospecies poorer for smaller than for larger organisms? For protists with complex and varied morphologies, such as the ciliated protozoa, this is probably not the case. However, the species taxonomy of the smaller organisms depends on a proportionately higher optical resolution. The species systematics of chrysomonad flagellates, for example, is necessarily based on electron microscopy. If protists in general demonstrated the degree of morphological richness normally perceived in the naked amoebae, we would have a major problem with the morphospecies concept. But naked amoebae were not included in our analysis for the detection of the cosmopolitan-biogeography transition (see above), and the morphological richness that can be perceived in the protists we did use – including the ciliates, diatoms, foraminifera, chrysomonads, desmids, heliozoans, choanoflagellates and testate amoebae, as well as microbial metazoans such as the rotifers, gastrotrichs and microcrustaceans – indicates that microfauna, meiofauna and macrofauna may share a similar degree of morphological richness. Do ‘cryptic species’ or ‘sibling species’ (i.e. species that are morphologically similar but reproductively isolated) show any large-scale geographical patterns in their distribution. This has been investigated in some protists, especially the ciliates, and the evidence indicates that members of sibling species complexes have worldwide distribution. Mating experiments with isolates of nominal species from different continents have also shown that the isolates are interfertile (for discussion and references, see Finlay, 2002; Finlay et al., 1996b). Cosmopolitan genotypes Recent molecular studies of isolates of the small ciliate morphospecies Cyclidium glaucoma from across the world have disclosed multiple, distinct (rDNA)
BODY SIZE AND BIOGEOGRAPHY
Figure 9.7 Intensive examination of grassland soils in Scotland (Finlay et al., 2000) revealed many rare species of ciliated protozoa. Amongst them was the ciliate shown here – Notohymena antarctica – which we found in re-wetted soil. This isolate from Scotland represents the first record of the species outside of the Antarctic continent, and in the northern hemisphere (Esteban et al., 2006). Protargol silver-impregnated specimen. Scale bar 20 mm.
genotypes, but there appears to be no evidence for any geographical pattern in their distribution (Finlay et al., unpublished), although particular genotypes do seem to be correlated with particular habitat types. Identical genotypes have been recovered from Great Salt Lake, and from hypersaline ponds in Spain. A single genotype has been isolated from brackish waters in Argentina, Japan, Morocco and Ukraine. A further single genotype was recorded from freshwater habitats in Guatemala and Ukraine, and another from France and Mongolia. These data are still being gathered, but they do suggest that protist genotypes can have cosmopolitan distribution, with no apparent biogeography. For asexual organisms with clonal evolution there is no theory-based species concept analogous to the biological species concept, so that in bacteria and the asexual microbial eukaryotes, a large degree of genotypic variation is expected
179
180
B. J. FINLAY AND G. F. ESTEBAN
within nominal species. We do not believe there is yet any convincing evidence for a biogeography of such variation in asexual protists. In the case of bacteria too, similar or identical genotypes occur worldwide (Glo¨ckner et al., 2000; Hagstro¨m et al., 2000, Massana et al., 2000). At the present time, the findings of identical rDNA genotypes in specific (and similar) habitat types worldwide is strong evidence supporting cosmopolitan distribution of microbial eukaryotes. ‘Biogeography’ of microbes? Some recent studies indicate that unicellular organisms may demonstrate ‘biogeography’. We do not subscribe to the view that microbial eukaryotes have ‘biogeography’ in the classical sense of, for example, geographically restricted tropical birds and mammals. It is unquestionably true that the distributions of microbes may show some structure and patterns (Green et al., 2004; HornerDevine et al., 2004; Noguez et al., 2005) that vary over time, and may demonstrate ‘at least some geographical differentiation’ (Green et al., 2004), but this is rather weak evidence for biogeography when compared with the stunning geographical patterns demonstrated by macroscopic organisms with real biogeography in established biogeographic regions. We resort to the key characteristic that free-living microbes thrive wherever their preferred habitat requirements are realized. Interestingly, Horner-Devine et al. (2004) provide evidence that this is so, using their example of bacteria. They discovered that habitats that are similar in environmental characteristics, are also similar in bacterial composition (i.e. even specific bacteria thrive wherever their habitat requirements are realized). On the other hand, if a microbial habitat becomes so degraded that it can no longer sustain a particular microbial community, the pattern of ‘biogeography’ may simply be ephemeral.
General validity Finlay and Fenchel (2004) revealed a cosmopolitan-biogeography transition in the region 1–10 mm for aquatic microbial eukaryotes, but is it in the right place? The data set used was the biggest available, but additional data, particularly for the 1–10 mm size range, would be especially useful. Unfortunately, the absence of any centralized, accessible source of data makes it extremely laborious to collect the information that would indicate the extent of geographical distribution of individual taxa. The two study sites used in the investigations by Finlay and Fenchel (2004) – Niva8 in Denmark and Priest Pot in the UK – are situated at roughly the same latitude in the northern hemisphere, and it is possible that the macrofauna would present a different picture if the field sites had been placed elsewhere, such as an area with a large number of endemic species. The existence of a large pool of undiscovered small tropical organisms cannot be ruled out, but the fact that Finlay and Fenchel (2004) found a substantial fraction of the global
BODY SIZE AND BIOGEOGRAPHY
inventory of <1 mm species in two small water bodies, suggests that a similar picture would have emerged irrespective of where the investigation had been carried out. This interpretation is supported by the finding that protist species richness is generally the same at all latitudes (Hillebrand & Azovsky, 2001). This is not to say that all protist species thrive at all latitudes. The conspicuous pantropical ciliate Neobursaridium gigas grows only within a narrow temperature range of a few degrees on either side of 25 8C (Dragesco, 1968) but it does so in more that one geographical region (i.e. Africa and South-East Asia). The coldwater planktonic foraminiferan species (Darling et al., 2000), and the dinoflagellate Polarella glacialis (Montresor et al., 2003) have bi-polar distributions. The vast majority of free-living protists probably tolerate fairly broad ranges of environmental factors (e.g. salinity, temperature) and the few taxa that have been investigated experimentally (e.g. Paraphysomonas, Cyclidium) indicate that they also have great capacity for physiological adaptation within genetically determined limits that are much broader than those of macrofauna. It is perhaps not surprising then, that an intensive study of ciliate morphospecies in an Australian crater-lake (Esteban et al., 2000) should reveal a species list consisting entirely of species already known from northern Europe.
Neutral theory – local:global species ratios Finlay and Fenchel (2004) showed that the size dependence of local:global species ratios in small aquatic organisms appears to exist independently of the taxonomic identity of the organisms concerned. The common factor underlying these patterns is organism size, because size is inversely correlated with population size, which largely determines the probability of dispersal (Finlay, 2002). To some extent, a neutral model of community structure (Bell, 2001; Hubbell, 2001) is supported. While we do not necessarily endorse all of the assumptions in current neutral community theories (e.g. that the aquatic species we have catalogued are ecologically identical), our results support some of the predictions of neutral theory – in particular that high absolute abundance of microbes drives their large-scale random dispersal, resulting in high local species richness, correlated local and global abundances of individuals within species, and the creation of metapopulations with global-scale distribution (Fenchel, 1993; Finlay et al., 2001). Cosmopolitan distribution means that the global species richness of freeliving micro-organisms will be relatively small because a microbial species carries out the same function in places that are geographically distant from each other, which explains why there are apparently relatively few small species, as originally highlighted by May (1988). These patterns support the dominating role of dispersal and local extinction in moulding the composition of biotic communities. We also note that in contrast to results obtained from higher plant communities (Condit et al., 2002), the application of neutral models
181
182
B. J. FINLAY AND G. F. ESTEBAN
to organisms with microbial dimensions will probably not break down at the global scale (Finlay et al., 2001).
Concluding remarks It may be easier to detect the footprints of cosmopolitan distribution (e.g. identical genotypes across the global scale) than to establish the cause of an apparent biogeographic pattern. Take for example the relationship between species richness and latitude. Most latitudinal gradients have been described for macrofauna and plants, and species richness tends to be significantly higher at low latitudes. In the case of (cosmopolitan) microbial eukaryotes, however, we do not expect any discernible latitudinal gradient in species richness. Hillebrand and Azovsky (2001) pursued an approach that nearly succeeded. Following expectations, no correlation was found between latitude and species richness; but in the southern hemisphere, there was a significant negative correlation – probably an artefact based on some very low richness values from the Antarctic. This whole area of science seems to have acquired the status of a ‘hot topic’, and we can expect to see previously unforeseen developments in the near future. We close here with three points: (1)
(2)
(3)
Protists thrive wherever their preferred habitat requirements are met and there are no barriers to dispersal. This is a key point. It is usually relatively straightforward to observe microfaunal species thriving in their preferred habitat type, but with few exceptions, this is not so straightforward with bacteria. Nevertheless, Horner-Devine et al. (2004) discovered that sites sharing similar environmental characteristics were also similar in bacterial taxonomic composition – i.e. bacteria, like protists, thrive wherever their preferred habitat is realized. With decreasing mean body size, local species richness becomes an increasing fraction of global species richness. This helps to explain the relatively flat taxon richness-area relationships in microbial eukaryotes (e.g. Green et al., 2004), and the observation that local species richness (body sizes 0.004 mm) can be almost as great as global species richness. The undersampling problem. Historically, undersampling has been a major problem, and in the ‘molecular age’ the almost infinitely variable ribosomal DNA sequences contribute little if anything to ease the problem. Now that species-specific primers can be developed routinely, our experience is that the scale of sampling protists and microfauna could potentially increase dramatically.
Acknowledgements This work was carried out with financial assistance from the Natural Environment Research Council (Marine and Freshwater Microbial Biodiversity
BODY SIZE AND BIOGEOGRAPHY
Thematic Programme). We are grateful to Professor Tom Fenchel (Copenhagen) for sharing ideas and contributing very useful suggestions.
References Abebe, E. & Coomans, A. (1995). Fresh-water nematodes of the Galapagos. Hydrobiologia, 299, 1151. Balech, E. (1941). Neobursaridium gigas n. gen. n. sp. de ciliado heterotrico. Physis (B. Aires), 19, 29–35. Bell, G. (2001). Neutral macroecology. Science, 293, 2413–2418. Berninger, U.-G., Finlay, B. J. & Kuuppo-Leinikki, P. (1991). Protozoan control of bacterial abundances in fresh water. Limnology and Oceanography, 36, 139–147. Condit, R., Pitman, N., Leigh Jr., E. G. et al. (2002). Beta-diversity in tropical forest trees. Science, 295, 666–669. Damuth, J. (1981). Population density and body size in mammals. Nature, 290, 699–700. Darling, K. F., Wade, C. M., Stewart, I. A. et al. (2000). Molecular evidence for genetic mixing of Arctic and Antarctic planktonic foraminifers. Nature, 405, 43–47. Dragesco, J. (1968). A propos de Neobursaridium gigas Balech, 1941: ste´nothermie, inclusions, ultrastructure des trichocystes. Protistologica, 4, 157–167. Dragesco, J. (1970). Cilie´s libres du Cameroun. Annals of the Faculty of Science of Yaounde´, (Hors se´rie), 1–141. Dragesco, J. & Dragesco-Kerne´is, A. (1986). Cilie´s Libres de l’Afrique Intertropicale. Collection Faune Tropicale no. 26, ORSTOM, Paris. Esteban, G. F. & Finlay, B. J. (2003). Cryptic freshwater ciliates in a hypersaline lagoon. Protist, 154, 408–411. Esteban, G. F. & Finlay, B. J. (2004). Marine ciliates (Protozoa) in central Spain. Ophelia, 58, 13–22. Esteban, G. F., Finlay B. J., Olmo, J. L. & Tyler, P. A. (2000). Ciliated protozoa from a volcanic
crater-lake in Victoria, Australia. Journal of Natural History, 34, 159–189. Esteban, G. F., Finlay, B. J., Charubhun, N. & Charubhun, B. (2001). On the geographic distribution of Loxodes rex (Protozoa, Ciliophora) and other alleged endemic species of ciliates. Journal of Zoology, London, 255, 139–143. Esteban, G. F., Clarke, K. J., Olmo, J. L. & Finlay, B. J. (2006). Soil protozoa – an intensive study of population dynamics and community structure in an upland grassland. Applied Soil Ecology, 33, 137–151. Fenchel, T. (1993). There are more small than large species? Oikos, 68, 375–378. Fenchel, T. & Finlay, B. J. (2003). Is microbial diversity fundamentally different from biodiversity of larger animals and plants? European Journal of Protistology, 39, 486–490. Fenchel, T. & Finlay, B. J. (2004). The ubiquity of small species: patterns of local and global diversity. Bioscience, 54, 777–784. Fenchel, T., Esteban, G. F. & Finlay, B. J. (1997). Local versus global diversity of microorganisms: cryptic diversity of ciliated protozoa. Oikos, 80, 220–225. Finlay, B. J. (2002). Global dispersal of free-living microbial eukaryote species. Science, 296, 1061–1063. Finlay, B. J. & Clarke, K. J. (1999a). Ubiquitous dispersal of microbial species. Nature, 400, 828. Finlay, B. J. & Clarke, K. J. (1999b). Apparent global ubiquity of species in the protist genus Paraphysomonas. Protist, 150, 419–430. Finlay, B. J. & Fenchel, T. (2001). Protozoan community structure in a fractal soil environment. Protist, 152, 203–218. Finlay, B. J. & Fenchel, T. (2004). Cosmopolitan metapopulations of free-living microbial eukaryotes. Protist, 155, 237–244.
183
184
B. J. FINLAY AND G. F. ESTEBAN
Finlay, B. J. & Maberly, S. C. (2000). Microbial Diversity in Priest Pot. Ambleside, UK: Freshwater Biological Association. Finlay, B. J., Esteban, G. F. & Fenchel, T. (1996a). Global diversity and body size. Nature, 383, 132–133. Finlay, B. J., Corliss, J. O., Esteban, G. & Fenchel, T. (1996b). Biodiversity at the microbial level: the number of free-living ciliates in the biosphere. The Quarterly Review of Biology, 71, 221–237. Finlay, B. J., Esteban, G. F. & Fenchel, T. (1998). Protozoan diversity: converging estimates of the global number of free-living ciliate species. Protist, 149, 29–37. Finlay, B. J., Esteban, G. F., Olmo, J. L. & Tyler, P. A. (1999). Global distribution of free-living microbial species. Ecography, 22, 138–144. Finlay, B. J. Black, H. I. J., Brown, S. et al. (2000). Estimating the growth of the soil protozoan community. Protist, 151, 69–80 (and Corrigendum: Protist, 151, 367). Finlay, B. J., Esteban, G. F., Clarke, K. J. & Olmo, J. L. (2001). Biodiversity of terrestrial protozoa appears homogeneous across local and global spatial scales. Protist, 152, 355–366. Finlay, B. J., Monaghan, E. B. & Maberly, S. C. (2002). Hypothesis: the rate and scale of dispersal of freshwater diatom species is a function of their global abundance. Protist, 153, 261–273. Finlay, B. J., Esteban, G. F. & Fenchel, T. (2004). Protist diversity is different? Protist, 155, 15–22. Foissner, W. (1999). Protist diversity: estimates of the near imponderable. Protist, 150, 363–368. Gajewskaja, N. (1933). Zur Oekologie, Morphologie und Systematik der Infusorien des Baikalsees. Zoologica, 32, 1–298. Glo¨ckner, F. O. Zaichikov, E., Belkova, N. et al. (2000). Comparative 16S rRNA analysis of lake bacterioplankton reveals globally distributed phylogenetic clusters including an abundant group of Actinobacteria. Applied
and Environmental Microbiology, 66, 5053–5065. Goodey, T. (1915). Notes on the remarkable retention of vitality by Protozoa from old stored soils. Annals of Applied Biology, 1, 395–399. Green, J. L., Holmes, A. J., Westoby, M. et al. (2004). Spatial scaling of microbial eukaryote diversity. Nature, 432, 747–750. Hagstro¨m, A8 ., Pinhassi, J. & Zweifel, U. L. (2000). Biogeographical diversity among marine bacterioplankton. Aquatic Microbial Ecology, 21, 231–244. Hillebrand, H. & Azovsky, A. I. (2001). Body size determines the strength of the latitudinal diversity gradient. Ecography, 24, 251–256. Horner-Devine M. C., Lage, M., Hughes, J. B. & Bohannan, J. M. (2004). A taxa-area relationship for bacteria. Nature, 432, 750–753. Hubbell, S. P. (2001). The unified neutral theory of biodiversity and biogeography. Monographs in Population Ecology, 32, Princeton, NJ: Princeton University Press. Lawton, J. H. (1998). Small is beautiful, and very strange. Oikos, 81, 3–5. Lee, W. J. & Patterson, D. J. (2000). Heterotrophic flagellates (Protista) from marine sediments of Botany Bay, Australia. Journal of Natural History, 34, 483–562. Mann, D. G. & Droop, S. J. M. (1996). Biodiversity, biogeography and conservation of diatoms. Hydrobiologia, 336, 19–32. Massana, R., DeLong, E. F. & Pedro´s-Alio´, C. (2000). A few cosmopolitan phylotypes dominate planktonic archaeal assemblages in widely different oceanic provinces. Applied and Environmental Microbiology, 66, 1777–1787. May, R. M. (1988). How many species are there on Earth? Science, 241, 1441–1449. Montresor, M., Lovejoy, C., Orsini, L., Procaccini, G. & Roy, S. (2003). Bipolar distribution of the cyst-forming dinoflagellate Polarella glacialis. Polar Biology, 26, 186–194.
BODY SIZE AND BIOGEOGRAPHY
Morgan, C. I. & King, P. E. (1976). British Tardigrades Synopses of the British Fauna No. 9. London: Linnean Society of London; and Academic Press. Nilsson, J. R. (1962). Observations on Neobursaridium gigas Balech, 1941 (Ciliata, Heterotrichida). Journal of Protozoology, 9, 273–276. Noguez, A. M., Arita, H. Y., Escalante A. E. et al. (2005). Microbial macroecology: highly structured prokaryotic soil assemblages in a tropical deciduous forest. Global Ecology and Biogeography, 14, 241–248. Nouzare`de, M. (1975). Sur un nouveau genre de protozoaires cilie´s ge´ants me´sopsammiques appartenant a` la famille des Geleidae, Kahl. Comptes Rendus de l’Acade´mie des Sciences, Se´rie D., Paris, 280, 625–628. Potapova, M. G. & Charles, D. F. (2002). Benthic diatoms in USA rivers: distributions along
spatial and environmental gradients. Journal of Biogeography, 29, 167–187. Schmid, P. E., Tokeshi, M. and Schmid-Araya, J. M. (2000). Relation between population density and body size in stream communities. Science, 289, 1557–1560. Shiel, R. J. & Green, J. D. (1996). Rotifera recorded from New Zealand, 1859–1995, with comments on zoogeography. New Zealand Journal of Zoology, 23, 193–209. Soinen, J., Paavola, R. & Muotka, T. (2004). Benthic diatom communities in boreal sreams: community structure in relation to environmental and spatial gradients. Ecography, 27, 330–342. Tyler, P. A. (1996). Endemism in freshwater algae. Hydrobiologia, 336, 127–135. Wilbert, N. & Kahan, D. (1981). Ciliates of Solar Lake on the Red Sea shore. Archiv fu¨r Protistenkunde, 124, 70–95.
185
CHAPTER TEN
By wind, wings or water: body size, dispersal and range size in aquatic invertebrates SIMON D . RUNDLE University of Plymouth
DAVID T . BILTON University of Plymouth
ANDREW FOGGO University of Plymouth
Introduction The past 15 years have seen a dramatic increase in the study of large-scale patterns and processes in ecology under the banner of macroecology (Brown & Maurer, 1989). Organismal body size is one of the key components of many of these studies, and the distribution of body size and its relationship with range size and abundance figure extensively in the macroecological literature (Gaston & Blackburn, 2000; Blackburn & Gaston, 2003; Gaston, 2003). Body size is also the central component of the ‘three-quarters scaling law’, which predicts that metabolism scales to body mass0.75 (e.g. Gillooly et al., 2001) and is seeing increasing use in ecological predictions, including those concerning trophic interactions (Woodward et al., 2005), population dynamics (Marquet et al., 2005), species diversity (Allen, Brown & Gillooly, 2002) and energy flow (Enquist et al., 2003); indeed, the scaling of metabolism with body mass has also recently been used in attempts to make macroecological predictions (e.g. Jetz et al., 2004). Despite the recent surge of interest in large-scale ecological patterns, aquatic ecologists have been slow to take up the concept of macroecology. It could perhaps be argued that much of the aquatic ‘community ecology’ over the past couple of decades, relating assemblage composition in aquatic systems to environmental variables, was macroecology of sorts. However, this research has rarely progressed to examine over-arching patterns and their potential underlying mechanisms, and is therefore somewhat limited in how it can inform general ecological theory. Hence, with the exception of some work on fishes (Taylor & Gotelli, 1994; Pyron, 1999; Rosenfield, 2002; Goodwin, Dulvy & Reynolds, 2005) and on eukaryotic microbes (see Finlay, 2002 and Finlay & Esteban, this volume), there is a dearth of research on the macroecological Body Size: The Structure and Function of Aquatic Ecosystems, eds. Alan G. Hildrew, David G. Raffaelli and Ronni Edmonds-Brown. Published by Cambridge University Press. # British Ecological Society 2007.
BODY SIZE, DISPERSAL AND RANGE SIZE IN AQUATIC INVERTEBRATES
importance of body size in the aquatic compared with the terrestrial realm, and little has been published explicity on the macroecology of aquatic invertebrates a deficit that this symposium was designed to rectify. A species’ dispersal ability has fundamental and far-reaching consequences for its biogeography, evolutionary persistence and, therefore, its macroecology. Yet dispersal has not figured heavily in macroecological investigations (Gaston, 2003), which almost certainly stems from the lack of good data on actual dispersal propensity for groups other than vertebrates (e.g. Paradis et al., 1998). At the same time, dispersal is an emergent property that is likely to be influenced by a combination of traits, which complicates investigations of its role in ecological terms (Table 10.1). It is likely, however, that some of the traits that influence the dispersal abilities of taxa are likely to be linked to body size and, hence, that dispersal is a key, emergent trait that may provide a link between organismal size and distribution (Table 10.1). Dispersal presents a variety of challenges for aquatic organisms, from freshwater taxa occupying discrete water bodies surrounded by an inhospitable terrestrial landscape mosaic to microscopic passive dispersers at the mercy of large-scale oceanic currents (Finlay & Esteban, this volume). Our aims in this chapter are, first, to review what is known about how invertebrates disperse in aquatic habitats, including the ways in which body size and dispersal may interact and, second, to explore whether there are potential links between body size, dispersal ability and macroecological patterns (i.e. range size). The difficulty of obtaining empirical data on the actual dispersal abilities of individual aquatic taxa makes investigation of the relationships between body size and dispersal per se intractable at present. Instead, we concentrate on the relationship between surrogate measures of dispersal that potentially relate to body size and range size.
Table 10.1 Traits potentially influencing dispersal in aquatic invertebrates: * indicates those traits most likely to be linked to body size. Life history Production of dispersive stages (e.g. resting eggs, cysts) Timing of reproduction Reproductive output* Ecological Habitat requirements (i.e. specialists vs. generalists) Metabolic/mechanical Flight efficiency*
187
S. D. RUNDLE ET AL.
Dispersal and body size in aquatic invertebrates Active dispersers in freshwaters Freshwater habitats are typically isolated, and relatively short-lived on geological timescales (Moss, 1998; Ribera, Foster & Vogler, 2003), and so freshwater organisms need to retain the ability to disperse, at least occasionally, if they are to persist. As noted by Darwin (1859), many freshwater taxa have relatively wide ranges, suggesting that they are capable of moving long distances. Movement between sites is achieved by active or passive means (Fig. 10.1; Bilton, Freeland & Okamura, 2001), although the dichotomy between these is not clear cut (see below). Active dispersal in freshwaters is achieved by self-generated movement of individuals, which may use sensory cues to seek out new areas of habitat or patches occupied by conspecifics. Insects constitute the overwhelming majority of actively dispersing freshwater organisms, and here movement between sites occurs during the adult stage. One obvious factor that could influence the dispersal ability of such flying taxa is their ability to travel large distances.
Rotifera Copepoda Hydrachnidia Cladocera Ostracoda Branchipoda Taxon
188
Diptera Plecoptera Ephemeroptera Trichoptera Hemiptera Coleoptera Odonata 0.1
1
102 103 10 Approx. body length (mm)
Figure 10.1 Approximate size ranges of adults in selected groups of freshwater invertebrates, divided into active and passive dispersers. Passive dispersers (open bars) generally have smaller modal body sizes than active dispersers (shaded bars). The shaded area indicates the size range in which the transition to cosmopolitanism has been predicted to occur (Finlay, 2002), whilst the dotted line indicates the size below which flying insects are frequent members of the aerial plankton (see text).
BODY SIZE, DISPERSAL AND RANGE SIZE IN AQUATIC INVERTEBRATES
Small, flying insects, particularly those <4 mm in length (Fig. 10.1) such as many Diptera and Coleoptera, ‘sail with the wind’ as aerial plankton during long-distance movements and control their direction largely during takeoff and landing (Johnson, 1969; Compton, 2002). Movement in such species clearly combines aspects of both active and passive dispersal, and the factors influencing long-range movement may have more in common with passively-dispersing non-insects (see below) than with larger flying animals. Larger insects associated with freshwaters have more control over their flight direction and speed, and achieve long-distance dispersal largely through active means (Fig. 10.1). Although active dispersal could be considered comparatively low risk, in that the dispersers have the ability to seek out appropriate habitats, there are other costs associated with active dispersal, including trade-offs with reproduction (Tanaka & Suzuki, 1998; Zera & Brink, 2000). Hence, flight efficiency may play an important role in dispersal ecology. The subject of how flight efficiency scales with body size and related morphological parameters has received considerable attention in the literature. In general terms, the muscles of larger animals perform the same mass-specific work for a lower metabolic cost and are thus more efficient than those of small animals (Ellington, 1991; Harrison & Roberts, 2000; Wakeling & Ellington, 1997). Aquatic insects broadly fit this model, although there are some subtleties that may determine the differences in dispersal potential among groups. Ellington (1991) suggested that flight performance in the closely related dragonflies and damselflies scaled positively with massþ 0.27 and negatively with mass 0.23, respectively, and Marden (1994) also suggested a negative scaling of body mass and maximum sustainable flight performance in damselflies. It seems that damselflies may be less efficient fliers than dragonflies. Based on estimates of lift and power from wings, Wakeling and Ellington (1997) also suggested that dragonflies and damselflies were adapted to different styles of flight, dragonflies being better adapted to fly rapidly in large spaces and damselflies for manoeuvring in small spaces. Based on these data it might be predicted that dispersal ability should be positively and negatively correlated with body size in dragonflies and damselflies, respectively, and that larger-bodied insects would generally have a greater dispersal potential than small. It is also clear, however, that other morphological characters, such as wing size (Guitie´rrez & Mene´ndez, 1997; McLachlan, 1985) and wing muscle mass and lever length scaling (e.g. Schilder & Marden, 2004), may also play a major role in dispersal dynamics of dragonflies and damselflies, and indeed other insects. At the same time, it is highly unlikely that flight efficiency alone explains dispersal ability in freshwater insects; other factors including levels of interspecific competition, habitat quality and persistence will also have an influence on successful dispersal (Travis & Dytham, 1999; Ferriere et al., 2000; Clobert et al., 2001; Ribera et al., 2003; Vogler & Ribera, 2003).
189
190
S. D. RUNDLE ET AL.
Passive dispersers in freshwaters In terms of overland, inter-site dispersal, both wind and animal vectors have been implicated in the passive transport of freshwater invertebrates (Dumont, 1994; Dahms, 1995; Korovchinsky & Boikova, 1996; Maguire, 1963). Anostracan eggs may be dispersed by the wind (Riddoch, Mpoloka & Cantrell, 1994; Brendonck & Riddoch, 1999), and such movement is likely to occur in other taxa with small, desiccation-resistant stages. Transport by mobile animal vectors has also been noted in a number of freshwater invertebrates (Bohonak, 1999; Bohonak, Smith & Thornton, 2004) including the movement of resting stages or individuals on and in waterfowl and other aquatic vertebrates (Proctor & Malone, 1965; Bilton et al., 2001; Green & Figuerola, 2005). Whilst being classed as passive, from the point of view of the dispersed organism, such movement differs fundamentally from wind-mediated transport in that the vector itself is dispersing actively, which may have important consequences for population structure and dynamics (Bohonak et al., 2004; Figuerola, Green & Michot, 2005). The majority of such passive dispersal in freshwater metazoans occurs during early life-history stages, and often involves resting eggs or cysts, which are resistant to desiccation and other adverse environmental conditions (Bilton et al., 2001). In many respects dispersal in such animals resembles dispersal in higher plants via seeds (De Stasio, 1989; Levin et al., 2003). Given the constraints of dispersal of propagules via wind, rain or animal vectors, these taxa would be expected to be relatively small. On the basis of the apparent ubiquity of many microbial eukaryotes Finlay (2002) and Finlay and Esteban (this volume) propose a transition between species with a biogeography and those which are ubiquitous at between 1 and 10 mm in length (see Fig. 10.1). In the case of testate amoebae, Wilkinson (2001) shows that this transition apparently occurs at around 100 mm, and indeed most of the organisms within the shaded area of Fig. 10.1 do have biogeographies, i.e. they are not cosmopolitan. As in the case of marine larval dispersers (see below), however, it is difficult here to identify the life-history stage where body size may be most relevant in shaping the evolution of dispersal. As with many wind-dispersed plant taxa (Fenner & Thompson, 2005), passive dispersers may be under selection for small propagule size, since these may be better dispersed. Unfortunately, data on propagule size in passively-dispersing freshwater metazoans are lacking or disparate (but see Poulin, 1995), making explicit tests of the relationship between propagule size and dispersal ability impossible at present. However, passive dispersal may also be related to the number of propagules produced, with fecundity related positively to dispersal potential. At the same time, if fecundity scales positively with body size in these passively dispersing freshwater taxa, we would predict a positive relationship between adult body size and dispersal ability, with larger-bodied taxa producing more propagules and being more widespread
BODY SIZE, DISPERSAL AND RANGE SIZE IN AQUATIC INVERTEBRATES
than small taxa, paralleling the scenario in marine environments but contrasting with Finlay (2002) and Finlay and Esteban’s (this volume) proposed model. Dispersal in marine systems In contrast to many of their freshwater counterparts, marine invertebrates are often benthic as adults and disperse as planktonic larvae. Three main types of developmental strategy have been described in marine taxa: (i) non-planktonic development, in which species are either viviparous, eggs and young develop demersally, or eggs are maternally brooded; (ii) lecithotrophy, in which a relatively large egg is produced and the yolk provisions the developing larva for at least part of its planktonic existence; and (iii) planktotrophy, in which the larvae derive their sustenance from exogenous sources, usually by predating other plankton. The division of development along such simple grounds, however, represents a considerable over-generalization (see Poulin, von Boletsky & Feral, 2001). For example, most planktotrophs develop feeding structures during their planktonic phase and therefore commence their development with maternal provisioning (Miner, McEdward & McEdward, 2005) whilst others which are primarily lecithotrophic may become facultatively planktotrophic (Emlet, McEdward & Strathmann, 1987). The relative frequency of different developmental modes differs both across clades of marine organisms, and across biogeographic regions within clades. Opisthobranch molluscs from the northeast Pacific are primarily planktotrophic (Goddard, 2004), whilst the relative proportions of lecithotrophs and brooders are higher in other groups such as echinoderms and prosobranchs in the same region. Groups such as peracarid crustaceans, meanwhile, are almost exclusively brooded, whilst most decapods have planktonic phases. The most profound patterns affecting dispersal mode are biogeographic trends in developmental pattern. Thorson’s rule (Mileikovsky, 1971) describes a trend towards decreased incidence of pelagic and planktotrophic development (towards direct demersal development) with increasing latitude (Collin, 2003). These trends mean that dominant modes and extents of dispersal may vary significantly with latitude. Most marine phyla exhibit a spectrum of modes of development, indicating that evolution of traits such as lecithotrophy and direct development have occurred separately in most of these groups (e.g. Jeffery & Emlet, 2003). It is difficult to justify the dichotomy of dispersal into active and passive modes in the marine environment. Whilst there are active components to planktonic existence, for example migrations within the water column, and small-scale dispersal by viviparous or brooding species can be highly effective (Kelaher, 2005), the majority of large-scale dispersal by both planktonic (both lecithotrophic and planktotrophic) and non-planktonic organisms is primarily passive. The duration (and therefore, potentially, scale) of pelagic transport may be influenced by a variety of both passive and active factors ranging from
191
192
S. D. RUNDLE ET AL.
current movements (Largier, 2003; Shanks, Largier & Brink, 2000; Siegel et al., 2003) to larval responses to their own energetic status or settlement cues (e.g. Marshall & Keough, 2003a). Non-planktonic species, meanwhile, may display both active dispersal by swimming or crawling, and passive dispersal, the latter being by means of rafting on floating objects (see Grantham, Eckert & Shanks, 2003). Given these considerations, it is possible to make some generalizations as to how body size could relate to dispersal ability in marine invertebrates. Some life-history models predict a dichotomy of egg sizes in planktonic marine invertebrates, with eggs giving rise to planktotrophic larvae having small optimal size and those giving rise to lecithotrophic larvae being optimally large (Vance, 1973, reviewed in Roff, 1992). However, empirical data do not support such a bimodal distribution of egg sizes, and instead often demonstrate a right-skewed or intermediate size distribution, even within wholly planktotrophic groups (Levitan, 2000). It remains possible, however, to make the generalization that, compared with lecithotrophic propagules (eggs and larvae), those of planktotrophic taxa tend to be smaller, spend longer suspended in the water column, and disperse further, especially in the case of specialized long-lived ‘teleplanic’ forms (Eckert, 1999; Largier, 2003; Shanks et al., 2000). Amongst planktotrophic dispersers, there is evidence for a positive correlation between adult body size and egg size within species (e.g. Marshall & Keough, 2003b; Miner et al., 2005), but little evidence for patterns between species, although a trade-off between size and number of eggs as a result of energetic and physical constraints might be expected. Second, there is a tendency within clades for small body size to be correlated with either brooding or lecithotrophy, especially amongst co-occurring species (Strathmann & Strathmann, 1982). A variety of hypothesized mechanisms exist to explain these relationships between life-history traits and dispersal patterns (Sewell, 1994). Chia (1974) hypothesized that limitations upon energetic reserves for reproduction in small-bodied taxa limit gamete production and hence favour viviparity, brooding or lecithotrophy. Strathmann and Strathmann (1982) cite three hypotheses to explain this trend: dispersal limitation, adult longevity and recruitment, and allometries in scaling of fecundity and space available for brooding. It is also possible that smaller, more specialized species may experience increased selection for retention of juveniles if their habitats are widely dispersed and the probability of successful dispersal to new habitats is low. Such a case has certainly been made for organisms inhabiting estuaries (Bilton, Paula & Bishop, 2002; Foggo, Frost & Attrill, 2003). Finally, it is probable that smallbodied organisms face a double jeopardy in employing broadcast spawning: quantities of sperm and eggs produced by small organisms may be inadequate to ensure fertilization (especially in denser waters at higher latitudes) (Yund, 2000). Also, the probability of successful development of relatively few
BODY SIZE, DISPERSAL AND RANGE SIZE IN AQUATIC INVERTEBRATES
planktotrophic propagules would be low given the attrition rates normally encountered in the plankton. Lecithotrophs and brooders might also be expected to face a higher extinction risk depending upon the availability of their habitats, and might be less likely to form widespread ‘open’ populations (Berryman, 2002; Camus & Lima, 2002; Grimm, Reise & Strasser, 2003). In general then, the trend is for increasing degrees of lecithotrophy, brooding and viviparity as organisms get smaller, with larger organisms in general tending towards planktotrophic development and broadcast spawning. We, therefore, would expect planktotrophic organisms to be the most widely dispersed marine forms with the greatest range sizes; this is supported for several published studies of both living (Scheltema, 1989; Kohn & Perron, 1994; Emlet, 1995) and fossil (Hansen, 1978; Jablonski, 1986) marine invertebrates. Lecithotrophic and brooding species should be less widely distributed, which is supported by the analyses presented by Emlet (1995). The picture is confused somewhat by the increased probability of passive transport in smaller lecithotrophic propagules and smaller brooding or direct-developing individuals, such that the overall particle size-range size distribution might be expected to be ‘U’- or ‘J’-shaped. However, if selection for intermediate egg size in planktotrophic species is indeed strong (sensu Levitan, 2000), then it is likely that the number, and planktonic duration, of dispersing propagules is the primary constraint upon dispersal in planktotrophic forms. Many taxa demonstrate trade-offs between egg number and size (see Llodra, 2002 for review), and significant relationships between adult female size and fecundity. Thus, larger individuals will tend to produce more eggs, and ultimately become more successful at dispersal through weight of numbers of dispersing propagules in a fashion similar to that proposed for passively-dispersing freshwater taxa.
Dispersal, macroecology and body size: case studies Active dispersers in freshwaters . . . several species of Sympetrum and Aeshna have most often been seen engaging in mass flights, sometimes settling in large numbers on ships at sea or flying in swarms, together with other migrating insects, through high mountain passes. (Askew, 1988.)
As the above quote suggests, the exemplary active dispersers in freshwater insects are the dragonflies and damselflies (Fig. 10.1), the adults of many species being known to disperse long distances. This ability to disperse is due to their highly aerodynamic wings, efficient flight muscles and very high flight muscle to body mass ratios (up to c. 60%; Marden, 1987). Despite the general efficiency of odonate flight, however, there are substantial differences between species in traits related to power production and efficiency that may have implications in terms of dispersal potential and could relate to body size or other morphological characteristics related to size (see above).
193
194
S. D. RUNDLE ET AL.
The odonates are ideally suited for investigating macroecological patterns and the potential role of body size and dispersal in influencing these patterns. First, they are used as models for exploring the biomechanics of flight (see above) and, hence, there is the potential for future studies to make robust macroecological predictions based on scaling relationships between body size and key traits related to dispersal ability (e.g. wing size and shape, mass of thoracic muscles, and ratios within muscle-lever systems). Second, the vivid markings and relatively large size of dragonflies and damselflies means that they are one of the groups of invertebrates that attract the most attention. Hence, they are well-known taxonomically and ecologically, and extensive distributional data exist for dragonflies and damselflies in many areas of the world, particularly the Holarctic. Data for North American damselflies and dragonflies have recently been made available electronically on the OdontataCentral database (Abbott, 2005) and these data were used to explore how range size related to patterns in relation to body size and wing size (see Fig. 10.2 for details). Based on some of the evidence presented above, significant positive relationships between size measures and distribution extent might have been expected. Although dragonflies were, on average, significantly larger than damselflies, they did not have significantly larger ranges (Fig. 10.2). However, within the dragonflies, there was some evidence for positive relationships between wing size and both range size and occupancy (Figs. 10.2c & d), no such relationships being evident for damselflies. Although these findings should be viewed with some caution, given the very high scatter around the derived regressions, they do suggest that the further exploration of such macroecological patterns, perhaps using more sophisticated regression techniques (e.g. McClain & Rex, 2001) may be instructive in revealing the potential role for traits linked to dispersal in macroecology. There have been a few previous attempts to relate dispersal potential to range size in freshwater invertebrates, with mixed findings. Ribera and Vogler (2000) consistently found that range size was significantly smaller in lotic versus lentic taxa in a comparison of clades of aquatic beetle differing in habitat preference. It was suggested that this macroecological pattern resulted from differing dispersal abilities in lentic versus lotic taxa, these differences being driven by differences in the duration of running (relatively long persistence) and standing waters (short persistence) over evolutionary timescales. Malmqvist (2000) examined the relationship of wing and body size with range size measures in mayflies and stoneflies in Sweden. He showed that there was no relationship between body size and range size, but that the ratio of wing length to body length was positively related to range extents in both groups. Hence, for at least two groups of freshwater taxa, there was evidence that a size-related trait might be related to distribution patterns. These results should be viewed with some caution, however, as the analysis was only partial (for most taxa studied, Sweden
BODY SIZE, DISPERSAL AND RANGE SIZE IN AQUATIC INVERTEBRATES
(b) 8
4
6
3
log occupancy
log range
(a)
4 2 0 1.3
1.5
1.7
2 1 0 1.3
1.9
1.5
log body length (c)
1.9
(d) 8
4
6
3
log occupancy
log range
1.7 log body length
4 2 0 1.1
1.3
1.5 log wing length
1.7
1.9
2 1 0 1.1
1.3
1.5
1.7
1.9
log wing length
Figure 10.2 The relationships of (a) body length (mm) and range size (km2), (b) body length and occupancy, (c) wing length and range size and (d) wing length and occupancy in dragonflies (closed circles) and damselflies (open circles) of the USA and Canada. Range size data were extracted from the OdonataCentral database (Abbott, 2005) by downloading distribution maps as JPEG files and then tracing and measuring ranges using image analysis software. Occupancy data were the number of site records for each species on this database. Stepwise regressions performed to explore the power of wing length and body length as predictors of area and occupancy demonstrated significant models for dragonflies only. For dragonfly range size and occupancy wing length was the only predictor entered into the model and the lines for these relationships are presented on the relevant panels: (c) R2 ¼ 0.14, y ¼ 1.31x þ 3.82, F1,213 ¼ 4.0, p ¼ 0.047; (d) R2 ¼ 0.19, y ¼ 0.993x þ 0.49, F1,213 ¼ 4.17, p ¼ 0.043.
represents only a small part of their range). In a recent analysis on a subset of the data collated from the OdonataCentral database Rundle et al. (2007) were able to demonstrate, for a much fuller analysis, a significant positive relationship between wing size and range size in North American species of the damselfly genus Enallagma (Fig. 10.3). The recent publication of a phylogeny for these damselflies (Turgeon et al., 2005) also facilitated an analysis that controlled for phylogenetic relatedness (Harvey & Pagel, 1991). This comparative analysis confirmed that evolutionary changes in wing size were correlated with those in range size (Fig. 10.3). As in Malmqvist’s (2000) study, and the analysis for North American dragonflies as a whole, there were no significant relationships
195
S. D. RUNDLE ET AL.
a)
b) 1.6 log range contrasts
7.5 7.0 log range
196
6.5 6.0 5.5 5.0 4.5 1.2
1.2 0.8 0.4 0
1.25
1.3 1.35 log wing length
1.4
0
0.01 0.02 0.03 0.04 log wing length contrasts
0.05
Figure 10.3 The relationship between: (a) wing length (mm) and range size (km2); and (b) wing length contrasts and range size contrasts from comparative analyses for Enallagma damselflies in North America. Analysis includes only species in the USA and Canada and range size data were collected as for Fig. 10.2. Contrast scores, representing independent, ancestral trait values were derived using Comparative Analysis by Independent Contrasts. Stepwise regressions (using body length and wing length and contrast scores for both as predictors) demonstrated significant relationships in both instances. The best model for the species comparison included both wing length and body length: R2 ¼ 0.57; y ¼ 16.1x 10.2x þ 0.71; F2,24 ¼ 16.9; p < 0.001, but the plot shown is for wing length only (a) y ¼ 7.84x 4.13, F1,24 ¼ 17.7, p < 0.001. For contrast scores only wing length was included in the model: (b) y ¼ 24.15x, F1,20 ¼ 60.1, p < 0.0001; here regression for contrast scores was forced through the origin (hence, there is no constant in the equation) and the ordinate axis was scaled to be positive.
with body length for either species comparisons or using contrast scores generated by independent contrasts analysis, which suggests that it is an increase in wing size rather than size per se that is correlated with increased range size in these freshwater invertebrates. Passive dispersers in freshwaters Empirical estimates of dispersal are either direct, involving mark-release-recapture, originally developed for population size estimation (Angelibert & Giani, 2003; Southwood & Henderson, 2000), or indirect, typically employing genetic analyses (see Bilton et al., 2001; Bullock, Kenward & Hails, 2002; Nathan et al., 2003 for reviews). Given the difficulty of obtaining direct estimates, many of the data on dispersal of aquatic invertebrates relate to the feasibility of dispersal rather than the frequency of successful dispersal events (Bilton et al., 2001). This is particularly true of passive freshwater dispersers, whose small, cryptic propagules do not lend themselves to direct study. In recent years, a combination of observational data and population genetic analysis have greatly expanded our understanding of dispersal in organisms such as bryozoans and zooplankton (see De Meester et al., 2002; Okamura & Freeland, 2002; Green & Figuerola, 2005; Figuerola et al., 2005), and several experimental studies have examined the relative ability of
BODY SIZE, DISPERSAL AND RANGE SIZE IN AQUATIC INVERTEBRATES
zooplankton species to disperse rapidly over local scales (Ca´ceres & Soluk, 2002; Cohen & Shurin, 2003; Largier, 2003; Louette & De Meester, 2005). Since the nineteenth century, many authors have considered the biogeography of selected passively-dispersing freshwater taxa – see for example accounts in Rundle, Robertson and Schmid-Araya, (2002a) for various meiofauna. Some of these studies have attempted to search for general macroecological patterns. Gillooly and Dodson (2000), for example, demonstrated that, for New World Cladocera, there was a peak in body size at around 60 degrees latitude in both the Northern and Southern Hemispheres and suggested that within these temperate regions body size was also maximal at intermediate temperatures during the year. However, there have been relatively few studies that attempt to assess macroecological patterns across taxa. Rundle, Bilton and Shiozawa (2000) and Rundle et al. (2002b) presented an analysis of distribution patterns at a variety of spatial scales (local-regional) employing parsimony analysis of endemicity to assess whether there were species assemblages characteristic of different areas of the globe. Analysis of the copepod family Canthocamptidae demonstrated that only 4% of described species were widespread (i.e. occurred in more than one biogeographical region – sensu Wallace, 1876). Galassi et al. (2002) suggest that even this may be an over-estimate, since recent taxonomic studies of microcrustacea (e.g. Galassi, Dole-Olivier & De Laurentiis, 1999; Stoch, 2001) tend to favour the splitting of species considered to be cosmopolitan on the basis of morphology alone, with genetic studies (e.g. Hebert & Finston, 1996) typically revealing additional phenotypically cryptic taxa. Despite the low number of widespread species, relationships between regions revealed from these analyses made biogeographic sense. At a global scale there was evidence of vicariant links, such as that between western Palaearctic and eastern Nearctic faunas, whereas within western Europe, there was a suggested role for post-glacial dispersal in shaping northern faunas. In relation to body size, it is worth noting that all species of Canthocamptidae are relatively small (lengths < 1000 mm), and that freshwater copepods are believed to disperse between water bodies largely via eggs, which are typically <50 mm in length. Such a size range is well below the transition to cosmopolitanism proposed by Finlay (2002) and Wilkinson (2001) (Fig. 10.1). Despite this, however, the vast majority of canthocamptids do indeed have a biogeography, and such a statement is true of most other meiofauna (Rundle et al., 2002a). Obviously, small dispersing propagules alone do not result in cosmopolitan distributions in freshwater taxa, and other factors such as population size, dispersal vector and the resistance of propagules must play an important role, as will the distribution of resulting dispersal distances themselves (Levin et al., 2003). Unfortunately, comparative data on such traits are lacking for freshwater copepods and, indeed, it is unclear how life-history features such as resting egg duration influence dispersal and range size. Many species of zooplankton
197
S. D. RUNDLE ET AL.
produce long-lived diapausing egg banks (De Stasio, 1989; Hairston et al., 1995; Ca´ceres, 1998; Brendonck & De Meester, 2003; Vandekerkhove et al., 2005), which lower local extinction risk. Such dormancy can be seen to represent dispersal in time, and theoretical and empirical data predict trade-offs between spatial and temporal dispersal (Venable & Lawlor, 1980; Levin, Cohen & Hastings, 1984; Hairston & Ca´ceres, 1996), suggesting that dispersal rate may actually be lower in taxa whose eggs can remain dormant for long periods. This may impact on the timing of local colonization events (Ca´ceres & Soluk, 2002), but over larger spatial and temporal scales the production of resting propagules is likely to lead to larger ranges due to increased propensity for long-distance dispersal. Adult body size is well documented for the canthocamptid species considered by Rundle et al. (2000; 2002b), and reanalysis of these data reveal a consistent tendency for widespread taxa to be larger than their restricted relatives, at both a global and European scale (Fig. 10.4). Mechanisms that may underlie this apparent trend are unclear, but may relate to relative reproductive output: if larger-bodied taxa have higher fecundity this may increase the likelihood of long-range dispersal. Studies of other groups such as Cladocera would be useful here, as would comparative data on egg size, survivorship and estimates of effective dispersal distances. As discussed above, the dispersal biology of many passively-dispersing freshwater taxa has similarities with that in seed plants. Levin et al. (2003) note that modelling studies of seed dispersal demonstrate that both mean dispersal distance and the overall distribution of dispersal distance are critical in
0.8
Adult body length (mm)
198
0.7
0.6
0.5
0.4
European
Global
Figure 10.4 Mean body size (SE) of freshwater canthocamptid copepods, which can be considered as geographically restricted, i.e. found in one area of endemicity (white bars), or widespread, i.e. found in >1 area of endemicity (grey bars), at global and Western European scales. Widespread species were significantly larger for both global (T40 ¼ 2.02, p ¼ 0.015) and European (T81 ¼ 1.99, p < 0.050) data.
BODY SIZE, DISPERSAL AND RANGE SIZE IN AQUATIC INVERTEBRATES
determining rates of spread. No attempt has been made explicitly to model spread in passively dispersing freshwater taxa, but future studies should focus on this, and on including the influence of dispersal vectors (cf. animal dispersal of seeds), fecundity and propagule resistance into rates of spread. Hard data on these parameters are lacking in freshwater taxa, and at present our empirical understanding of the comparative dispersal biology of passive dispersers in freshwaters is cursory, limiting our ability to explore the possible influence of body size. Dispersal in marine systems There have been few studies of body size and range size in marine systems (Gaston, 2003) and fewer still that consider invertebrates and their modes of development and dispersal. For marine invertebrates, passive dispersal via planktotrophic larvae represents the most effective means of achieving long-distance transport. However, the probability of successful recruitment to adulthood from such meroplanktonic or teleplanic lifestages is low, and broadcast spawners adopting such a strategy must compensate by producing larger numbers of dispersing propagules. For brooding and viviparous species such as Amphipoda, active movement, whilst effective at small spatial scales, is likely to contribute little to their large-scale dispersal (Costa et al., 2004; Kelaher, 2005), with rafting and other passive means of dispersal dominating (Thiel & Gutow, 2005). We might draw comparisons, therefore, between all marine invertebrates (given that passive dispersal dominates in this realm) and passively dispersing freshwater taxa, with the caveat that size spectra in these two groups overlap little, if at all, with the marine organisms representing orders of magnitude greater adult body sizes than their freshwater counterparts. Given that selection for egg sizes in planktotrophic taxa tends to produce intermediate sizes of egg across species (Levitan, 2000), and that there is no overall observed relationship between egg size and adult body size (Eckert, 1999), we might expect no relationship to result between range size and body size in passively dispersing marine taxa. Larger propagules should develop faster and have a shorter pelagic larval duration and, thus, we might expect an egg size-range size relationship to exist amongst planktotrophic taxa. However, larger species are generally longer lived and more fecund (Llodra, 2002), and therefore generate more dispersing individuals. Goodwin et al. (2005) demonstrate that in the case of marine egg-laying teleosts, geographic range indeed scales positively with body size, and Reaka (1980) reported a similar relationship for Stomatopods (most species initially brood their eggs, which then hatch and the larvae disperse planktonically). These trends are probably related to adult fecundity rather than to the size of the dispersing particle.
199
S. D. RUNDLE ET AL.
600 500 Egg/larval size (µm)
200
400 300 200 100 0 0
20
40 Adult shell length (mm)
60
80
Figure 10.5 Relationships between adult shell length and egg size (open circles, dashed line) and hatching larvae (closed circles, solid line) for a set of planktotrophically dispersing calyptraeid gastropods: eggs R2 ¼ 0.342, F1,28 ¼ 14.58, P ¼ 0.001; larvae R2 ¼ 0.328, F1,19 ¼ 9.28, P ¼ 0.007. Data from Collin (2003).
Re-analysis of data from Collin (2003) describing sizes and distributions of planktotrophically dispersing calyptraeid gastropods, shows a weak relationship between adult body size and both eggs and hatching larvae (Fig. 10.5), although there is no evidence for an effect of adult size upon range. A similar pattern linking egg size to adult (i.e. test) size is evident in re-analysis of data describing regular echinoids (composite data from Emlet, 1995; Levitan, 2000; Moyse & Tyler, 1990; Meinkoth, 1981) (Fig. 10.6a). However, there is again no apparent trend in either of these cases for an effect of egg or adult size upon overall geographic range (Figs. 10.6b & c), although Emlet (1995) does demonstrate a clear effect of reproductive/dispersive strategy upon range size, with planktotrophs having significantly greater geographic ranges than non-feeding planktonic dispersers. Foggo et al. (unpublished) summarize trends in abundance, distribution and body size for British marine macroinvertebrates (see Foggo et al., 2003 for methods and data sources). These data were examined for evidence of patterns linking adult body size to a measure of range (proportion of sampled sites occupied). Polychaetes represent a group with a range of reproductive/dispersive strategies and, although there was no evidence of linear trends in range size with increasing body size, there were clear differences between non-planktonic, lecithotrophic and planktotrophic developers in terms of both the mean proportion of sites they occupied (Fig. 10.7a) or their mean body lengths (Fig. 10.7b). Thus, there appears to be a general trend for smaller passively dispersed marine invertebrates to be less widespread, echoing the trends encountered in some freshwater passive dispersers (see above). The comparison between passive dispersal in marine and freshwater systems seems at the outset to be somewhat contrived. Freshwaters are highly disjunct
BODY SIZE, DISPERSAL AND RANGE SIZE IN AQUATIC INVERTEBRATES
(a)
0.2
Egg diameter (mm)
0.16 0.12 0.08 0.04 0 0
50 100 Adult test diameter (mm)
150
(b)
25 000
Range size (km)
20 000 15 000 10 000 5000 0 0.05
0.07
0.09
0.11 0.13 0.15 Egg diameter (mm)
0.17
0.19
(c)
25 000
Range size (km)
20 000 15 000 10 000 5000 0 0
50 100 Adult test diameter (mm)
150
Figure 10.6 Relationships between: (a) adult test diameter and egg size (R2 ¼ 0.322, F1,18 ¼ 8.56, P ¼ 0.009), (b) egg diameter and adult range size, and (c) adult test diameter and range size in regular echnoids. Data from various sources (see text).
201
S. D. RUNDLE ET AL.
(a)
Proportional site occupancy
0.07 0.06 0.05 0.04 0.03 0.02 0.01 0 Planktotrophs
Lecithotrophs
Non-planktonic
Lecithotrophs
Non-planktonic
(b) 120 100 Body length (mm)
202
80 60 40 20 0
Planktotrophs
Developmental mode
Figure 10.7 Mean (SE) proportional site occupancy (F2,125 ¼ 1.66, ns) (a) and body length (F2,125 ¼ 0.69, ns) (b) of planktotrophic (n ¼ 66), lecithotrophic (n ¼ 34) and non-planktonic (n ¼ 26) polychaetes. Planktotrophic taxa are more widespread (ANOVA p < 0.05, data subjected to p Asin (x) transformation to normalize variances) than non-planktonic dispersers, and tend to be larger. Data from Foggo et al. (unpublished).
systems, discrete and surrounded by inimical habitat. They have relatively low persistence on both ecological and geological timescales compared with marine systems, which are also highly connected (Moss, 1998). In addition, marine populations might be viewed as effectively open systems, with lower probabilities of local extinctions occurring, especially in species dispersing planktonically (Eckert, 2003). Marine habitats have historically been more stable and less disturbed than terrestrial or freshwater systems (Moss, 1998), and may therefore be less shaped overall by dispersal patterns modes and hence by body sizes. However, the parallel between the two systems may be most valid at the level of dispersal viewed as a lottery with a high risk associated. The greater the number of tickets a species buys, the greater the probability of success, thus
BODY SIZE, DISPERSAL AND RANGE SIZE IN AQUATIC INVERTEBRATES
the key effect of body size in these passive dispersing species is upon fecundity. In marine systems it is also possible that the main consequences of body size for dispersal act through dictating reproductive-dispersive strategies.
Conclusions In the case of freshwater environments the data we present suggest that there may be some positive relationships between range size and body size or traits that may be linked to dispersal. For passive dispersers, rather than metabolic costs associated directly with dispersal there may be a link between propagule output and body size – a possibility worth further investigation. For at least some active dispersers, there is evidence that wing size rather than body size per se is of particular importance. We would suggest that aquatic insects could provide a valuable model system in which scaling laws for flight (e.g. Schilder & Marden, 2004) could be used alongside those for metabolism for making robust macroecological predictions. In the marine environment links between dispersal, range and body size are most likely to be moderated by developmental mode and lifetime reproductive output. Smaller organisms, tending towards viviparity or brooding, invest more in each individual offspring and have more rapid generation times. Such a strategy is likely to be less successful for dispersing over long distances than the possession of a planktonic stage. Body size is likely to manifest itself more through relationships with fecundity, as larger organisms will be able to produce greater numbers of eggs. This is, however, less likely to result in strong range size–body size correlations due to the kind of allometries in fecundity and size of area for brooding described by Strathmann and Strathmann (1982). In intermediate size organisms that produce lecithotrophic larvae the size and number of dispersing particles will dictate range, with the trade-off between maternal investment (size, and hence developmental time spent in the plankton; Levitan, 2000) and fecundity being key. Larger organisms will be able to generate larger numbers of eggs irrespective of egg size, and this effect of fecundity is likely to result in any observed trends between body size and range. In marine planktotrophs there is both theoretical (Levitan, 2000) and empirical (Eckert, 1999) evidence that no general body size–egg/larval size relationship exists, although some taxa do seem to indicate such a trend (see Figs. 10.5 & 10.6). No inherent ‘dispersability–body size’ relationship can be predicted for individual species groups, as is the case of freshwater dispersers. Body size–range size relationships in these taxa may instead be determined by fecundity, which in turn is likely to scale directly with body size, invoking the same basis in metabolic rules as proposed for other macroecological patterns (Gillooly, et al., 2001; Marquet et al., 2005). In summary, we hope that we have demonstrated that the exploration of macroecological patterns in aquatic invertebrates can be profitable and that it
203
204
S. D. RUNDLE ET AL.
may be possible to use dispersal as an organismal property that links body size and distribution patterns. Certainly, the production of databases such as OdonataCentral for other groups of aquatic invertebrates and for other regions will allow the continued search for links between dispersal propensity, body size and range size. It is key, however, that the exploration of these patterns is driven by robust predictions based on traits that are likely to influence dispersal ability and that some attempts are made to gather empirical data on dispersal.
References Abbott, J. C. (2005) OdonataCentral: an online resource for the Odonata of North America. Austin, Texas. Available at http://www. odonatacentral.com. (Accessed April 18, 2005.) Allen, A. P., Brown, J. H. & Gillooly, J. F. (2002). Global biodiversity, biochemical kinetics and the energetic-equivalence rule. Science, 297, 1545–1548. Angelibert, S. & Giani, N. (2003). Dispersal characteristics of three odonate species in a patchy habitat. Ecography, 26, 13–20. Askew, R. R. (1988). The Dragonfiles of Europe. Colchester: Harley Books. Berryman, A. A. (2002). Population: a central concept for ecology? Oikos, 97, 439–442. Bilton, D. T., Freeland, J. R. & Okamura, B. (2001). Dispersal in freshwater invertebrates. Annual Review of Ecology and Systematics, 32, 159–181. Bilton, D. T., Paula, J. & Bishop, J. D. D. (2002). Dispersal, genetic differentiation and speciation in estuarine organisms. Estuarine, Coastal and Shelf Science, 55, 937–952. Blackburn, T. M. & Gaston, K. J. (2003). Macroecology Concepts and Consequences. Oxford: Blackwell Science. Bohonak, A. J. (1999). Effect of insect-mediated dispersal on the genetic structure of postglacial water mite populations. Heredity, 82, 451–461. Bohonak, A. J., Smith, B. P. & Thornton, M. (2004). Distributional, morphological and genetic consequences of dispersal for temporary pond water mites. Freshwater Biology, 49, 170–180.
Brendonck, L. & De Meester, L. (2003). Egg banks in freshwater zooplankton: evolutionary and ecological archives in the sediment. Hydrobiologia, 491, 65–84. Brendonck, L. & Riddoch, B. J. (1999). Wind-borne short-range egg dispersal in anostracans (Crustacea: Branchiopoda). Biological Journal of the Linnean Society, 67, 87–95. Brown, J. H. & Maurer, B. A. (1989). Macroecology: the division of food and space among species on continents. Science, 243, 1145–1150. Bullock, J. M., Kenward, R. E. & Hails, R. S. (eds.) (2002). Dispersal Ecology. Oxford: Blackwell Science. ´ Caceres, C. E. (1998). Interspecific variation in the abundance, production, and emergence of Daphnia diapausing eggs. Ecology, 79, 1699–1710. Ca´ceres, C. E. & Soluk, D. A. (2002). Blowing in the wind: a field test of overland dispersal and colonization by aquatic invertebrates. Oecologia, 131, 402–408. Camus, A. P. & Lima, M. (2002). Populations, metapopulations, and the open-closed dilemma: the conflict between operational and natural population concepts. Oikos, 97, 433–438. Chia, F. S. (1974). Classification and adaptive significance of developmental patterns in marine invertebrates. Thalassia Jugoslavica, 10, 121–130. Clobert, J., Danchin, E., Dhondt, A. A. & Nichols, J. D (eds.) (2001). Dispersal. Oxford: Oxford University Press.
BODY SIZE, DISPERSAL AND RANGE SIZE IN AQUATIC INVERTEBRATES
Cohen, G. M. & Shurin, J. B. (2003). Scaledependence and mecanisms of dispersal in freshwater zooplankton. Oikos, 103, 603–617. Collin, R. (2003). Worldwide patterns in mode of development in calyptraeid gastropods. Marine Ecology Progress Series, 247, 103–122. Compton, S. G. (2002). Sailing with the wind: dispersal by small flying insects. In Dispersal Ecology, ed. J. M. Bullock, R. E. Kenward and R. S. Hails. Oxford: Blackwell Science, pp. 113–133. Costa, F. O., Neuparth, T., Theodorakis, C. W., Costa, M. H. & Shugart, L. R. (2004). RAPD analysis of southern populations of Gammarus locusta: comparison with allozyme data and ecological inferences. Marine Ecology Progress Series, 277, 197–207. Dahms, H.-U. (1995). Dormancy in the Copepoda – an overview. Hydrobiologia, 306, 199–211. Darwin, C. (1859). On the Origin of Species by Means of Natural Selection. London: John Murray. De Meester, L., Gomez, A., Okamura, B. & Schwenk, K. (2002). The monopolization hypothesis and the dispersal-gene flow paradox in aquatic organisms. Acta Oecologica, 23, 121–135. De Stasio, B. T. (1989). The seed bank of a freshwater crustaean: copepodology for the plant ecologist. Ecology, 70, 1377–1389. Dumont, H. J. (1994). The distribution and ecology of the fresh- and brackish-water medusae of the world. Hydrobiologia, 272, 1–12. Eckert, G. L. (1999). A comparative analysis of egg size in marine invertebrates: relationships with development mode, planktonic period and adult size. American Zoologist, 39, 39. Eckert, G. L. (2003). Effects of the planktonic period on marine population fluctuations. Ecology, 84, 372–383. Ellington, C. P. (1991). Limitation on animal flight performance. Journal of Experimental Biology, 160, 71–91. Emlet, R. B. (1995). Developmental mode and species geographic range in regular
sea-urchins (Echinodermata, Echinoidea). Evolution, 49, 476–489. Emlet R. B., McEdward L. B. & Strathmann R. R. (1987). Echinoderm larval ecology viewed from the egg. In Echinoderm Studies, vol. 2, ed. M. Jangoux and J. M. Lawrence, Rotterdam: Balkema, pp. 55–136. Enquist, B. J., Economo, E. P., Huxman, T. E. et al. (2003). Scaling metabolism from organisms to ecosystems. Nature, 423, 639–642. Fenner, M. & Thompson, K. (2005). The Ecology of Seeds. Cambridge: Cambridge University Press. Ferriere, R., Belthoff, J. R., Olivieri, I. & Krackow, S. (2000). Evolving dispersal: where to go next? Trends in Ecology and Evolution, 15, 5–7. Figuerola, J., Green, A. J. & Michot, T. C. (2005). Invertebrate eggs can fly: evidence of waterfowl-mediated gene flow in aquatic invertebrates. American Naturalist, 165, 274–280. Finlay, B. J. (2002). Global dispersal of free-living microbial eukaryote species. Science, 296, 1061–1063. Foggo, A., Frost, M. T. & Attrill, M. J. (2003). Abundance-occupancy patterns in British estuarine macroinvertebrates. Marine Ecology Progress Series, 265, 297–302. Galassi, D. M. P., Dole-Olivier, M-J. & De Laurentiis, P. (1999). Phylogeny and biogeography of the genus Pseudectinosoma and description of Pseudectinosoma janineae sp. N. (Crustacea, Copepoda, Ectinosomatidae). Zoologica Scripta, 28, 289–303. Galassi, D. M. P., Marmonier, P., Dole-Olivier, M-J. & Rundle, S. D. (2002). Microcrustacea. In Freshwater Meiofauna Biology and Ecology, ed. S. D. Rundle, A. L. Robertson and J. M. Schmid-Araya. Leiden: Backhuys Publishers, pp. 135–175. Gaston, K. J. (2003). The Structure and Dynamics of Geographic Ranges. Oxford: Oxford University Press.
205
206
S. D. RUNDLE ET AL.
Gaston, K. J. & Blackburn, T. M. (2000). Pattern and Process in Macroecology. Oxford: Blackwell Science. Gillooly, J. F. & Dodson, S. I. (2000). Latitudinal patterns in the size distribution and seasonal dynamics of new world, freshwater cladocerans. Limnology and Oceanography, 45, 22–30. Gillooly, J. F., Brown, J. H., West, G. B., Savage, V. M. & Charnov, E. L. (2001). Effects of size and temperature on metabolic rate. Science, 293, 2248–2251. Goddard, J. H. R. (2004). Developmental mode in benthic opisthobranch molluscs from the northeast Pacific Ocean: feeding in a sea of plenty. Canadian Journal of Zoology, 82, 1954–1968. Goodwin, N. B., Dulvy, N. K. & Reynolds, J. D. (2005). Macroecology of live-bearing in fishes: latitudinal and depth range comparisons with egg-laying relatives. Oikos, 110, 209–218. Grantham, B. A., Eckert, G. L. & Shanks, A. L. (2003). Dispersal potential of marine invertebrates in diverse habitats. Ecological Applications, 13, Supplement: S108–S116. Green, A. J. & Figuerola, J. (2005). Recent advances in the study of long-distance dispersal of aquatic invertebrates by birds. Diversity and Distributions, 11, 149–156. Grimm, V., Reise, K. & Strasser, M. (2003). Marine metapopulations: a useful concept? Helgoland Marine Research, 56, 222–228. Guitie´rrez, D. & Mene´ndez, R. (1997). Patterns in the distribution, abundance and body size of carabid beetles (Coleoptera: Caraboidea) in relation to dispersal ability. Journal of Biogeography, 24, 903–914. Hairston, N. G. J. & Ca´ceres, C. E. (1996). Distribution of crustacean diapause: micro- and macroevolutionary pattern and process. Hydrobiologia, 320, 27–44. Hairston, N. G. J., Van Brunt, R. A., Kearns, C. M. & Engstrom, D. R. (1995). Age and
survivorship of diapausing eggs in a sediment egg bank. Ecology, 76, 1706–1711. Hansen, T. A. (1978). Larval dispersal and species longevity in lower tertiary gastropods. Science, 199, 885–887. Harrison, J. F. & Roberts, S. P. (2000). Flight respiration and energetics. Annual Review of Physiology, 62, 179–205. Harvey, P. H. & Pagel, M. D. (1991). The Comparative Method in Evolutionary Biology. Oxford: Oxford University Press. Hebert, P. D. N. & Finston, T. L. (1996). Genetic differentiation in Daphnia obtusa: a continental perspective. Freshwater Biology, 35, 311–321. Jablonski, D. (1986). Larval ecology and macroevolution in marine-invertebrates. Bulletin of Marine Science, 39, 565–587. Jeffery, C. H. & Emlet, R. B. (2003). Macroevolutionary consequences of developmental mode in temnopleurid echinoids from the tertiary of southern Australia. Evolution, 57, 1031–1048. Jetz, W., Carbone, C., Fulford, J. & Brown, J. H. (2004). The scaling of animal space use. Science, 306, 266–268. Johnson, C. G. (1969). Migration and Dispersal of Insects by Flight. London: Methuen & Co. Kelaher, B. P. (2005). Does colonization contribute to spatial patterns of common invertebrates in coralline algal turf ? Austral Ecology, 30, 40–48. Kohn, A. J. & Perron, F. E. (1994). Life History and Biogeography: Patterns in Conus. Oxford: Oxford University Press, pp. 57–67. Korovchinsky, N. M. & Boikova, O. S. (1996). The resting eggs of Ctenopoda (Crustacea: Branchiopoda): a review. Hydrobiologia, 320, 131–140. Largier, J. L. (2003). Considerations in estimating larval dispersal distances from oceanographic data. Ecological Applications, 13, S71–S89. Levin, S., Cohen, D. & Hastings, A. (1984). Dispersal strategies in patchy environments. Journal of Population Biology, 26, 165–191.
BODY SIZE, DISPERSAL AND RANGE SIZE IN AQUATIC INVERTEBRATES
Levin, S. A., Muller-Landau, H. C., Nathan, R. & Chave, J. (2003). The ecology and evolution of seed dispersal: a theoretical perspective. Annual Review of Ecology and Systematics, 34, 575–604. Levitan, D. R. (2000). Optimal egg size in marine invertebrates: theory and phylogenetic analysis of the critical relationship between egg size and development time in echinoids. American Naturalist, 156, 175–192. Llodra, E. R. (2002). Fecundity and life-history strategies in marine invertebrates. Advances in Marine Biology, 43, 87–170. Louette, G. & De Meester, L. (2005). High dispersal capacity of cladoceran zooplankton in newly founded communities. Ecology, 86, 353–359. Maguire, B. Jr. (1963). The passive dispersal of small aquatic organisms and their colonisation of isolated bodies of water. Ecological Monographs, 33, 161–185. Malmqvist, B. (2000). How does wing length relate to distribution patterns of stoneflies (Plecoptera) and mayflies (Ephemeroptera)? Biological Conservation, 93, 271–276. Marden, J. H. (1987). Maximum lift production during takeoff in flying animals. Journal of Experimental Biology, 130, 235–258. Marden, J. H. (1994). From damselflies to pterosaurs: how burst and sustainable flight performance scale with size. American Journal of Physiology, 266, R1077–R1084. Marquet, P. A., Quinones, R. A., Abades, S. et al. (2005). Scaling and power-laws in ecological systems. Journal of Experimental Biology, 208, 1749–1769. Marshall, D. J. & Keough, M. J. (2003a). Variation in the dispersal potential of non-feeding invertebrate larvae: the desperate larva hypothesis and larval size. Marine Ecology Progress Series, 255, 145–153. Marshall, D. J. & Keough, M. J. (2003b). Sources of variation in larval quality for free-spawning marine invertebrates: egg size and the local sperm environment. Invertebrate Reproduction and Development, 44, 63–70.
McClain, C. R. & Rex, M. A. (2001). The relationship between dissolved oxygen concentration and maximum size in deep-sea turrid gastropods: an application of quantile regression. Marine Biology, 139, 681–685. McLachlan, A. (1985). The relationship between habitat predictability and wing length in midges. Oikos, 44, 391–397. Meinkoth, N. A. (1981). National Audubon Society Field Guide to North American Seashore Creatures. New York: A. A. Knopf. Mileikovsky, S. A. (1971). Types of larval development in marine bottom invertebrates, their distribution and ecology significance: a re-evaluation. Marine Biology, 10, 193–213. Miner, B. G., McEdward, L. A. & McEdward, L. R. (2005). The relationship between egg size and the duration of the facultative feeding period in marine invertebrate larvae. Journal of Experimental Marine Biology and Ecology, 321, 135–144. Moss, B. (1998). Ecology of Fresh Waters, Man and Medium, Past to Future. Oxford: Blackwell. Moyse, J. & Tyler, P. A. (1990). Echinodermata. In The Marine Fauna of the British Isles and NorthWest Europe, Vol. 2., ed. P. J. Hayward and J. S. Ryland. Oxford: Clarendon Press. Nathan, R., Perry, G., Cronin, J. T., Strand, A. E. & Cain, M. L. (2003). Methods for estimating long-distance dispersal. Oikos, 103, 261–273. Okamura, B. & Freeland, J. R. (2002). Gene flow and the evolutionary ecology of passively dispersing aquatic invertebrates. In Dispersal Ecology, ed. J. M. Bullock, R. E. Kenward and R. S. Hails. Oxford: Blackwell Science, pp. 194–216. Paradis, E., Baillie, S. R., Sutherland, W. J. & Gregory, R. D. (1998). Patterns of natal dispersal in birds. Journal of Animal Ecology, 67, 518–536. Poulin, R. (1995). Clutch size and egg size in free-living and parasitic copepods: a comparative analysis. Evolution, 49, 325–336. Poulin, E., von Boletzky, S. & Feral, J. P. (2001). Combined ecological factors permit
207
208
S. D. RUNDLE ET AL.
classification of developmental patterns in benthic marine invertebrates: a discussion note. Journal of Experimental Marine Biology and Ecology, 257, 109–115. Proctor, V. W. & Malone, C. (1965). Further evidence of the passive dispersal of small aquatic organisms via the intestinal tract of birds. Ecology, 46, 728–729. Pyron, M. (1999). Relationships between geographical range size, body size, local abundance, and habitat breadth in North American suckers and sunfishes. Journal of Biogeography, 26, 549–558. Reaka, M. L. (1980). Geographic range, life history patterns, and body size in a guild of coral-dwelling mantis-shrimps. Evolution, 34, 1019–1030. Ribera, I. & Vogler, A. P. (2000). Habitat type as a determinant of species range sizes: the example of lotic-lentic differences in aquatic Coleoptera. Biological Journal of the Linnean Society, 71, 33–52. Ribera, I., Foster, G. N. & Vogler, A. P. (2003). Does habitat use explain large scale species richness patterns of aquatic beetles in Europe? Ecography, 26, 145–152. Riddoch, B. J., Mpoloka, S. W. & Cantrell, M. (1994). Genetic variation and localized gene flow in the fairy shrimp, Branchipodopsis wolfi, in temporary rainwater pools in southern Botswana. In Genetics and Evolution of Aquatic Organisms, ed. A. R. Beaumant. London: Chapman and Hall, pp. 96–102. Roff, D. A. (1992). The Evolution of Life Histories: Theory and Analysis. London: Chapman and Hall. Rosenfield, J. A. (2002). Pattern and process in the geographical ranges of freshwater fishes. Global Ecology and Biogeography, 11, 323–332. Rundle, S. D., Bilton, D. T. & Shiozawa, D. K. (2000). Global and regional patterns in lotic meiofauna. Freshwater Biology, 44, 123–134. Rundle, S. D., Robertson, A. L. & Schmid-Araya, J. M. (eds.) (2002a). Freshwater Meiofauna
Biology and Ecology. Leiden: Backhuys Publishers. Rundle, S. D., Bilton, D. T., Galassi, D. & Shiozawa, D. K. (2002b). The geographical ecology of freshwater meiofauna. In Freshwater Meiofauna Biology and Ecology, ed. S. D. Rundle, A. L. Robertson and J. M. Schmid-Araya. Leiden: Backhuys Publishers, pp. 279–293. Rundle, S. D., Bilton, D. T., Abbott, J. C. & Foggo, A. (2007). Range size in North American Enallagma damselflies correlates with wing size. Freshwater Biology, 52, 471–477. Scheltema, R. S. (1989). Planktonic and non-planktonic development among prosobranch gastropods and its relationship to the geographic range of species. In Reproduction, Genetics and Distributions of Marine Organisms, ed. J. S. Ryland and P. A. Tyler. Fredensborg: Olsen and Olsen, pp. 183–188. Schilder, R. J. & Marden, J. H. (2004). A hierarchical analysis of the scaling of force and power production by dragonfly flight motors. Journal of Experimental Biology, 207, 767–776. Sewell, M. A. (1994). Small-size, brooding, and protandry in the apodid sea-cucumber Leptosynapta clarki. Biological Bulletin, 187, 112–123. Shanks, A., Largier, J. & Brink, L. (2000). Demonstration of the onshore transport of larval invertebrates by the shoreward movement of an upwelling front. Limnology and Oceanography, 45, 230–236. Siegel, D. A., Kinlan, B. P., Gaylord, B. & Gaines, S. D. (2003). Lagrangian descriptions of marine larval dispersion. Marine Ecology Progress Series, 260, 83–96. Southwood, T. R. E. & Henderson, P. A. (2000). Ecological Methods, 3rd edn. Oxford: Blackwell Science. Stoch, F. (2001). How many species of Diacyclops? New taxonomic characters and species richness in a freshwater cyclopid genus
BODY SIZE, DISPERSAL AND RANGE SIZE IN AQUATIC INVERTEBRATES
(Copepoda, Cyclopoida). Hydrobiologia, 453, 525–531. Strathmann, R. R. & Strathmann, M. F. (1982). The relationship between adult size and brooding in marine invertebrates. American Naturalist, 119, 91–101. Tanaka, S. & Suzuki, Y. (1998). Physiological trade-offs between reproduction, flight capability and longevity in a wing-dimorphic cricket, Modicogryllus confirmatus. Journal of Insect Physiology, 44, 121–129. Taylor, C. M. & Gotelli, N. J. (1994). The macroecology of Cyprinella: correlates of phylogeny, body size, and geographical range. American Naturalist, 144, 549–569. Thiel, M. & Gutow, L. (2005). The ecology of rafting in the marine environment-I – The floating substrata. Oceanography and Marine Biology: An Annual Review, 42, 181–263. Travis, J. M. J. & Dytham, C. (1999). Habitat persistence, habitat availability and the evolution of dispersal. Proceedings of the Royal Society of London B, 266, 723–728. Turgeon, J., Stoks, R., Thum, R. A., Brown, J. M. & McPeek, M. A. (2005). Simultaneous quaternary radiations of three damselfly clades across the Holarctic. American Naturalist, 165, E78–E107. Vance, R. R. (1973). On reproduction strategies in marine benthic invertebrates. American Naturalist, 107, 339–352. Vandekerkhove, J., Declerck, S., Bendonck, L., et al. (2005). Hatching of cladoceran resting eggs: temperature and photoperiod. Freshwater Biology, 50, 96–104.
Venable, D. L. & Lawlor, L. (1980). Delayed germination and dispersal in desert annuals: escape in space and time. Oecologia, 46, 272–282. Vogler, A. P. & Ribera, I. (2003). Evolutionary analysis of species richness patterns in aquatic beetles: why macroecology needs a historical perspective. In Macroecology Causes and Consequences, ed. T. M. Blackburn and K. J. Gaston. Oxford: Blackwell Science, pp. 17–30. Wakeling, J. M. & Ellington, C. P. (1997). Dragonfly flight III. Lift and power requirements. Journal of Experimental Biology, 200, 583–600. Wallace, A. R. (1876). The Geographical Distribution of Animals. London: MacMillan. Wilkinson, D. M. (2001). What is the upper size limit for cosmopolitan distribution in free-living microorganisms? Journal of Biogeography, 28, 285–291. Woodward, G., Ebenman, B., Emmerson, M., et al. (2005). Body size in ecological networks. Trends in Ecology and Evolution, 20, 402–409. Yund, P. O. (2000). How severe is sperm limitation in natural populations of marine free-spawners? Trends in Ecology and Evolution, 15, 10–13. Zera, A. J. & Brink, T. (2000). Nutrient absorbtion and utilization by wing and flight muscle morphs of the cricket Gryllus firmu: implications for the trade-off between flight capability and early reproduction. Journal of Insect Physiology, 46, 1207–1218.
209
CHAPTER ELEVEN
Body size and diversity in marine systems RICHARD M . WARWICK Plymouth Marine Laboratory
Introduction Much has been written concerning the relationship between body size and biological traits, mostly concerning the terrestrial situation. There is no reason to suppose that many of these relationships will be different in the sea; for example quarter-power scaling with body mass applies to virtually all organisms (West, Brown & Enquist, 1999). For marine animals, metabolic rate and production scales at three-quarters power (e.g. Brey, 1990; Warwick & Price, 1979), while it is likely that life span increases in proportion to body mass raised to the power of one quarter, although so little is known about the natural history of marine animals that this latter relationship cannot yet be established. On the other hand, the very different phyletic composition of terrestrial and marine faunas, and the big differences in life-history characteristics, suggest that relationships between body size and diversity will differ between these two realms. The relationship between body size and diversity is fraught with uncertainties and inconsistencies. Hutchinson (1959) suggested that ‘. . . small size, by permitting animals to become specialised to the conditions offered by small diversified elements of the environmental mosaic, clearly makes possible a degree of diversity quite unknown among groups of larger organisms’. However, it is now suggested that the spatial and temporal structure of the physical environment is fractal (Bell et al., 1993 and references therein; see Schmid & Schmid-Araya, this volume), and if habitat complexity largely determines species diversity this leads to the prediction (for a single perfect fractal) that all organisms, regardless of size, will perceive the environment as equally complex and should have equivalent diversity. A comprehensive study by Orme et al. (2002) did not support any correlation between the median body size of a phylum and the number of species in that phylum, contrary to the Hutchinsonian view that the smallest bodied taxa are the most diverse (Van Valen, 1973; May, 1986; Kochmer & Wagner, 1988) or more recent suggestions that diversity is highest in taxa of intermediate size (Dial & Marzluff, 1988; Fenchel, 1993; Siemann, Tilman & Harrstad, 1996; Etienne & Olff, 2004). Uncertainty also exists as to whether local diversity is controlled by regional processes, involving biogeographic and evolutionary mechanisms, or Body Size: The Structure and Function of Aquatic Ecosystems, eds. Alan G. Hildrew, David G. Raffaelli and Ronni Edmonds-Brown. Published by Cambridge University Press. # British Ecological Society 2007.
BODY SIZE AND DIVERSITY IN MARINE SYSTEMS
whether intrinsic local processes involving ecological mechanisms are more important (Cornell & Lawton, 1992; Witman, Etter & Smith, 2004). Translated to the body size/diversity problem, the question arises as to whether the local numbers of species of different sizes, in taxonomic/functional guilds or integral assemblages, are simply a more or less random selection from the regional species pool, or whether local environmental constraints and species interactions are more important in determining this pattern. Problems of definition then arise: what is a ‘guild’ (Adams & Shorrocks, 1985) and how is the ‘regional species pool’ defined (Dupre, 2000)? This chapter attempts to resolve some of the above uncertainties with respect to marine faunas, with particular emphasis on the benthos, which has a much higher diversity than the plankton, and a higher phyletic diversity than any other realm on Earth (Warwick, 1996).
Species guilds Two generalized concepts with respect to species guilds are relevant here. First, the controversial concept that there is a minimum size difference between competing species that will permit their coexistence (Huxley, 1942), that ratio being shown empirically to be 1:28 (Hutchinson, 1959). It has been argued that this ‘Hutchinsonian ratio’ may be more to do with the generalities of assembling sets of tools than with anything directly biological (Horn & May, 1977), or indeed may be an artefact (Eadie, Broekhoven & Colgan, 1987; Greene, 1987). Also, tools or musical instruments with which analogies have been drawn (Horn & May, 1977) are fixed in size, whereas animals increase in size during development and are potentially in competition for resources with their congeners of different sizes and also with other species in the guild (Woodward & Warren, this volume). For taxa with distinct developmental stages (instars) the increase in size between each moult (Dyar’s Constant) is the same as the Hutchinsonian ratio of 1:3 (Dyar, 1890), and this also applies to many marine taxa (see for example Table 11.1). Dyar’s Constant is much better substantiated than the Hutchinsonian ratio, but Lawton and Strong (1981) argue that it is ‘nothing to do with competition or partitioning of food resources’. The second concept relates to the number of species in a guild that are typically found to co-occur, usually seven to eight and never more than ten (Adams & Shorrocks, 1985). Again, the explanation for this also involves competition for resources: more species would intensify competition for resources beyond the limits of the number of species that could coexist. Reverting to the analogy of inanimate objects there are instances where both these ‘rules’ may apply. A new violin family introduced at a conference on musical acoustics in Cambridge in 1997 had a ‘rich, even tone quality; particularly shines in contrapuntal writing’ (Darius, 1977). There were eight viols in the consort, with size ratios of 1:2, 1:2, 1:3, 1:3, 1:3, 1:3, 1:3 (Horne & May, 1977).
211
212
R. M. WARWICK
Table 11.1 Dyar’s Constants (increase in length between successive moults) for seven species of free-living marine nematodes. Species
Constant
Calculated from
Monhystera denticulata Chromadora macrolaimoides Steineria ericia Steineria pilosa Subsphaerolaimus litoralis Sphaerolaimus gracilis Parasphaerolaimus paradoxus Mean
1.28 1.41 1.37 1.34 1.32 1.22 1.31 1.32
Tietjen & Lee (1972) Tietjen & Lee (1973) Lorenzen (1978) Lorenzen (1978) Lorenzen (1978) Lorenzen (1978) Lorenzen (1978)
For marine animals very little is known about the species numbers and bodysize distributions of guilds of species that co-occur in specific locations or habitats. A much simpler exercise is to analyze these patterns in regional species pools because comprehensive regional inventories are available for many taxa. In Fig. 11.1 the numbers of species of different sizes in various taxa that might reasonably be defined as guilds, in what might reasonably be defined as regional species pools, are plotted. The term ‘guild’ here is taken to mean a taxonomically coherent group of animals with similar trophic roles. The taxa have been chosen to cover the whole spectrum of marine metazoan body sizes from tardigrades to toothed whales, and the body-length axis is divided into 1.28 geometric size classes. A remarkable feature of these plots is that each guild spans almost the same range of sizes, from nine size classes (tardigrades, shrimps and prawns, regular echinoids) to 14 (mysids), with a mean of 10.6. Thus, there seems to be a rather fixed range of body sizes that a given body plan will allow, irrespective of the nature of that body plan. If local assemblages were a random selection of species from the regional pool, the resulting guilds would not be inconsistent with the size ratio and guild size ‘rules’ noted above, and local competition or partitioning of food resources need not be invoked to explain them.
Adult-body size distributions in integral benthic assemblages The species numbers/adult body size relationship for marine benthic metazoan assemblages is quite unlike that described for any terrestrial or freshwater situation. Warwick (1984) found a highly conservative pattern for temperate soft-bottom benthic assemblages, with two separate lognormal distributions corresponding to the size categories traditionally regarded as macrofauna and meiofauna (Fig. 11.2). The meiofaunal mode occurs at a dry body mass 0.64 mg and the macrofaunal mode at 3.2 mg, with a trough between them at 45 mg. Rather few subsequent studies have attempted to count species numbers across the whole
BODY SIZE AND DIVERSITY IN MARINE SYSTEMS
8 7 6 5 19
Number of species
4
18 17 16 15
3
14 13 12 11 2
1
10 9
10–1
100
101
102
103
104
Length (mm)
Figure 11.1 Numbers of species in regional species pools of guilds of marine animals divided into size categories on a 1:28 geometric scale. 1 – Tardigrades, Britain (Morgan & King, 1976); 2 – Selective deposit feeding (bactivorous) nematodes, Chile (Wieser, 1959); 3 – Non-selective deposit feeding nematodes, Chile (Wieser, 1959); 4 – Epistrate (diatom) feeding nematodes, Chile (Wieser, 1959); 5 – Carnivorous/omnivorous nematodes, Chile (Wieser, 1959); 6 – Harpacticoid copepods, Norway (Sars, 1911); 7 – Calanoid copepods, Norway (Sars, 1903); 8 – Ostracods, Norway (Sars, 1928); 9 – Gammarid amphipods, Britain (Lincoln, 1979); 10 – Cumaceans, Britain (Jones, 1976); 11 – Mysids, Britain (Tattersall & Tattersall, 1951); 12 – Shrimps and prawns, Britain (Smaldon, 1979); 13 – Filter-feeding bivalves, Britain (Tebble, 1966); 14 – Ophiuroids, Britain (Mortensen, 1927); 15 – Regular echinoids, Britain (Mortensen, 1927); 16 – Irregular echinoids, Britain (Mortensen, 1927); 17 – Flatfish, NW Europe (Wheeler, 1969); 18 – Gadoids, NW Europe (Wheeler, 1969); 19 – toothed whales, Europe (Van den Brink, 1967).
meiofauna–macrofauna size spectrum, but these have confirmed the bimodal species size distribution, with many meiofaunal and many macrofaunal species, but few of intermediate size (Warwick et al., 1986; Warwick & Joint, 1987; Kendall, Warwick & Somerfield, 1997). An intuitive explanation for this follows Schwinghamer’s (1981) explanation for the ataxonomic benthic biomass spectrum concerning the constraint on body size of sediment granulometry.
213
R. M. WARWICK
24 Number of species
214
Northumberland mud
20
Carmarthen Bay sand
Algoa Bay sand
16 12 8 4 0 2
6 10 14 18 22 26 30
2
6 10 14 18 22 26 30
2
6 10 14 18 22 26 30
X2 geometric weight classes
Figure 11.2 Benthic species-size distributions in subtidal soft-sediment assemblages off NW England (Northumberland), South Wales (Carmarthen Bay) and South Africa (Algoa Bay) showing the conservative bimodal pattern (after Warwick, 1984).
Schwinghamer (1981) found a biomass trough at exactly the same body size as the species body-size trough (between 0.5 and 1 mm equivalent spherical diameter) and argued that this size defines the upper limit of the interstitial meiofauna and is the region of a shift from interstitial to burrowing lifestyles, with a class of intermediate-sized animals capable of neither. This hypothesis, involving the importance of habitat architecture (Holling, 1992), is not only intuitively appealing but is also open to empirical test (Stead et al., 2005). However, the only experimental test on soft-bottom marine benthos is that of Leaper et al. (2001) who found no support for it. Furthermore, this explanation does not bear critical examination based on comparative studies, since the bimodal species body-size distribution is apparent in thixotropic muddy sediments where the interstitial mode is not possible (Warwick, 1984), and also in sediment-free algal habitats where such sedimentary constraints do not exist (Gee & Warwick, 1994). Further, in freshwater sediments, evidence suggests that the meiofauna/ macrofauna dichotomy does not exist. Strayer (1986) found that, in the sediments of Mirror Lake, New Hampshire, USA, both the biomass spectrum and the species size distribution were unimodal, in marked contrast to the marine situation, and this unimodal pattern of taxon richness in body-size classes has also been found in freshwater streams (Stead et al., 2005) (Fig. 11.3). Physical sedimentary constraints are clearly the same in the freshwater and marine situations, but the reproductive characteristics of the species involved are very different. Most shallow-water temperate marine macrobenthos have planktonic larvae, whereas many species in the freshwater benthos are the larval stages of flying insects. The conservative nature of the species bodysize spectrum in marine habitats therefore adds credence to the suggestion that it results from evolutionary adaptations to the spatial and temporal structure of the marine environment, which will affect the regional species pool, rather than ecological constraints imposed by the physical nature of particular habitats. The meiobenthos are considered to be the first metazoans to evolve in the Middle Precambrian, with the Plathelminthes (including Gnathostomulida) and
BODY SIZE AND DIVERSITY IN MARINE SYSTEMS
(a)
Proportion of species (%)
16
4
1
0.25
(b)
Number of taxa
20
15 10
5 0 10–9
10–6 10–3 Body size (g dry weight)
100
Figure 11.3 (a) Species body-size distribution for the benthos of Mirror Lake, New Hampshire, USA (triangles and dotted line) compared with a typical shallow-water temperate marine benthic distribution (dots and solid line) (from Strayer, 1986). (b) Taxon body-size distribution in Lone Oak stream, SE England, drawn on the same x-axis scale as (a) (after Stead et al., 2005). Note the close similarity of the two unimodal freshwater distributions and the contrast with the marine bimodal distribution.
‘Aschelminthes’ (Nematoda, Gastrotricha, Kinorhyncha) being considered the most primitive forms (Boaden, 1989). These all have reproductive adaptations associated with small size: direct benthic development, dispersal as adults, short generation times (<1 year), semelparity, and reaching an asymptotic body size after which growth stops and reproduction commences. It is probable that these early meiofauna had the full range of trophic specializations before the major macrobenthic groups (Annelida, Arthropoda, Mollusca, Echinodermata) appeared. They were, and are, motile forms seeking food particles in a highly discriminate manner. Macrofaunal species have an alternative set of size related life-history and feeding traits: planktonic larval development and dispersal, long generation times (>1 year), iteroparity (usually), and continuing growth between successive spawnings. They may be either sedentary or motile, and feed less selectively on food particles. These contrasting traits are summarized in Table 11.2. Warwick (1984) suggested that the bimodal pattern of species-size distribution is apparent because there is a particular size at which meiofaunal life-history and feeding traits can be optimized, and another for macrofauna traits, compromises either being non-viable or disadvantageous. As size departs in eithes direction (larger or smaller) from these optima, fewer and fewer species
215
216
R. M. WARWICK
Table 11.2 Size related biological traits that switch abruptly at about 45 g dry weight in temperate shallow water marine benthos (from Warwick, 1984).
Development Dispersal Generation time Reproduction Feeding Resource partitioning Growth Mobility
Smaller than 45 mg
Larger than 45 mg
Direct benthic As adults Less than one year Semelparous Discriminate use of particles Particle selection (size, shape, quality) Reach asymptotic adult size Motile
Planktonic Planktonic larvae More than one year Iteroparous (usually) Indiscriminate use of particles Spatial segregation and particle size selection Continue growth throughout life Sedentary or motile
are able to coexist. Warwick (1989) went on to suggest that it had been ‘necessary’ for larger animals to evolve a planktonic larva to avoid competition with and predation by the permanent meiobenthos, which constitute a highly efficient consumer unit. Whilst it is difficult to invoke the strength of present-day interactions between the meiobenthos and macrobenthic larvae as evidence for selection pressure in the past, there is strong present-day evidence that the meiobenthos can significantly depress the densities of newly settled macrobenthic larvae. This is mainly by predation but also by competition and disturbance (Bell & Coull, 1980; Watzin, 1983, 1986). Meiobenthic predators can take rather large prey relative to their own body size, even larger than themselves (Bilio, 1967; Straarup, 1970; Watzin, 1985), including the young of traditionally macrobenthic taxa (Fig. 11.4). Eggs produced by macrobenthic species, and consequently the larvae hatching from them, fall exactly in the same size range as the adult meiobenthos (Fig. 11.5), and settle out of the plankton to the bottom at a size corresponding to the position of the trough in the bimodal species body-size distribution (Figs. 11.5 & 11.6). It had already been suggested that the planktotrophic larval phase in macrobenthos is principally a migration for feeding and safety (with dispersal as an inevitable secondary by-product), and that ‘Life-history theories for marine animals cannot ignore a strong historical component stretching back to the origin of the metazoa’ (Strathmann, 1985).
All-animal body-size distributions in integral benthic assemblages The above discussion concerns the diversity of metazoans divided among adult body-size categories. Of course, natural assemblages also comprise the smaller developmental stages of these species that are potentially interacting with the adults. Warwick, Dashfield and Somerfield (2006) explicitly addressed the issue of all-animal size distributions by determining the degree to which infaunal
BODY SIZE AND DIVERSITY IN MARINE SYSTEMS
MEIOFAUNA
LARVAL MACROFAUNA Syllidae <8 setigers Nephtyiidae
Macrostomum beaufortensis
Nereidae <7 setigers Polydora sp. Streblospio benedicti
Turbellarian A
Prionospiop sp. Convoluta sp.
Neochildia fusca
Capitellidae Orbiniidae Cirratuldae
Acoel B
Archiloa wilsoni Clymenella torquata
Other harpacticoid copepods
Terebellidae Oligochaeta <1 mm
Enhydrosoma sp. Nereidae 8–12 setigers
Nematoda 1 mm
Amphipoda Bivalvia (shell broken)
Oligochaeta >1 mm
Figure 11.4 The meiofaunal and young macrofaunal diets of two species of meiobenthic turbellarians, Neochildia fusca and Archiloa wisoni. All animals drawn to scale. Note the large size of many prey items relative to the body size of the predators (after Watzin, 1985).
diversity assemblage structure varied with the mesh sizes of the sieves used to extract the animals from the sediment. Within an apparently homogeneous area of coarse intertidal sand in the Isles of Scilly, UK, samples were extracted using a standard range of five mesh sizes (63, 125, 250, 500, 1000 mm), with the sample areas and distances between samples scaled to the mesh size. All metazoans in each of the sample sets were identified to species level. Diversity and dominance patterns showed a dramatic stepwise change between the 250 mm and 500 mm mesh-size samples, being relatively constant in the <500 mm and >500 mm categories, with diversity higher in the former (Figs. 11.7 & 11.8). This suggested a fractal structure within but not between the <500 mm and >500 mm body-size categories (cf. Schmid & Schmid-Araya, this volume). This clear meiofauna/ macrofauna dichotomy, Warwick et al. (2006) argued, supports the view that the important relationships are those between body size and various biological characteristics such as feeding behaviour, reproductive mode and life history as they are affected by the spatial and temporal structure of the environment.
217
R. M. WARWICK
(b) Proportion of species Number of species
(a) Proportion of species Number of species
218
0.1 benthic adults
0.1 benthic adults
0.05
0.05
0 10
eggs
0 0
10
20
30
X2 geometric weight class
0 3
Thorson’s polychaete larvae
0 4 Northumberland larvae 0 0
10
20
X2 geometric weight class
Figure 11.5 (a) Histogram of the size distribution of eggs of macrobenthic species that have planktotrophic larvae, compared with a generalized benthic species body-size distribution curve (from Warwick et al., 1986). (b) Histogram of the sizes of the polychaete larvae in Figure 11.6 and for a variety of newly settled macrobenthic larvae off the coast of Northumberland, compared with the adult species body-size distribution curve for the Northumberland station (from Warwick, 1984).
Pelagic assemblages For comparison with benthic species size distributions, Warwick et al. (1986) made a comprehensive collection of pelagic animals at station CS2 in the Celtic Sea, using a range of sampling devices with mesh sizes of 10 mm, 1 mm, 280 mm and 53 mm, and determined the adult body weights of all species in the samples. The centre of the size distribution of pelagic species closely corresponded with the trough in the bimodal benthic curve, with a few species extending as a tail across the macrobenthic size range (Fig. 11.9). Since the larvae of the macrobenthos (the meroplankton) are in the plankton when they occupy the meiobenthic size range, resource partitioning on a size basis between the meroplankton and holoplankton may have been an important evolutionary driving force in determining the pelagic size-distribution pattern (Warwick et al., 1986; Warwick & Joint, 1987).
Effects of pollution and disturbance Disturbed soft-bottom assemblages, and particularly those subjected to organic enrichment, become dominated by a group of species that have become known as ‘pollution indicators’ (Pearson & Rosenberg, 1978). These species are the smallest representatives of the macrofauna, mainly small annelids, and the largest species of the meiofauna, mainly oncholaimid nematodes and harpacticoid copepods of the genera Tisbe and Bulbamphiascus (Table 11.3). The body size of these species falls within the trough between the meiofauna and macrofauna
30
BODY SIZE AND DIVERSITY IN MARINE SYSTEMS
Polydora caeca
Nerine foliosa Pygospio elegans
Magelona papillicornis
Polydora ciliata 1 mm
Lagis koreni Spiophanes bombyx
Lanice conchilega
Figure 11.6 Late stage marine polychaete larvae at the size they begin to settle from the plankton to the seabed (after Thorson, 1946).
Shannon diversity (H’loge)
4.5 4 3.5 3 2.5 2 1.5 63
125
250
500
Sieve mesh size (µm)
1000
Figure 11.7 Species diversity in samples extracted from the sediment using a range of sieve mesh sizes on a sandflat in the Isles of Scilly, UK (after Warwick et al., 2006).
219
R. M. WARWICK
Cumulative % dominance
100 80 Mesh size (µm)
63 125 250 500 1000
60 40 20 0 1
10
100
1000
Species rank
Figure 11.8 k-dominance curves averaged over all replicates for each mesh size in the same samples as Figure 11.7 (after Warwick et al., 2006).
30
Number of species
220
20
benthic
10 pelagic 0 5 Pollution indicators 0 0
10
20
X2 geometric weight class
30
Figure 11.9 Species body-size distribution of the metazoan pelagos (solid histogram) compared with the metazoan benthos (upper curve) at a station in the Celtic Sea. Also shown (dotted histogram) is the size distribution of the ‘pollution indicator’ species listed in Table 11.3 (after Warwick et al.,1986).
adult body-size distributions, and within the size range of adult pelagic animals (Fig. 11.9). In these ‘enrichment communities’ of opportunistic species the meiofauna/macrofauna traits dichotomy (Table 11.2) breaks down; all the species have direct benthic development and are non-selective deposit feeders with approximately the same lifespan. This may also be the reason why the bimodal body-size distribution also breaks down.
Discussion and conclusion The mechanisms usually invoked to explain species size distributions, as with most other attributes of multispecies assemblages, have generally involved local ecological processes; species interactions (resource partitioning, predation, etc.) or environmental constraints (habitat architecture, disturbance, etc.). This chapter advocates the alternative, or additional, view that it is evolutionary processes acting over geological time that shape the structure of the regional species pool,
BODY SIZE AND DIVERSITY IN MARINE SYSTEMS
Table 11.3 Adult dry weight of species that become very abundant in organically enriched habitats. For sources of data see Warwick et al. (1986). Species Annelids Capitella capitata Polydora ciliata*/ligni Streblospio benedicti/shrubsolii* Tubificoides benedeni Ophryotrocha hartmanni Ophryotrocha puerilis Raphidrilus sp. Protodorvillea kefersteini Pholoe minuta Nematodes Metoncholaimus albidus Metoncholaimus scanicus Pontonema spp. Copepods Tisbe spp. Bulbamphiascus imus
Dry weight (mg)
505 125* 188* 306 50 140 158 108 190 12 23 64 12 3
and that the species occurring in any particular location are a more or less random selection from that pool. This would account for the conservative basic patterns encountered in marine habitats, which of course will be modified to some degree by ecological constraints. Whilst the effects of these ecological constraints can be examined experimentally, we cannot do manipulative experiments over evolutionary time and we must rely on so-called ‘natural experiments’, studying comparative situations where the invoked evolutionary mechanisms differ. For example, it would be a good test for the importance of the developmental mode if situations where the benthic/pelagic larval development dichotomy does not exist, e.g. the deep-sea, could be compared. Although no comprehensive size spectra yet exist for the deep sea, the model predicts a unimodal species-size distribution if the developmental mode is the key trait responsible for the size dichotomy.
Acknowledgements The ideas summarized here result from interactions with a large number of colleagues, and financial support from a large number of funding agencies over the course of a long career (resulting in an embarrassing proportion of selfcitations). These are far too numerous to list, but I thank them all. The writing of
221
222
R. M. WARWICK
this chapter was done as a contribution to the biodiversity element of the Plymouth Marine Laboratory’s core strategic research programme. It was supported by the UK Natural Environment Research Council (NERC) and the UK Department for Environment, Food and Rural Affairs (DEFRA) through the AMBLE project ME3109, and I acknowledge my position as an honorary fellow of the Plymouth Marine Laboratory.
References Adams, J. & Shorrocks, B. (1985). Competing species come in sevens. New Scientist, 107(1465), 42–44. Bell, G., Lechowicz, M. J., Appenzeller, A. et al. (1993). The spatial structure of the physical environment. Oecologia, 96, 114–121. Bell, S. S. & Coull, B. C. (1980). Experimental evidence for a model of juvenile macrofauna – meiofauna interactions. In Marine Benthic Dynamics, ed. K. R. Tenore and B. C. Coull. Columbia: University of South Carolina Press, pp. 179–192. Bilio, M. (1967). Nahrungsbeziehungen der Turbellarien in Ku¨stensalzweisen. Helgola¨nder Wissenschaftliche Meeresuntersuchungen, 15, 602–621. Boaden, P. J. S. (1989). Meiofauna and the origin of the Metazoa. Zoological Journal of the Linnean Society, 96, 217–227. Brey, T. (1990). Estimating productivity of macrobenthic invertebrates from biomass and mean individual weight. Meeresforschung, 32, 329–343. Cornell, H. V. & Lawton, J. H. (1992). Species interactions, local and regional processes, and limits to the richness of ecological communities – a theoretical perspective. Journal of Animal Ecology, 61, 1–12. Darius, J. (1977). Catgut in Cambridge. New Scientist, 75(1061), 175. Dial, K. P. & Marzluff, J. M. (1988). Are the smallest organisms the most diverse? Ecology, 69, 1620–1624. Dupre, C. (2000). How to determine a regional species pool: a study in two Swedish regions. Oikos, 89, 128–136.
Dyar, H. G. (1890). The number of molts of lepidopterous larvae. Psyche, 5, 420–422. Eadie, J. McA., Broekhoven, L. & Colgan, P. (1987). Size ratios and artefacts: Hutchinson’s rule revisited. American Naturalist, 129, 1–17. Etienne, R. S. & Olff, H. (2004). How dispersal limitation shapes species-body size distributions in local communities. American Naturalist, 163, 69–83. Fenchel, T. (1993). There are more small than large species? Oikos, 68, 375–378. Gee, J. M. & Warwick, R. M. (1994). Body-size distribution in a marine metazoan community and the fractal dimensions of macroalgae. Journal of Experimental Marine Biology and Ecology, 178, 247–259. Greene, E. (1987). Sizing up size ratios. Trends in Ecology and Evolution, 2, 79–81. Holling, C. S. (1992). Cross-scale morphology, geometry, and the dynamics of ecosystems. Ecological Monographs, 62, 447–502. Horne, H. S. & May, R. M. (1977). Limits to similarity among coexisting competitors. Nature, 270, 660–661. Hutchinson, G. E. (1959). Homage to Santa Rosalia, or why are there so many kinds of animals. American Naturalist, 93, 117–125. Huxley, J. (1942). Evolution: the Modern Synthesis. New York: Harper. Jones, N. S. (1976). British Cumaceans. Synopses of the British Fauna (New Series) No. 7. London: Academic Press. Kendall, M. A., Warwick, R. M. & Somerfield, P. J. (1997). Species size distributions in Arctic benthic communities. Polar Biology, 17, 389–392.
BODY SIZE AND DIVERSITY IN MARINE SYSTEMS
Kochmer, J. P. & Wagner, R. H. (1988). Why are there so many kinds of passerine birds? Because they are small. A reply to Raikow. Systematic Zoology, 37, 68–69. Lawton, J. H. & Strong, D. R. (1981). Community patterns and competition in folivorous insects. American Naturalist, 118, 317–338. Leaper, R., Raffaelli, D., Emes, C. & Manly, B. (2001). Constraints on body-size distributions: an experimental test of the habitat architecture hypothesis. Journal of Animal Ecology, 70, 248–259. Lincoln, R. J. (1979). British Marine Amphipoda: Gammaridea. London: British Museum (Natural History). Lorenzen, S. (1978). Postembyonalentwicklung von Steineria- und Sphaerolaimidenarten (Nematoden) und ihre Konsequenzen fu¨r die Systematic. Zoologische Anzeiger, Jena, 200, 53–78. May, R. H. (1986). The search for patterns in the balance of nature: advances and retreats. Ecology, 67, 278–279. Morgan, C. I. & King, P. E. (1976). British Tardigrades. Synopses of the British Fauna (New Series) No. 9. London: Academic Press. Mortensen, T. H. (1927). Handbook of the Echinoderms of the British Isles. London: Oxford University Press. Orme, C. D. L., Quicke, D. L. J., Cook, J. M. & Purvis, A. (2002). Body size does not predict species richness among the metazoan phyla. Journal of Evolutionary Biology, 15, 235–247. Pearson, T. H. & Rosenberg, R. (1978). Macrobenthic succession in relation to organic enrichment and pollution of the marine environment. Oceanography and Marine Biology: an Annual Review, 16, 229–311. Sars, G. O. (1903). An Account of the Crustacea of Norway, Volume 4 – Copepoda, Calanoida. Bergen: Bergen Museum. Sars, G. O. (1911). An Account of the Crustacea of Norway, Volume 5 – Copepoda, Harpacticoida. Bergen: Bergen Museum.
Sars, G. O. (1928). An Account of the Crustacea of Norway, Volume 9 – Ostracoda. Bergen: Bergen Museum. Schwinghamer, P. (1981). Characteristic size distributions of integral benthic communities. Canadian Journal of Fisheries and Aquatic Sciences, 38, 1255–1263. Siemann, E., Tilman, D. & Haarstad, J. (1996). Insect species diversity, abundance and body size relationships. Nature, 380, 704–706. Smaldon, G. (1979). British Coastal Shrimps and Prawns. Synopses of the British Fauna (New Series) No. 15. London: Academic Press. Stead, T. K., Schmid-Araya, J. M., Schmid, P. E. & Hildrew, A. G. (2005). The distribution of body size in a stream community: one system, many patterns. Journal of Animal Ecology, 74, 475–487. Straarup, B. J. (1970). On the ecology of turbellarians in a sheltered brackish shallowwater bay. Ophelia, 7, 185–216. Strathmann, R. R. (1985). Feeding and nonfeeding larval development and life-history evolution in marine invertebrates. Annual Review of Ecology and Systematics, 16, 339–361. Strayer, D. (1986). The size structure of a lacustrine zoobenthic community. Oecologia, 69, 513–516. Tattersall, W. M. & Tattersall, O. S. (1951). The British Mysidacea. London: The Ray Society. Tebble, N. (1966). British Bivalve Seashells. British Museum (Natural History), London. Thorson, G. (1946). Reproduction and larval development of Danish marine bottom invertebrates. Medd. Komm. Danm. Fisk. -og Havunders, ser Plankton, 4, 1–523. Tietjen, J. H. & Lee, J. J. (1972). Life cycles of marine nematodes. Influence of temperature and salinity on the development of Monhystera denticulata Timm. Oecologia, 10, 167–176. Tietjen, J. H. & Lee, J. J. (1973). Life history and feeding habits of the marine nematode Chromadora macrolaimoides Steiner. Oecologia, 12, 303–314.
223
224
R. M. WARWICK
Van den Brink, F.-H. (1967). Guide des mammife`res sauvages d’Europe. Neuchaˆtel: Delachaux et Niestle´. Van Valen, L. (1973). Body size and numbers of plants and animals. Evolution, 27, 27–35. Warwick, R. M. (1984). Species size distributions in marine benthic communities. Oecologia, 61, 32–41. Warwick, R. M. (1989). The role of meiofauna in the marine ecosystem: evolutionary considerations. Zoological Journal of the Linnean Society, 96, 229–241. Warwick, R. M. (1996). Biodiversity and production on the sea floor. In The Oceans and the Poles, ed. G. Hempel. Jena: Gustav Fischer Verlag, pp. 217–227. Warwick, R. M. & Joint, I. R. (1987). The size distribution of organisms in the Celtic Sea: from bacteria to Metazoa. Oecologia, 73, 185–191. Warwick, R. M. & Price, R. (1979). Ecological and metabolic studies on free-living nematodes from an estuarine mud-flat. Estuarine Coastal and Shelf Science, 9, 257–271. Warwick, R. M., Collins, N. R., Gee, J. M. & George, C. L. (1986). Species size distributions of benthic and pelagic Metazoa: evidence for interaction? Marine Ecology Progress Series, 34, 63–68. Warwick, R. M., Dashfield, S. L. & Somerfield, P. J. (2006). The integral structure of a benthic
infaunal assemblage. Journal of Experimental Marine Biology and Ecology, 330, 12–18. Watzin, M. C. (1983). The effects of meiofauna on settling macrofauna: meiofauna may structure macrofauna communities. Oecologia, 59, 163–166. Watzin, M. C. (1985). Interactions among temporary and permanent meiofauna: observations on the feeding behaviour of selected taxa. The Biological Bulletin, 169, 397–416. Watzin, M. C. (1986). Larval settlement into marine soft-sediment systems: interactions with the meiofauna. Journal of Experimental Marine Biology and Ecology, 98, 65–113. West, G. B., Brown, J. H. & Enquist, B. J. (1999). The fourth dimension of life: fractal geometry and allometric scaling of organisms. Science, 284, 1677–1679. Wheeler, A. (1969). Fishes of the British Isles and North-West Europe. London: Macmillan. Wieser, W. (1959), Reports of the Lund University Chile expedition 1948–49, 34. Free-living marine nematodes IV. General part. Lunds Universitets A˚ rsskrift (2), 55 (5), 1–111. Witman, J. D., Etter, R. J. & Smith, F. (2004). The relationship between regional and local species diversity in marine benthic communities: a global perspective. Proceedings of the National Academy of Sciences of the United States of America, 101, 15664–15669.
CHAPTER TWELVE
Interplay between individual growth and population feedbacks shapes body-size distributions LENNART PERSSON Umea8 University, Sweden
ANDRE´ M . DE ROOS University of Amsterdam, The Netherlands
Body size in contemporary ecology Body size and variation in body size have formed the focus of many studies in ecology, ranging from the study of individual performance to large-scale communities and ecosystems (Werner & Gilliam, 1984, Gaston & Lawton, 1988, Werner, 1988, Cohen, Johnson & Carpenter, 2003, Brown et al., 2004, Loeuille & Loreau, 2005). This focus is well-founded given the large variation in body size that exists among organisms from micro-organisms to large mammals (Gaston & Lawton, 1988; Werner, 1988). Body size is also the most important trait that affects the performance of individuals. Basic ecological capacities such as foraging rate and metabolic requirements are close functions of body size (Peters, 1983; Kooijmann, 2000; Brown, et al., 2004) affecting, for example, competitive abilities of differently sized organisms (Wilson, 1975; Persson, 1985; Werner, 1994). Body size strongly influences the diet of consumers with mean prey size, but also the variation in the size of prey eaten, increasing with predator size (Wilson, 1975; Werner & Gilliam, 1984; Cohen et al., 2005; Woodward & Warren, this volume; Humphries, this volume). Furthermore, the risk for an organism being preyed upon is heavily influenced by its own body size as well as the body size of its potential predator (Polis, 1988; Werner, 1988; Claessen, De Roos & Persson, 2000). Given its influence on basic individual ecological processes, body size has been an important variable in the investigation of larger ecological entities including communities, food webs and ecosystems. For example, predator–prey size ratios have formed the basis for food-web models such as the cascade model (Chen & Cohen, 2001), and for estimating interaction strengths in food webs (Emmerson & Raffaelli, 2004). Body size has also been the key variable in the analysis of food-web patterns with regard to numerical and biomass abundance at different trophic positions (Cohen et al., 2003; Cohen, this Body Size: The Structure and Function of Aquatic Ecosystems, eds. Alan G. Hildrew, David G. Raffaelli and Ronni Edmonds-Brown. Published by Cambridge University Press. # British Ecological Society 2007.
226
L. PERSSON AND A. M. DE ROOS
volume). Another example where body size is a key variable is in size spectra analyses of the organization of trophic dynamics among populations of organisms (Kerr & Dickie, 2001; Shin et al., 2005). Finally, during the last decade the ‘metabolic theory of ecology’ (West, Brown & Enquist, 1997; Brown et al., 2004) has become very popular and is heavily founded on body-size variation. This theory has been advanced by its proponents to form a conceptual basis for ecology comparable to that of genetic theory for evolution and to ‘link the performance of individual organisms to the ecology of populations, communities, and ecosystems’ (Brown et al., 2004; Brown, Allen & Gillooly, this volume).
Neglected aspects of body size in contemporary ecology Although body size plays a central role in ecology, an important aspect of body size in many ecological communities has been largely neglected in the theoretical and empirical research mentioned above. In fact, the ecological entities upon which patterns have been analyzed are based on ‘average individuals’ (of different body sizes), an approach that basically is in conflict with a Darwinian view stressing variation among the individual organisms (see De Roos & Persson, 2005a). A major part of observed body-size variation is related to within-species variation, as individuals grow over a substantial part of their life cycle, whereas most contemporary ecological studies restrict their attention to between-species variation. To consider ontogenetic variation among individuals seems essential for any conceptual synthesis, given that the overwhelming majority of the Earth’s taxa exhibit some degree of size/stage structure (Werner, 1988) and the ecological effects of intraspecific variation in body size are well represented in the other chapters of this volume (e.g. Woodward & Warren, this volume; Warwick, this volume). Actually a whole body of theory on ontogenetic development and food-dependent growth of individuals was developed during the 1980s (Sebens, 1982; Werner & Gilliam, 1984; Sebens, 1987; Sauer & Slade, 1987; Ebenman & Persson, 1988), a literature that has been largely neglected by more recent ecological studies on body size. The purpose of this chapter is to give first a short historical overview of studies considering patterns of development and growth in organisms and to link size-dependent individual performance to community patterns. Second, we give an overview of how to progress towards an explicit and rigorous link between individual body size and population and community processes. Our focus will be on how size-dependent interactions shape the dynamics and structure of ecological communities including body-size distributions.
Development and growth – a retrospective overview As already mentioned, considering individual growth and development is important, because the majority of animals exhibit substantial changes in size and/or morphology over their ontogeny (Werner, 1988). Further, for most plant
INDIVIDUAL GROWTH AND BODY SIZE
species growth and development is a major aspect of their life history. An overview of different animal taxa shows that major changes in body morphology as a result of metamorphosis are present in more than 85% of all taxa (25 of 33 phyla) (Werner, 1988). This pattern largely results because of the very many species of insects. Nevertheless, even if only vertebrates are considered, individuals of 75% of all taxa show substantial growth for much of their lives, which is due to the taxonomic dominance of fish, amphibians and reptiles. Actually, it is only among altricial birds and some mammals where the young are close to the adult body size when they become independent of the parents (Werner, 1988). Scaling constraints and growth patterns It has been suggested that large changes in body size due to ontogenetic development and growth impose a number of constraints on the body morphology of organisms related to physical, chemical and biological processes (Peters, 1983; Calder, 1984; Werner, 1988; Stearns, 1992; Humphries, this volume). When increasing in size, scaling properties – depending on both physical and ecological constraints – will set limits over which size range a particular lifestyle can be exploited (Calder, 1984; Werner, 1988). For example, physical parameters acting on small and large organisms are very different exemplified/illustrated by the effects of different Reynolds numbers on small and large aquatic organisms, respectively (Humphries, this volume). For small organisms, the low Reynolds number means that they swim with friction as the propulsive mode. In contrast, large organisms use the inertia of the water to propel themselves (Werner, 1988). Within the broader limits set on morphology by physical constraints, ecological constraints are also present related to, for example, which prey types an organism with a specific body morphology and size can efficiently utilize (Werner, 1988; Woodward & Warren, this volume). In particular, the morphologies that can evolve to efficiently handle different prey sizes during different parts of ontogenesis are constrained by genetic additative covariance in the genotype (Werner, 1988; Ebenman, 1992). Werner (1988) argued that allometric growth in organisms is only partly sufficient to cope with the different demands on body morphology made during different parts of the life cycle. These constraints imposed by allometric growth therefore result in an ‘allometric scaling problem’ for performance over the life cycle. He argued that, if scaling imposes a problem during the ontogeny, there should be patterns among the variety of life-history strategies by which animals cope with this problem (Cohen, 1985; Werner, 1988). Four particular tactics were discussed by Cohen (1985) and Werner (1988). The first represents organisms that largely avoid the problem of substantial size change by specializing as a very small adult (for example protozoans). The second tactic represents organisms in which the basic (small) trophic apparatus
227
228
L. PERSSON AND A. M. DE ROOS
remains intact but is multiplied (e.g. coral polyps). The third tactic represents the situation where the adults extensively provide the egg and juvenile with transformable mass (e.g. yolk, maternal fluids, bodies of prey) until the young has reached a size where it can take up the parental lifestyle (birds and mammals). Fourth, the organism may adopt a succession of complex life histories that accommodate the increase in size (insects, amphibians, fish). Here, complex life cycles in the form of metamorphosis represent a way to break up genetic covariances between sizes/stages (Werner, 1988; Ebenman, 1992). One common trait among the two groups (birds and mammals) where the parents provide the egg and juvenile with transformable mass until it can take up the parental lifestyle is endothermy, including a high body temperature (Case, 1979; Stearns, 1992). This observation suggests that the rapid development from juvenile to adult in these groups not only requires the ability to provide eggs/juveniles with extensive amounts of energy, but also the ability to transform that energy rapidly into growth. This leads to the hypothesis that there is a relationship between endothermy and individual growth rate, an assumption that is supported by empirical data, because mammals and birds have a growth rate that is an order of magnitude higher than that of ectotherms such as reptiles and fishes (Ricklefs, 1973; Case, 1979). In summary, species for which individual growth and development plays a smaller role are primarily found among unicellular organisms and endotherms. In other organisms, substantial growth and development after the juvenile becomes independent of its parents is the rule. The different growth patterns observed among different organisms have formed the basis for different classifications of growth types (see Sebens, 1987). Without giving a more detailed description of these classification schemes, they have focused on two aspects of ontogenetic growth: the extent to which growth and development is plastic/indeterminate and hence food dependent, and the extent to which the asymptotic size is fixed or food dependent. In several instances these two aspects of growth have been mixed. For example, Stearns (1992) defined determinate growth as the situation where individuals do not grow in size after maturation. This definition of determinate growth is in our opinion not satisfactory, as it totally neglects whether growth up to maturation is food dependent or not. As considered above, ontogenetic growth and development take very different forms in different groups of taxa. Food-dependent growth over ontogeny can be continuous, as in fish and plants, or discrete where the development time between stages is food dependent, as in many invertebrates. We argue that it is food-dependent development per se that forms one dividing line for how ontogenetic development will affect individual performance, population and community processes and the biomass structures.
INDIVIDUAL GROWTH AND BODY SIZE
Individual-level formulations for how individuals grow – linkage to community patterns The main message to be drawn from the above overview is that ontogenetic growth and development are characteristic of many organisms on Earth and that individual performance over ontogeny is constrained by both physical and ecological factors. At a broad scale, organisms showing substantial growth after becoming independent of their parents and those that do not can be separated along the endothermy–ectothermy gradient. At the same time, the different growth patterns described above are limited to broad categories, and a more quantitative link between individual performance and growth, with its consequences for community attributes such as body-size patterns, is therefore needed. A number of attempts were also undertaken during the 1980s to link individual body-size dependent performance and the dynamics of ecological communities. First, the scaling of foraging rate and metabolism with body size was used to determine the competitive ability of differently sized organisms and to predict niche shifts over ontogeny based on energy maximization (Mittelbach, 1981; Werner & Gilliam, 1984; Werner, 1988). Second, as the risk of being eaten is also a function of body size, the literature on individual size-dependent performance also came to include the effects of predation using optimal control theory (Werner & Gilliam, 1984). This individual-based framework using explicit body-size scalings of different rates was quite successfully applied to predict and understand the distribution of species and size classes within and between systems, primarily in freshwater fish communities (Mittelbach, 1981; Werner, 1986; Persson, 1988). Implicitly this approach assumed that community patterns could be predicted from individual-level traits ignoring population-level dynamics and feedbacks. An exception is the study by Hamrin and Persson (1986) on population cycles in cisco (Coregonus albula) where the population dynamics were explained from size-dependent foraging and metabolic rate including feedbacks on the resource. Modelling methods to address the dynamics of size-structured dynamics (Sinko & Streifer, 1967) were already being discussed at this time (Werner & Gilliam, 1984). However, more complete modelling formulations to address size-structured dynamics were first developed during the second half of the 1980s (Metz & Diekmann, 1986; De Roos et al., 1990) and their efficient use in ecological theory started first in the 1990s, which is the focus of the rest of this chapter (De Roos et al., 1990; Persson et al., 1998; Claessen, De Roos & Persson, 2000).
Developments of an explicit link from individual body size to population dynamics Brown et al. (2004) envisage metabolic theory to link the performance of individual organisms to the ecology of populations, communities and ecosystems.
229
230
L. PERSSON AND A. M. DE ROOS
This might be true for some elements of the hierarchy from individuals to ecosystems, but a number of key elements, especially at the level of populations, are inadequately considered. Although we agree that there are constraints on ecological performance, such as individual metabolic rate, population maximum growth rate and ecosystem turnover as a function of body size (Brown et al., 2004), this theory addresses how ecological interactions shape body-size distributions in ecological communities only to a limited extent. Moreover, given that ontogenetic development is a major feature in most organisms, the effect of ontogeny on the development of body-size distributions is also a major aspect to take into account. In the following, we briefly describe a modelling framework that (i) explicitly links individual-level processes, including bodysize scaling, to population-size distributions and, (ii) considers ontogenetic development. We will subsequently discuss how food-dependent development rate gives rise to both dynamical and structural patterns not present in unstructured theory and how this shapes body-size distributions. It will become evident that body size in different ecological configurations is both an input to (by determining individual performance) and outcome of (as a result of population feedbacks) ecological interactions. Modelling framework The modelling approach we consider are physiologically structured population models (PSPMs) (Metz & Diekmann, 1986; De Roos et al., 1990) referred to as i-state distribution models. They are based on two different state concepts, the individual or i-state and the population or p-state. The i-state represents the state of the individual in terms of a collection of characteristic physiological traits, such as size, age and energy reserves, while the p-state is the frequency distribution over space of all possible i-states. The model formulation process consists of deriving a mathematical description of how individual performance (growth, survival, reproduction) depends on the physiological characteristics of the individual and the condition of the environment (i-state description). Handling the population-level (p-state) dynamics is subsequently just a matter of bookkeeping of all individuals in different states without making any further model assumption at this level (Fig. 12.1). The core of PSPMs is thus the individual state and the modelling of the individual life history. The derivation of the PSPM proceeds by writing down the equations describing the i-state dependent processes of energy gathering (attack rate, digestive capacity), metabolism, energy channelling between somatic and gonad growth and survival (generally a function of energy status and size-dependent mortality from predators) (see De Roos et al., 1990 and Persson et al., 1998 for examples). Energy allocated into gonad tissue may be spent continuously or discretely constrained by, for example, season. Bookkeeping provides the link from the individual to the population level, which also includes calculations of the impact of the total population on its
INDIVIDUAL GROWTH AND BODY SIZE
Individual Life History
Modelling
Environmental condition (e.g. food density)
Reproduction
Growth
Bookkeeping
Survival
Individual state (e.g. size)
Dynamics of population abundance and composition
Ecological interactions
Figure 12.1 Schematic representation of the philosophy behind the framework of physiologically structured population models. Based on the state of the individual and the environmental condition, we can model foraging rate, growth, survival and reproduction of the individual, while the population state is merely a bookkeeping of all individuals in all i-states.
environment. The change in the environment resulting from this impact represents the population feedback on individual life history and/or behaviour (Fig. 12.1). In a consumer-resource system, for example, the population influence on consumer life history operates through an increased or decreased density of resource, which affects individual growth, mortality and reproduction. In addition to i- and p-states, an environmental (E) state is defined, which in the consumer-resource system is the resource. In a system including predators of the consumer, the E-state also includes all potential predators on a consumer of a specific i-state. Ontogenetic development – dynamical aspects As discussed above, the size scalings of foraging and metabolic rate were recognized as basic variables to determine the competitive ability of differently sized individuals (Mittelbach, 1981; Lundberg & Persson, 1993; Werner, 1994). We extend this individual-level argumentation to analyze its populationlevel consequences using two case studies, a size-structured consumer-resource and a cannibalistic system, with the purpose of showing how size-structured dynamics may induce temporal variation in body-size distributions. For a consumer-resource model, the body-size scalings of foraging rate (attack rate, digestive capacity) and metabolism have been shown to have major effects on the population dynamics observed (Persson et al., 1998). Combining the size scaling of attack rate, digestive rate and metabolic rate allows us to calculate
231
L. PERSSON AND A. M. DE ROOS
1E8 Consumers
1E7 1 000 000 100 000 10 000 1000 0
2
4
6
8
10
0
2
4
6
8
10
0
2
4
6
8
10
12
1E-4 Resources
232
1E-5
1E-6 0
2
4
6
8
10
Year
Figure 12.2 Changes in the numbers of young-of-the-year (dotted black lines), juveniles from an age of one year (solid grey lines) and adult (black solid thick lines) consumers and resource levels in the two types of cohort cycles discussed in the text. Left panels: cohort cycle driven by recruiting individuals. When a strong cohort is born, it almost immediately depresses the resource to low levels and out-competes older cohorts. The pattern repeats itself when the dominating recruiting cohort matures and gives rise to a new strong reproductive pulse. Right panels: cohort cycle driven by larger juveniles. When a strong cohort is born it causes a decrease in the resource for several years driving new cohorts to starvation death (vertical dotted grey lines) despite that adult reproduction is present for several years (years 2–4).
the resource level that differently sized individuals need to just meet maintenance (critical resource density, Persson et al., 1998). If the critical resource density increases monotonically with body size, cohort cycles driven by competitively superior recruiting cohorts will prevail (Fig. 12.2). In contrast, when critical resource density at first decreases with body size but increases thereafter, cohort cycles driven by larger individuals will occur (Fig. 12.2). Finally, in a narrow parameter range where critical resource density is relatively independent of body size, equilibrium conditions with many coexisting size cohorts will prevail (Persson et al., 1998). The different dynamics observed can thus be predicted from the form of the critical resource-density function. The cycle length of the resulting cohort cycles, which are driven by cohort interactions is, in contrast to predator–prey cycles, set by the time it takes individuals to reach maturation (generation time cycle length). A literature review shows that single generation cycles (37% of all cycling populations) are as
INDIVIDUAL GROWTH AND BODY SIZE
common as traditional predator–prey cycles (38% of all cycling populations) among fluctuating populations (Murdoch et al., 2002). Together with another type of stage-based dynamics, delayed-feedback cycles (25% of all cycling populations), size/stage-based cycles thus are the dominant type of cycle observed in populations. Moreover, predator–prey cycles are found almost exclusively in specialist predators. Two further things can be stated about size-dependent consumer-resource dynamics. First, the observed size scalings of foraging rate and metabolism (that determine the competitive ability of differently sized individuals) suggest that cyclic dynamics should dominate over equilibrium dynamics, with the consequence that the body-size distributions of consumer populations will vary over time. Second, cohort (‘single species’) cycles are common in many species systems (Murdoch et al., 2002), suggesting that intraspecifically driven cycles cannot be ignored in many species systems. Typical examples of cohort-dominated cycles include population oscillations in Daphnia and planktivorous fish (McCauley et al., 1999; Persson et al., 1998). For planktivorous fish, estimated critical-resource demands show that these cycles are recruit-driven with an observed cycle length of 2–5 years. For shorter cycles (2–3 years), age cohorts may coexist and a cycling in growth rates of the different age cohorts is observed (Hamrin & Persson, 1986; Cryer, Peirson & Townsend, 1986; Townsend, Sutherland & Perrow, 1990). For longer cycles, strong age cohorts become more dominant and major shifts in size distributions can be observed over the years due to growth of dominant age cohorts (Fig. 12.3) (Sanderson et al., 1999). Cannibalism represents an extension of consumer-resource interactions to include predation among consumers. Because many cannibals share a common resource with their victims, the latter may subject the cannibals to severe exploitative competition for this shared resource (Polis, 1988; Persson, 1988; Polis, Myers & Holt, 1989). Experiments and field data, particularly on cannibalistic fish, show that a positive cannibalistic attack rate is constrained between a lower boundary below which the cannibal does not encounter a victim because of difficulties in seeing it, and an upper boundary above which the escape ability of victims and gape-size constraints prevents cannibalism (Claessen et al., 2000, Juanes, 2003). Analyses of the dynamics of cannibalistic systems show that three aspects of the size scaling of the cannibalistic attack function are important in determining population dynamics and cannibal-size distributions. The maximum victim/cannibal size ratio has less effect on population dynamics, but a strong effect on the ultimate size that an individual reaches and thereby the size distribution of the cannibal population (Claessen et al., 2002). In contrast, the overall rate by which the cannibalistic attack rate increases with cannibal size and the lower victim/cannibal size ratio have major effects on population dynamics and therefore on the degree of temporal variation in size distributions of cannibalistic populations (Claessen et al., 2000, 2002).
233
L. PERSSON AND A. M. DE ROOS
(a)
(b) 160
40
140
35
120
30 Percentage
Length (mm)
234
100 80
1988 1992
25 20 15
60 10 40 5 20 0 1982
1984
1986
1988 Year
1990
1992
40
60
80
100 120 140 Length (mm)
160
Figure 12.3 (a) Growth (mean 1 SD) of two dominating cohorts of yellow perch (Perca flavescens) born in 1980 and 1985, respectively. (b) Shift in the size distribution of yellow perch from 1988 to 1992 as a result of growth of the dominating cohort born in 1985 (data from Sanderson et al., 1999).
Three different dynamical outcomes can be found: (i) low-amplitude/fixed-point dynamics, where cannibals control victims by inducing a high mortality on them, (ii) high-amplitude cohort cycles of the kind discussed above, in which victims outcompete cannibals before the latter can start cannibalizing, and (iii) a dynamics involving a mix of the previous two. Studies of cannibalistic perch (Perca fluviatilis) populations show that this species with a relatively high minimum victim–cannibal size ratio agrees with theoretical expectations and undergoes shifts in dynamics between a ‘stunted’ cannibal-driven phase and a phase driven by inter-cohort competition resulting in dramatic shifts in size distributions (Fig. 12.4b). The phases are also characterized by very different life-history trajectories with a periodic appearance of giant individuals (Fig. 12.4a). In conclusion, the two examples (consumer-resource and cannibalistic interactions) used here to illustrate the implications of size-structured interactions on dynamics show that different size scalings of individual rates have strong influences on distribution of body sizes over time. As the dynamics of species such as fish, feeding near the top of food web, feed back on lower trophic components, the changes in size distribution at these higher trophic positions will also cascade down to cause shifts in overall size distribution of the food web (Cryer et al., 1986; Shiomoto et al., 1997; Persson et al., 2003).
INDIVIDUAL GROWTH AND BODY SIZE
(a)
(b) 400 1000 350 1993 1995
800
250
Density (no/ha)
Size (mm)
300
200 150
600
400
100 200
50 0
0 1986 1988 1990 1992 1994 1996 1998 2000 Year
0
50
100 150 200 250 300 Length (mm)
Figure 12.4 (a) Examples of growth curves (mean 1 SD) of perch becoming stunted (born in 1986) and becoming giants (born in 1990) in Lake Abborrtja¨rn 3. (b) Shift in perch size distributions between stunted (1993) and giant phases (1995). The main part of the size distribution in 1993 consisted of stunted mature perch with a median size of 156 mm. The main part of the size distribution in 1995 consisted of small immature perch on which a few large mature perch are feeding (data from Persson et al., 2003).
Ontogenetic development and community structure Analyses of the dynamics of consumer-resource and cannibalistic systems show how different body-size scalings impact population dynamics and temporal changes in body-size distributions. Size-structured dynamics in these systems also have the potential to result in alternative states, which, particularly in cannibalistic systems, may result in very different size distributions (Claessen & De Roos, 2003). An extension of physiologically structured population models to more complex trophic configurations will further increase the likelihood of alternative states. In the following, we discuss alternative states and body-size distributions of populations exemplified by tritrophic food chains and tritrophic configurations with life-history omnivory. Since the classical paper by Brooks and Dodson (1965) it is generally accepted that size-dependent predation by top predators has a strong structuring impact on prey size distributions. Size-dependent predation in combination with fooddependent development in prey has the potential to lead to alternative stable states with vastly different size distributions of the prey (De Roos & Persson, 2002). In the tritrophic food chain studied by De Roos and Persson (2002), two
235
236
L. PERSSON AND A. M. DE ROOS
alternative states were found under otherwise identical conditions, one consumer-resource, the other a predator-consumer-resource state. It is noteworthy that the invasion boundary of the predator occurred at a higher productivity level than its persistence boundary and, similarly, the harvesting persistence boundary for the predator was positioned at higher harvesting levels than the invasion boundary, making these systems prone to catastrophic collapses. The mechanistic explanation behind this catastrophic behaviour of the system is that a predator selecting small size classes of the consumer will increase the development rate of the remaining consumers into larger adult size classes, leading to increased population fecundity of the consumer and, therefore, counter-intuitively to an increased number in individuals of the size class that the predator feeds on (Fig. 12.5a) (De Roos & Persson, 2002). This overcompensatory effect has been termed an emergent Allee effect as it, in contrast to other mechanisms accounting for Allee effects, is purely based on exploitation of resources. The emergent Allee effect is also present if the predators select the largest size classes in the consumer population (De Roos, Persson & Thieme, 2003b; De Roos & Persson, 2005b). Size-structured dynamics in tritrophic food chains have also been shown to lead to the possibility of predator facilitation, where one predator may allow
Figure 12.5 (a) Size distributions of the consumer species in the absence (black bars) and presence (white bars) of size-selective predators in the tritrophic model with size structure in the consumer population. (b) Size structure of the char population in Takvatnet in 1985 and 1998 (data from Klemetsen et al., 2002).
INDIVIDUAL GROWTH AND BODY SIZE
the invasion of another predator by altering the size structure of the consumer (De Roos & Persson, 2005b). The emergent Allee effect may be one explanation for the collapse and lack of recovery that has been observed in stocks of marine top predators (Carscadden, Frank & Leggett, 2001; De Roos & Persson, 2002). In agreement with model predictions, capelin, the main prey fish of cod, has been observed to show decreased per capita fecundity and mean size following the collapse of the North Atlantic cod (Carscadden et al., 2001). A large-scale removal of a fish stock in the Norwegian Lake Takvatn provides another example suggesting alternative states induced by the emergent Allee effect. At the beginning of the study, the population of the numerically dominant fish species (Arctic char) was stunted, with very few individuals growing larger than 25 cm (Fig. 12.5b) (Klemetsen et al., 2002). Following the removal of more than 70% of the Arctic char during the late 1980s, individual growth rate increased drastically, leading to a shift towards larger individuals in the size distribution that still persists (Fig. 12.5b). Despite the reduction of the Arctic char population, population fecundity increased substantially to more than six times the original population fecundity, thus showing strong elements of an overcompensatory effect. Following the reduction in char numbers, its main predator, brown trout, which was very rare at the start of the experiment, has increased in numbers by 30 times. Thus, the data for both predator and prey all support the contention that the dynamics of this system may involve an emergent Allee effect. The trait that individuals grow over their life cycle will affect the dominant interaction they experience between competitive and predatory interactions resulting in life-history omnivory (Werner & Gilliam, 1984; Lasenby, Northcote & Fu¨rst, 1986; Persson, 1988; Wilbur, 1988). For size-structured omnivorous (intraguild predation) systems, food-dependent development has been shown to reduce the scope for coexistence between top predator and intermediate consumer compared to unstructured models or structured models where transitions between stages is not food (or density) dependent (van de Wolfshaar, 2006). In systems with life-history omnivory where the top predator competes with the intermediate consumer at small sizes (, effect) but preys on it at larger sizes (þ, effect), model results and field data suggest that very different size distributions of the predator and prey will develop depending on environmental conditions (for example, productivity). Modelling results suggest that the growth of young (young-of-the-year) top predators will generally be slower in the presence of the intermediate consumer than in its absence, showing an interspecific competitive effect of the intermediate consumer on juvenile predators (Figs. 12.6a, b). In contrast, the effects on the growth of larger size classes of the top predator can be qualitatively different, with both an increase as well as a decrease in the maximum size of the top predator in the presence of the intermediate consumer compared with in its
237
L. PERSSON AND A. M. DE ROOS
(a)
(b)
OYO
Numbers
OYO 10 000
10 000
1000
1000
TP alone
100
100 TP alone
10
10 50
100
150
200
250
50
300
100
150
200
250
300
100
150
200
250
300
(d)
(c) 10 Percentage
238
10 1 1 0.1 0.1 50
100
150
200
250
50 300 Length (mm)
Figure 12.6 Above: model predictions of size distributions of top predator (TP) where the maximum size achieved is increased (a) or decreased (b) in the presence of the intermediate consumer compared with when the intermediate consumer is absent (arrows TP alone). In both cases the sizes of one-year-old (OYO) top predators are smaller in the presence of the intermediate consumer than in its absence (arrows OYO) (from van de Wolfshaar, 2006). Below: different size distributions of perch in moderately productive Lake N. Bolmen (c) and highly productive Lake So¨vdeborg (d) (data from Persson, 1983).
absence (Figs. 12.6a, b) (van de Wolfshaar, 2006). Varying size distributions of perch (top predator) populations coexisting with roach (Rutilus rutilus) (intermediate consumer) have also been observed and attributed to the relative strength of predatory and competitive interactions, respectively (Figs. 12.6c, d) (Persson, 1983; Bystro¨m, Persson & Wahlstro¨m, 1998). The empirical relationship between intermediate consumer abundance and size distribution and top predator abundance and size distribution is presently under investigation. To summarize, size-dependent multispecies interactions exemplified in the tritrophic food chain and in systems with life-history omnivory potentially cause substantially different body-size distributions of communities. Furthermore, size-dependent dynamics clearly affect the stability properties of ecological communities and promote the presence of alternative states. This
INDIVIDUAL GROWTH AND BODY SIZE
suggests that an explicit consideration of the dynamics of size-dependent interactions may be essential to understand and predict body-size distributions in many systems.
Extensions to more complex configurations The results presented above show that size-dependent processes have important implications for the dynamics and structure of ecological communities. Sizestructured interactions may promote the likelihood of alternative states in terms of both species composition (presence/absence) as well as size distributions of coexisting species. Still, the model complexity and parameter richness of the consumer-resource model considered above that forms the basis for many of the results discussed here, will limit the extent to which these models can be expanded to multispecies configurations. Since many of the questions raised in the body-size literature deal with multi-species systems, there is a need for simpler modules that can handle more speciose systems but still incorporate essential aspects of the individual’s life history, especially food-dependent growth. Recently, a model-building block based on stages, termed a ‘structured biomass community module’ has been developed that accounts for fooddependent development and maturation (De Roos et al., unpublished). This modelling approach parallels the bioenergetics population models developed by Yodzis and Innes (1992) for non-structured populations. Compared with the consumer-resource model considered above, the dimensionality of the system is heavily reduced to a two-stage (juveniles, adults) model. Both juvenile and adult consumption is assumed to follow a Type II functional response with a maximum ingestion rate Imax and a half saturation constant H. The net production per unit of biomass for adult ( a) and juveniles (vj), respectively, is given by: R va ¼ qImax HþR T R T vj ¼ Imax HþR
(12:1)
q represents the difference between adults and juveniles in time spent foraging on the resource. It is assumed that ingestion and metabolic rate scale linearly with body mass. Ingested energy is assimilated with an efficiency , and assimilated energy is first used to cover maintenance T. The net energy left is transformed into consumer biomass. For a system with juveniles (J) investing all their net energy into growth and adults (A) investing all their net energy into reproduction, the dynamics of the system takes the following form:
239
240
L. PERSSON AND A. M. DE ROOS
dJ ¼ aþ A þ j J jþ J J J dt dA þ ¼ va A v þ a A þ v j J A A dt dR R ¼ R ðImax J þ qImax AÞ dt HþR
(12:2)
Juvenile biomass increases through recruitment (aþ A) and growth in biomass ( jJ) and decreases due to maturation to the adult stage ( jþ J) and mortality (JJ). Adult biomass increases through maturation of juveniles ( jþ J) and decreases due to mortality (AA). aþ A can only take positive values and ensures that no reproduction in adults takes place if resource levels are too low to sustain maintenance. Similarly, jþ J ensures that no maturation takes place if resource levels are too low. is resource productivity and resource turn-over rate. The maturation function is dependent on juvenile mortality, newborn and adult size, and the net production per unit body mass of juveniles, j. The structured biomass community model is derived through the formulation of a physiologically structured population model and also yields the same equilibrium results as the latter. Analyses show that the multiple equilibria resulting from the emergent Allee effect, emergent facilitation and ontogenetic niche shifts are also present in this more aggregated biomass community model (De Roos et al., unpublished). With a body-size perspective, this more aggregated module thus has the potential to allow an investigation of size-structured induced shifts in body-size distributions in many species systems.
Conclusions As evident in our retrospective overview, growth and development is a major feature of many organisms on Earth, hence the implications of this for individual, population, community and ecosystem processes can hardly be neglected. This circumstance has also been recognized in the most recent body-size literature (Woodward et al., 2005; Woodward & Warren, this volume). As discussed in this chapter, growth and development take different forms in different organisms, but one dividing line can be seen along the ectothermy–endothermy gradient. Theoretical studies suggest that a critical element giving rise to the various size-dependent patterns is food-dependent development per se and not whether organisms continue to grow in somatic size over their whole life period (De Roos et al., 2003b; De Roos et al., unpublished). That is, whether individuals continue to grow in somatic tissue or allocate all their net energy into reproduction after maturation is not a factor affecting the presence of, for example, the emergent Allee effect. The recent body-size literature has emphasized the need to document sizedependent interactions at the level of individuals, with the main purpose of reducing biases in estimates of predator/prey (parasitoid/host) ratios (Cohen
INDIVIDUAL GROWTH AND BODY SIZE
et al., 2005). The investigation of individual size scaling of foraging rates in particular, was a significant area of research 20 years ago including the development of individual level models (Werner & Gilliam, 1984). The development of the framework of physiologically structured population models with its two-level (individual, population) representation has made it possible formally to connect size-dependent individual-level processes to the population level, which, in turn, has allowed us to start to answer questions about the effects of body-size scaling for population and community dynamics and its feedbacks on body-size distributions. The insights gained so far from these analyses concern both dynamical aspects and equilibrium properties. As an example of the former, studies of size-structured consumer-resource systems show that cohort cycles are the most common outcome of consumerresource interactions (Persson et al., 1998). Furthermore, these cohort cycles are not restricted to systems with few species but are also present in manyspecies systems; hence the non-equilibrium dynamics resulting from cohort dynamics cannot be ignored in a food-web context (Murdoch et al., 2002). Correspondingly, the presence of temporal variation in body-size distribution is increasingly recognized in the body-size literature (Stead et al., 2005; Woodward et al., 2005). Studies of multitrophic systems show that size-dependent processes will give rise to alternative equilibrium states with major ramifications for overall bodysize distributions (De Roos & Persson, 2002; De Roos et al., 2003a). Given the common observation of food-dependent development in organisms, we argue that approaches not incorporating food-dependent development may have a limited capacity to yield both understanding of, and useful predictions, about the dynamics and structure of ecological communities. A major challenge for the future is to develop approaches that allow the analyses of more complex configurations in terms of the number of species present, but still encompass major processes resulting from ontogenetic development (De Roos et al., unpublished). Such approaches should also make it possible to provide new insights about the community dynamics based on a relatively limited number of intraspecific size-scaling parameters along the lines of the body-size based trophic dynamic models developed for interspecific interactions by Yodzis and Innes (1992).
Acknowledgements We thank the organizers for inviting us to the British Ecological Society meeting on the effects of body size in aquatic systems. The research based on which this overview was written has been supported by the Swedish Research Council and the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning to L. Persson, and the Netherlands Organization for Scientific Research to A. M. De Roos.
241
242
L. PERSSON AND A. M. DE ROOS
References Brooks, J. L. & Dodson, S. I. (1965). Predation, body size and composition of plankton. Science, 150, 28–35. Brown, J. H., Gillooly, J. F., Allen, A. P. Savage, V. M. & West, G. B. (2004). Towards a metabolic theory of growth. Ecology, 85, 1771–1789. Bystro¨m, P., Persson, L. & Wahlstro¨m, E. (1998). Competition between predator and prey – competitive juvenile bottlenecks in whole lake experiments. Ecology, 79, 2153–2167. Calder, W. A. III. (1984). Size, Function and Life History. Cambridge, MA: Harvard University Press. Carscadden, J. E., Frank, K. T. & Leggett, W. C. (2001). Ecosystem changes and the effects on capelin (Mallotus villosus), a major forage species. Canadian Journal of Fisheries and Aquatic Sciences, 58, 73–85. Case, T. J. (1979). Optimal body size and an animal’s diet. Acta Biotheoretica, 28, 54–69. Chen, X. & Cohen, J. E. (2001). Transient dynamics and food-web complexity in the Lotka Volterra cascade model. Proceedings of the Royal Society of London Series B, 268, 869–877. Claessen, D. & De Roos, A. M. (2003). Bistability in a size-structured population model of cannibalistic fish – a continuation study. Theoretical Population Biology, 64, 49–65. Claessen, D., De Roos, A. M. & Persson, L. (2000). Dwarfs and giants – cannibalism and competition in size-structured populations. American Naturalist, 155, 219–237. Claessen, D., van Oss, C., De Roos, A. M. & Persson. L. (2002). The impact of sizedependent predation on population dynamics and individual life history. Ecology, 83, 1660–1675. Cohen, J. E. (1985). Metamorphosis: introduction, usages, and evolution. In Metamorphosis, ed. M. Bulls and M. Brownes. Oxford: Clarendon, pp. 1–19. Cohen, J. E., Jonsson, T. & Carpenter, S. R. (2003). Ecological community description using the
food web, species abundance, and body size. Proceedings of the National Academy of Science (USA), 100, 1781–1786. Cohen, J. E., Jonsson, T., Mu¨ller, C. B., Godfray, H. C. J. & Savage, V. M. (2005). Body sizes of hosts and paraitoids in individual feeding relationships. Proceedings of the National Academy of Science (USA), 102, 684–689. Cryer, M., Peirson, G. & Townsend, C. R. (1986). Reciprocal interactions between roach, Rutilus rutilus, and zooplankton in a small lake: Prey dynamics and fish growth and recruitment. Limnology and Oceanography, 31, 1022–1038. De Roos, A. M. & Persson, L. (2002). Sizedependent life-history traits promote catastrophic collapses of top predators. Proceedings of the National Academy of Science (USA), 99, 12907–12912. De Roos, A. M. & Persson, L. (2005a). Unstructured population models: do general assumptions yield general theory? In Ecological Paradigms Lost: Routes to Theory Change, ed. B. Beissner and K. Cuddington. Amsterdam: Academic Press, pp. 31–62. De Roos, A. M. & Persson, L. (2005b). The influence of individual growth and development on the structure of ecological communities. In Dynamic Food Webs – Multispecies Assemblages, Ecosystem Development and Environmental Change, ed. P. C. de Ruiter, V. Wolters and J. C. Moore. Amsterdam: Academic Press, pp. 89–100. De Roos, A. M., Metz, J. A. J., Evers, E. & Leipoldt, A. (1990). A size dependent predator-prey interaction: who pursues whom? Journal of Mathematical Biology, 28, 609–643. De Roos, A. M., Persson, L. & McCauley, E. (2003a). The influence of size-dependent life history traits on the structure and dynamics of populations and communities. Ecological Letters, 6, 473–487. De Roos, A. M., Persson, L. & Thieme, H. (2003b). Emergent Allee effects in top predators
INDIVIDUAL GROWTH AND BODY SIZE
feeding on structured prey populations. Proceedings of the Royal Society of London Series B, 270, 611–618. Ebenman, B. (1992). Evolution in organisms that change their niches during the life-cycle. American Naturalist, 139, 990–1021. Ebenman, B. & Persson, L. (1988). Dynamics of size-structured populations – an overview. In Size-Structured Populations – Ecology and Evolution, ed. B. Ebenman & L. Persson. Berlin: Springer, pp. 3–9. Emmerson, M. & Raffaelli, D. (2004). Predator prey body size, interaction strength and the stability of a real food web. Journal of Animal Ecology, 63, 399–409. Gaston, K. J. & Lawton, J. H. (1988). Patterns in the abundance and distribution of insect populations. Nature, 331, 709–712. Hamrin, S. F. & Persson, L. (1986). Asymmetrical competition between age classes as a factor causing population oscillations in an obligate planktivorous fish. Oikos, 47, 223–232. Juanes, F. (2003). The allometry of cannibalism in piscivorous species. Canadian Journal of Fisheries and Aquatic Sciences, 60, 594–602. Kerr, S. R. & Dickie, L. M. (2001). The Biomass Spectrum. A Predator-Prey Theory of Aquatic Production. New York: Columbia University Press. Klemetsen, A., Amundsen, P. A., Grotnes, P. E. et al. (2002). Takvatn through 20 years: Long-term effects of an experimental mass removal of Arctic charr, Salvelinus Alpinus, from a subarctic lake. Environmental Biology of Fishes, 64, 1–3. Kooijman, S. A. L. M. (2000). Dynamic Energy and Mass Budgets in Biological Systems, 2nd edn. Cambridge: Cambridge University Press. Lasenby, D. C., Northcote, T. G. & Fu¨rst, M. (1986). Theory, practice and effects of Mysis relicta introductions to North American and Scandinavian lakes. Canadian Journal of Fisheries and Aquatic Sciences, 43, 1277–1284. Loeuille, N. & Loreau, M. (2005). Evolutionary emergence of size-structured food webs.
Proceedings of the National Academy of Science (USA), 102, 5761–5766. Lundberg, S. & Persson, L. (1993). Optimal body size and resource density. Journal of Theoretical Biology, 164, 163–180. McCauley, E., Nisbet, R. M., Murdoch, W. W., De Roos, A. M. & Gurney, W. S. C. (1999). Largeamplitude cycles of Daphnia and its algal prey in enriched environments. Nature, 402, 653–656. Metz, J. A. J. & Diekmann, O. (1986). The Dynamics of Physiologically Structured Populations. Springer lecture notes in biomathematics, Vol. 68. Heidelberg: Springer. Mittelbach, G. G. (1981). Foraging efficiency and body size: a study of optimal diet and habitat use by bluegills. Ecology, 62, 1370–1386. Murdoch, W. W., Kendall, B. E., Nisbet, R. M. et al. (2002). Single-species models for manyspecies food webs. Nature, 417, 541–543. Persson, L. (1983). Food consumption and competition between age classes in a perch (Perca fluviatilis) population in a shallow eutrophic lake. Oikos, 40, 197–207. Persson, L. (1985). Asymmetrical competition: are larger animals competitively superior? American Naturalist, 126, 261–266. Persson, L. (1988). Asymmetries in competitive and predatory interactions in fish populations. In Size-Structured Populations – Ecology and Evolution, ed. B. Ebenman and L. Persson. Berlin: Springer, pp. 203–218. Persson, L., Leonardsson, K., Gyllenberg, M., De Roos, A. M. & Christensen, B. (1998). Ontogenetic scaling of foraging rates and the dynamics of a size-structured consumerresource model. Theoretical Population Biology, 54, 270–293. Persson, L., De Roos, A. M., Claessen, D. et al. (2003). Gigantic cannibals driving a whole lake trophic cascade. Proceedings of the National Academy of Science (USA), 100, 4035–4039. Peters, R. H. (1983). The Ecological Implications of Body Size. Cambridge: Cambridge University Press.
243
244
L. PERSSON AND A. M. DE ROOS
Polis, G. A. (1988). Exploitation competition and the evolution of interference, cannibalism, and intraguild predation in age/size-structured populations. In Size-Structured Populations – Ecology and Evolution, ed. B. Ebenman and L. Persson. Berlin: Springer, pp. 185–202. Polis, G. A., Myers, C. & Holt, R. (1989). The ecology and evolution of intraguild predation: potential competitors that eat each other. Annual Review in Ecology and Systematics, 20, 297–330. Ricklefs, R. E. (1973). Patterns of growth in birds. II. Growth rate and mode of development. Ibis, 115, 177–201. Sanderson, B. L., Hrabik, T. R., Magnuson, J. J. & Post, D. D. (1999). Cyclic dynamics of a yellow perch (Perca flavescens) population in an oligotrophic lake: evidence for the role of intraspecfic interactions. Canadian Journal of Fisheries and Aquatic Sciences, 56, 1534–1542. Sauer, J. R. & Slade N. A. (1987). Size-based demography of vertebrates. Annual Review in Ecology and Systematics, 18, 71–80. Sebens, K. P. (1982). The limits to indeterminate growth: an optimal model applied to passive suspension feeders. Ecology, 63, 209–222. Sebens, K. P. (1987). The ecology of indeterminate growth in animals. Annual Review in Ecology and Systematics, 18, 371–407. Shin, Y. J., Rochet, M., Jennings, S., Field, J. G. & Gislason, H. (2005). Using size-based indicators to evaluate ecosystem effects of fishing. ICES Journal of Marine Science, 62, 384–396. Shiomoto, A., Tadokoro, K., Nagasawa, K. & Ishida, Y. (1997). Trophic relations in the subArctic North Pacific ecosystem: possible feeding effect from pink salmon. Marine Ecology Progress Series, 150, 75–85. Sinko, J. W. & Streifer, W. (1967). A new model for age-size structure of a population. Ecology, 48, 910–918. Stead, T. K., Schmid-Arya, J. M., Schmid, P. E. & Hildrew, A. G. (2005). The distribution of body size in a stream community: one system, many patterns. Journal of Animal Ecology, 74, 475–487.
Stearns, S. C. (1992). The Evolution of Life Histories. Oxford: Oxford University Press. Townsend, C. R., Sutherland, W. J. & Perrow, M. R. (1990). A modelling investigation of population cycles in the fish Rutilus rutilus. Journal of Animal Ecology, 59, 469–485. van de Wolfshaar, K. E. (2006). Population persistence in the face of size-dependent predation and competition interactions. PhD thesis, University of Amsterdam. Werner, E. E. (1986). Species interactions in freshwater fish communities. In Community Ecology, ed. J. Diamond and T. J. Case. New York: Harper & Row, pp. 344–358. Werner, E. E. (1988). Size, scaling and the evolution of complex life cycles. In Size-Structured Populations – Ecology and Evolution, ed. B. Ebenman and L. Persson. Berlin: Springer, pp. 60–81. Werner, E. E. (1994). Ontogenetic scaling of competitive relations: size-dependent effects and responses in two Anuran larvae. Ecology, 75, 197–231. Werner, E. E. & Gilliam, J. F. (1984). The ontogenetic niche and species interactions in size-structured populations. Annual Review in Ecology and Systematics, 15, 393–425. West, G. B., Brown, J. H. & Enquist, B. J. (1997). A general model for the origin of allometric scaling laws in biology. Science, 276, 122–126. Wilbur, H. M. (1988). Interactions between growing predators and growing prey. In SizeStructured Populations – Ecology and Evolution, ed. B. Ebenman and L. Persson. Berlin: Springer, pp. 157–172. Wilson, D. S. (1975). The adequacy of body size as a niche difference. American Naturalist, 109, 769–784. Woodward, G., Ebenman, B., Ernmerson, M. et al. (2005). Body size in ecological networks. Trends in Ecology and Evolution, 20, 402–409. Yodzis, P. & Innes, S. (1992). Body size and consumer resource dynamics. American Naturalist, 139, 1151–1175.
CHAPTER THIRTEEN
The consequences of body size in model microbial ecosystems OWEN L . PETCHEY University of Sheffield
ZACHARY T . LONG University of North Carolina at Chapel Hill Virginia Institute of Marine Science
PETER J . MORIN Rutgers University
Introduction Patterns in the sizes of coexisting organisms have always intrigued ecologists (Hutchinson, 1961). Some kinds of regularities are well known for some systems (Sheldon, Prakash & Sutcliffe, 1972), and are less appreciated or rediscovered for others (Enquist & Niklas, 2001; Cohen, Jonsson & Carpenter, 2003). One example of this kind of pattern is the apparent constancy of total biomass within different size fractions of organisms living in aquatic communities (Sheldon, et al. 1972; Cyr, 2000; Kerr & Dickie, 2001; Cohen et al., 2003; Sheldon, Sutcliffe & Paranjape, 1977; Tilman et al., 2001; Mulder et al., 2005). Essentially, over many orders of magnitude of organism size, any particular size class holds about the same total biomass of organisms per unit volume. The result is an inverse relationship between the log of organism size and the log of organism abundance per unit volume, with a slope of 1. The apparent constancy of this relationship has even led some workers to suggest, tongue in cheek, that it could be used to estimate the population size of some organisms that have proven to be notoriously difficult to observe, once assumptions about their average size were made (Sheldon & Kerr, 1972, 1973). Whether or not the elusive Loch Ness Monster (to which these calculations were rather whimsically applied) actually exists, it appears that the total biomass of organisms in some habitats is fixed by certain features of the habitat, most likely the abundance of incoming energy and nutrients that drive productivity (Sheldon et al., 1977; Cyr, 2000; Kerr & Dickie, 2001; Cohen et al., 2003; Mulder et al., 2005). A recent flurry of research on the allometry of metabolism seeks to explain such patterns mechanistically as a consequence of the ways that organisms capture and transport energy and material (Brown, 2004). It is worth noting that none of these patterns suggest that diversity, the number of different organisms that total biomass is divided among, should necessarily influence the total standing stock of biomass supported by a particular environment and its energetic regime. Body Size: The Structure and Function of Aquatic Ecosystems, eds. Alan G. Hildrew, David G. Raffaelli and Ronni Edmonds-Brown. Published by Cambridge University Press. # British Ecological Society 2007.
246
O. L. PETCHEY ET AL.
Independently of this line of inquiry, ecologists have suggested that the total biomass of particular subgroups of organisms, plants for instance, can increase as the diversity of organisms within that group increases (Hector et al., 1999; Tilman et al., 2001). For this pattern to square with the pattern of constant total biomass observed in natural systems, one of several things must hold. (i) Natural communities may be at the maximum diversity that a particular habitat will support – and the communities where diversity-dependent biomass increases are seen must be below that level of diversity. (ii) An increase in the diversity and biomass within a particular size fraction of the community will produce a corresponding increase in the biomass of size fractions in the remainder of the community – an interesting but little explored phenomenon. Some of the work that we describe here was initially conceived and conducted to evaluate critiques of early biodiversity experiments, specifically to determine whether what are now known as selection effects caused by the chance occurrence of larger organisms can explain patterns of increased biomass in more diverse systems. While it seems likely that such transient effects could arise as a consequence of initial conditions, the size invariance of productivity and biomass noted above make it seem unlikely that such differences would persist for long in communities after population densities of organisms respond and adjust to the availability of energy and nutrients. To address these ideas we first ask if diversity per se can change the size–population abundance relationship within communities and among experiments. Second, we ask whether communities initially constructed of organisms of very different size would rapidly converge in biomass. Finally, we investigate the consequences of body size for an ecosystem-level process: whole community metabolism. Specifically, we test the hypothesis that the metabolism of a given biomass will depend on whether the biomass is divided up among lots of small individuals or just a few large ones. Why investigate these consequences using data from model microbial ecosystems? Microbial microcosms can contain organisms with a wide range (about seven orders of magnitude) of organism size. The organisms used in our microcosms also fill a wide range of trophic roles, including photoautotrophs, decomposers, primary consumers and secondary consumers. This allowed us to evaluate effects of size across multiple trophic levels, thereby extending the analysis beyond primary producers (Duffy, 2002; Petchey et al., 2004). Additionally, these organisms have generation times ranging from hours to days. This allows for rapid population dynamics, such that an experiment encompassing many generations of population dynamics can take place in a few weeks. Properly testing the predictions of allometric theory may require multiple generations because it may take many generations for changes in population sizes due to thinning or density compensation to occur, so that initial conditions, transient dynamics, and potential selection effects may be less important. We could also limit the effect of the environment on abundance and yield by holding environmental conditions
CONSEQUENCES OF BODY SIZE IN MODEL MICROBIAL ECOSYSTEMS
constant (Gillooly et al., 2001; Savage et al., 2004). Conversely, controlled manipulation of environmental conditions, particularly temperature, allows investigation of how it alters the consequences of body size.
Methods General methods The analyses presented here draw on data from four experiments that share much methodology in common. In all, experimental communities were 240–250 ml glass or pyrex jars each containing 100 ml of nutrient medium (1 litre of well water, 0.55 g Carolina Biological Supply Protozoan Pellet, and 2–6 wheat seeds) housed in incubators at 188C unless otherwise stated below. This medium was first autoclaved and then inoculated with three or four species of bacteria (Serratia marcescens, Bacillus subtilus, B. cereus and sometimes Proteus vulgaris). To this basal food-web level were added various combinations of the protist (algae, ciliates and sarcodines) and small metazoans (rotifers) that were kept in laboratory stock cultures. Bacterivores and autotrophs were typically added before predators and herbivores, to allow resources to grow to levels that could sustain consumers. The experiments lasted six or seven weeks and were sampled between two and six times. All analyses are of only the last week of collected data. Sampling involved swirling the microcosms to homogenize the distribution of organisms, removal of first 0.3 ml and then potentially 5 ml of media. The number of individuals of each species was recorded in the smaller volume and if some species were not observed the larger volume was searched. Autotrophs were sampled by a third method: enumeration using a haemocytometer, with a minimum detection threshold of approximately 104 ml1. All counts were transformed into number of individuals per ml to give the population density of each species in each community. Cell size (mass here) was estimated from linear dimensions of up to ten cells of each species and the rough shape of the cells (Wetzel & Likens, 1991). Transformation of the linear dimensions to volume, and assumption of 1 mm3 ¼ 1 milligram resulted in the mass of each cell and is reported in milligrams here. Some variation in cell size occurs during the growth of protist populations and can also result from changes in environmental conditions that occur during an experiment (e.g. depletion of available resources) or from experimental treatments (e.g. a cool versus warm environment). Though here we ignore this intraspecific variation in cell mass, it seems unlikely that it could invalidate the analyses of many orders of magnitude of interspecific variation in cell mass presented. Choice of species One might argue that studying the consequences of body size in communities with artificial body-size distributions is meaningless. We do not believe this is the case. First, all of the species in our laboratory stocks potentially co-occur in natural aquatic communities and so are not necessarily evolutionarily and ecologically
247
248
O. L. PETCHEY ET AL.
naı¨ve to each other. Consequently, the community and ecosystem properties that emerge result from similar ecological interactions as might occur in natural systems. Second, in each of the four experiments analyzed, the species present in particular communities were selected at random from the laboratory pool. The frequency distribution of body sizes in this pool is approximately log-normally distributed (unpublished data). It therefore seems unlikely that we have biased body-size distributions by choice of particular species. However, extinctions often occur during an experiment and alter community composition (Weatherby, Warren & Law, 1998; McGrady-Steed & Morin, 2000). Presumably this change in composition results from ecological interactions and influences body-size distributions and their consequences. Third, because the experiments are long relative to the generation time of the organisms, great changes in population size occur. For example, a bacterivores species may initially occur at ten individuals per ml, but in a week have multiplied to thousands of individuals per ml. So although cellsize distributions are somewhat constrained, population densities, population biomasses and ecosystem process rates are emergent. Description of the four experiments and data sets
Experiment 1 This experiment was designed to address two questions (Long & Morin, 2005). First, whether diversity or the size of dominant species affects ecosystem functioning, and second, how effects changed in importance through time. Food webs were assembled using two size fractions (small species or large species) and two levels of diversity (four or nine species) using autotrophic and heterotrophic species of protist and small metazoan. The two size fractions of organisms differed in average cell mass by approximately two orders of magnitude. There were two different community compositions in the low diversity food webs. The resulting six food webs were replicated five times. Population densities were measured once a week for each of the six weeks of the experiment. Experiment 2 The experiment was designed to investigate the relationship between biodiversity and ecosystem functioning and the variability of ecosystem functioning, and is described in detail elsewhere (McGrady-Steed, Harris & Morin, 1997). There were 10 different communities with from 3 to 31 autotrophic and heterotrophic species of protists and small metazoans distributed among multiple trophic levels. There were five replicates of each community, and population densities were recorded at least once a week for six weeks. Experiment 3 This experiment was designed to examine the effects of environmental warming on ecological communities (Petchey et al., 1999). There were four different
CONSEQUENCES OF BODY SIZE IN MODEL MICROBIAL ECOSYSTEMS
communities (two compositions at two diversity levels) and an environmental warming treatment (constant temperature or warmed). Five replicates of each treatment combination made a total of 40 microcosms. Population densities were sampled in the second and sixth week of the experiment.
Experiment 4 This experiment tested the insurance hypothesis: that biodiversity stabilizes aggregate community and ecosystem properties (Petchey et al., 2002). There were four diversity levels (two, four, six, and eight species) and four compositions at each level. This is the only experiment that contained a single trophic level of protist species (bacterivores). In addition, there were three environmental variability treatments (constant temperature, fast temperature fluctuations and slow temperature fluctuations). Population densities and CO2 flux were estimated in the second and sixth week of the experiment. Population consequences of body size We analyzed the relationship between cell mass and population size within and across the four experiments using three separate analysis of covariance models. In all analyses, data points were averages of the values of a species across the replicates of each treatment combination in which it occurred. That is, there is only one data point per species per treatment combination. All analyses included log 10 abundance as the response variable, and log 10 body size as a continuous explanatory variable. The first model included three data sets (Experiments 1, 2, 3) but only constant temperature regimes. The explanatory variables were trophic group (categorical, five levels, autotroph, bacterivores, herbivores, omnivores, predators), experiment (categorical, three levels, 1, 2, 3) and species richness (continuous). The second model included only the data set from Experiment 3 and explanatory variables were environment (categorical, two levels, constant temperature and higher temperature) and species richness (continuous). The third model included only the data set from Experiment 4 and explanatory variables of environment (categorical, three levels, constant temperature, fast varying temperature, slow varying temperature) and species richness (continuous). Significant interaction terms between body mass and another explanatory variable would indicate that variables affect the population consequences of body size. All interaction terms were initially included and backwards elimination by Akaike Information Criterion (AIC) and F-test were used to find the minimum adequate model (Crawley, 2002). Community consequences of body size Food webs that receive similar amounts of energy should reach similar levels in total biomass, regardless of differences in organism size found among different food webs (Damuth, 1981, 1987; Brown & Gillooly, 2003; Brown, 2004; Mulder
249
250
O. L. PETCHEY ET AL.
et al., 2005). If this occurs, distinct food webs that experience similar abiotic conditions should fall on the single constant yield line described by log Ntotal ¼ log b log Maverage, where Ntotal is the total number of individuals in a community, Maverage is the average mass of an individual in a community, and b reflects environmental influences like resource supply rates and temperature. The slope of 1 means that a community containing cells twice the average mass of those in another community will contain half the number of individuals. Consequently, differences in the mass of individuals are accompanied by opposite and perfectly compensatory changes in the number of individuals. To test this hypothesis we used the same three analyses as described earlier except: log 10 Ntotal was the response variable, log 10 Maverage was the continuous explanatory variable, datapoints were communities (rather than species), and there was no ‘trophic group’ categorical variable. Ecosystem consequences of body size Metabolic rate scales with individual mass0.75 (Peters, 1983; Damuth, 1987; West, Brown & Enquist, 1997) and consequently lots of small individuals should have higher total metabolism than just a few large ones, even if they represent the same amount of total biomass. Of course, the microbial communities that we analyze are composed of a mixture of small or large individuals, so we predict community metabolism from the size and abundance observed for each of the species in the community. This prediction is compared with observed flux of CO2, a measure of whole community metabolism. Observed ecosystem level CO2 flux was estimated using a closed-circuit microrespirometer (Micro-oxymax, Columbus Instruments, OH, USA). This measures the CO2 concentration in the gas above the liquid in the sealed culture vessels at multiple discrete times, and flux can be calculated from these measures. Concentrations were measured during at least an eight-hour period and flux averaged across that. Can the total biomass or body sizes of the organisms in the community in a particular vessel predict ecosystem-level CO2 flux? To test this we assumed that CO2 flux is proportional to community metabolism (CM) and this is the sum of the metabolism of all the individuals in the community. We limit the analyses to the data from the constant temperature replicates of the only experiment that included only heterotrophic species (Experiment 4). The other three experiments all contained autotrophs. If the individuals are grouped into species, community metabolism is the sum of the species’ metabolisms, where a species’ metabolism is the sum of the metabolisms of the individuals of that species:
CM ¼
s X i¼1
Ni Mi ;
(13:1)
CONSEQUENCES OF BODY SIZE IN MODEL MICROBIAL ECOSYSTEMS
where Ni is the number of individuals of species i, Mi is the metabolism of an individual of species i, and s is the number of species in the community. An individual’s metabolism is related to its body size by allometry such that Mi ¼ b B0:75 ; i
(13:2)
were Bi is the body size of species i and b is a constant converting mass to energy per second. Consequently, CM ¼ b
s X
Ni B0:75 i
(13:3)
i¼1
This prediction for community metabolism (CO2 flux) was compared to the explanatory power of total biomass using linear regression by ordinary least squares.
Results Population consequences of body size – results In each of the three analyses of the relationship between cell mass and population density (the mass–density relationship), cell mass was very strongly and relatively linearly related to population size (Figs. 13.1, 13.2 and 13.3). However, the slope differed greatly between the first two analyses and the third analysis (0.93, 0.95 and 0.52, respectively) (Tables 13.1, 13.2 and 13.3). Communities in the third analysis contained only one trophic level, while all of the communities in the first and second contained species at several trophic levels. However, effects of trophic group on the mass-density relationship within the first two analyses were inconsistent with this between analysis difference. In the first, slopes differed between trophic groups, with the slope becoming shallower from autotrophs, to bacterivores, to omnivores (Table 13.1). In the second analysis, trophic group had a marginally significant effect on the intercept (but not the slope), with bacterivores having the steepest slope and omnivores the shallowest (Table 13.2). The species richness of the community had a significant negative effect on the intercept of the mass–density relationship in analysis one (Figure 13.1, Table 13.1). This is a signature of density compensation, in which populations within less species-rich communities are on average denser than when in richer communities (McGrady-Steed & Morin, 2000; Petchey et al., 2002); density compensation can occur because species compete for similar resources and are therefore somewhat redundant. Density compensation was not evident in analysis two or three where there was no significant effect of species richness on the slope or intercept of the mass–density relationship (Figs. 13.2 and 13.3, Tables 13.2 and 13.3). Perhaps surprisingly, there was no evidence that environment conditions altered the slope or intercept of the mass–density relationship. In analysis
251
O. L. PETCHEY ET AL.
(b) 2 species 6 species 10 species 14 species 18 species
6
4
2
Log10(Density per ml)
Log10(Density per ml)
(a)
Autotroph Bacterivore Omnivore
6
4
2
0
0 –8
–7
–6
–5
–4
–3
–2
Log10(Cell mass (mg))
–8
–7
–6
–5
–4
–3
–2
Log10(Cell mass (mg))
(c)
Log10(Density per ml)
252
Expt. 1 Expt. 2 Expt. 3
6
4
2
0 –8
–7
–6
–5
–4
–3
Log10(Cell mass (mg))
–2
Figure 13.1 Size-abundance relationship for analysis one, which tests for influence of (a) species richness, (b) species trophic position, and (c) experiment on the slope of relationship.
two, the 12 8C difference in temperature between normal and warmed communities had no significant effect (Fig. 13.2, Table 13.2), and in analysis three there was no apparent effect of the different temporal patterns of temperature variation (Fig. 13.3, Table 13.3). Community consequences of body size – results The slope of the relationship between total density and mean cell mass was 0.93 in analysis one (Fig. 13.4), 0.81 in analysis two (Fig. 13.5) and 0.35 in analysis three (Fig. 13.6). Again, results were therefore more similar in communities that contained species from multiple trophic levels (analyses one and two) compared with communities containing only one trophic level. The exponent for all communities in analysis one was not different from unity, indicating that the
CONSEQUENCES OF BODY SIZE IN MODEL MICROBIAL ECOSYSTEMS
(b) 2 species 6 species 10 species 14 species
6
4
2
Log10(Density per ml)
Log10(Density per ml)
(a)
Autotroph Bacterivore Omnivore
6
4
2
0
0 –8
–7
–6
–5
–4
–3
–2
Log10(Cell mass (mg))
–8
–7
–6
–5
–4
–3
–2
Log10(Cell mass (mg))
(c)
Log10(Density per ml)
Low temperature High temperature 6
4
2
0 –8
–7
–6
–5
–4
–3
Log10(Cell mass (mg))
–2
Figure 13.2 Size-abundance relationship for analysis two, which tests for influence of (a) species richness, (b) species trophic position, and (c) environmental temperature on the slope of the relationship.
communities reside on the constant yield line and total biomass was not a function of the size distribution of individuals. However, the slopes in analysis two and three were significantly less than unity, indicating that communities containing larger individuals contained greater total biomass. Species richness tended to increase the intercept of the relationship between average mass and total density in both analysis one and three (Figs. 13.4 and 13.6; Tables 13.4 and 13.6). This indicates that more speciose communities contain greater total biomass than less rich communities and has been observed directly in these communities before (McGrady-Steed & Morin, 2000; Petchey et al., 2002). This could result from a variety of mechanisms, including complementarity among species and more probabilistic processes (Loreau et al., 2001).
253
O. L. PETCHEY ET AL.
Table 13.1 Summary of analysis one of the population consequences of body size on population density, the response variable. M ¼ body mass, S ¼ species richness, T ¼ trophy, TB ¼ bacterivores, TO ¼ omnivores. M:T ¼ interaction term between M and T. * ¼ p < 0.05; ** ¼ p < 0.01; *** ¼ p < 0.001. Analysis of variance (ANOVA) table
M S T M:T Residuals
Df
Sum sq
Mean sq
F value
Pr(>F)
1 1 2 2 186
482.36 4.34 8.35 5.26 111.43
482.36 4.34 4.17 2.63 0.60
805.1509 7.2508 6.9687 4.3883
<2.2e 16*** 0.007734** 0.001207** 0.013734*
Coefficients
(Intercept) M S TB TO M:TB M:TO
Estimate
Std. error
t value
Pr(>|t|)
0.84994 0.92967 0.04101 0.22936 1.25460 0.13027 0.43899
0.42476 0.06802 0.01331 0.50877 0.54686 0.09105 0.14999
2.001 13.668 3.081 0.451 2.294 1.431 2.927
0.04685* <2e 16*** 0.00237** 0.65265 0.02290* 0.15420 0.00385**
2 species 4 species 6 species 8 species
4 3 2
Log10(Density per ml)
(b)
(a)
Log10(Density per ml)
254
Constant Fast variation Slow variation
4 3 2 1
1
–7
–6
–5
–4
–3
Log10(Cell mass (mg))
–2
–7
–6
–5 –4 –3 Log10(Cell mass (mg))
Figure 13.3 Size-abundance relationship for analysis three, which tests for influence of (a) species richness, (b) environmental variability on the slope of the relationship.
–2
CONSEQUENCES OF BODY SIZE IN MODEL MICROBIAL ECOSYSTEMS
Table 13.2 Summary of analysis two of the population consequences of body size on population density, the response variable. M ¼ body mass, T ¼ trophy, TB ¼ bacterivores, TO ¼ omnivores. ANOVA table
M T Residuals
Df
Sum sq
Mean sq
F value
Pr(>F)
1 2 181
625.23 5.25 142.27
625.23 2.62 0.79
795.4230 3.3367
<2e 16*** 0.03775*
Coefficients Estimate 1.38488 0.94759 0.26031 0.24933
(Intercept) M TB TO
Std. error 0.27125 0.04251 0.14711 0.24332
t value 5.105 22.291 1.769 1.025
Pr(>|t|) 8.3e 07*** <2e 16*** 0.0785 0.3069
Table 13.3 Summary of analysis three of the population consequences of body size on population density, the response variable. M ¼ body mass. ANOVA table
M Residuals
Df
Sum sq
Mean sq
F value
Pr(>F)
1 173
123.164 89.290
123.164 0.516
238.63
<2.2e 16***
Coefficients
(Intercept) M
Estimate
Std. error
t value
Pr(>|t|)
0.15208 0.52241
0.16945 0.03382
0.898 15.448
0.371 <2e 16 ***
There was an interaction between this effect and ‘experiment’ in analysis one: there was no biomass increase due to species richness in the warming experiment data. Analysis two showed a significant three-way interaction between average mass, species richness and environment, indicating that the slope of the relationship depended on the interaction of species richness and environmental temperature. In the warming experiment, the large and directional changes in average cell mass that occurred when warmed communities lost their large top predator
255
O. L. PETCHEY ET AL.
(b)
2 species 6 species 10 species 14 species 18 species
7
6
5
4
Log10(Total density per ml)
Log10(Total density per ml)
(a)
Expt. 1 Expt. 2 Expt. 3
7 6 5
4
3
3 –7
–6
–5
–4
–7
Log10(Average cell mass (mg))
–6
–5
–4
Log10(Average cell mass (mg))
Figure 13.4 Relationships between total density and average cell size of the individuals in a community in analysis one, coded by (a) species richness and (b) experiment.
(a)
(b) 2 species 6 species 10 species 14 species
7.0
6.5
6.0
–8.0
–7.5
–7.0
–6.5
–6.0
Log10(Average cell mass (mg))
7.5 Log10(Total density per ml)
7.5 Log10(Total density per ml)
256
7.0
6.5
Low temperature High temperature
6.0
–8.0
–7.5
–7.0
–6.5
–6.0
Log10(Average cell mass (mg))
Figure 13.5 Relationships between total density and average cell size of the individuals in a community in analysis two, coded by (a) species richness and (b) temperature.
species suggest that analysis of covariance (ANCOVA) may not be the most appropriate method of analysis, because of covariation between explanatory variables. Ecosystem consequences of body size – results Predicted community metabolism was positively related with total community biomass (Fig. 13.7a). However, there were three unusual communities with greater than 1 mg biomass per ml of media (this equates to about 0.1% of the culture volume occupied by organisms), while the remaining communities
CONSEQUENCES OF BODY SIZE IN MODEL MICROBIAL ECOSYSTEMS
Table 13.4 Summary of analysis one of the community consequences of body size on total community biomass. The response variable is total community density. M ¼ body mass, S ¼ species richness, E ¼ Experiment 1, E2 ¼ Experiment 2, E3 ¼ Experiment 3. ANOVA table
M S E S:E Residuals
Df
Sum sq
Mean sq
F value
Pr(>F)
1 1 2 2 93
82.631 3.085 3.058 0.828 5.610
82.631 3.085 1.529 0.414 0.060
1369.7640 51.1433 25.3445 6.8662
<2.2e 16*** 1.923e 10*** 1.638e 09*** 0.001655**
Coefficients
(Intercept) M S E2 E3 S:E2 S:E3
Estimate
Std. error
t value
Pr(>|t|)
0.37009 0.91873 0.12799 0.13487 0.74098 0.06112 0.12586
0.24455 0.03881 0.05476 0.23196 0.28869 0.05607 0.05855
1.513 23.670 2.337 0.581 2.567 1.090 2.150
0.1336 <2e 16*** 0.0216* 0.5624 0.0119* 0.2786 0.0342*
(a)
(b) 4.5 Log10(Density per ml)
Log10(Density per ml)
4.5 4.0 3.5 2 species 4 species 6 species 8 species
3.0 2.5 –6.5
–6.0
–5.5
4.0 3.5 3.0 Constant Fast variation Slow variation
2.5 –5.0
–4.5
Log10(Cell mass (mg))
–4.0
–6.5
–6.0
–5.5 –5.0 –4.5 –4.0 Log10(Cell mass (mg))
Figure 13.6 Relationships between total density and average cell size of the individuals in a community in analysis three, coded by (a) species richness and (b) environmental fluctuations. Arrows show the communities in Fig. 13.3 that have unusually high (>1 mg ml1) total biomass.
257
258
O. L. PETCHEY ET AL.
Table 13.5 Summary of analysis two of the community consequences of body size on total community biomass. The response variable is total community density. M ¼ body mass, S ¼ species richness, E ¼ environmental warming treatment, Ew ¼ warmed. ANOVA table
M S E M:S M:E S:E M:S:E Residuals
Df
Sum sq
Mean sq
F value
Pr(>F)
1 1 1 1 1 1 1 32
11.0975 0.0022 0.2561 0.1847 0.1634 0.0432 0.2138 0.8428
11.097 0.0022 0.2561 0.1847 0.1634 0.0432 0.2138 0.0263
421.3570 0.0843 9.7256 7.0116 6.2034 1.6387 8.1185
<2.2e 16*** 0.773437 0.003829** 0.012466* 0.018131* 0.209700 0.007601**
Coefficients
(Intercept) M S Ew M:S M:Ew S:Ew M:S:Ew
Estimate
Std. error
t value
Pr(>|t|)
0.53928 0.91835 0.13457 10.74478 0.02363 1.44316 1.03664 0.14428
1.76310 0.2545 0.24425 2.88291 0.03724 0.39465 0.35082 0.05064
0.306 3.608 0.551 3.727 0.635 3.657 2.955 2.849
0.761685 0.001039** 0.585502 0.000749*** 0.530111 0.000909*** 0.005827** 0.007601**
contained less than 0.1 mg per ml (Fig. 13.7a). While the positive relationship between biomass and predicted metabolism persisted among the communities with less than 1 mg per ml, there were also communities with very similar total biomass but different predicted metabolism (Fig. 13.7b). That two communities with equal total biomass have very different levels of metabolism must result from differences in the size distributions between those communities. Higher metabolism for a given biomass results if a community contains mostly small individuals in contrast to communities that contain mostly large individuals. Across all of the communities, total biomass and predicted metabolism explained very similar amounts of variation in observed metabolism (measured by respirometry) (Figs. 13.7c & 13.e). They both explained about 45% of the variation in observed metabolism and were both significant at p ¼ 0.01. Analyzing on the communities with less than 1 mg per ml revealed different results. Here, total community biomass had very little explanatory power
CONSEQUENCES OF BODY SIZE IN MODEL MICROBIAL ECOSYSTEMS
Table 13.6 Summary of analysis three of the community consequences of body size on total community biomass. The response variable is total community density. M ¼ body mass, S ¼ species richness. ANOVA table
M S Residuals
Df
Sum sq
Mean sq
F value
Pr(>F)
1 1 69
3.2283 2.8076 9.9136
3.2283 2.8076 0.1437
22.469 19.541
1.109e 05*** 3.575e 05***
Coefficients
(Intercept) M S
Estimate
Std. error
t value
Pr(>|t|)
1.62628 0.37154 0.12205
0.39459 0.07407 0.02761
4.121 5.016 4.421
0.000103*** 3.93e 06*** 3.58e 05***
(Fig. 13.7d, r2 ¼ 0.08, p ¼ 0.38) while predicted community metabolism explained 35% of the variance in observed community metabolism and was marginally significant (p ¼ 0.04).
Discussion Population consequences of body size – discussion Brown and Gillooly (2003) hypothesize that the slope of the relationship between body size and population abundance will be shallower when measured across species that share a common resource pool (e.g. with an exponent of 0.75) and steeper when species acquire resources from different sources, in particular, when species occupy multiple trophic levels and therefore forage on each other (e.g. with an exponent of 1). The inefficiency of energy transfer from a lower to higher trophic level means that less energy is available to species at the higher level, and consequently their abundance should be lower relative to their biomass. We found some support for this hypothesis, but there was also evidence that was inconsistent with this idea. Analyses of communities that contained multiple trophic levels (analysis one and two) revealed much steeper mass–density relationships than in the communities that contained only species that consumed bacteria (analysis three). This is consistent with the hypothesis. In contrast, analyses of mass–density relationships within communities (i.e. within analysis one and two), did not consistently show that species at higher trophic levels (classed as omnivores here) had lower intercepts (and the same slope) as species at lower trophic levels (the bacterivores). We are uncertain about what
259
O. L. PETCHEY ET AL.
(b) 4
Predicted CO2 flux
Predicted CO2 flux
(a)
3 2 1 0
1.5 1.0 0.5 0.0
0.0
0.4
0.8
1.2
0.00
Community biomass (d)
r2 = 0.48; p = 0.004
4.0
3.0
2.0 0.0
0.4
0.8
3.0
2.5
2.0
1.2
0.0
(f)
Observed CO2 flux
4.0
3.0
2.0 1
2
3
Predicted CO2 flux
0.04
0.08
Community biomass
r2 = 0.43; p = 0.008
0
0.08
r2 = 0.08; p = 0.38
Community biomass (e)
0.04
Community biomass
Observed CO2 flux
Observed CO2 flux
(c)
Observed CO2 flux
260
4
r2 = 0.35; p = 0.04 3.0
2.5
2.0
0.0
0.5
1.0
1.5
Predicted CO2 flux
Figure 13.7 Relationships between total community biomass, total community metabolism and predicted community metabolism (measured as CO2 flux). The left column of graphs contains all available data, the right column excludes the three communities with very high biomass that are obvious in (a). The explanatory power (r2) and statistical significance of linear regression is given for the lower four graphs only.
CONSEQUENCES OF BODY SIZE IN MODEL MICROBIAL ECOSYSTEMS
caused this difference in results between the two levels of comparison (between communities versus within communities). It is possible that classification of the species as bacterivores and omnivores is too simplistic and that accounting for more fine-scale differences in their trophic strategies would change the results of within-community level analyses. It is also possible that some of our findings can be explained by processes that are not included in classic allometric theory, such as the effect body size has on the diets of consumers, and therefore their potential sources of energy (Marquet, Navarette & Castilla, 1995; Carbone & Gittleman, 2002; Loeuille & Loreau, 2006). Theory predicts that equilibrium abundances will be lower at higher temperatures (Allen, Brown & Gillooly, 2002; Brown, 2004). Our results are not consistent with theory, because there are no significant effects of environmental temperature in analysis two. Though the analysis of covariance should be interpreted with caution, because of incomplete overlap of explanatory categories on the covariate axis, there was no indication that population density (given body mass) was lowered by a warming event of greater than 10 8C. The explanation for this difference between theoretical expectation and experimental result seems to be that the theory assumes no change in resource supply rate accompanies a change in temperature (Brown & Gillooly, 2003). In our experiments the resources are bacteria, which decompose the organic matter present in the ecosystem. It seems likely that increasing temperatures will increase rates of decomposition, bacterial production and resource supply rates to the protist species. If the scaling of bacterial production with temperature is the same as the scaling of consumer metabolism with temperature, there should be no decrease in equilibrium abundance with increases in temperature. A factorial experimental manipulation of temperature and resource supply rates could separate the effects of both on the population level consequences of body size, but this experiment remains to be done. Community consequences of body size – discussion Links between biodiversity and ecosystem functioning have been alternatively interpreted as consequences of traits of dominant species or diversity per se. Some aspects of functioning (such as productivity or total community biomass) may increase with diversity because more diverse communities may by chance contain larger species, and not because of mechanisms associated with diversity. However, allometric theory predicts that communities receiving the same amount of energy will eventually support the same total amount of biomass, regardless of organism size. Thus, the size of dominant organisms may determine functioning early in community development, but with enough time thinning or increases in density may eliminate size-specific influences on functioning. These hypotheses were supported in a recently published experimental manipulation of body-size distributions in microbial microcosms (Long &
261
262
O. L. PETCHEY ET AL.
Morin, 2005). Early in the experiment, total community biomass was lower in communities containing species with small individuals than in communities containing larger individuals. However, after 25–50 generations, there were no detectable differences between the biomass of the two types of community. The results of the analyses presented suggest that certain ecological differences among communities and experiments might affect the relationship between individual size and density, so that the total biomass of all communities is not equal, even if energy supply to the communities is. The first two analyses showed much greater levels of biomass equivalence than the third, in which communities with larger individuals tended to have greater total biomass. One finding that is recovered in our analyses is that communities with greater numbers of species tend to have greater total biomass than those with fewer species (McGrady-Steed & Morin, 2000; Petchey et al., 2002; Hooper et al., 2005). Data in Fig. 13.7 also clearly show that factors other than energy supply must influence total community biomass, because there are three communities with nearly an order of magnitude greater biomass than others. Ecosystem consequences of body size – discussion For a subset of the communities in one of the experiments (Experiment 3) there was evidence that the size distribution of individuals influenced community metabolism, as measured by the amount of CO2 produced by a community during a fixed time. Communities that contained many small individuals tended to have higher metabolism than communities with fewer individuals. It is perhaps surprising that any significant relationships between biomass and community metabolism, and predicted and observed metabolism, were detected considering that a major taxonomic component of the microcosms was completely ignored (the bacteria). Their small size might suggest a very large contribution to community metabolism and future analyses would benefit from taking their size and abundance into account. The evidence was equivocal, however, and further studies are needed to test the robustness and generality of the importance of size distributions for community metabolism. In general, the analyses presented above should be considered preliminary in nature and can perhaps be best used to make recommendations for future experiments. These could include predictions based on the size distribution of individuals within species, as opposed to the assumption made here that all individuals of the same species are the same mass. Furthermore, experiments could manipulate the size distribution of species (and thereby individuals) present in different communities (e.g. Long & Morin, 2005). Here, communities were assembled mostly from species with small sized individuals, or mostly from species with large sized individuals. We did not analyze Long and Morin’s (2005) data in this context because their communities contained photoautotrophs and including their contribution to
CONSEQUENCES OF BODY SIZE IN MODEL MICROBIAL ECOSYSTEMS
community metabolism would have required additional model complexity. This could be included in future models that predict ecosystem process rates from the size and abundance of organisms at multiple trophic levels. Closing remarks This volume contains many valuable contributions towards the understanding of the ecological consequences of body size that use a range of methodologies, including observations of natural systems, experiments with natural systems, theoretical modelling and experiments with artificial ecosystems. We present examples of the last approach, using new analyses of data from our published experiments. These analyses have demonstrated effects of body size at three levels of organisation, and how these effects can be altered by community properties such as trophic complexity and species richness. We have illustrated the utility of experimental microbial systems for evaluating allometric theories of community structure and metabolism, and shown that patterns based on observations across many natural systems readily emerge in experimental systems where attributes such as organism size and trophic complexity are subject to rigorous experimental control. These results also point to an important role for biodiversity in influencing community biomass and metabolism that remains incompletely integrated into allometric theories of community and ecosystem properties. Discrepancies between the predictions of allometric theory and our findings could be due to any of many properties of real organisms, for example, evolution can produce a mass–density relationship (Loeuille & Loreau, 2006). Integration of biodiversity, trophic structure and evolution remains an important challenge for ecologists working at the interface of population, community and ecosystem ecology.
Acknowledgements Jill McGrady-Steed kindly made the data from her experiment available in electronic format. Those data and others were originally collected with the support of the US NSF under grant 9806427 to PJM and Timothy Casey. Discussion with participants of the British Ecological Society Special Symposium on Body Size and the Organisation and Function of Aquatic Ecosystems, and the comments of two anonymous reviewers improved the content of this manuscript. OLP is a Royal Society University Research Fellow.
References Allen, A. P., Brown, J. H., & Gillooly, J. F. (2002). Global biodiversity, biochemical kinetics, and the energy-equivalence rule. Science, 297, 2545–1548. Brown, J. H. (2004). Toward a metabolic theory of ecology. Ecology, 85, 1771–1789.
Brown, J. H. & Gillooly, J. F. (2003). Ecological food webs: high-quality data facilitate theoretical unification. Proceedings of the National Academy of Sciences of the United States of America, 100, 1467–1468.
263
264
O. L. PETCHEY ET AL.
Carbone, C. & Gittleman, J. L. (2002). A common rule for the scaling of carnivore density. Science, 295, 2273–2276. Cohen, J. E., Jonsson, T. & Carpenter, S. R. (2003). Ecological community description using the food web, species abundance, and body size. Proceedings of the National Academy of Sciences of the United States of America, 100, 1781–1786. Crawley, M. J. (2002). Statistical Computing. An Introduction to Data Analysis using S-Plus. Chichester: John Wiley & Sons, Ltd. Cyr, H. (2000). Individual energy use and the allometry of population density. In Scaling in Biology, ed. J. H. Brown and G. B. West. Oxford: Oxford University Press, pp. 267–295. Damuth, J. (1981). Population density and body size in mammals. Nature, 290, 699–700. Damuth, J. (1987). Interspecific allometry of population density in mammals and other animals: the independence of body mass and population energy use. Biological Journal of the Linnean Society, 31, 193–246. Duffy, J. E. (2002). Biodiversity and ecosystem function: the consumer connection. Oikos, 99, 201–219. Enquist, B. J. & Niklas, K. J. (2001). Invariant scaling relations across tree-dominated communities. Nature, 410, 655–660. Gillooly, J. F., Brown, J. H., West, G. B., Savage, V. M. & Charnov, E. L. (2001). Effects of size and temperature on metabolic rate. Science, 293, 2248–2251. Hector, A., Schmid, B., Beierkuhnlein, C. et al. (1999). Plant diversity and productivity experiments in European grassland. Science, 286, 1123–1127. Hooper, D. U., Chapin, F. S., Ewel, J. J. et al. (2005). Effects of biodiversity on ecosystem functioning: a consensus of current knowledge. Ecological Monographs, 75, 3–35. Hutchinson, G. E. (1961). The paradox of the plankton. The American Naturalist, 95, 137–145.
Kerr, S. R. & Dickie, L. M. (2001). The Biomass Spectrum. NewYork: Columbia University Press. Loeuille, N. & Loreau, M. (2006). Evolution of body size in food webs: does the energy equivalence rule hold? Ecology Letters, 9, 171–178. Long, Z. T. & Morin, P. J. (2005). Effects of organism size and community composition on ecosystem functioning. Ecology Letters, 8, 1271–1282. Loreau, M., Naeem, S., Inchausti, P. et al. (2001). Biodiversity and ecosystem functioning: current knowledge and future challenges. Science, 294, 804–808. Marquet, P. A., Navarette, S. A. & Castilla, J. C. (1995). Body size, population density, and the energetic equivalence rule. Journal of Animal Ecology, 64, 325–332. McGrady-Steed, J. & Morin, P. J. (2000). Biodiversity, density compensation, and the dynamics of populations and functional groups. Ecology, 81, 361–373. McGrady-Steed, J., Harris, P. M. & Morin, P. J. (1997). Biodiversity regulates ecosystem predictability. Nature, 390, 162–165. Mulder, C., Cohen, J. E., Seta¨la¨, H., Bloem, J. & Breure, A. M. (2005). Bacterial traits, organism mass, and numerical abundance in the detrital soil food web of Dutch agricultural grasslands. Ecology Letters, 8, 80–90. Petchey, O. L., McPhearson, P. T., Casey, T. M. & Morin, P. J. (1999). Environmental warming alters food-web structure and ecosystem function. Nature, 402, 69–72. Petchey, O. L., Casey, T. J., Jiang, L., McPhearson, P. T. & Price, J. (2002). Species richness, environmental fluctuations, and temporal change in total community biomass. Oikos, 99, 231–240. Petchey, O. L., Downing, A. L., Mittelbach, G. G. et al. (2004). Species loss and the structure and functioning of multitrophic aquatic ecosystems. Oikos, 104, 467–478.
CONSEQUENCES OF BODY SIZE IN MODEL MICROBIAL ECOSYSTEMS
Peters, R. H. (1983). The Ecological Implications of Body Size. Cambridge: Cambridge University Press. Savage, V. M., Gillooly, J. F., Brown, J. H., West, G. B. & Charnov, E. L. (2004). Effects of body size and temperature on population growth. American Naturalist, 163, E429–E441. Sheldon, R. W. & Kerr, S. R. (1972). The population density of monsters in Loch Ness. Limnology and Oceanography, 17, 796–798. Sheldon, R. W. & Kerr, S. R. (1973). The Loch Ness Monster: reply to comments of C. H. Mortimer. Limnology and Oceanography, 18, 345–346. Sheldon, R. W., Prakash, A. & Sutcliffe, W. H. (1972). The size distribution of particles in the ocean. Limnology and Oceanography, 17, 327–340. Sheldon, R. W., Sutcliffe, W. H. & Paranjape, M. A. (1977). Structure of pelagic food-chain
and relationship between plankton and fish production. Journal of the Fisheries Research Board of Canada, 34, 2344–2353. Tilman, D., Reich, P. B., Knops, J. et al. (2001). Diversity and productivity in a long-term grassland experiment. Science, 294, 843–845. Weatherby, A. J., Warren, P. H. & Law, R. (1998). Coexistence and collapse: an experimental investigation of the persistent communities of a protist species pool. Journal of Animal Ecology, 67, 554–566. West, G. B., Brown, J. H. & Enquist, B. J. (1997). A general model for the origin of allometric scaling laws in biology. Science, 276, 122–126. Wetzel, R. G. & Likens, G. E. (1991). Limnological Analyses, 2nd edn. New York: Springer-Verlag.
265
CHAPTER FOURTEEN
Body size, exploitation and conservation of marine organisms SIMON JENNINGS Centre for Environment, Fisheries and Aquaculture Science (CEFAS)
JOHN D . REYNOLDS Simon Fraser University, Burnaby
Introduction Aquatic ecologists and conservationists have long been obsessed with trying to understand links between body size, exploitation and conservation (e.g. Adams, 1980; Dickie, Kerr & Schwinghamer, 1987a). There are several reasons for this interest. First, at the individual level, fisheries management for both vertebrate and invertebrate populations tries to minimize the mortality of smaller individuals, in order to increase the probability that individuals have reproduced before they are caught (Jennings, Kaiser & Reynolds, 2001a). Even if other management methods are used, practically every stock assessment that has ever been done on an indeterminately growing species has included size as a key input parameter. Second, at the population level, large-bodied species have life-history traits that lead to slow rates of population turnover, with clear implications for productivity, resilience and recovery potential (Hutchings, 2001; Denney, Jennings & Reynolds, 2002; Reynolds et al., 2005; Goodwin et al., 2006). Third, at the community level, predator–prey relationships are strongly linked to size (Cohen et al., 1993; Woodward & Warren, this volume; Persson & De Roos, this volume), leading to the potential for understanding how fishing mortality may have wider ecosystem impacts through food-web dynamics (Dickie et al., 1987a). At the population level, studies of the effects of exploitation have long focused on the direct effects of fishing mortality on a single stock, and this is often a pragmatic response to limited information on the indirect effects of mortality on one species affecting other species in the ecosystem (Hilborn & Walters, 1992). However, renewed concerns about the environmental impacts of fishing have encouraged new research on fishing effects on communities and ecosystems, where an understanding of the direct and indirect effects of mortality, and the capacity to partition them, is essential. Since the structure of aquatic food webs is often strongly size based, we hope to show that understanding interactions between body size and exploitation provides the basis for describing and managing the effects of fishing. Body Size: The Structure and Function of Aquatic Ecosystems, eds. Alan G. Hildrew, David G. Raffaelli and Ronni Edmonds-Brown. Published by Cambridge University Press. # British Ecological Society 2007.
BODY SIZE AND EXPLOITATION
In this chapter we consider linkages between body size, exploitation and aquatic conservation at all three scales: individuals, populations and communities. Our aim is to understand, from first principles of metabolic scaling rules and life histories, how population and community dynamics are linked to the sizes of individuals. Although much of the research that has been done in this field has focused on fish as the main animals caught, a nice feature of working with body-size distributions at the community level is their continuity from primary producers to top predators, which often transcend taxonomic boundaries (e.g. Warwick, this volume).
Setting: the magnitude of fishing effects Some brief highlights of the state of aquatic biodiversity can make for depressing reading. The Food and Agriculture Organization of the United Nations (FAO) (2005) report on the state of the world’s fisheries points out that 52% of the world’s stocks are fully exploited, 17% are over exploited, 7% depleted and 1% recovering. The proportion of stocks that are over-exploited, depleted and recovering has increased from 10% in the mid 1970s to 25% in the early 1990s. World marine fisheries catches levelled out in the late 1980s, with fluctuations since then driven by changes in some highly productive areas in the Pacific (Watson & Pauly, 2001; FAO, 2005). This limit to marine catches has occurred in spite of a continuing increase in fishing effort and efficiency. When we break these global figures down to the level of specific populations, we can identify many reductions in abundance over the past two to three decades. A review of data from 232 exploited fish populations revealed a median maximum rate of decline of 83% from known historical levels (Hutchings & Reynolds, 2004). Over half (58%) the populations declined by more than 80%. Note that these are maximum declines, based on time series of at least ten years, and the timeseries usually began well after the onset of fishing. The declines are also much greater than those associated with taking a large sustainable yield from these populations. Even if fishing mortality can be substantially reduced, some stocks, in particular those of bottom dwelling fishes, do not show recovery if they have been pushed to very low abundance (Hutchings, 2001, 2002; Hutchings & Baum, 2005). There is evidence for extinction of some populations of fishes and other marine species. Dulvy, Sadovy and Reynolds (2003) compiled a list of 133 cases of local, regional or global extinctions of marine species. Of these, 55 cases involved fishes, while the rest included birds, mammals and invertebrates. This is a preliminary assessment: we still know very little about the status of the vast majority of aquatic organisms. Indeed, 80% of the extinctions were discovered through historical comparisons rather than real-time detections, with a median 53-year lag between disappearance and the reporting of
267
268
S. JENNINGS AND J. D. REYNOLDS
that disappearance. Thus, the near-extinction of the common skate, Dipturus batis, from the Irish Sea was brought to the world’s attention only several decades after the decline had occurred (Brander, 1981). As we scale up from populations to communities, the principal effects of fishing on size and species composition are well known. Both spatial comparisons between areas subject to different fishing intensities and temporal comparison within areas where fishing effort has increased over time, indicate broadly predictable effects. As fishing mortality rises, the mean size of individuals in the community falls, and species with larger body sizes form a smaller proportion of community biomass. Since larger species and individuals often, but not exclusively, feed at higher trophic levels, so changes in size structure are often paralleled by a reduction in the trophic level of the community – an effect dubbed ‘fishing down the food web’ by Pauly et al. (1998). While much of the theory and data on fishing effects refers to target species, we have also learned a great deal in the past 20 years about impacts of fishing gear on the seabed (Kaiser & de Groot, 2000). Impacts on bottom fauna are most severe in habitats that do not naturally receive much disturbance. As we will see below, we can use the same body of theory developed for linking body size, life histories and responses to mortality for predicting the responses of benthic invertebrate species and communities to trawling disturbance.
Size-related responses to exploitation The responses of populations and communities to human activities depend on interactions between the extrinsic mortality rate and the intrinsic aspects of the species’ biology that affect population growth rates. Body size plays a key role in both elements of vulnerability. Fisheries are selective, typically targeting the large-bodied, high-value individuals and species that are favoured for consumption and sale. Even when fisheries do not deliberately pursue larger individuals and species, there will still be higher mortality for such animals because they are often caught as by-catch due to their greater likelihood of being retained by gears such as trawls. While there are a few notable exceptions to the rule of larger animals being more vulnerable to capture, such as gill nets and traps that select individuals and species of intermediate size, almost all population-based assessments show that large size classes suffer higher mortality. There is clear evidence for the size-relatedness of fishing effects in populations and communities. Thus, in one early study, we compared trends in the abundance of populations of the same species that had different maximum sizes, after accounting for differences in fishing mortality. Populations with larger maximum sizes consistently showed greater rates of decline in response
BODY SIZE AND EXPLOITATION
to mortality (Jennings, Reynolds & Mills, 1998). In a second study, using a phylogenetic comparative approach to account for relatedness, we compared trends in pairs of species of different body size. In eight out of nine comparisons, the larger species declined more in abundance during a period of increasing fishing mortality (Jennings, Greenstreet & Reynolds, 1999). There are many other recent examples of trends in abundance that can be related to body size (Reynolds, 2003). The overall trend of targeting large species has had the same impacts on large, profitable prey on land as in the water (Reynolds & Peres, 2006). Thus, we have seen the sequential loss of species from the largest to the smallest, in groups as diverse as whales in the Southern Ocean, abalones along the coasts of western North America, and primates in the Brazilian Amazon. The effects of fishing on communities are a consequence of interactions between the direct and indirect effects of fishing. The direct effects are due to mortality on component populations and tend to result in a reduction in the mean size of animals in the community. The indirect effects are due to the changes in predator–prey relationships which occur when predator and/or prey abundance changes. Direct and indirect effects usually result in a reduction in the mean body size of animals in the community and a decrease in the proportion of animals with larger body size (Shin et al., 2005). Fishing effects on communities have often been reported in terms of changes to the slope of abundance (y) vs. body mass (x) relationships, otherwise known as size spectra. The slopes of size spectra tend to become steeper with increased fishing (Bianchi et al., 2000; Duplisea & Kerr, 1995; Gislason, 2002; Pope et al., 1988; Rice & Gislason, 1996). Intercepts were often reported to increase as the slope declined, but this could have been a real effect or a statistical artefact resulting from the correlation between intercept and slope. To avoid this, Daan et al. (2005) re-scaled size spectra to the midpoint size class and expressed midpoint heights rather than intercepts. Results suggested that changes in the spectrum can be driven both by the loss of large fish and proliferation of small fish as their larger predators are depleted (Daan et al., 2005). The physical impacts of fishing gears also cause size-selective mortality. When trawl gears are towed across the seabed they may kill differentially many larger and more fragile animals, as small ones may be pushed aside by the pressure wave in front of the gear (Gilkinson et al., 1998). The larger species are also less able to withstand a given rate of mortality. As a result, benthic communities in trawled areas comprise smaller individuals and species (Kaiser et al., 2000), and the slopes of benthic invertebrate size spectra become steeper in more heavily trawled areas (Jennings et al., 2002; Warwick, this volume).
Linking body size, life histories and population dynamics Rates of metabolism, the biological processing of energy and materials, are systematically related to body size and ultimately correlate with the life histories
269
270
S. JENNINGS AND J. D. REYNOLDS
of animals and their responses to exploitation. Relationships between metabolism and size are further modified by chemical composition and temperature (Brown et al., 2004). These factors consequently drive most other biological rates and timings, such as lifespans and growth (Gillooly, 2000; Gillooly et al., 2001, 2002; Brown, Allen & Gillooly, this volume; Atkinson & Hirst, this volume). The combined effect of body size and temperature on individual wholeorganism metabolic rate has been approximated as: I ¼ i0 M0:75 eE=kT
(14:1)
Where I is individual metabolic rate, i0 is a normalization constant, M is body mass, T is temperature (Kelvin), E is the activation energy of metabolism and k is Boltzmann’s constant. Since the mass-specific rate of metabolism (R) is 1/M, R will scale with M as R / M0:25 eE=kT
(14:2)
This demonstrates that large organisms must require more resources and flux them more slowly than smaller ones (Brown et al., 2004). Not surprisingly, intrinsic rates of increase and production scale similarly, and turnover time as the inverse of production; approximately W0.25. Changes in resource requirements with size will limit the resources available for allocation to life histories, and species have evolved many ways of allocating these limited resources to maximize reproductive output. In a former British Ecological Society symposium, Law (1979) described the Darwinian demon, an organism in which all the problems of maximizing reproductive output had been solved. This animal began reproducing immediately after birth, producing large numbers at frequent intervals as it got older. It experienced no mortality and its capacity for dispersal and finding mates knew no bounds. Of course, as Law pointed out, no such animal exists because of trade-offs. Thus, species have followed a wide diversity of paths through these trade-offs in arriving at the combinations of life-history traits that we see today and which maximize individual fitness (Atkinson & Hirst, this volume). Relationships among traits can be described with invariants, which reflect general life-history patterns among species after removing the dimensions of mass and time (Charnov, 1993). Examples of invariants are relationships between size at maturity and maximum size, lifespan and age at maturity, natural mortality and growth rate (Beverton, 1992). Two aspects of life histories, which result from trade-offs, are critical in determining the response of a population to additional mortality: the intrinsic rate of increase and the strength of compensation. The intrinsic rate of natural increase is often denoted rmax, to make it clear that this is the maximum rate that populations could achieve in the absence of density dependence, which we can usually expect to apply to small populations that are far from their carrying
BODY SIZE AND EXPLOITATION
capacity. There is a strong theoretical literature that links body size to rmax, through links with key components of rmax such as age at maturity and natural lifespan (Roff, 1992). The strength of compensation, or density dependence, indicates how the production of recruits (juvenile fish at a specified age) per spawner changes with the abundance of spawners. Since the abundance of recruits can be measured at different ages for different populations, estimates of population-specific mortality in subsequent years can be used to express recruits per spawner in a consistent way for comparative purposes: as spawners per spawner (Myers, Mertz & Fowlow, 1997).
Population dynamics and responses to mortality Intrinsic rates of increase Species with high rmax will have fast population turnover, and should therefore be better able to compensate for increased mortality due to fishing. This has been suggested by a variety of theoretical analyses, which have focused on body size as well as its correlated life-history traits, such as age at maturity and natural lifespan (Adams, 1980; Stokes, McGlade & Law, 1993). The prediction that large-bodied species should decline more quickly when exposed to a given rate of fishing mortality has been upheld in many studies of marine fishes in both tropical and temperate waters (reviewed by Reynolds, 2003). Many of these studies have included comparisons between closely related species, to avoid ‘apples and oranges’ comparisons between species that differ greatly in other aspects of behaviour or life histories (Harvey & Pagel, 1991). The prediction that large-bodied species should have lower recovery potential has also been upheld in comparative studies of marine invertebrates (Fenchel, 1974) and fishes (Denney et al., 2002). For example, Denney et al. (2002) measured the slope at the origin of recruitment plotted against adult stock size for fish populations in the northeast Atlantic. This provided a practical metric for rmax, the number of spawners produced per spawner in the absence of density dependence (Myers et al., 1997), which could be plotted against various life-history traits. Body size was negatively related to rmax, the implication being that, all else being equal, small-bodied species should be able to bounce back from small population size faster than larger ones. When the length-based measures of body size reported by Denney et al. (2002) are converted to mass (Maxwell & Jennings, 2005), it is notable that the slope of the relationship between log10 spawner per spawner and log10 maximum body mass does not significantly differ from the M0.25 scaling of intrinsic rate and body mass predicted from theory (Eq. (14.2); Savage et al., 2004). Density dependence The strength of density dependence is critical for determining the ability of populations to compensate for increased mortality. Yet, by definition, rmax and
271
S. JENNINGS AND J. D. REYNOLDS
the surrogates used to measure it, such as the slope at the origin of a spawnerrecruitment relationship, apply to small populations in the absence of density dependence. For example, if the population follows a logistic growth pattern, the rate of change in number of individuals per unit time, dN/dt, will depend on population growth rate r and the carrying capacity, K. dN N ¼ rN 1 (14:3) dt K Only at the smallest population size will r ¼ rmax. Cynics might argue that this is the case for the majority of fisheries today, so we need go no further! Even in heavily fished populations, however, we cannot completely ignore densitydependent interactions between individuals, predator–prey relationships and other aspects of ecology and behaviour that may determine intrinsic aspects of vulnerability. These features of biology translate the conceptually simple rmax into realized population growth rates. Recently, Goodwin et al. (2006) have examined relationships between density dependence and life history. They used stock-recruitment curves to measure the compensation ratio of a population (Fig. 14.1). The compensation ratio (CR; Goodyear, 1977) was defined as: CR ¼
Maximum recruit survivalðÞ Recruit survival at SSBF¼0
(14:4)
Where a is the slope at the origin of a spawner–recruit relationship and SSBF¼0 is the spawning stock biomass at equilibrium in the absence of fishing. Goodwin et al. (2006) demonstrated that large-bodied species with low maximum rates of increase and long generation times tended to have strong compensation. Their results help us to understand how populations respond to fishing. Those populations with small body size and low spawners per recruit in the absence of fishing (SPRF¼0) have weak density-dependence and high annual production. At the outset α SF=0 Recruit abundance (R)
272
Spawner abundance (S)
Figure 14.1 A spawner–recruit relationship where is the maximum recruits per spawner at low population size and SF¼0 is equilibrium spawner abundance in the absence of fishing. The relationship between maximum recruits per spawner and recruit survival at SF¼0 defines the compensation ratio. After Goodwin et al. (2006).
BODY SIZE AND EXPLOITATION
of fishing even low fishing mortality could rapidly deplete these populations since they would show little compensatory response. However, they would show high resilience to extinction at very low abundance. Conversely, populations with strong density-dependence have a greater initial compensatory reserve, but can collapse owing to low resilience once they reach low abundance.
Communities and their responses to mortality Size-based structuring of marine communities To understand and predict how fishing affects communities, it is necessary to know how communities are structured in the absence of fishing. Strong sizebased structuring is ubiquitous in marine ecosystems (Sheldon, Prakash & Sutcliffe, 1972) and trophic level increases more or less continuously with body size (Fry & Quinones, 1994). Why food webs of this type have evolved in aquatic environments when many terrestrial webs have clear trophic levels is not clear. Loeuille and Loreau (2005) recently explored one possibility, based on a model where food webs evolve from a single ancestor and assuming that adaptation acts on body size, which has a well-established impact on metabolism and interactions between organisms. Based on parameters defined at the organism scale, the model predicts emergent properties at the food-web scale. If niche width and competition intensity are small then distinct trophic levels evolve. When niche width or competition intensity are large then clear trophic levels do not emerge, consistent with patterns seen in many aquatic food webs. One feature of many marine species is their large scope for individual growth. Thus, individuals of most species begin life as larvae feeding at the base of food chains, but can end life as large terminal predators. Size-based predation and the large scope for growth in aquatic animals are significant because, at least over the course of a species’ life cycle, body size will be a better indicator of trophic level than species identity (see Woodward & Warren, this volume). This observation provides a compelling reason to adopt size rather than species-based analysis of aquatic food webs; treating small individuals of a large species as functionally equivalent to large individuals of small species (Kerr & Dickie, 2001). While such size-based analyses are necessarily simplifications, they provide an excellent means of understanding the development of size structure and assessing the effects of mortality (Jennings, 2005; Persson & De Roos, this volume). Attempts to model the processes that lead to the emergence of size spectra have included simple models based on fundamental ecological principles (Kerr & Dickie, 2001; Brown & Gillooly, 2003) and detailed process-based models of predator–prey interactions (Shin & Cury, 2004). Almost all models are underpinned by the recognition that the scaling of metabolism with body size determines the energy requirements of animals in different size classes. The time-averaged slopes of abundance–body mass relationships in size-based food webs are principally determined by the availability of energy to animals in
273
274
S. JENNINGS AND J. D. REYNOLDS
each size class (Kerr & Dickie, 2001). Availability of energy at size, and hence the scaling of abundance and body mass, can be predicted by extending the theory of energetic equivalence that applies to communities that share energy. For these communities, such as phytoplankton using sunlight, numerical abundance (N) typically scales with body mass (M) as M0.75. Since the scaling of individual metabolic rate and body size can be approximated as Mþ0.75 (Eq. (14.1)), the rate of energy use is expected to be independent of body size (Damuth, 1981), a prediction that has since been referred to as energetic equivalence (Nee et al., 1991). In the case of phytoplankton communities, for example, the sunlight energy used by plankton cells in a size class is the same as in any other size class (Li & Charnov, 2001). Food chains are not characterized by the sharing of energy, but by larger predators eating smaller prey. Thus the availability of energy falls with size and trophic level (Ware, 2000). The rate at which available energy falls will depend on the efficiency of energy transfer and the number of predator–prey interactions that transfer energy from small to large body-size classes. The latter depends on the ratio of the mean predator size to mean prey size (PPMR), which in most aquatic ecosystems ranges from 100 to 1000:1 by mass (Cushing, 1975). Cyr (2000) and Brown and Gillooly (2003) proposed that knowledge of PPMR and transfer efficiency (TE) could be used to predict the changes in energy available to animals of different body sizes in a complete size spectrum and hence the slope. Their analysis helped to explain the remarkable consistency in the observed slopes of size spectra (Boudreau & Dickie, 1992). This is because PPMR and TE place significant constraints on the slope of abundance–body mass relationships and because PPMR and TE are remarkably consistent in different ecosystems (Jennings, 2005). Brown and Gillooly’s (2003) analysis, as further developed in Brown et al. (2004), was based on a series of trophic levels, each of which extended over a range of body sizes. In most aquatic communities trophic level actually rises continuously with body size (Jennings et al., 2002). Jennings and Mackinson (2003) formalized the approach of Brown and Gillooly (2003) for application to such a community, and showed that the method provided good predictions of the slope of the size spectrum. These methods for predicting the slope of size spectra are helpful in understanding the size structure of communities and for providing a baseline for assessing the relative effects of exploitation. However, they do not allow prediction of the consequences of various levels of exploitation. To achieve this, models that account for the growth and mortality of individuals are required. This is the interface between understanding the dynamics of populations and understanding how population dynamics contribute to community structuring. Describing and predicting responses to mortality It can be difficult to link changes in fishing mortality to changes in the slopes of size spectra empirically, because mortality data for communities are hard to
BODY SIZE AND EXPLOITATION
obtain. However, available evidence suggests that the slopes respond principally to fishing rather than to other drivers such as temperature. Blanchard et al. (2005), for example, considered the effects of fishing and climate variation on size-based metrics, including slope of the size spectrum, in the Celtic Sea. Their analysis suggested that size-based metrics responded clearly to the effects of fishing in variable environments, reflecting the ubiquity of size-based processes in defining community structure and responses to mortality. This is consistent with theory. While temperature will have a marked effect on rates of biomass turnover and energy flux in the spectrum (Eqs. 14.1 & 14.2), TE and PPMR are largely temperature independent. Much work on size spectra has focused on fishing effects on fish communities, but benthic invertebrate communities are also directly impacted by bottom fishing gears such as trawls and show strong size-based responses to fishing. Comparisons among areas subject to different levels of trawling disturbance have shown that the frequency of bottom trawling disturbance in the central North Sea had a greater effect on the size structure of the fauna in a soft-sediment benthic community than other environmental variables such as sediment particle size and depth (Duplisea et al., 2002). Size spectra became steeper and their heights decreased with increased trawling disturbance. As a result, the total production of infaunal invertebrates fell with increasing trawling disturbance while relative production of the infaunal community rose significantly. The increases in relative production were largely attributable to the dominance of smaller animals in the trawled community and did not compensate for the loss of biomass and production of larger animals (Jennings et al., 2001b). Differences between the slopes of size spectra in exploited communities and those that are predicted for unexploited communities can be used to assess the effects of fishing on abundance in different ecosystems. In addition, given wellestablished scaling relationships between body size and biological properties, the size spectrum can be parameterized to estimate the effects of fishing on production (P) or turnover time. To assess the overall effects of fishing on the North Sea fish community, Jennings and Blanchard (2004) attempted to compare a theoretical abundance–body mass relationship for the unexploited North Sea with an abundance–body mass relationship estimated from contemporary data (Fig. 14.2). The slope of the unexploited size spectrum was predicted from PPMR and TE (using the methods introduced above under Size-based structuring of marine communities) since there was little evidence that either parameter was affected by exploitation to the same extent as biomass (B). By comparing the unexploited theoretical and observed size spectra they predicted that the scaling of B with M had changed from M0.10 (unexploited) to M1.0 (exploited). This suggested that the current biomass of large fishes weighing 4–16 kg and 16–66 kg was 97.4% and 99.2%, respectively, lower than would be expected in the absence of fisheries exploitation. The mean
275
S. JENNINGS AND J. D. REYNOLDS
(a) 6
Trophic level
5
4
3 1
2
3
4
5
Body mass (log10)
Trophic level (b)
4.0
4.2
4.4
4.6
4.8
5.0
5.2
1 TE = 0.150
Biomass (log10)
276
TE = 0.125 0
TE = 0.100
–1
–2 1
2
3 Body mass (log10)
4
5
Figure 14.2 The relationship between trophic level, as estimated with nitrogen stable isotope analysis, and body mass for North Sea fishes in 2001 (a); 95% confidence intervals for the mean reflect among-site variation, and the slope of this relationship implies a mean predator–prey mass ratio (PPMR) of 390:1. Predicted slopes of abundance–body mass relationships for the unexploited community (b) were calculated from PPMR and estimates of transfer efficiency (TE) of 0.100–0.150. The observed slope of the abundance–body mass relationship in 2001 is shown with a broken line. From Jennings and Blanchard (2004).
turnover time of the exploited community (1/P:B) was almost twice as fast as that of the unexploited community, falling from 3.5 to 1.9 years. Comparisons such as these are useful for comparing the magnitude of fishing effects in different ecosystems and for assessing the relative impacts of fishing on a range of community and ecosystem properties. However, such static descriptions cannot be used to guide management and to modify fishing to achieve desirable community structures, such as those that provide good yields of target species while ensuring that large and vulnerable species do not go extinct.
BODY SIZE AND EXPLOITATION
Models of the effects of varying fishing mortality on the slope of size spectra have focused on the various processes that lead to changes in slope, specifically: (1) the differential reduction in the abundance of larger species that are more vulnerable to a given rate of mortality, (2) within-population changes in mean body size and life history due to the direct effects of fishing on the population, (3) genetic changes in life history and (4) predator–prey relationships in the community that result in proliferation of small species and individuals that are better able to tolerate a given rate of fishing mortality and benefit from depletion of their predators. Responses (1) and (2) are governed by the links between body size, life histories and response to mortality that we considered above. Existing theoretical models that describe the structure of size spectra and their response to fishing have principally focused on (1), (2) and (4) (e.g. Boudreau et al., 1991; Dickie, Kerr & Boudreau, 1987b; Kerr, 1974; Thiebaux & Dickie, 1993). Response (3) is arguably as important as the others in the medium- to long-term and there is a developing body of theory and empirical observation at the population level that would allow genetic effects to be addressed (Law, 2000; Law & Stokes, 2000; Heino & Godo, 2002). Some recently published modelling results suggest that the effects may not be intuitive due to the interplay between the selection caused by fishing and changes in predation pressure on different size classes (Thygesen et al., 2005; Persson & De Roos, this volume). In models of fishing effects on community size structure, the capacity to partition the direct and indirect effects of fishing and to understand how they interact is critical in providing insight into the effects of fishing. Gislason and Rice (1998) used a combination of existing length-based fishery models and multispecies models to predict the slopes and intercepts of size spectra for a community consisting of 11 fish species that account for much of the fisheries production in the North Sea. Fishing mortality increased the slope of the size spectrum. Building on this approach, Pope et al. (2006) generalized the model for a community of species defined by their maximum body sizes (asymptotic length L1 as defined in the von Bertalanffy growth model, in 10 cm length classes from 10 cm to 130 cm). We discuss this model in a little detail, as it gives important insights into the links between size-based population and community responses to fishing and could provide a basis for predicting and managing the impacts of fishing on size-based communities. For each species (S), the number (N) surviving from one length class (L) to the next, where successive length classes are denoted 1 and 2 was given as: NL2;S ¼ NL1;S
L1 ðSÞ L2 L1 ðSÞ L1
ZðL1;SÞ kðSÞ (14:5)
based on the fisheries assessment method of length cohort analysis (Jones, 1974), where Z(L1,S) is the total mortality rate (y1) at length L1 for the species and k is the Brody growth parameter from the von Bertalanffy growth model
277
278
S. JENNINGS AND J. D. REYNOLDS
that reflects the rate of growth towards asymptotic length (L1). One limitation of the application of this approach was that the assumption of a fixed relationship between L1 and growth rate, means that the modelled rates of growth persist regardless of the level of available resources. The average number of fish L1;S ) in the L1 length group by species is given by: per year (N L1;S ¼ ðNL1;S NL2;S Þ=ZðL1; SÞ N
(14:6)
L1;S may then be summed across species at each length group L from 10 to The N L;# : Consistent with standard practice 130 cm to give the overall size spectrum N in multispecies fishery models, the components of Z considered were fishing mortality rate (F), non-predation natural mortality rate M1 and predation mortality rate M2. Thus: Z ¼ F þ M1 þ M2
(14:7)
Predation mortality was set proportional to the sum of the power of biomass in larger size classes, modified by a size-preference function. Maturity was based on life-history invariants. Once mature, fishes produced recruits in proportion to their total abundance and length. A power term in the spawner–recruit relationship provided compensation, the density dependent reduction in recruitment at high spawner abundance that we discussed previously. The slope of the size spectrum for fishes from 20–100 cm was broadly linear and became increasingly steep as a function of fishing mortality. Density dependent controls, due to predation mortality and the extent of compensation in the spawner–recruit relationship, were key mechanisms in maintaining the slope. When compensation exerted strong control, the model suggested that the role of fishing quickly dominated the effect of predation. The model also provided insights into the relative role of predation mortality and fishing in different size classes, building on the observations of Dickie (1976) who demonstrated that: (1) the ratio of biomass at successive trophic levels must be independent of sizes of individuals and, (2) the ratio of biomass at successive trophic levels is the ecological efficiency (ratio of food intake at two successive trophic levels) corrected by the ratio of predation rates. Thus Dickie (1976) showed that predation rate must be relatively higher at lower trophic levels, and that a given amount of fisheries yield from a lower trophic level is a much smaller proportion of predation mortality rate than at a higher trophic level. The Pope et al. (2006) model suggests that the reduction in the biomass of larger species and individuals in the community leads to the proliferation of smaller individuals. Using a method developed by Daan et al. (2005) to assess changes in the abundance of fishes in different size classes over time, there have now been three empirical demonstrations of increases in the steepness of size
BODY SIZE AND EXPLOITATION
spectra being partially driven by the proliferation of fish in smaller size classes: for the North Sea (Daan et al., 2005), Fijian reefs (Dulvy et al., 2004) and the Celtic Sea (Blanchard et al., 2005) (Fig. 14.3). The model of Pope et al. (2006) suggests that the relationship between the strength of compensation in the spawner–recruit relationship and body size play an important role in controlling the extent of prey release. The Pope et al. (2006) size-based model is an important development because they have captured the dynamics of individual species within the size spectrum,
(a) 0.4
Relative change in biomass
0.2 0 –0.2 –0.4 –0.6 –0.8 –1.0 2
4
6
8
10
12
Body mass class (log2) (b) 0.4
Relative change in biomass
0.2 0 –0.2 –0.4 –0.6 –0.8 –1.0 20
30
40
Body mass class (cm)
50
60
Figure 14.3 Relative changes in biomass as a function of body size across (a) a temporal gradient in multispecies fishing mortality in the Celtic Sea and (b) a spatial gradient in fishing effort on Fijian reefs. Both examples show evidence for prey release following predator depletion. Data from Blanchard et al. (2005) and Dulvy et al. (2004).
279
S. JENNINGS AND J. D. REYNOLDS
1
Spawner biomass (log10)
280
0
10 cm
–1 20 cm
–2
30 cm
–3
50 cm 70 cm
–4 0.0
130 cm
0.5
1.0 1.5 2.0 2.5 Fishing mortality
3.0
Figure 14.4 The effects of fishing mortality on the relative abundance of fishes as a function of their maximum potential size. From Pope et al. (2006).
by associating species identity with a maximum length. This allows the model to be used to predict the proportion of species with different ultimate body sizes that ‘populate’ any size class in the spectrum (Fig. 14.4). In the same way that Cohen, Jonsson and Carpenter (2003) and Cohen (this volume) uniquely linked species identity, body mass and abundance in a new approach to conventional food-web analysis, the Pope et al. (2006) approach attempts to bring species identity into size-based models and provides the basis for starting to investigate links between community processes and the diversity and persistence of populations. Size-based models have also been used to predict fishing effects on benthic invertebrate communities. In benthic communities, the structure of the size spectrum is driven more by competition rather than predation, since the animals present in many size classes often share the same energy sources (Warwick, this volume). Duplisea et al. (2002) developed a competition-based model to assess the effects of size-selective mortality due to trawling disturbance on the size-structure of the benthic community. By applying size-related mortality, they showed that fishing increased the slope of benthic size spectra, consistent with empirical observation.
Conclusions There are clear links between the metabolic processes that drive rates of living and dying, life histories, and the structure and function of populations and communities. Thus body size and body-size distributions, as easily measured attributes of individuals, populations and communities, can provide significant insight into responses to exploitation. It seems fitting that fisheries ecologists have made many contributions to the development and testing of theory that links body size, exploitation and conservation, since this reinvigorates a long but intermittent history of fisheries science contributing to wider ecological thinking (Frank & Leggett, 1994; Shuter & Abrams, 2005). Indeed,
BODY SIZE AND EXPLOITATION
some of the most influential scientists working on the development of fisheries assessment methods were responsible for theoretical developments that proved fundamental in understanding life histories (Beverton & Holt, 1957, 1959). While fisheries scientists have consistently been aware of links between population biology and the ecosystem, the politics and practicalities of management have led to the development of management systems that were dominated by the assessment of single populations. This approach was accepted by most management agencies from the 1960s to the present, but two events in recent years have provided a strong driver for moving toward management systems that also focus on communities and ecosystems. First, there has been near universal commitment within international fora that fisheries must be managed in an ecosystem context to take account of conservation concerns (Sinclair & Valdimarsson, 2003). Second, there have been major advances in the development of food-web theory that may have management applications, and in the development of connected theory that helps to formalize many of the observed relationships between individual, population and community processes (Brown & Gillooly, 2003; Cohen et al., 2003; Brown et al., this volume; Cohen, this volume). The changing drivers for management, and these new developments in ecology, suggest that theoretical development should reinvigorate the development of new management and assessment methods.
Acknowledgements We thank Nick Goodwin for providing the materials to prepare Fig. 14.1 and Julia Blanchard and Nick Dulvy for providing data to prepare Fig. 14.3.
References Adams, P. B. (1980). Life history patterns in marine fishes and their consequences for management. Fishery Bulletin, 78, 1–12. Beverton, R. J. H. (1992). Patterns of reproductive strategy parameters in some marine teleost fishes. Journal of Fish Biology, 41 (Supplement B), 137–160. Beverton, R. J. H. & Holt, S. J. (1957). On the Dynamics of Exploited Fish Populations, Rep. No. 19. London: Ministry of Agriculture, Fisheries and Food. Beverton, R. J. H. & Holt, S. J. (1959). A review of the lifespan and mortality rates of fish in nature and their relationship to growth and other physiological
characteristics. Ciba Foundation Colloquim on Ageing, 5, 142–180. Bianchi, G., Gislason, H., Graham, K. et al. (2000). Impact of fishing on size composition and diversity of demersal fish communities. ICES Journal of Marine Science, 57, 558–571. Blanchard, J. L., Dulvy, N. K., Jennings, S. et al. (2005). Do climate and fishing influence size-based indicators of Celtic Sea fish community structure? ICES Journal of Marine Science, 62, 405–411. Boudreau, P. R. & Dickie, L. M. (1992). Biomass spectra of aquatic ecosystems in relation to fisheries yield. Canadian Journal of Fisheries and Aquatic Sciences, 49, 1528–1538.
281
282
S. JENNINGS AND J. D. REYNOLDS
Boudreau, P. R., Dickie, L. M. & Kerr, S. R. (1991). Body-size spectra of production and biomass as system-level indicators of ecological dynamics. Journal of Theoretical Biology, 152, 329–339. Brander, K. (1981). Disappearance of common skate Raia batis from Irish Sea. Nature, 290, 48–49. Brown, J. H. & Gillooly, J. F. (2003). Ecological food webs: high-quality data facilitate theoretical unification. Proceedings of the National Academy of Sciences USA, 100, 1467–1468. Brown, J. H., Gillooly, J. F., Allen, A. P., Savage, V. M. & West, G. B. (2004). Towards a metabolic theory of ecology. Ecology, 85, 1771–1789. Charnov, E. L. (1993). Life History Invariants. Oxford: Oxford University Press. Cohen, J. E., Pimm, S. L., Yodzis, P. & Saldan˜a, J. (1993). Body sizes of animal predators and animal prey in food webs. Journal of Animal Ecology, 62, 67–78. Cohen, J. E., Jonsson, T. & Carpenter, S. R. (2003). Ecological community description using the food web, species abundance and body size. Proceedings of the National Academy of Sciences, USA, 100, 1781–1786. Cushing, D. H. (1975). Marine Ecology and Fisheries. Cambridge: Cambridge University Press. Cyr, H. (2000). Individual energy use and the allometry of population density. In Scaling in Biology, ed. J. H. Brown and G. B. West. Oxford: Oxford University Press, pp. 267–295. Daan, N., Gislason, H., Pope, J. G. & Rice, J. C. (2005). Changes in the North Sea fish community: evidence of the indirect effects of fishing? ICES Journal of Marine Science, 62, 177–188. Damuth, J. (1981). Population density and body size in mammals. Nature, 290, 699–700. Denney, N. H., Jennings, S. & Reynolds, J. D. (2002). Life history correlates of maximum population growth rates in marine fishes. Proceedings of the Royal Society: Biological Sciences, 269, 2229–2237.
Dickie, L. M. (1976). Predation, yield and ecological efficiency in aquatic food chains. Journal of the Fisheries Research Board of Canada, 33, 313–316. Dickie, L. M., Kerr, S. P. & Schwinghamer, P. (1987a). An ecological approach to fisheries assessment. Canadian Journal of Fisheries and Aquatic Sciences, 44 (Supplement 2), 68–74. Dickie, L. M., Kerr, S. R. & Boudreau, P. R. (1987b). Size-dependent processes underlying regularities in ecosystem structure. Ecological Monographs, 57, 233–250. Dulvy, N. K., Polunin, N. V. C. Mill, A. C. & Graham, N. A. J. (2004). Size structural change in lightly exploited reef fish communities: evidence for weak indirect effects. Canadian Journal of Fisheries and Aquatic Sciences, 61, 466–475. Dulvy, N. K., Sadovy, Y. & Reynolds, J. D. (2003). Extinction vulnerability in marine populations. Fish and Fisheries, 4, 25–64. Duplisea, D. E. & Kerr, S. R. (1995). Application of a biomass size spectrum model to demersal fish data from the Scotian shelf. Journal of Theoretical Biology, 177, 263–269. Duplisea, D. E., Jennings, S., Warr, K. J. & Dinmore, T. A. (2002). A size-based model to predict the impacts of bottom trawling on benthic community structure. Canadian Journal of Fisheries and Aquatic Sciences, 59, 1785–1795. FAO (2005). Review of the State of World Marine Fishery Resources. FAO Fisheries Technical Paper 457. Rome. Fenchel, T. (1974). Intrinsic rate of natural increase: the relationship with body size. Oecologia, 14, 317–326. Frank, K. T. & Leggett, W. C. (1994). Fisheries ecology in the context of ecological and evolutionary theory. Annual Review of Ecology and Systematics, 25, 401–422. Fry, B. & Quinones, R. B. (1994). Biomass spectra and stable-isotope indicators of trophic level in zooplankton of the northwest Atlantic. Marine Ecology Progress Series, 112, 201–204.
BODY SIZE AND EXPLOITATION
Gilkinson, K., Paulin, M., Hurley, S. & Schwinghamer, P. (1998). Impacts of trawl door scouring on infaunal bivalves: results of a physical trawl door model/dense sand interaction. Journal of Experimental Marine Biology and Ecology, 224, 291–312. Gillooly, J. F. (2000). Effect of body size and temperature on generation time in zooplankton. Journal of Plankton Research, 22, 241–251. Gillooly, J. F., Brown, J. H., West, G. B., Savage, V. M. & Charnov, E. L. (2001). Effects of size and temperature on metabolic rate. Science, 5538, 2248–2250. Gillooly, J. F., Charnov, E. L., West, G. B., Savage, V. M. & Brown, J. H. (2002). Effects of size and temperature on developmental time. Nature, 417, 70–73. Gislason, H. (2002). The effects of fishing on nontarget species and ecosystem structure and function. In Responsible Fisheries in the Marine Ecosystem, ed. M. Sinclair & G. Valdimarsson. CAB International, Wallingford, pp. 255–274. Gislason, H. & Rice, J. C. (1998). Modelling the response of size and diversity spectra of fish assemblages to changes in exploitation. ICES Journal of Marine Science, 55, 362–370. Goodwin, N. B., Grant, A., Perry, A. L., Dulvy, N. K. & Reynolds, J. D. (2006). Life history correlates of density-dependent recruitment in marine fishes. Canadian Journal of Fisheries and Aquatic Sciences, 63, 494–509. Goodyear, C. P. (1977). Assessing the impact of power plant mortality on the compensatory reserve of fish populations. In Proceedings of the Conference on Assessing the Effects of Power Plant Induced Mortality on Fish Populations, ed. W. Van Winkle. New York: Pergamon Press, pp. 186–195. Harvey, P. H. & Pagel, M. D. (1991). The Comparative Method in Evolutionary Biology. Oxford: Oxford University Press.
Heino, M. & Godo, O. R. (2002). Fisheries-induced selection pressures in the context of sustainable fisheries. Bulletin of Marine Science, 70, 639–656. Hilborn, R. & Walters, C. J. (1992). Quantitative Fisheries Stock Assessment: Choice, Dynamics and Uncertaincy. New York: Chapman and Hall. Hutchings, J. A. (2001). Influence of population decline, fishing, and spawner variability on the recovery of marine fishes. Journal of Fish Biology, 59 (Supplement A), 306–322. Hutchings J. A. (2002). Life histories of fish. In Handbook of Fish Biology and Fisheries. Vol. 1., ed. P. J. B. Hart and J. D. Reynolds. Oxford: Blackwell Publishing, pp. 149–174. Hutchings, J. A. & Baum, J. K. (2005). Measuring marine fish biodiversity: temporal changes in abundance, life history and demography. Philosophical Transactions of the Royal Society B, 360, 315–338. Hutchings, J. A. & Reynolds, J. D. (2004). Marine fish population collapses: consequences for recovery and extinction risk. Bioscience, 54, 297–309. Jennings, S. (2005). Size-based analysis of aquatic food webs. In Aquatic Food Webs: An Ecosystem Approach, ed. A. Belgrano, U. M. Scharler, J. Dunne and R. E. Ulanowicz. Oxford: Oxford University Press, pp. 86–97. Jennings, S. & Blanchard, J. L. (2004). Fish abundance with no fishing: predictions based on macroecological theory. Journal of Animal Ecology, 73, 632–642. Jennings, S. & Mackinson, S. (2003). Abundancebody mass relationships in size structured food webs. Ecology Letters, 6, 971–974. Jennings, S., Reynolds, J. D. & Mills, S. C. (1998). Life history correlates of responses to fisheries exploitation. Proceedings of the Royal Society: Biological Sciences, 265, 333–339. Jennings, S., Greenstreet, S. P. R. & Reynolds, J. D. (1999). Structural change in an exploited fish community: a consequence of differential fishing effects on species with
283
284
S. JENNINGS AND J. D. REYNOLDS
contrasting life histories. Journal of Animal Ecology, 68, 617–627. Jennings, S., Kaiser, M. J. & Reynolds, J. D. (2001a). Marine Fisheries Ecology. Oxford: Blackwell Science. Jennings, S., Dinmore, T. A., Duplisea, D. E., Warr, K. J. & Lancaster, J. E. (2001b). Trawling disturbance can modify benthic production processes. Journal of Animal Ecology, 70, 459–475. Jennings, S., Pinnegar, J. K., Polunin, N. V. C. & Warr, K. J. (2002). Linking size-based and trophic analyses of benthic community structure. Marine Ecology Progress Series, 226, 77–85. Jones, R. (1974). Assessing the long-term effects of changes in fishing effort and mesh size from length composition data. ICES CM 1974, pp. 33–47. Kaiser, M. J. & de Groot, S. J. (eds.) (2000). The Effects of Fishing on Non-Target Species and Habitats: Biological, Conservation and SocioEconomic Issues. Oxford: Blackwell Science. Kaiser, M. J., Ramsay, K., Richardson, C. A., Spence, F. E. & Brand, A. R. (2000). Chronic fishing disturbance has changed shelf sea benthic community structure. Journal of Animal Ecology, 69, 494–503. Kerr, S. R. (1974). Theory of size distribution in ecological communities. Journal of the Fisheries Research Board of Canada, 31, 1859–1862. Kerr, S. R. & Dickie, L. M. (2001). The Biomass Spectrum: a Predator-Prey Theory of Aquatic Production. New York: Columbia University Press. Law, R. (1979). Ecological determinants in the evolution of life histories. In Population Dynamics, ed. R. M. Anderson, B. D. Turner and L. R. Taylor. Oxford: Blackwell Scientific Publications, pp. 81–103. Law, R. (2000). Fishing, selection and phenotypic evolution. ICES Journal of Marine Science, 57, 659–668. Law, R. & Stokes, T. K. (2000). Evolutionary impacts of fishing on target populations. In
Marine Conservation Biology: the Science of Maintaining Biodiversity, ed. L. Crowder and E. Norse. Island Press. Li, B. L. & Charnov, E. L. (2001). Diversity-stability relationships revisited: scaling rules for biological communities near equilibrium. Ecological Modelling, 140, 247–254. Loeuille, N. & Loreau, M. (2005). Evolutionary emergence of size-structured food webs. Proceedings of the National Academy of Sciences, 102, 5761–5766. Maxwell, D. & Jennings, S. (2005). Power of monitoring programmes to detect decline and recovery of rare and vulnerable fish. Journal of Applied Ecology, 42, 25–37. Myers, R. A., Mertz, G. & Fowlow, P. S. (1997). Maximum population growth rates and recovery times for Atlantic cod Gadus morhua. Fishery Bulletin, 95, 762–772. Nee, S., Read, A. F., Greenwood, J. J. D. & Harvey, P. H. (1991). The relationship between abundance and body size in British birds. Nature, 351, 312–313. Pauly, D., Christensen, V., Dalsgaard, J., Froese, R. & Torres, F. (1998). Fishing down marine food webs. Science, 279, 860–863. Pope, J. G., Stokes, T. K., Murawski, S. A. & Idoine, S. I. (1988). A comparison of fish size composition in the North Sea and on Georges Bank. In Ecodynamics: Contributions to Theoretical Ecology, ed. W. Wolff, C. J. Soeder and F. R. Drepper. Berlin: Springer Verlag, pp. 146–152. Pope, J. G., Rice, J. C., Daan, N., Jennings, S. & Gislason, H. (2006). Modelling an exploited marine fish community with 15 parameters – results from a simple size-based model. ICES Journal of Marine Science, 63, 1029–1044. Reynolds, J. D. (2003). Life histories and extinction risk. In Macroecology, ed. T. M. Blackburn and K. J. Gaston. Oxford: Blackwell Publishing, pp. 195–217. Reynolds, J. D. & Peres, C. A. (2006). Overexploitation. In Principles of Conservation Biology, ed. M. Groom, G. Meffe and
BODY SIZE AND EXPLOITATION
R. Carroll. Massachusetts: Sinauer, Sunderland. Reynolds, J. D., Dulvy, N. K., Goodwin, N. B. & Hutchings, J. A. (2005). Biology of extinction risk in marine fishes. Proceedings of the Royal Society London, B, 272, 2337–2344. Rice, J. & Gislason, H. (1996). Patterns of change in the size spectra of numbers and diversity of the North Sea fish assemblage, as reflected in surveys and models. ICES Journal of Marine Science, 53, 1214–1225. Roff, D. A. (1992). The Evolution of Life Histories: Theory and Analysis. New York: Chapman & Hall. Savage, V. M., Gillooly, J. F., Brown, J. H., West, G. B. & Charnov, E. L. (2004). Effects of body size and temperature on population growth. American Naturalist, 163, 429–441. Sheldon, R. W., Prakash, A. & Sutcliffe, W. H. (1972). The size distribution of particles in the ocean. Limnology and Oceanography, 17, 327–340. Shin, Y. J. & Cury, P. (2004). Using an individualbased model of fish assemblages to study the response of size spectra to changes in fishing. Canadian Journal of Fisheries and Aquatic Sciences, 61, 414–431. Shin, Y.-J., Rochet, M.-J., Jennings, S., Field, J. & Gislason, H. (2005). Using size-based indicators to evaluate the ecosystem effects
of fishing. ICES Journal of Marine Science, 62, 384–396. Shuter, P. J. & Abrams, B. J. (eds.) (2005). Building on Beverton’s legacy: life history variation and fisheries management. Canadian Journal of Fisheries and Aquatic Sciences, 62, 725–902. Sinclair, M. & Valdimarsson, G. (2003). Responsible Fisheries in the Marine Ecosystem. Rome: FAO. Stokes, T. K., McGlade, J. M. & Law, R. eds. (1993). The Exploitation of Evolving Resources. Berlin: Springer-Verlag. Thiebaux, M. L. & Dickie, L. M. (1993). Structure of the body size spectrum of the biomass in aquatic ecosystems: a consequence of allometry in predator-prey interactions. Canadian Journal of Fisheries and Aquatic Sciences, 50, 1308–1317. Thygesen, U. H., Farnsworth, K. D., Anderson, K. H. & Beyer, J. E. (2005). How optimal life history changes with the community sizespectrum. Proceedings of the Royal Society B, 272, 1323–1331. Ware, D. M. (2000). Aquatic ecosystems: properties and models. In Fisheries Oceanography: An Integrative Approach to Fisheries Ecology and Management, ed. P. J. Harrison and T. R. Parsons. Oxford: Blackwell Science, pp. 161–194. Watson, R. & Pauly, D. (2001). Systematic distortions in world fisheries catches. Nature, 414, 534–536.
285
CHAPTER FIFTEEN
How body size mediates the role of animals in nutrient cycling in aquatic ecosystems ROBERT O . HALL , JR . University of Wyoming, USA
BENJAMIN J . KOCH University of Wyoming, USA
MICHAEL C . MARSHALL University of Wyoming, USA
BRAD W . TAYLOR University of Wyoming, USA
LUSHA M . TRONSTAD University of Wyoming, USA
Introduction Aquatic ecosystems have been fertile ground for understanding the extent to which animals can alter nutrient cycling. Although animals have been included in ecosystem models for years (for example, Teal, 1962), it is only more recently that investigators have looked at animals, either as individuals, single species, or assemblages, as agents regulating nutrient cycling (Kitchell et al., 1979; Meyer, Schultz & Helfman, 1983; Grimm, 1988; Jones & Lawton, 1995). A recent review details how animals can affect nutrient cycling in freshwater ecosystems (Vanni, 2002), but the next step is to understand the controls on which animals are important regulators of nutrient dynamics in ecosystems. One controlling factor is determined by attributes of the animals themselves, such as their body size. Animals can regulate nutrient cycling directly or indirectly (Kitchell et al., 1979; Vanni, 2002). Direct regulation is the transformation and transportation of nutrients by animal ingestion, egestion, production and excretion. For example, animal excretion can constitute the largest source of plant-available nitrogen (N) within an ecosystem (Hall, Tank & Dybdahl, 2003) and animals can move nutrients between habitats (Meyer et al., 1983). Perhaps more common are indirect controls, whereby animals alter nutrient cycling by changing the biomass, production or distribution of the plants or microbes that take up nutrients. For example, predatory fish can regulate phosphorus (P) dynamics or nitrogen retention via a trophic cascade (Elser et al., 1998; Simon et al., 2004). Body Size: The Structure and Function of Aquatic Ecosystems, eds. Alan G. Hildrew, David G. Raffaelli and Ronni Edmonds-Brown. Published by Cambridge University Press. # British Ecological Society 2007.
BODY SIZE AND NUTRIENT CYCLING
In this paper we consider only direct effects of animals on nutrient cycling, because predicting indirect effects in food webs contains much more uncertainty (Wootton, 1994). A point mentioned by Vanni (2002), that we expand on here, is the role of body size in controlling the degree to which animals contribute to ecosystem nutrient fluxes. Body size may control animal-mediated nutrient cycling by three main mechanisms. First, small animals have higher metabolism and, therefore, higher mass-specific excretion rates (Wen & Peters, 1994; Brown, Allen & Gillooly, this volume). Thus, total biomass being equal, an assemblage with small animals may have higher animal-mediated nutrient fluxes than one with large animals. Second, as body size increases, allometric variation in structural tissue (e.g. P-rich bone) may alter ratios of excreted nutrients. Third, large animals have larger home ranges and are more likely to migrate long distances, so nutrient translocation by animals may also be a function of body size. In this chapter we first address how body size controls nutrient fluxes in the context of the first two mechanisms described above by using published and unpublished data to examine the relationship between nutrient excretion and body size. We also consider the spatial and temporal translocation of nutrients by animal movements as a function of body size. In the second part of the chapter we apply these findings to address how ecosystem-level nutrient cycling will change as a function of variation in animal body size. In short, we know excretion can vary as a function of body size, but does this variation matter in ecosystems? We explore other factors that affect animal-mediated nutrient cycling, such as variation in the biomass of animal assemblages and their taxonomic composition, so that we can compare their influence to the effects of body size. Lastly, predators, especially humans, may alter the size structure of animal assemblages, and we consider how loss of large-bodied organisms may indirectly alter nutrient cycling (see also Jennings & Reynolds, this volume).
Body size and nutrient excretion Rates Aquatic animals excrete N and P in mostly mineral forms which are readily taken up by microbes. The primary form of N is ammonium, which is excreted via the gut in insects, or diffuses across the integument and gills of other animals. Animals primarily excrete P in the form of PO43 . Nutrient excretion rates vary with body size. In general, excretion rates (E) scale allometrically with body mass (M): E ¼ aMb
(15:1)
where a and b are constants (Huxley, 1932; Peters, 1983; Wen & Peters, 1994; Gillooly et al., 2001). For most aquatic animals, the relationship has an exponent b < 1 indicating that excretion rates increase at a rate less than isometric with
287
288
R. O. HALL ET AL.
increasing body size (Wen & Peters, 1994). For example, individual ammonium excretion rates for stream invertebrates, from at least 18 taxonomic orders, scale to the 0.85 power of body mass (Fig. 15.1), implying that larger taxa excrete at a lower rate for their size than smaller taxa. The mechanism for the less than proportional increase in excretion rate is probably linked to metabolism, which scales as the 3/4 power of body size (Brown et al., 2004; Brown et al., this volume). However, for many specific groups of animals, b can be higher or lower than 3/4. For example, b ¼ 0.67 for N and 0.54 for P in zooplankton (Wen & Peters, 1994), whereas macroinvertebrates (Fig. 15.1) have higher values. However, body size is not the sole factor controlling variation in animal excretion rate. It is worth considering the influence of other variables, that may alter or interact with the effects of body size, on animal-mediated nutrient cycling in aquatic ecosystems. For example, temperature influences metabolic processes, such as excretion rate (Peters, 1983; Fukuhara & Yasuda, 1989; Zhuang, 2005). Metabolic theory (Gillooly et al., 2001; Brown et al., 2004; Brown et al., this volume) provides a mechanistic framework for incorporating the effects of both temperature and body size on excretion rate. Fed animals have higher excretion rates than unfed animals (Gardner & Scavia, 1981; Grimm,
Figure 15.1 Ammonium excretion rates increase less than proportionally with body size (b < 1) for many benthic stream invertebrate taxa, indicating that larger invertebrates excrete ammonium at a lower rate per mg of body mass than do smaller invertebrates. The regression line (log10 [excretion rate] ¼ 1.057 þ 0.853log10 [mean individual body mass]; n ¼ 320, r2 ¼ 0.381, 95% CI on b [0.776 0.937]) was estimated using type II, reduced major axis linear regression (Bohonak & van der Linde, 2004). Data points were gathered using identical methods on field-caught animals from six streams and represent total excretion rates computed from one or more similarly-sized individuals of the same taxon within the same incubated container (Hall et al., 2003; R. O. Hall, unpublished data; Koch, 2005; M. C. Marshall, unpublished data).
BODY SIZE AND NUTRIENT CYCLING
1988), probably because of active metabolism of recently digested and assimilated N compounds. The use of fed or unfed animals may have contributed to the variability in Fig. 15.1 and make studies where methods differ less comparable. Taxonomic differences can also explain some of the variation in invertebrate nutrient excretion rates (Wen & Peters, 1994). For example, Conroy et al. (2005) found taxonomic differences in excretion rates of P, but not of N, between two species of mussels in the genus Dreissena. Interestingly, in contrast to nearly all other freshwater invertebrate taxa studied to date, N and P excretion rates for Dreissena increase disproportionately with body size (b ¼ 1.379), such that larger individuals excrete nutrients at a higher mass-specific rate than do smaller individuals (Conroy et al., 2005). The mechanisms behind this relationship are unclear, although it highlights the importance of recognizing taxonomy in studies of animal-mediated nutrient cycling. To examine taxonomic and size variation in excretion rates among fishes, we compared published fish excretion rates of individuals (n ¼ 156 for P and 163 for N) and species means (n ¼ 30 species for P and N) among freshwater representatives of 14 families, including Anostomidae, Aspredinidae, Catostomidae, Cetopsidae, Characidae, Characidiidae, Cichlidae, Clupeidae, Curimatidae, Loricariidae, Parodontidae, Pimelodidae, Salmonidae and Trichomycteridae (Schaus et al., 1997b; Gido, 2002; Vanni et al., 2002; Andre, Hecky & Duthie, 2003; B. J. Koch, unpublished data). Individual rates are the excretion of a single fish (many individuals in a species were measured) and species means were calculated by averaging the excretion and size of all the individuals in that species. All studies measured ammonium, but Schaus et al. (1997a) and Andre et al. (2003) measured soluble reactive P, Vanni et al. (2002) estimated total dissolved P, while Gido (2002) measured total reactive P. We converted wet mass to dry mass by assuming dry mass was 25% of wet mass (Schaus et al., 1997a; Gido, 2002; Andre et al., 2003), or used measured values directly (Vanni et al., 2002; B. J. Koch, unpublished data). Excretion of individuals within a species scaled with body size similarly, and were higher or lower than the species means (Table 15.1). The P excretion of three species scaled less than 1 (Table 15.1; Fig. 15.2a), meaning that the massspecific excretion rates declined with increasing size. However, N excretion of species showed greater variability (b < 1, b ¼ 1, b > 1), indicating that both size and taxonomy influence rates (Table 15.1; Fig. 15.2b). When fish species means were considered, excretion scaled proportionally with dry mass (that is b 1; Table 15.1, Figs. 15.2c, d). These data cannot disentangle the relative contribution of phylogeny vs. size because they are not independent. However, comparing species means to individual species, we can conclude that fish scale similarly to each other. Additionally, measurements were collected by different researchers under different conditions, which may cause high variation in excretion rates among all fishes.
289
290
R. O. HALL ET AL.
Table 15.1 Reduced major axis regression estimates for nitrogen and phosphorus excretion (log10 g N or P fish1 h1) and size (log10 dry mass, g) in five groups of fish (see Fig. 15.2). Data for Mbuna Cichlidae, Carpiodes carpio, Ictiobus bubalus and Dorosoma cepedianum estimates are excretion rates from individual fish within a taxon. We also calculated the mean excretion rate and mean size of 30 fish species taken from Gido (2002); Vanni et al. (2002); Andre et al. (2003); Schaus et al. (1997b); and Koch (unpublished data) and regressed mean excretion rate on mean body size. The bootstrapped 95% confidence intervals of the exponents are in parentheses. Taxa
Nitrogen n
Mbuna* Cichlidae Carpiodesx carpio Ictiobusx bubalus Dorosomayx cepedianum Species means
Intercept Exponent
Phosphorus r2
n
Intercept Exponent
r2
40 2.04
0.759 (0.664–0.886) 0.769 37 0.975
0.886 (0.567–1.24)
10 2.96
0.789 (0.525–1.06)
0.770 10 1.39
0.733 (0.543–0.848) 0.875
16 2.27
0.983 (0.764–1.43)
0.633 16 0.803
0.568 (0.347–0.906) 0.301
93 2.13
1.14 (1.04–1.24)
0.883 93 1.40
0.921 (0.844–0.997) 0.781
30 2.41
0.953 (0.851–1.05)
0.930 30 0.916
1.07 (0.903–1.32)
x
Gido (2002) Andre et al. (2003) y Schaus et al. (1997b) *
The relative importance of taxonomy, body size and temperature in controlling nutrient excretion rates is only just beginning to be explored, and adequately testing the interactions among these factors will require richer data sets and resolved molecular phylogenies. In addition, determining the basis of taxonomic variation in excretion rates remains a challenge. Body nutrient composition and diet may both play roles. Given that ammonium excretion rates for stream invertebrates are higher for fed than unfed animals (Grimm, 1988), predators, which feed sporadically, may have more variable excretion rates over time than continuously feeding grazers and detritivores. Stoichiometric differences in animal nutrient use might also drive taxonomic variation (Elser & Urabe, 1999). Predators, with relatively N-rich diets, may have higher N excretion rates than other feeding groups. Understanding when to account for taxonomic variation and when body size alone is sufficient for studies of animal-mediated nutrient cycling is central to predict successfully the role of animals in the nutrient dynamics of aquatic ecosystems.
0.176
0.831
BODY SIZE AND NUTRIENT CYCLING
Figure 15.2 Phosphorus (a, c) and nitrogen (b, d) excretion rates (mg P or N fish1 h1) versus dry mass (g) for individual fish of several taxa (a, b) and means of 30 fish species (c, d) from the literature (Schaus et al., 1997b; Gido, 2002; Vanni et al., 2002; Andre et al., 2003; B. J. Koch, unpublished data). See Table 15.1 for regression coefficients.
Ratios of N and P Not only will the amount of N and P excreted by animals be important in ecosystem nutrient cycling, but the ratio of these nutrients may also drive microbial assemblage structure and productivity (Elser et al., 1988). Nutrient ratios in food sources, animal composition and excretion (that is, ecological stoichiometry) have received much attention in aquatic ecology (Sterner & Elser, 2002). Stoichiometric theory predicts that the N:P in excretion is a positive function of the N:P of ingested food, and a negative function of the N:P requirement of the consumer (Sterner, 1990). Data show that the link between N:P in the zooplankton body and excreted N:P is not nearly as strong as the link with the N:P of their food (Elser & Urabe, 1999); that is, most of the variance in excreted N:P is accounted for by variation in the food. Few analyses show how body size drives the N:P in excretion in animals; indeed, there is little information on animal C:N:P content solely as a function of body size (Sterner & Elser, 2002). One hypothesis might be that aquatic animals should increase their N:P content as size increases, because increased size should lead to decreased demand for P as growth rate declines (Elser et al., 1996). Given higher body N:P, big animals should have lower excreted N:P than small ones. However,
291
292
R. O. HALL ET AL.
this effect may be hidden by phylogeny and allometric constraints, as taxonomy correlates with body size because large animals are often vertebrates that have high P storage in bone apatite, and presumably a high N:P in excreta. Data on aquatic animals suggest that excreted N:P increases with body size. Wen and Peters (1994) showed that log N excretion rate (mg N/d) increased more steeply with body mass than did excreted P for zooplankton. The difference in the exponents is 0.13, which corresponds to the exponent for N:P of excretion vs. body mass. Thus the N:P of excretion increases with body mass, suggesting that mechanisms other than growth rate control the relationship of excreted N:P with body size. Data from some vertebrates also suggest increases in the N:P excreted with body size. Excretion N:P in fishes and amphibians from a Piedmont stream in Venezuela was positively related to body size, which agrees with qualitative predictions based on a decreasing body N:P with increasing body mass in vertebrates (Vanni et al., 2002). For example, bony-scaled armoured catfishes (Loricariidae) had particularly low body N:P and therefore high N:P in excretion (Vanni et al., 2002). Tadpoles (families Bufonidae and Ranidae) had low excreted N:P; because they do not have ossified bones (low skeletal demand for P). These studies, although few, suggest that not only will body size determine the rates of nutrient regeneration, but it will also determine the ratio of these nutrients, with the data so far suggesting mostly increasing N:P with body size. Mechanisms for this increase are unclear, and certainly vary across taxa. For example, vertebrates will have proportionally more bone as their size increases (Sterner & Elser, 2002), which will increase P demand (lowering P excretion) with body size.
Body size and nutrient translocation Aquatic animals can alter nutrient cycling by moving nutrients from one location to another, thus subsidizing the receiving habitat (Kitchell et al., 1979; Vanni, 2002). In some instances this nutrient movement is between habitats within an ecosystem such as, for example, benthic feeding fish that excrete nutrients in the pelagic zone (Vadeboncoeur, Vander Zanden & Lodge, 2002) or haemulid grunts that feed in seagrass beds at night and rest above coral heads during the day, where they release nutrients that stimulate coral growth (Meyer et al., 1983). In other cases, animals move nutrients between ecosystems on a daily basis; e.g. ocean-foraging river otters (Lontra canadensis) excrete nutrients in discrete locations in terrestrial habitat (Ben-David et al., 2005). Less mobile or small-sized animals may actually concentrate nutrients at high levels in localized areas (Reinertsen et al., 1986). In contrast, Pacific salmon (Onchorhynchus spp.) transport nutrients from the ocean to rivers via an annual long-distance spawning migration (Gende et al., 2002). The degree of movement will be determined in part by the speed at which animals move and the behavioural
BODY SIZE AND NUTRIENT CYCLING
constraints on their home range. Both of these controls on movements should scale with body size. The distance moved by aquatic animals will depend on their body size because swimming speed scales with animal body size (Peters, 1983). For a given time travelled, a big animal can migrate further than a small one. Over large size ranges, an animal’s Reynolds number constrains movement (e.g. zooplankter versus a salmon). Small animals (e.g. rotifers) move very slowly because their short length confers a low Reynolds number, and therefore viscous forces are much higher than inertial forces. Within fishes that have high Reynolds numbers, swimming speed scales at about M0.14 (Weihs, 1977) assuming M / length2.6 (Peters, 1983). These modelled swimming speeds include both Reynolds number effects plus allometric scaling of swimming force and metabolic costs. Animals with lower Reynolds numbers have a steeper positive relationship between body mass and swimming speed, probably because of the more pronounced effects of viscous forces at small sizes. Swimming speed in diving beetles (Dytiscidae), increases as M0.36 (Nachtigall, 1977) assuming M / L2.5 (Benke et al., 1999). Thus, the decline in swimming speed for small animals probably decreases more quickly with body size than it does for fish. Behavioural constraints on home-range size and migration will also control nutrient movement by animals. Home range scales with body size in mammals at roughly M1 (Jetz et al., 2004). Home-range sizes of fishes are similar to mammals, scaling as M1.1, while insects and crustaceans are at M0.7 and molluscs at M0.55 (Alimov, 2003). Given that distance moved will scale as the squareroot of area, distance moved for fishes should then scale as approximately M0.5. This rate of increase with body size in the actual distance moved by animals is higher than that for speed alone, because home range is determined by many more attributes than is speed. These include, for example, resource requirements and interactions with conspecifics (Jetz et al., 2004). Animals that transport substantial nutrients among habitats are likely to be large, as in Pacific salmon (Gende et al., 2002), river otters (Ben-David et al., 2005), and the longdistance migratory fish, sapuara (Semaprochilodus kneri) (Winemiller & Jepsen, 2004). It is important to consider the strong effect of behaviour; the much smaller sapuara migrates long distances along rivers, and therefore transfers nutrients much further than does the coastal river otter. Coral reef fishes are large enough to travel long distances, but many stay in one spot on the reef all their lives. Thus, while large animals are more likely to move nutrients, behavioural characteristics also control this distance.
Consequences of size-varying nutrient cycling Variation in body-size distributions Because excretion rates typically increase less than proportionally with animal body size, variation in size distributions can partially control animal-driven
293
294
R. O. HALL ET AL.
nutrient mineralization and storage in ecosystems. Here we ask to what degree does variation in animal size distribution regulate nutrient mineralization? Researchers have described a wide variety of biomass-size distributions (also called size spectra) for aquatic animal assemblages, including flat or smooth, uni-, bi- and poly-modal, and step or asymptotic functions. Size distributions can vary considerably in space and time within and among aquatic habitats (Hanson, Prepas & Mackay, 1989; Stead et al., 2005), complicating generalizations (see Warwick, this volume). The diversity of methods in body-size estimations (for example, Morin & Nadon, 1991; Ramsay et al., 1997; Baca & Threlkeld, 2000) and analytical techniques, such as different sieve or size classes, further complicate size-spectra summaries (Cyr & Pace, 1993; Robson, Barmuta & Fairweather, 2005). However, when only the invertebrate portions of published aquatic assemblage spectra are included (that is, smaller and larger portions excluded), clearer patterns of shape categories emerge. Most size spectra have biomass peaks that are skewed left, meaning larger animals generally account for most of the total biomass, even though they may be outnumbered by smaller ones. Size distributions in lakes vary as a function of habitat. Studies that include multiple habitat types within the same lake suggest that pelagic and littoral assemblages tend to have bimodal distributions of invertebrates (Hanson et al., 1989; Cyr & Pace, 1993; Rasmussen, 1993) and polymodal distributions when fishes are included (Gaedke, 1992), whereas profundal (and sublittoral) distributions tend to be unimodal (Hanson et al., 1989). The magnitude and locations of biomass peaks and troughs also vary among habitats within lakes; littoral habitats have peaks at larger body sizes. For example, the two biomass maxima for littoral habitats tended to occur between 1–4 mg and 64–256 mg wet mass (Rasmussen, 1993), whereas the two peak densities of pelagic zooplankton occurred between 0.044–0.125 mg and 2.0–11.3 mg dry mass for small and large animals, respectively (Cyr & Pace, 1993). Streams generally have unimodal biomass size distributions (Cattaneo, 1993; Bourassa & Morin, 1995; Mercier et al., 1999; Schmid, Tokeshi & Schmid-Araya, 2002). Body-size maxima, as equivalent to a spherical diameter, were between 2–4 mm in streams (Cattaneo, 1993), and the average individual biomass increased slightly with increasing trophic status from 24–40 mg dry mass in oligotrophic to urban eutrophic streams, respectively (Bourassa & Morin, 1995). Overall, although unimodality is robust across many streams, total biomass can vary by an order of magnitude (for example, Bourassa & Morin, 1995) suggesting possible dramatic differences in animal driven nutrient fluxes within a stream system. Estimating nutrient flux from biomass size distributions Animal assemblages with different size distributions should have different nutrient supply rates to ecosystems, all else being equal. To illustrate this
BODY SIZE AND NUTRIENT CYCLING
Figure 15.3 (a–c) Representative animal size spectra from three littoral ecosystems: (a) Lake Brome, (b) Lake Waterloo, (c) Lake Bromont (Rasmussen, 1993). Total biomass (mg dry mass m2) has been normalized to 1000 mg dry mass m2 for the three communities. Size classes are Log2 (mg dry mass). (d–f) Modelled P fluxes (mg P m2 h1) supplied by excretion for the three assemblages, assuming a negative relationship between mass-specific excretion rate and body size. Total nutrient flux varies nearly two-fold for the three communities (60, 115 and 104 mg P m2 h1 for panels d, e and f, respectively) and the shapes of nutrient flux distributions changed relative to size spectra.
point we used data-capturing software to extract published size spectra from plots. We gathered three representative aquatic animal size spectra: a bimodal distribution with proportionally more large individuals (Fig. 15.3a), a strongly peaked bimodal distribution (Fig. 15.3b) and a unimodal distribution (Fig. 15.3c, Rasmussen, 1993). We assumed dry mass was 25% of wet mass (Feller & Warwick, 1988) and normalized the literature spectra data to have equivalent total biomasses (1000 mg dry mass m2) while preserving the same distribution shape in the original data sets. For each of these three animal assemblages, we then calculated the P flux supplied by excretion for each size class (Figs. 15.3d–f ), using a negative relationship between mass-specific excretion rate and body size (mg P mg dry mass1 h1 ¼ 0.0954[dry mass](0.541); Wen & Peters, 1994). While this analysis accounts for variation in animal excretion rate due to body size, it does not incorporate the effects of potentially different temperatures or taxonomic composition among animal assemblages. Nevertheless, despite total biomass being the same for the three communities, total nutrient flux (cumulative area of rectangles) from each of the three animal communities is not equivalent, varying by almost a factor of two in this example (60, 115 and 104 mg P m2 h1 for Figs. 15.3d, e & f, respectively). Furthermore, the shapes of the
295
296
R. O. HALL ET AL.
nutrient flux distributions differ from their respective biomass size spectra. For example, although larger animals comprise most of the total biomass in Fig. 15.3a, small- and medium-sized animals supply the bulk of the nutrient flux from this assemblage (Fig. 15.3d). Thus the size spectra of animal communities have important consequences on the supply and cycling of nutrients, and those size classes that contribute most to total assemblage nutrient flux are not necessarily the most biomass-rich size classes in the assemblage. Predator control of prey body size and nutrient cycling The well-known impact of predators on prey size structure may alter nutrient cycling in aquatic ecosystems. Fish predators can decrease average size of prey by eating large zooplankton (for example, Brooks & Dodson, 1965; Li, Wetterer & Hairston, 1985) and large benthic invertebrates in lakes (Blumenshine, Lodge & Hodgson, 2000). Alternatively, planktonic invertebrate predators, such as Chaoborus, select small zooplankton (for example, Dodson, 1974), increasing average prey body size. In streams, predatory invertebrates, fish and mammals tend to consume the largest individuals of their prey (Quinn & Kinnison, 1999; Allan, 2001; Woodward & Warren, this volume). In addition to changes in size structure via consumptive effects, the presence of predators can alter prey-size distribution simply through non-consumptive effects, such as chemical cues (for example, Tollrian, 1995; Peckarsky et al., 2002) and excretion (Ramcharan, France & McQueen, 1996). Simultaneous to their effects on body size, predators can also affect prey physiology by increasing the allocation of nutrients to structural cells, (for example, Lively, 1986; Vanni, 1987; Crowl & Covich, 1990; Stibor, 1992; Barry, 1994; Arendt & Wilson, 2000; Dahl & Peckarsky, 2002), which may change the composition of consumer-mineralized nutrients. Altered size structure of the prey assemblage may change nutrient cycling, because mass-specific excretion rate decreases with increasing animal size. Additionally, body size affects the nutrient ratios at which animals excrete. Changes in excretion N:P can alter the supply of the nutrient that limits primary producers. Elser et al. (1988) suggest that phytoplankton communities are more likely P-limited when the zooplankton assemblage includes large-bodied individuals and N-limited when the zooplankton assemblage is mainly small-bodied individuals. Understanding how changes in the size structure of prey can affect nutrient cycling is not straightforward, because predators can simultaneously alter prey abundance and biomass, and regenerate nutrients by consuming prey. Bartell (1981) modelled P cycling under differing levels of predation using previously published data on zooplankton size and biomass in lakes, and a mass-specific excretion model for zooplankton. Nutrient fluxes from zooplankton did not always increase when the assemblage switched from large-bodied to the
BODY SIZE AND NUTRIENT CYCLING
small-bodied individuals that have greater mass-specific excretion rates. In fact, P fluxes either remained similar, increased or decreased when lakes were altered from low to high planktivore abundance, depending on changes in total zooplankton biomass. However, nutrients may be more available in lakes with abundant zooplanktivorous fish, because smaller zooplankton turn P over faster than larger-bodied zooplankton (Henry, 1985). In addition to zooplankton, fish can also be an important source of nutrients to primary producers. Some studies have reported that the nutrient flux from zooplankton is much larger than fluxes from fish (Ramcharan et al., 1996), while others found the reverse (Vanni & Findlay, 1990; Carpenter et al., 1992). Boers, Vanballegooijen & Uunk (1991) showed that the main P source switched from zooplankton to fish as planktivore biomass increased. Regardless of which animal supplies more nutrients, their body size can affect nutrient cycling. To illustrate how size structure can change nutrient supply and demand we use lakes with low and high planktivorous fish abundance. In lakes with low planktivore abundance, both large and small zooplankton may be present (Fig. 15.4a), but the assemblage is mainly composed of small zooplankton when planktivores are abundant (solid line, Fig. 15.4c). Compensatory increases in the number of small zooplankton may result when fish are present (dashed line, Fig. 15.4c); however, most studies show an overall decrease in total zooplankton biomass (for example, Vanni & Findlay, 1990). When the density of planktivorous fish is low (that is, both large and small zooplankton are present), zooplankton excrete at a range of N:P ratios (grey line is N; black line is P; Figure 15.4b); however, zooplankton excrete at a lower N:P ratio when planktivorous fish are abundant (causing N to be potentially limiting). Based on modelling by Bartell (1981), changes in zooplankton size structure may either increase, decrease or not change lake nutrient fluxes (Fig. 15.4d), depending on compensatory changes in assemblage biomass. In contrast to planktivorous fish, planktonic-invertebrate predators selectively consume small zooplankton, resulting in a large-bodied prey assemblage excreting at a high N:P ratio. Depending on biomass, prey nutrient fluxes could change in either direction but may cause P to be limiting. The effect of predators on zooplankton body size in temperate lakes is well known; however, to our knowledge no studies have investigated how shifts in body size of stream invertebrates could alter nutrient cycling. Because stream predators selectively consume large-bodied prey, similar to planktivores feeding on zooplankton, we suggest that a decline in N:P mineralization and an increase in mineralization rates may hold for streams. However, even with the advances in methods to estimate pools and fluxes of nutrients in streams, the effects of predators on prey body size and nutrient cycling has not been investigated, even though in certain cases stream invertebrates can be an important source of ammonium (Grimm, 1988; Hall et al., 2003; Koch, 2005).
297
298
R. O. HALL ET AL.
Figure 15.4 Harvesting and predators can alter the size structure of their prey, which can change nutrient cycling. (a) In an assemblage with low harvesting or planktivore density (plankton-eating fish), large animals persist but the small animals are most abundant. (b) Nitrogen (grey line) and phosphorus (black line) mass-specific excretion rates are inversely related to body size, thus smaller animals excrete at a lower N:P than larger animals. (c) When harvesting or planktivore density is high, only small animals will be abundant, which may cause compensatory increases in density (dashed line). (d) Nutrient cycling by the small-bodied assemblage may result in compensatory increases, decreases or no change in nitrogen and phosphorus supply by animals, but nutrient ratios will probably be altered.
The effect of harvesting-induced changes in animal size structure on nutrient cycling Harvesting by humans affects the size structure of aquatic animal assemblages, and these altered size distributions may affect the rates and types of nutrients mineralized by animals (Jennings & Reynolds, this volume; Persson & De Roos, this volume). Similar to many other animals, humans selectively harvest large individuals and species (Pauly et al., 1998; Jackson et al., 2001; Roy et al., 2003; Allan et al., 2005). Size-selective harvesting can substantially change species composition and food-web structure (for example, removal of predators), leading to fishing down the food web – a process by which larger species, often predators, with slower growth rates are successively removed from the assemblage, leaving smaller species with faster growth rates (and thus higher massspecific nutrient excretion) that occupy lower trophic levels (Pauly et al., 1998; Welcomme, 1999). In addition, size-selective harvesting can decrease body size indirectly, by causing earlier maturation at smaller sizes via rapid evolutionary
BODY SIZE AND NUTRIENT CYCLING
change or increased resource availability that accelerates growth and decreases time to maturity of the remaining individuals (Trippel, 1995). Taken together, human harvesting generally decreases or eliminates the biomass of large animals from an ecosystem (Myers & Worm, 2003; Allan et al., 2005; Ward & Myers, 2005). There are several mechanisms by which harvest-induced changes in animal body size may alter the role of aquatic animals in mineralizing nutrients. Foremost, size-selective harvesting results in the loss of large-bodied individuals and species with high excretion rates per individual, but low mass-specific excretion. There are also important differences in the ratios at which limiting nutrients, such as N and P, are released by animals of different size (Wen & Peters, 1994; Schindler & Eby, 1997; Sterner & Elser, 2002; Vanni et al., 2002). As a result, the removal of large individuals may disproportionately reduce the amount of N relative to P supplied by animal assemblages (Fig. 15.4a, b), assuming there is no compensatory increase in abundances of smaller individuals or species (solid line; Fig. 15.4c). If there are compensatory increases in abundance of smaller individuals or species (dashed line; Fig. 15.4c) with higher massspecific mineralization rates, then the total supply of nutrients by the assemblage experiencing harvesting may equal or surpass the amount supplied by the assemblage before harvesting (Fig. 15.4d). In addition, because home-range size and migration distance increases with body size (Brown, 1995; Alimov, 2003; Jetz et al., 2004), reduced body size due to harvesting could also decrease the spatial scale over which nutrients are distributed by animals. This impact has been realized; harvesting of large, migratory salmon may have decreased marine nutrient loads to inland rivers, potentially lowering their productivity (Thomas et al., 2003). Moreover, the larger animals, which are often the first and most intensely harvested, generally have longer lifespans and more stable population cycles than the smaller, short-lived species that are less frequently harvested. Therefore, the removal of large, long-lived animals could increase the fluctuations of nutrients mineralized by animal populations. Overharvesting of large animals is a hallmark of all aquatic environments (Myers & Worm, 2003; Allan et al., 2005). However, surprisingly little is known about how the removal of larger animals alters the type or supply rate of nutrients mineralized by animal assemblages and, more importantly, whether such changes in nutrients are large enough to alter ecosystem-level processes. In the Baltic sea, Hjerne and Hansson (2002) estimated the removal of N and P in fish biomass by harvesting to be 1.4–7% of the total nutrient load, although the nutrient loss due to decreased mineralization by fish was not quantified. Although information is available on how predators can mediate nutrient mineralization rates by altering the size-structure of their prey, the process and long-term effects of harvesting by humans are likely to be very different. Humans typically remove the biomass of the largest animals, rarely switch
299
300
R. O. HALL ET AL.
prey until populations are severely reduced or regulatory restrictions are imposed, and harvest at maximal rates, which are often supported by external factors such as economic subsidies. In freshwater, species removals for biomanipulation (Horppila, 1998; Tarvainen, Sarvala & Helminen, 2002), and declines in introduced species, affect nutrient fluxes (Kraft, 1993); however, few studies have documented the direct effects of size-selective harvesting on nutrient fluxes. One reason is the mismatch in the data that are available on nutrient mineralization rates and harvesting rates of aquatic animals between marine and freshwater ecosystems. There are comparatively better data on catch size and body size of marine animals (Pauly et al., 1998; Myers & Worm, 2003; Ward & Myers, 2005) than freshwater animals (Allan et al., 2005), whereas there are more empirical data on nutrient regeneration rates for freshwater animals (Sterner & Elser, 2002). In marine systems, it may be useful to apply bioenergetic models to estimate the amount and type of nutrients lost from these systems as a result of having removed 80% of the large predatory fish biomass (Myers & Worm, 2003). Predicting the effects of harvesting-induced changes in body size on nutrient cycling is a new challenge that could improve our understanding of the role of animals in ecosystem functioning, and provide urgently needed guidance for managing and restoring these systems.
The next steps? Given that animals can often be important regenerators, storers and transporters of nutrients in ecosystems (Kitchell et al., 1979; Gende et al., 2002; Vanni, 2002; Koch, 2005), body size may be the single most important trait of the animals themselves in controlling these processes. There are plenty of avenues in which to explore further the role of body size in conjunction with other animal attributes (for example, phylogeny), and ecosystem processes. Below we give some of these examples. 1.
2.
3.
Taxonomic identity probably determines a large fraction of variation in excretion rates, and taxonomy covaries with body size. To what degree does size alone determine nutrient excretion rates? Can we integrate size and phylogeny to improve predictions of nutrient excretion rate? Body size allows us to examine how traits of animals impact ecosystem processes, but we cannot forget that the attributes of the ecosystems themselves will, in part, determine the impact. For example, plant nutrient demand, disturbance and hydrologic flushing rates are certainly important. How important is animal assemblage structure relative to physical controls and plant/microbial demand for nutrients? We can only speculate as to the potential role of many fisheries on changes to nutrient cycling. Some are well known (for example, salmon), but most are unknown (for example, groundfish stocks). These human-induced
BODY SIZE AND NUTRIENT CYCLING
changes present an opportunity to examine how changes in aquatic animal assemblages affect ecosystem processes and may provide the means to compare the relative importance of direct versus indirect effects of assemblage and size structure on nutrient cycling.
Acknowledgements Mike Vanni, Emidio Andre, Keith Gido and Maynard Schaus kindly provided tables of their published data for analysis. Two anonymous reviewers provided useful comments on an earlier draft of this manuscript. Financial support was provided by National Science Foundation; Environmental Protection Agency; and the Juneau Pacific Northwest Research Station, USDA Forest Service.
References Alimov, A. F. (2003). Territoriality in aquatic animals and their sizes. Biology Bulletin, 30, 79–86. Allan, J. D. (2001). Stream Ecology: Structure and Function of Running Waters, Boston: Kluwer Academic Publishers. Allan, J. D., Abell, R., Hogan, Z., Revenga, C., Taylor, B. W., Welcomme, R. L. & Winemiller, K. (2005). Overfishing of inland waters. Bioscience, 55, 1041–1051. Andre, E. R., Hecky, R. E. & Duthie, H. C. (2003). Nitrogen and phosphorus regeneration by cichlids in the littoral zone of Lake Malawi, Africa. Journal of Great Lakes Research, 29, 190–201. Arendt, J. D. & Wilson, D. S. (2000). Population differences in the onset of cranial ossification in pumpkinseed (Lepomis gibbosus), a potential cost of rapid growth. Canadian Journal of Fisheries and Aquatic Sciences, 57, 351–356. Baca, R. M. & Threlkeld, S. T. (2000). Using size distributions to detect nutrient and sediment effects within and between habitats. Hydrobiologia, 435, 197–211. Barry, M. J. (1994). The costs of crest induction for Daphnia carinata. Oecologia, 97, 278–288. Bartell, S. M. (1981). Potential impact of sizeselective planktivory on phosphorus release by zooplankton. Hydrobiologia, 80, 139–145.
Ben-David, M., Blundell, G. M., Kern, J. W. et al. (2005). Communication in river otters: creation of variable resource shed for terrestrial communities. Ecology, 86, 1331–1345. Benke, A. C., Huryn, A. D., Smock, L. A. & Wallace, J. B. (1999). Length-mass relationships for freshwater macroinvertebrates in North America with particular reference to the southeastern United States. Journal of the North American Benthological Society, 18, 308–343. Blumenshine, S. C., Lodge, D. M. & Hodgson, J. R. (2000). Gradient of fish predation alters body size distributions of lake benthos. Ecology, 81, 374–386. Boers, P., Vanballegooijen, L. & Uunk, J. (1991). Changes in phosphorus cycling in a shallow lake due to food web manipulations. Freshwater Biology, 25, 9–20. Bohonak, A. J. & van der Linde, K. (2004). RMA: Software for reduced major axis regression, Java version. http://www.kimvdlinde.com/ professional/rma.html. Bourassa, N. & Morin, A. (1995). Relationships between size structure of invertebrate assemblages and trophy and substrate composition in streams. Journal of the North American Benthological Society, 14, 393–403.
301
302
R. O. HALL ET AL.
Brooks, J. L. & Dodson, S. I. (1965). Predation, body size, and composition of plankton. Science, 150, 28–35. Brown, J. H. (1995). Macroecology. Chicago: University of Chicago Press. Brown, J. H., Gillooly, J. F., Allen, A. P., Savage, V. M. & West, G. B. (2004). Toward a metabolic theory of ecology. Ecology, 85, 1771–1789. Carpenter, S. R., Kraft, C. E., Wright, R. et al. (1992). Resilience and resistance of a lake phosphorus cycle before and after food web manipulation. American Naturalist, 140, 781–798. Cattaneo, A. (1993). Size spectra of benthic communities in Laurentian streams. Canadian Journal of Fisheries and Aquatic Sciences, 50, 2659–2666. Conroy, J. D., Edwards, W. J., Pontius, R. A. et al. (2005). Soluble nitrogen and phosphorus excretion of exotic freshwater mussels (Dreissena spp.): potential impacts for nutrient remineralisation in western Lake Erie. Freshwater Biology, 50, 1146–1162. Crowl, T. A. & Covich, A. P. (1990). Predatorinduced life-history shifts in a fresh-water snail. Science, 247, 949–951. Cyr, H. & Pace, M. L. (1993). Allometric theory: extrapolations from individuals to communities. Ecology, 74, 1234–1245. Dahl, J. & Peckarsky, B. L. (2002). Induced morphological defenses in the wild: predator effects on a mayfly, Drunella coloradensis. Ecology, 83, 1620–1634. Dodson, S. I. (1974). Zooplankton competition and predation: an experimental test of the size-efficiency hypothesis. Ecology, 55, 605–613. Elser, J. J. & Urabe, J. (1999). The stoichiometry of consumer-driven nutrient recycling: theory, observation and consequences. Ecology, 80, 735–751. Elser, J. J., Elser, M. M., McKay, N. A. & Carpenter, S. R. (1988). Zooplankton mediated
transitions between N and P limited algal growth. Limnology and Oceanography, 33, 1–14. Elser, J. J., Dobberfuhl, D. R., MacKay, N. A. & Schampel, J. H. (1996). Organism size, life history, and N:P stoichiometry: toward a unified view of cellular and ecosystem processes. Bioscience, 46, 674–684. Elser, J. J., Chrzanowski, T. H., Sterner, R. W. & Mills, K. H. (1998). Stoichiometric constraints on food-web dynamics: a wholelake experiment on the Canadian Shield. Ecosystems, 1, 120–136. Feller, R. J. & Warwick, R. M. (1988). Energetics. In Introduction to the Study of Meiofauna, ed. R. P. Higgins and H. Thiel. Washington, DC: Smithsonian Institution Press, pp. 181–196. Fukuhara, H. & Yasuda, K. (1989). Ammonium excretion by some freshwater zoobenthos from a eutrophic lake. Hydrobiologia, 173, 1–8. Gaedke, U. (1992). The size distribution of plankton biomass in a large lake and its seasonal variability. Limnology and Oceanography, 37, 1202–1220. Gardner, W. S. & Scavia, D. (1981). Kinetic examination of nitrogen release by zooplankters. Limnology and Oceanography 26, 801–810. Gende, S. M., Edwards, R. T., Willson, M. F. & Wipfli, M. S. (2002). Pacific salmon in aquatic and terrestrial ecosystems. Bioscience, 52, 917–928. Gido, K. B. (2002). Interspecific comparisons and the potential importance of nutrient excretion by benthic fishes in a large reservoir. Transactions of the American Fisheries Society, 131, 260–270. Gillooly, J. F., Brown, J. H., West, G. B., Savage, V. M. & Charnov, E. L. (2001). Effects of size and temperature on metabolic rate. Science, 293, 2248–2251. Grimm, N. B. (1988). Role of macroinvertebrates in nitrogen dynamics of a desert stream. Ecology, 69, 1884–1893.
BODY SIZE AND NUTRIENT CYCLING
Hall, R. O., Tank, J. L. & Dybdahl, M. F. (2003). Exotic snails dominate nitrogen and carbon cycling in a highly productive stream. Frontiers in Ecology and the Environment, 1, 407–411. Hanson, J. M., Prepas, E. E. & Mackay, W. C. (1989). Size distribution of macroinvertebrate community in a freshwater lake. Canadian Journal of Fisheries and Aquatic Sciences, 46, 1510–1519. Henry, R. L. (1985). The impact of zooplankton size structure on phosphorus cycling in field enclosures. Hydrobiologia, 120, 3–9. Hjerne, O. & Hansson, S. (2002). The role of fish and fisheries in Baltic Sea nutrient dynamics. Limnology and Oceanography, 47, 1023–1032. Horppila, J. (1998). Effects of mass removal and variable recruitment on nutrient excretion by a planktivorous roach stock. Journal of Fish Biology, 52, 951–961. Huxley, J. S. (1932). Problems of Relative Growth, London: Methuen. Jackson, J. B. C., Kirby, M. X., Berger, W. H. et al. (2001). Historical overfishing and the recent collapse of coastal ecosystems. Science, 293, 629–638. Jetz, W., Carbone, C., Fulford, J. & Brown, J. H. (2004). The scaling of animal space use. Science, 306, 266–268. Jones, C. G. & Lawton, J. H. (1995). Linking Species and Ecosystems. New York: Chapman & Hall. Kitchell, J. F., O’Neil, R. V., Webb, D. et al. (1979). Consumer regulation of nutrient cycling. Bioscience, 29, 28–34. Koch, B. J. (2005). Invertebrate-mediated nitrogen cycling in three connected aquatic ecosystems, M. S. thesis, Laramie: University of Wyoming, p. 54. Kraft, C. E. (1993). Phosphorus regeneration by Lake Michigan Alewives in the mid-1970s. Transactions of the American Fisheries Society, 122, 749–755.
Li, K. T., Wetterer, J. K. & Hairston, N. G. (1985). Fish size, visual resolution, and prey selectivity. Ecology, 66, 1729–1735. Lively, C. M. (1986). Competition, comparative life histories, and maintenance of shell dimorphism in a barnacle. Ecology, 67, 858–864. Mercier, V., Vis, C., Morin, A. & Hudon, C. (1999). Patterns in invertebrate and periphyton size distributions from navigation buoys in the St. Lawrence River. Hydrobiologia, 394, 83–91. Meyer, J. L., Schultz, E. T. & Helfman, G. S. (1983). Fish schools: an asset to corals. Science, 220, 1047–1049. Morin, A. & Nadon, D. (1991). Size distribution of epilithic lotic invertebrates and implications for community metabolism. Journal of the North American Benthological Society, 10, 300–308. Myers, R. & Worm, B. (2003). Rapid worldwide depletion of predatory fish communities. Nature, 423, 280–283. Nachtigall, W. (1977). Swimming mechanics and energetics of locomotion in variously sized water beetles-Dytiscidae, body length 2 to 35 mm. In Scale Effects in Animal Locomotion, ed. T. J. Pedley. London: Academic Press, pp. 269–283. Pauly, D., Christensen, V., Dalsgaard, J., Froese, R. & Torres, F., Jr. (1998). Fishing down marine food webs. Science, 279, 860–863. Peckarsky, B. L., McIntosh, A. R., Taylor, B. W. & Dahl, J. (2002). Predator chemicals induce changes in mayfly life history traits: a wholestream manipulation. Ecology, 83, 612–618. Peters, R. H. (1983). The Ecological Implications of Body Size. Cambridge: Cambridge University Press. Quinn, T. P. & Kinnison, M. T. (1999). Sizeselective and sex-selective predation by brown bears on sockeye salmon. Oecologia, 121, 273–282. Ramcharan, C. W., France, R. L. & McQueen, D. J. (1996). Multiple effects of planktivorous fish
303
304
R. O. HALL ET AL.
on algae through a pelagic trophic cascade. Canadian Journal of Fisheries and Aquatic Sciences, 53, 2819–2828. Ramsay, P. M., Rundle, S. D., Attrill, M. J. et al. (1997). A rapid method for estimating biomass size spectra of benthic metazoan communities. Canadian Journal of Fisheries and Aquatic Sciences, 54, 1716–1724. Rasmussen, J. B. (1993). Patterns in the size structure of littoral zone macroinvertebrate communities. Canadian Journal of Fisheries and Aquatic Sciences, 50, 2192–2207. Reinertsen, H. A., Jensen, A., Langeland, A. & Olsen, Y. (1986). Algal competition for phosphorus: the influence of zooplankton and fish. Canadian Journal of Fisheries and Aquatic Sciences, 43, 1135–1141. Robson, B. J., Barmuta, L. A. & Fairweather, P. G. (2005). Methodological and conceptual issues in the search for relationship between animal body-size distributions and benthic habitat architecture. Marine and Freshwater Research, 56, 1–11. Roy, K., Collins, A. G., Becker, B. J., Begovic, E. & Engle, J. M. (2003). Anthropogenic impacts and historical decline in body size of rocky intertidal gastropods in southern California. Ecology Letters, 6, 205–211. Schaus, M. H., Vanni, M. J., Wissing, T. E. et al. (1997). Nitrogen and phosphorus excretion by detritivorous gizzard shad in a reservoir ecosystem. Limnology and Oceanography, 42, 1386–1397. Schindler, D. E. & Eby, L. A. (1997). Stoichiometry of fishes and their prey: implications for nutrient recycling. Ecology, 78, 1816–1831. Schmid, P. E., Tokeshi, M. & Schmid-Araya, J. M. (2002). Scaling in stream communities. Proceedings of the Royal Society of London B, 269, 2587–2594. Simon, K. S., Townsend, C. R., Biggs, B. J. F., Bowden, W. B. & Frew, R. D. (2004). Habitat-specific nitrogen dynamics in New Zealand streams containing native and invasive fish. Ecosystems, 7, 777–792.
Stead, T. K., Schmid-Araya, J. M., Schmid, P. E. & Hildrew, A. G. (2005). The distribution of body size in a stream community: one system, many patterns. Journal of Animal Ecology, 74, 475–487. Sterner, R. W. (1990). The ratio of nitrogen to phosphorus resupplied by herbivores – zooplankton and the algal competitive arena. American Naturalist, 136, 209–229. Sterner, R. W. & Elser, J. J. (2002). Ecological Stoichiometry. Princeton: Princeton University Press. Stibor, H. (1992). Predator induced life-history shifts in a fresh-water cladoceran. Oecologia, 92, 162–165. Tarvainen, M., Sarvala, J. & Helminen, H. (2002). The role of phosphorus release by roach Rutilus rutilus (L.) in the water quality changes of a biomanipulated lake. Freshwater Biology, 47, 2325–2336. Teal, J. M. (1962). Energy flow in the salt marsh ecosystem of Georgia. Ecology, 43, 614–649. Thomas, S. A., Royer, T. V., Minshall, G. W. & Snyder, E. (2003). Assessing the role of marine derived nutrients in Idaho streams. In Nutrients in Salmonid Ecosystems: Sustaining Productivity and Biodiversity, ed. J. G. Stockner. Bethesda, Maryland: American Fisheries Society, pp. 41–55. Tollrian, R. (1995). Predator-induced morphological defenses: costs, life history shifts, and maternal effects in Daphnia pulex. Ecology, 76, 1691–1705. Trippel, E. A. (1995). Age at maturity as a stress indicator in fisheries. Bioscience, 45, 759–771. Vadeboncoeur, Y., Vander Zanden, M. J. & Lodge, D. M. (2002). Putting the lake back together: reintegrating benthic pathways into lake food web models. Bioscience, 52, 44–54. Vanni, M. J. (1987). Effects of nutrients and zooplankton size on the structure of a phytoplankton community. Ecology, 68, 624–635.
BODY SIZE AND NUTRIENT CYCLING
Vanni, M. J. (2002). Nutrient cycling by animals in freshwater ecosystems. Annual Review of Ecology and Systematics, 33, 341–370. Vanni, M. J. & Findlay, D. L. (1990). Trophic cascades and phytoplankton community structure. Ecology, 71, 921–937. Vanni, M. J., Flecker, A. S., Hood, J. M. & Headworth, J. L. (2002). Stoichiometry of nutrient recycling by vertebrates in a tropical stream: linking species idenitity and ecosystem processes. Ecology Letters, 5, 285–293. Ward, P. & Myers, R. A. (2005). Shifts in openocean fish communities coinciding with the commencement of commercial fishing. Ecology, 86, 835–847. Weihs, D. (1977). Effects of size on sustained swimming speeds of aquatic organisms. In Scale Effects in Animal Locomotion, ed. T. J. Pedley. London: Academic Press, pp. 299–313.
Welcomme, R. L. (1999). A review of a model for qualitative evaluation of exploitation levels in multi-species fisheries. Fisheries Management and Ecology, 6, 1–19. Wen, Y. H. & Peters, R. H. (1994). Empirical models of phosphorus and nitrogenexcretion rates by zooplankton. Limnology and Oceanography, 39, 1669–1679. Winemiller, K. & Jepsen, D. B. (2004). Migratory neotropical fish subsidize food webs of oligotrophic blackwater rivers. In Food Webs at the Landscape Level, ed. G. A. Polis, M. E. Power and G. R. Huxel. Chicago: University of Chicago Press, pp. 115–132. Wootton, J. T. (1994). Predicting direct and indirect effects: an integrated approach using experiments and path analysis. Ecology, 75, 151–165. Zhuang, S. (2005). The influence of body size and water temperature on metabolism and energy budget in Laternula marilina Reeve. Aquaculture Research, 36, 768–775.
305
CHAPTER SIXTEEN
Body sizes in food chains of animal predators and parasites JOEL E . COHEN Rockefeller and Columbia Universities, New York
Introduction Food chains in which animal predators are bigger than their animal prey are called predator chains; those in which the consumers are smaller are called parasite chains (Elton, 1927; Hutchinson, 1959, p. 147). The purpose of this chapter is to display and test empirically some consequences, for predator chains and parasite chains, of assuming that the average mass of a consumer species (predator or parasite) is related to the average mass of its animal resource species (prey or host) by a power law with an exponent less than 1. In 1858, as part of his development of the theory of evolution, Wallace (1858, p. 54) noted that animal predators are generally larger and less numerous than their prey. Among the many echoes of Wallace’s remark, Elton (1927) observed anecdotally that animal predators weigh more than their prey in terrestrial food chains, Hutchinson (1959) analyzed some of the theoretical consequences of predators weighing more than their prey, and Sheldon, Prakash and Sutcliffe (1972) and others posited that marine animal predators outweigh their marine animal prey (see also Humphries, this volume; Woodward & Warren, this volume). Only recently have body sizes been studied empirically in parasite chains (Memmott, Martinez & Cohen, 2000; Leaper & Huxham, 2002) and parasitoid chains (Cohen et al., 2005). The study of parasitoid chains (e.g. Rott & Godfray, 2000; Memmott et al., 2000) appears not to have been considered by Elton (1927) and Hutchinson (1959). Predator and parasite chains are not the only possibilities observed in nature. Other relations between mass and feeding arise from social hunting and metaphoetesis. Among animals that hunt socially (such as wolves and army ants), the aggregate mass of the hunting group may be a more appropriate index of size than the mass of an individual predator. In animals where mass or feeding habit or both change dramatically with the stage of the life cycle (as in many insects and fishes), it is misleading to represent the masses of all stages by a typical adult Body Size: The Structure and Function of Aquatic Ecosystems, eds. Alan G. Hildrew, David G. Raffaelli and Ronni Edmonds-Brown. Published by Cambridge University Press. # British Ecological Society 2007.
BODY SIZES IN FOOD CHAINS
mass (Hutchinson, 1959; Cohen et al., 2005). Hutchinson (1959) called a change in diet with changing stage of the life cycle metaphoetesis. When Christine Mu¨ller measured the body lengths of individual aphid hosts and of their parasitoid wasp consumers (Cohen et al., 2005, raw data available online), one nymphal aphid Capitophorus carduinis parasitized by a male wasp Aphidius matricariae was longer than the wasp when the wasp emerged from its aphid host, while another nymphal aphid Capitophorus carduinis parasitized by a male wasp Aphidius matricariae was shorter than the wasp when the wasp emerged from its aphid host. Both aphids were found on the same plant species Cirsium palustre. Even controlling for the life stage of the aphid, for the sex of the parasitoid wasp, and for the plant, the ordering of body sizes may differ from one consumer–resource pair of individuals to another. This single instance is illustrative of the variability in the relationship of host and parasitoid body length found in many comparable observations. In most real food webs, predator chains and parasite chains are tightly interwoven because animal predators of almost all sizes support viruses, bacteria and other microscopic and macroscopic parasites. The analysis here of the typical patterns of predator chains and parasite chains makes no claim to cover all kinds of food chains. The relation between the average masses of animal predator species and the average masses of animal prey species has been approximated empirically as a power law (Schoener, 1968; Peters, 1983, p. 277; Ve´zina, 1985; Warren & Lawton, 1987). The power law also applies to individual body masses of aphids and parasitoid wasps (Cohen et al., 2005), but the theoretical implications parallel to those derived here for species-average masses remain to be studied. The analysis below predicts, first, that in predator chains, there is an upper limit to the mass of possible predators and prey; and that this upper limit is independent of the number of trophic links in the chain and independent of the mass of the smallest prey. Conversely, in parasite chains, there is a lower limit to the mass of the smallest host and parasite; this limit is independent of the number of trophic links in the chain and independent of the mass of the largest host. A second consequence is that, in a predator chain, the ratio of predator mass to prey mass decreases according to a power law, with an exponent one less than that for predators and prey masses, as the trophic level and the mass of the prey increase. (In a single food chain in which no species occurs more than once, the trophic level of a species may be unambiguously defined as the number of links between it and the basal species in the chain; thus the basal species has trophic level 0, its consumer has trophic level 1, and the top species in a chain of n trophic links and n þ 1 species has trophic level n.) Conversely, in a parasite chain, the ratio of parasite mass to host mass increases as the trophic level of the host increases (and the mass of the host decreases).
307
308
J. E. COHEN
This theoretical and exploratory analysis of body sizes and feeding is part of a larger picture that includes numerical abundance (Cohen, Jonsson & Carpenter, 2003).
Theory Maximal and minimal body masses Consider a food chain based on an animal (prey or host) of mass M0. Let M1 ¼ f (M0) be the typical (e.g. geometric mean) mass of a consumer (predator or parasite) of that prey, ignoring variation in the mass of consumers that eat prey of a given mass. Let M2 ¼ f (M1) ¼ f ( f (M0)) ¼ f 2(M0) be the typical mass of a consumer that eats the consumer of typical mass M1. The notation M2 ¼ f 2(M0) indicates that M2 results from applying two iterations of f to M0; f 2(M0) does not denote the square of f (M0), which would be written [ f (M0)]2. Similarly, f nþ 1(M0) ¼ f ( f n(M0)) is the typical mass of a consumer n þ 1 trophic links above the basal animal of mass M0. When the typical mass Y of predators on animal prey of mass X is a power function Y ¼ f ðXÞ ¼ AXB ; A > 0;
(16:1) b
then by induction (letting ^ denote exponentiation so that a^b means a ) ! n1 X n n n Bn ^ m f ðM0 Þ ¼ M0 ½A B ¼ M0B Að1B Þ=ð1BÞ : (16:2) m¼0
The equality on the left of Eq. (16.2) is valid for any B. The equality on the right of n1 P m B ¼ ð1 Bn Þ=ð1 BÞ. Were the Eq. (16.2) is valid when B 6¼ 1, since then m¼0
consumer’s mass directly proportional to the resource’s mass according to Y ¼ AX, i.e. were B ¼ 1, then the mass of the consumer species would change by a factor of A with each additional link in the food chain and then f n(M0) ¼ AnM0. Were B ¼ 0, the mass of consumers would be constant and equal to A, regardless of their position in a food chain. Assume henceforth that 0 < B < 1, in addition to the previous assumption that A > 0. Then according to Eq. (16.1), consumer and resource would have equal body mass X ¼ f(X) when X ¼ A1/(1 B), and this mass is strictly positive. This positivity guarantees that the intersection of the power law Eq. (16.1) with the diagonal line where Y ¼ X lies in the positive quadrant. In this model, a chain is a predator chain or a parasite chain according to whether M0 < A1/(1 B) or M0 > A1/(1 B). With increasing trophic level, the masses of successive consumers approach the finite limit A1/(1 B) > 0 (Fig. 16.1a) because the assumption 0 < B < 1 implies Bn # 0 as n"1 and hence lim f n ðM0 Þ ¼ A1=ð1BÞ : n"1
(16:3)
BODY SIZES IN FOOD CHAINS
The limit Eq. (16.3) is a maximum if each consumer has a bigger mass than its resource, as assumed in a predator chain. The limit Eq. (16.3) is a minimum if each consumer has a smaller mass than its resource, as assumed in a parasite chain. The predicted maximal mass of a top predator is independent both of the number of links leading up to that predator and of the mass M0 of the basal animal prey in the food chain. The predicted minimal mass of a parasite is independent both of the trophic level of that parasite and of the mass M0 of the basal animal host in the food chain. The limit A1/(1 B) is very sensitive to the values estimated for A and B. As B"1, 1/(1 B) A "1. The values of A and B of course are not known exactly. They are usually estimated by a least-squares fit of the coefficients of the linear relation y ¼ a þ bx where y ¼ log10 Y and x ¼ log10 X. The parameters are connected by A ¼ 10a but B ¼ b. For a given value of A, the closer B is to 1, the slower the approach to the limiting size as one proceeds along a food chain from successive resource to successive consumer (Mark Huxham, personal communication, 9 September 1995). So the limit A1/(1 B) may not be closely approached in reality when there are other limitations on food chain length. According to this model of species-average body mass in food chains, in very long chains, the predators are mostly big, close in mass to the limiting maximum, and the parasites are mostly small, close in mass to the limiting minimum (Fig. 16.1a). The removal from a predator chain of top predators shifts the size distribution of species-average body masses from one concentrated near the upper maximum to a more widely spaced distribution across the lower portions of the possible range of average body masses. This prediction could be compared with quantitative data on the body size distributions of North American vertebrate species before and after the major extinction of the megafauna and with quantitative data on the body size distributions of marine fauna before and after widespread industrial fishing. This allometric model of species-average body masses has an implication for predator–parasite cycles. Assume that Mt þ 1 ¼ AMtB along a predator chain of n links, t ¼ 0, . . ., n 1, that the top predator is the starting point for a parasite chain of n links, i.e. V0 ¼ Mn > a1/(1 b) and Vtþ 1 ¼ aVtb, t ¼ 0, . . ., n 1, with a > 0, 0 < b < 1 along the parasite chain. Then it turns out that M0 can be less than, equal to, or greater than Vn. More generally, dropping the assumption that the predator chain and the parasite chain are of equal lengths, it is still possible for M0 to be less than, equal to, or greater than Vn, as long as each chain is sufficiently long. The case where Vn ¼ M0, i.e. where the basal prey of the predator chain weighs the same as the top parasite of the parasite chain, is illustrated numerically in Fig. 16.1b. In this case, if the basal prey and the top parasite were the same
309
J. E. COHEN
(a) 100 predator food chain 10 consumer weight
310
1
0.1 parasite food chain 0.01
0.001 0.001
0.01
0.1 1 resource weight
10
100
Figure 16.1 (a) Theoretical progression of body masses along a predator chain and a parasite chain. Arrows go from resource to consumer. Body masses are on logarithmic scales. The power-law relation Eq. (16.1) appears as a straight dashed line with slope B; here, B ¼ 1/2 for both predator chain (A ¼ 10) and parasite chain (A ¼ 0.1). On the solid diagonal line, consumer mass equals resource mass. Predator chains appear above the diagonal; parasite chains appear below the diagonal. In the predator chain, an arbitrary small basal prey mass is chosen (in this example, M0 ¼ 0.001) and the corresponding predator mass M1 is found by moving upward to the upper dash line. This predator is the prey of the predator at the next trophic level of the chain. The mass M1 is located on the abscissa by moving horizontally right to the diagonal line. Then vertical upward motion to the upper dashed straight line gives the mass M2 of the predator two links above the basal prey. Alternating horizontal right and vertical upward motions intersect the power function (upper dashed straight line) at the masses of successively higher predators in the predator chain. All such trajectories converge where the upper dashed line and the diagonal line intersect. In the parasite chain, an arbitrary large basal host mass is chosen (in this example, M0 ¼ 100) and the corresponding parasite mass M1 is found by moving down to the diagonal line. This parasite is the host of the parasite at the next trophic level of the chain. The mass M1 is located on the abscissa by moving horizontally left to the diagonal line. Then moving down to the lower dashed straight line gives the mass M2 of the parasite two links above the basal host. All such trajectories converge where the lower dashed line and the diagonal line intersect. (b) Theoretical masses of species in a predator chain (open bars) and in a parasite chain (filled bars) where the top predator is the basal host, and the top parasite has the same mass as the basal prey. In the predator chain, Mnþ1 ¼ 10 Mn0.5 and the upper limit of mass is 100. In the parasite chain, Vnþ1 ¼ Vn0.5 and the lower limit of mass is 1. M0 ¼ V10 ¼ 1.004503 and M10 ¼ V0 ¼ 99.55172.
BODY SIZES IN FOOD CHAINS
(b)
100 90 80
predator chain
body weight
70 60 50 40 30 20 parasite chain
10 0 0
1
2 3 4 5 6 7 8 9 trophic level in predator chain
10
Figure 16.1 (cont.)
species, the predator chain and the parasite chain would be linked in a predator– prey cycle. On the linear scale of mass used in Fig. 16.1b, after the first few trophic levels in both the predator chain and the parasite chain, the consumers are near in mass to the limiting mass.
Predicted value of the exponent The exponent B may be computed exactly for simple models of the distribution of the pairs (x, y), where x ¼ log10 X and y ¼ log10 Y are log prey (or host) mass and log predator (or parasite) mass, respectively. Suppose xmin is the log10 minimal observed species average mass and xmax is the log10 maximal species average mass. The previous theory predicts that xmax ¼ log10(A1/(1 B)) but the following calculations hold whether or not that prediction is true. The slope of any linear relation between y and x will be unaffected if both x and y are replaced by the identical linear transformation of x and y, so no generality is lost by assuming that xmin ¼ 0 and xmax ¼ 1. Then each trophic link from resource to consumer may be represented by a dot in a square in the (x, y) plane with lower left corner at the origin (0, 0) and upper right corner at (1, 1). The diagonal of the square is the locus of points where consumer body mass equals resource body mass. Suppose that trophic links are uniformly and independently distributed over this square, and that all links above the diagonal are in predator chains and all links below the diagonal are in parasite chains. Then, in a predator chain, for a given x (between 0 and 1), the expected y is halfway between the diagonal and the upper horizontal edge of the square, that is, E(y|x) ¼ x þ (1/2)(1 x) ¼ 1/2 þ x/2. Thus the slope of average y as a linear function
311
312
J. E. COHEN
of x is predicted to be B ¼ 1/2. Similarly, in a parasite chain, for a given x (between 0 and 1), the expected y is halfway between the diagonal and the lower horizontal edge of the square, that is, E(y|x) ¼ x/2. The slope of average y as a linear function of x is again predicted to be B ¼ 1/2. In the part of this model that pertains to a predator chain, the assumption that each trophic link is uniformly and independently distributed in the triangle above the diagonal follows from the cascade model (Cohen, Briand & Newman, 1990) in the continuous limit (possibly with additional assumptions) of many species of predators and prey. The cascade model assumes that all species are ordered by something interpreted here as body mass, and that each consumer species consumes with equal probability and independently every species smaller than it. (The cascade model does not attempt to describe parasite chains because it was intended to account for food-web data that ignored parasites.) The continuous limit of the joint distribution of prey-to-predator links posited in the cascade model is (possibly with additional assumptions) a twodimensional distribution of trophic links that is uniform in the upper triangle above the diagonal from (0, 0) to (1, 1) in the unit square in the plane where x ¼ log10 X and y ¼ log10 Y, as supposed in the previous paragraph. For parasite chains, to derive a distribution of trophic links in the (x, y) plane that is uniform over the triangle below the diagonal, as supposed above, all that is required is to reverse the ordering by body size in the argument just given for predator chains. When the pairs (x, y) are not distributed uniformly but lie in a band parallel to the diagonal, the predicted slope B will move from 1/2 toward 1. Such a band parallel to the diagonal would arise if there were a nearly constant ratio of average body mass between consumer (predator or parasite) and resource. When the pairs (x, y) lie in a band parallel to the x axis (because most predator species are roughly the same size, or most parasite species are roughly the same size), the predicted slope B will move from 1/2 toward 0.
Ratios and differences of consumer mass and resource mass Let R ¼ Y/X ¼ AXB 1 be the ratio of consumer mass to resource mass in a single trophic link. Then R is a decreasing power-law function of X. The exponent B 1 is negative because B < 1. A regression of log R on log X is predicted to have a slope exactly one less than the slope of a regression of log Y on log X, for the same set of data. The ratio R decreases (towards a limit of 1) with increasing trophic level of the prey in predator chains. In parasite chains, because body masses decrease with increasing trophic level, the ratio R increases towards a limit of 1 with increasing trophic level of the host. The difference in masses behaves in a more complex way than the ratio of masses, as the following analysis shows. Let D ¼ Y X ¼ (R 1)X be the difference between the consumer mass Y and the resource mass X in a single
BODY SIZES IN FOOD CHAINS
trophic link. (In predator chains, D > 0. In parasite chains, D < 0.) Because dD/dX ¼ RB 1, the difference D increases with increasing resource mass X if and only if R > 1/B. The smaller B is, the bigger R must be for D to increase with X. As B < 1, a necessary condition for D to increase with X is that R > 1, and this happens only in predator chains. Thus, in predator chains, the difference D in mass between consumer and resource may increase with increasing trophic position (if initially R > 1/B); but once R 1/B, the difference D will thereafter decrease (towards a limit of 0) with increasing trophic position. By contrast, in parasite chains, where R < 1 and B < 1, it follows that RB 1 < 0 always; hence with increasing trophic level (and therefore decreasing body mass), D is always increasing (from negative values towards a limit of 0), that is, host mass minus parasite mass is always positive and decreases towards a limit of 0.
Data The data presented here deal only with food webs (cross-linked food chains), rather than with isolated food chains. The theory is relevant to these food webs in so far as food chains are a first approximation to more complex food webs. First, two examples of data on the masses of animal predators and their animal prey in a particular community will be analyzed. Then some data will be examined from literature surveys of pooled communities of specified habitat types (terrestrial and coastal). A recent database of the masses of consumers and resources (Brose et al., 2005) has been analyzed by Brose et al. (2006).
Studies of a well-defined community Menge et al. (1986) described the food web and the masses of the animals of a tropical Panamanian rocky intertidal community. From 31 data points (Fig. 16.2a), hand-read in part from their published graphs, linear regression of log10 masses yielded a ¼ 2.2334 (with 95% confidence interval (1.80, 2.67)), and b ¼ 0.4819 (with 95% confidence interval (0.19, 1.15)). The geometric mean mass Y (kg) of animal predators on animal prey of mass X would be estimated from these data as Y ¼ 0.1712X0.4819 and the upper limit in mass A1/(1 B) for the largest predator would be nearly 20.4 kg. The largest observed predator in the data weighed just under 2 kg. The 95% confidence interval for B includes both 0 and 1. If the data satisfy the assumptions of the underlying regression model well enough to justify the conclusion that the asserted confidence interval really has probability 95%, then these data do not specify an allometric relation with sufficient precision to have the predictive upper limit falsified by any finite maximal predator mass. A simple sensitivity calculation, referred to below as ‘the 10% sensitivity range,’ confirms a wide range of uncertainty in the upper limit. If the regression intercept log A and the regression slope B are both replaced by 90% of their estimated values, the maximal predator mass A1/(1 B) is 3.5 kg. If the regression
313
J. E. COHEN
(a) 3.5 3 log10 weight (g) of predator
314
2.5 2 1.5 1 0.5 0 –0.5 –1 –1
–0.5
0
0.5
1
1.5
2
log10 weight (g) of prey
Figure 16.2 (a) Predator masses and prey masses in trophic links of a tropical rocky intertidal food web described by Menge et al. (1986). The solid diagonal line indicates where predator and prey masses are equal; all but one of the trophic links fall above this line. The upper dashed line is the regression line: log10(predator mass, g) ¼ 2.2334 þ 0.4819 log10(prey mass, g). The regression line is obtained by ordinary least squares using the log-transformed masses; the standard error of the slope coefficient is 0.3268. The data are read from Menge et al. (1986); the analysis and figure are original. (b, c) Prey and predator body mass (kg) in Tuesday Lake in (b) 1984 and (c) 1986, one marker for every trophic link in the unlumped food web. Cannibalistic links are excluded. Dotted line indicates equal prey and predator body mass. The links are coded according to the prey: circles ¼ phytoplankton, squares ¼ zooplankton, stars ¼ fish. For 1984, the regression coefficients were a ¼ 1.5598, b ¼ 0.8445, with correlation r ¼ 0.7859 and 263 trophic links (Jonsson et al., 2005, p. 34). For 1986, the regression coefficients were a ¼ 1.4108 and b ¼ 0.5928, with correlation r ¼ 0.6094 and 233 trophic links. (d) Regression lines from (b, solid diamonds) and (c, open squares) plotted over the approximate range from the mass of the smallest observed organism to the mass where predator and prey are equal. Figure 16.2b is reprinted from Cohen et al. (2003) with permission from the National Academy of Sciences. Figure 16.2c is reprinted from Jonsson et al. (2005), copyright 2005 by T. Jonsson, J. E. Cohen, S. R. Carpenter. Figure 16.2d is original.
intercept log A and the regression slope B are both replaced by 110% of their estimated values, the maximal predator mass is 169.2 kg. Combining 90% of log A with 110% of B and vice versa yields a narrower range of uncertainty from 19.0–21.8 kg. When a plausible range of the predicted maximal size is as large as the 10% sensitivity range, from 3.5–169.2 kg, only order-of-magnitude agreement between predictions and observations should be expected, at best. If seals or sea lions are occasionally part of the rocky intertidal community, the average body mass of those consumers could be compared with the limit
BODY SIZES IN FOOD CHAINS
(b)
10−2
predator body mass (kg)
10−4 10−6 10−8 10−10 10−12 10−14 −14 10
10−12
10−10 10−8 10−6 prey body mass (kg)
10−4
10−2
(c)
100
predator body mass (kg)
10−2 10−4 10−6 10−8 10−10 10−12 10−14 −14 10
10−12
10−10 10−8 10−6 10−4 prey body mass (kg)
10−2
100
Figure 16.2 (cont.)
predicted here. For example, adult Gala´pagos fur seal females Arctocephalus galapagoensis average about 28 kg in body mass (Horning & Trillmich, 1997). Gala´pagos fur seal bulls average about 70 kg in body mass (http://www.tamug. tamu.edu/labb/Galapagos/GFSwork/GFS_work.htm, accessed 27 August 2005). Cohen et al. (2003), Reuman and Cohen (2004), Jonsson, Cohen and Carpenter (2005), Reuman and Cohen (2005), and Cohen and Carpenter (2005) analyzed the community food web, the numerical abundance and the average body size of species in the pelagic community of a small lake, Tuesday Lake, in Michigan. The raw data on the food web, average body mass and numerical abundance by
315
J. E. COHEN
(d) 15 10 predator body mass (log10 kg)
316
5 0 –5 –10 –15 –10 –14
–9
–4 1 prey body mass (log10 kg)
6
11
Figure 16.2 (cont.)
species are given by Jonsson et al. (2005). Tuesday Lake was intensively sampled during the summers of 1984 and 1986. During the summer of 1985, the three resident species of fishes were largely removed and replaced by a fourth fish species, which consumed the remaining individuals of the original three fish species. In addition to the complete turnover of the fish species present in Tuesday Lake between 1984 and 1986, the other species in the lake also changed dramatically between 1984 and 1986 (Jonsson et al., 2005, p. 23). Almost all predators had larger average body mass than their prey in 1984 (Fig. 16.2b) and 1986 (Fig. 16.2c). The calculated upper limit in 1984 exceeds 10.7 109 kg, far in excess of the largest average species mass observed in 1984, namely, 1.29 103 kg, or 1.29 g. The upper limit in 1986, 0.34 103 kg, or 0.34 g, was exceeded by the average body mass, 1.95 101 kg, of the largest species, the introduced fish Micropterus salmoides. The average body mass of this fish also exceeded the upper limit of the 10% sensitivity range. The allometric model of the relationship between predator and prey mass was probably less adequate in 1986 than in 1984: the correlation (on log-log scales) between predator and prey masses dropped notably from 1984 to 1986. Following the complete manipulation of the fish fauna in 1985, the pelagic community may have been observed in 1986 during a transient response to the manipulation. The regression lines before and after the manipulation appear in Fig. 16.2d. While the predator–prey pairs of Cohen et al. (1993) could reasonably be seen as uniformly distributed in the upper triangle of the square in the (x, y) plane, these trophic links were pooled from a variety of different communities. The
BODY SIZES IN FOOD CHAINS
predator–prey pairs from Tuesday Lake (Figs. 16.2b, c) and some other individual communities indicate that the pairs (x, y) may sometimes lie in a band parallel to the diagonal (Reuman & Cohen, 2004, p. 857). The non-parasite data from Ythan estuary (Leaper & Huxham, 2002, p. 447) seem to be intermediate between lying in an upper triangle and lying in a band parallel to the diagonal; they are distributed in a wedge shape.
Studies that pool multiple communities Numerous studies have collected masses of organisms in defined taxonomic groups from scattered published sources and identified trophic links based on reports of feeding habits, without reference to whether the organisms would be likely to occur within a single habitat at any single time (e.g. Schoener, 1968; Peters, 1983; Ve´zina, 1985; Hansen, Bjı´rnsen & Hansen, 1994). Other studies have combined community studies and literature surveys (e.g. Warren & Lawton, 1987; Cohen et al., 1993; Jonsson & Ebenman, 1998a; Brose et al., 2006). Cohen et al. (1993) presented two independently collected sets of data on the sizes of animal predators and prey in multiple community food webs. Data set A gave average adult masses of predators and prey in 354 trophic links from 18 community food webs. Data set B gave lengths of prey and predators in 478 trophic links from 30 webs of a compendium of sink, source and community webs. In roughly 90% of the trophic links reported in food webs from terrestrial, coastal, freshwater and marine habitats, the body mass of an animal predator exceeded that of its animal prey. Figure 16.3a compares the estimated regression lines of 109 trophic links from coastal webs in data set A, ten trophic links from coastal webs in data set B, and 31 trophic links from Menge et al. (1986). All three regression slopes are less than 1/2. The predicted largest predator from coastal data sets A and B would weigh, respectively, 0.4 kg (10% sensitivity range 0.2–0.8 kg) and 54.0 kg (10% sensitivity range 11.6–293 kg). Figure 16.3b compares the estimated regression lines of 48 trophic links from terrestrial webs in data set A, 162 trophic links from terrestrial webs in data set B, and a line hand-fitted to graphed terrestrial data from Ve´zina (1985). All three regression slopes are greater than 1/2. The predicted largest predator from terrestrial data set A would weigh 23.8 kg (10% sensitivity range 3.6–247 kg). Because the slope for data set B is so close to 1, the predicted maximal predator mass is meaningless. For the estimates from Ve´zina’s data, the largest predator would weigh 168 kg (10% sensitivity range 13.5–4660 kg). For comparison, the heaviest terrestrial vertebrate predator in Ve´zina’s data is the East African lion (Panthera leo), which weighs 160 kg. The eight empirical regression lines of log predator weight as a function of log prey weight plotted in Figs. 16.2 and 16.3 have slopes ranging from 0.1463 to 0.9443, with median value 0.5489, not far from the predicted value of 1/2.
317
J. E. COHEN
(a) 6
log10 weight (g) of predator
318
4
coastal B
2 Menge 1986
coastal A
0 –2 –4 –6 –6
–4
–2 0 2 log10 weight (g) of prey
4
6
Figure 16.3 (a) Regression lines of predator masses and prey masses in trophic links in three sets of coastal food webs. The regression line for Menge et al. (1986) is taken from Fig. 16.2a. The regression lines for coastal A and coastal B are computed from the statistics of Cohen et al. (1993, p. 71, Table 2). Lengths reported in data set B were converted to masses, assuming spherical geometry, by log10(mass) ¼ log10(p/6) þ 3 log10(length). Although the regression lines for coastal A and coastal B extend to the right of the diagonal line where predator and prey masses are equal, roughly 90% of the data points fell above and to the left of the diagonal. The coastal A regression line is: log10(predator mass, g) ¼ 2.2114 þ 0.1463 log10(prey mass, g). The coastal B regression line is: log10(predator mass, g) ¼ 3.1985 þ 0.3241 log10(prey mass, g). (b) Regression lines of predator masses and prey masses in trophic links in three sets of terrestrial food webs. Procedures of Fig. 16.3a apply here. The terrestrial A regression line is: log10(predator mass, g) ¼ 2.1105 þ 0.5177 log10(prey mass, g). The terrestrial B regression line is: log10(predator mass, g) ¼ 1.9924 þ 0.9443 log10(prey mass, g). Ve´zina (1985) gave numerically only a range of masses for predators and no masses for their prey. Here a single straight line was fitted by hand to Ve´zina’s graph of the data for insectivores, piscivores and carnivores, and the hand-fitted straight line for the data of Ve´zina (1985) is: log10(predator mass, g) ¼ 2.19 þ 0.58 log10(prey mass, g).
Jonsson and Ebenman (1998a) computed the ratio of predator mass to prey mass for all trophic links for which both masses were known using 768 consumer species (697 trophic species) in 52 community food webs. Their finding that, for most food webs, the higher the trophic level of the predator, the smaller the predator–prey mass ratio, agrees with the theory developed here for isolated food chains if higher trophic level correlates closely with higher body mass. It will be desirable to re-examine this asserted pattern using the data of Brose et al. (2005). Data on parasite and parasitoid food webs and body sizes appear to be scarce. Memmott et al. (2000) reported a source food web of a broom community that
BODY SIZES IN FOOD CHAINS
(b) 8 terrestrial B log10 weight (g) of predator
6 4
Vezina 1985
2 0 terrestrial A
–2 –4 –6 –6
–4
–2 0 2 log10 weight (g) of prey
4
6
Figure 16.3 (cont.)
contained one plant species, 19 herbivores, 66 parasitoids, 60 predators, five omnivores and three pathogens. They plotted log consumer-species length as a function of log resource-species length with separate symbols for parasitoids, pathogens and predators (their Fig. 7) but they did not report estimates of allometric relations nor list the length data, though they did provide the foodweb data. According to Leaper and Huxham (2002), the web reported by Memmott et al. (2000) was the first and until 2002 the only published food web to present body masses for both parasites (in fact, they were parasitoids) and other consumers. Leaper and Huxham (2002) calculated but did not publish body masses of 160 of the 171 taxa in the food web of the Ythan estuary: 113 average adult body weights and mean weights for the given life-history stage of the remaining 47 taxa. They reported Pearson’s r2 for log10 consumer and log10 resource body masses for six versions of the web: including non-parasites only, parasites only, and all taxa; and for each group of taxa, with and without distinguishing life stages according to their trophic relations. No regression coefficients of log10 consumer body mass on log10 resource body mass were given. The correlations were positive in all cases and were statistically significantly different from zero at the 0.001 level except for parasites only, disregarding differences among life stages.
Discussion Hutchinson (1959, p. 147) examined ‘the order of magnitude of the diversity that a single food chain can introduce into a community’. It is worthwhile to revisit his influential calculations in the light of data and theory available since he wrote. Hutchinson assumed that ‘in general 20 per cent of the energy passing
319
320
J. E. COHEN
through one [species] can enter the next [species] in the chain’ and that ‘each predator has twice the mass (or 1.26 the linear dimensions) of its prey, which is a very low estimate of the size difference between links . . .’ This assumption may be represented in the model Eq. (16.1) by putting A ¼ 2 and B ¼ 1. This model led Hutchinson (1959, p. 147) to envisage the ‘ultimate predator’ at trophic level 49, with an individual body size ‘vastly greater than the volume of the world ocean’. Hutchinson then implicitly assumed that the numerical abundance N, or population size, of each animal species in a predator food chain equals the total energy available divided by the typical body size M, which is tantamount to assuming that the energy consumption of each animal species is directly proportional to its typical body size M. With each increase in the trophic level of species in a predator chain, according to Hutchinson’s assumptions, 20% as much energy has to be divided among the organisms each twice as big. The population size therefore is reduced by a factor of 0.2/2 ¼ 0.1, i.e. decreases by 90%. Consequently, Hutchinson concluded, the population size or numerical abundance N4 of the fifth animal species will be 10 4 times the population size N0 of the first. In this hypothetical world, food chains cannot be very long. Hutchinson’s assumptions imply an allometric relation between numerical abundance (or population size) and average body mass. Along a trophic link from any species 1 to any species 2, Hutchinson assumes that M2 ¼ 2M1 (mass doubles) while N2 ¼ (1/10)N1 (numerical abundance falls by 90%). The slope of the allometric relation between numerical abundance and body mass is then: D log10 ðNÞ=D log10 ðMÞ ¼ ½log10 ðN2 Þ log10 ðN1 Þ=½log10 ðM2 Þ log10 ðM1 Þ ¼ log10 ð10Þ= log10 ð2Þ ¼ 3:32:
(16:4)
Each step in Hutchinson’s argument has been re-examined. Pauly and Christensen (1995) estimated a mean trophic transfer efficiency of 10% (half Hutchinson’s estimate of 20%). Rather than doubling with each trophic link, animal body size in a predator chain is more likely to be described by Eq. (16.2) with A > 0, 0 < B < 1, neglecting the substantial variability in the size of predators on prey of a given size. Animal metabolic energy requirements increase approximately in proportion to M3/4 rather than to M (Kleiber, 1961). In Tuesday Lake, Michigan, the regression of log10(N) on log10(M) had slope 08413 (with 99% confidence interval 098, 071) in 1984 and slope 07461 (with 99% confidence interval 091, 059) in 1986 (Reuman & Cohen, 2004). These slopes are far from the slope of 332 that follows from Hutchinson’s assumptions. Cohen and Carpenter (2005) showed that the statistical assumptions underlying linear regression were justified for Tuesday Lake data in regressions of log10(N) on log10(M) but not vice versa. If animal population size were constrained by available energy alone, as Hutchinson supposed, and if the food chain were isolated from all other food chains to or from which energy might be diverted, then, in principle, a better
BODY SIZES IN FOOD CHAINS
formula than Hutchinson’s for the ratio of the population size or numerical abundance Nn at trophic level n to the numerical abundance N0 of the basal animal in a predator chain would appear (for the moment) to be: Nn =N0 ¼ ðM0 =Mn Þ3=4 ð0:1Þn
(16:5)
and the slope of the relation between numerical abundance and body mass is predicted by these assumptions to be: D log10 ðNÞ=D log10 ðMÞ ¼ ½log10 ðNnþ1 Þ log10 ðNn Þ=½log10 ðMnþ1 Þ log10 ðMn Þ ¼ 3=4 þ 1= log10 ðMnþ1 =Mn Þ:
(16:6)
For large n, M0/Mn approaches a constant (less than 1) and the ratio Eq. (16.5) declines by a factor of 0.1 with each increase in trophic level. Apparently by coincidence, this is exactly the behaviour Hutchinson calculated. That is the good news. The rest of the news is bad, and gets worse. For small n, Eq. (16.5) predicts a slower-than-exponential decline, unlike Hutchinson’s calculation. For large n, Mnþ 1/Mn ! 1 so log10(Mnþ 1/Mn) ! 0 and the right side of Eq. (16.6) diverges to infinity, clearly an unrealistic prediction. Evidently the assumptions stated just before Eq. (16.5) do not hold in the real world. One weak assumption is that the predator chain is energetically isolated from all other food chains. In addition, the population sizes of species, especially species with small body sizes, are often not limited by energy (Blackburn, Lawton & Pimm, 1993; Blackburn & Lawton, 1994). While large-bodied animal species are usually rare, small-bodied animal species commonly have a wide range of population sizes, from abundant to rare. Overall, Hutchinson’s argument that a predator chain (and by his off-hand extension, a parasite chain) ‘clearly . . . of itself cannot give any great diversity’ founders in the face of more recent facts and models. For three collections of data from coastal communities, 0 < b < 1/2, while for three collections of data from terrestrial communities, 1/2 < b < 1. Is this difference true in general? If confirmed by data of better quality from more communities, then a kilogram of resource supports a predator of larger body mass in a terrestrial community than in a coastal community. Why is this? The starting hypothesis here is that the mass of the consumer (predator or parasite) is related to the mass of the animal resource (prey or host) by a power law with exponent less than 1. This hypothesis is at best an approximation to reality, on both empirical and theoretical grounds (Cohen et al., 1993). Empirically, large predators sometimes eat prey of a wide range of masses while small predators eat prey with a narrower range of masses (as in Figs. 1 and 2 of Cohen et al., 1993). However, in Tuesday Lake, observed trophic links appear to fall in a band above and parallel to the diagonal line where predator mass equals prey mass, rather than in a triangular region in the (x, y) plane (Reuman & Cohen, 2004). Approximating both such relations by a power-law
321
322
J. E. COHEN
function ignores the apparent differences between them in how the variance in predator mass changes with the mass of the prey. The only direct evidence on how well a power-law relation describes body masses in parasite chains is Fig. 1 of Leaper and Huxham (2002, p. 447). Their scatter plots for parasites only in the Ythan estuary provide weak support for the usefulness of a power-law approximation. For log parasite size and log host size of parasites only, without trophospecies r2 ¼ 0.015 was not significant, and with trophospecies r2 ¼ 0.125 was statistically significant but still small. An approximate power law with exponent less than 1 has been derived theoretically from models of food-web structure, species abundance distributions, and the distribution of biomass across species mass categories (Cohen, 1991, pp. 5–8). Cohen et al. (1993) suggested that the logarithm of animal species masses may be approximately normally distributed, and that E(y|x) is the mean of a normal distribution censored below x (i.e. retaining only that portion of the normal distribution to the right of x). Unpublished numerical calculations show that, under this model, E(y|x) is a convex nonlinear function (always with slope less than 1) of x, rather than a strictly linear function as expected by the powerlaw relation Eq. (16.1). With the observed distribution of body mass reported by Cohen et al. (1993, p. 73, their Table 4), the power law approximates reasonably the convex nonlinear function in the range of animal body masses from 106 g to 10þ6 g. Terrestrial vertebrate predators far larger than contemporary top carnivores lived in the past (Burness, Diamond & Flannery, 2001). It would be interesting to determine whether predator and prey masses during the Cretaceous and late Pleistocene are consistent with a power law Eq. (16.1); if so, whether the coefficients A and B had different values from those estimated here; and if so, whether the maximum predator mass at that time could be predicted from the predator–prey body mass relations then in effect. Jonsson and Ebenman (1998b) suggested that the decrease they observed (Jonsson & Ebenman, 1998a) in the ratio of predator mass to prey mass with increasing trophic level in predator chains has significant consequences for stability in dynamic models of food chains. This suggestion could be extended to parasite chains, and merits further analysis and testing. The derivation of maximal body mass from the phenomenology of body sizes in trophic links is only one among many possible approaches. Other constraints on maximal body mass include mechanical or design constraints, energetics of food supply and metabolism, land area (for terrestrial consumers), natural selection of life histories and the processes of development (e.g. Bonner, 1988; Yoshimura & Shields, 1995; Burness, Diamond & Flannery, 2001; Gomer, 2001). It remains to be demonstrated whether, and if so how, these approaches are compatible. To summarize, food chains in which animal predators are bigger than their animal prey are called predator chains. Food chains in which the consumers are
BODY SIZES IN FOOD CHAINS
smaller than their animal prey are called parasite chains. If the mass of the consumer (predator or parasite) is related to the mass of the animal resource (prey or host) by a power law with exponent less than 1, then, in predator chains, there is an upper limit to the mass of the largest predator and prey, and in parasite chains, there is a lower limit to the mass of the smallest host and parasite. These limits are independent of the number of trophic links in the chain and independent of the mass of the basal animal species. In a predator chain that obeys this allometric relation of predator and prey masses, the ratio of predator mass to prey mass decreases as the trophic level and mass of the prey increase. In a parasite chain that obeys this allometric relation of predator and prey masses, the ratio of parasite mass to host mass increases as the trophic level of the host increases and the mass of the host decreases. In the data on predator chains here, predator masses generally exceed prey masses. The regression of the logarithm of predator mass on the logarithm of prey mass has slope b less than 1 in all cases. While it is possible to calculate maximal predator sizes from these regressions, estimates of maximal predator size are highly sensitive to uncertainty in the parameters of the regression lines. For three collections of data from coastal communities, 0 < b < 1/2, while for three collections of data from terrestrial communities, 1/2 < b < 1. A model of the joint distribution of consumer and resource body masses predicts a slope of 1/2 for both predator and parasite chains, and specifies conditions under which the slope should deviate up or down from 1/2. The theory developed here pertains to isolated chains, but all the data are drawn from webs with interconnecting chains. An ideal test of the theory would describe the full frequency distribution of body sizes of each species in a more or less isolated chain, if such can be found in nature. It would also be useful to extend the theory from isolated chains to more complex food webs and to analyze the consequences in the variability of body sizes of both resources and consumers.
Acknowledgements I am grateful for helpful, critical comments on previous drafts from John Tyler Bonner, Benni Hansen, Mark Huxham, Tomas Jonsson, Mark Laska, Robert H. Peters, Dave Raffaelli, Thomas W. Schoener, Alain Ve´zina and two reviewers; the support of US National Science Foundation grants BSR92-07293, DEB 9981552 and DMS-0443803; the assistance of Priscilla K. Rogerson; and the hospitality of Mr and Mrs William T. Golden during this work.
References Blackburn, T. M. & Lawton, J. H. (1994). Population abundance and body size in animal assemblages. Philosophical Transactions of the Royal Society of London Series B, 343, 33–39.
Blackburn, T. M., Lawton, J. H. & Pimm, S. L. (1993). Non-metabolic explanations for the relationship between body size and animal abundance. Journal of Animal Ecology, 62, 694–702.
323
324
J. E. COHEN
Bonner, J. T. (1988). The Evolution of Complexity. Princeton, NJ: Princeton University Press. Brose, U., Cushing, L., Berlow, E. L. et al. (2005). Body sizes of consumers and their resources. Ecology, 86, 2545. Brose, U., Jonsson, T., Berlow, E. L. et al. (2006). Consumer-resource body-size relationships in natural food webs. Ecology, 87, 2411–2417. Burness, G. P., Diamond, J. & Flannery, T. (2001). Dinosaurs, dragons, and dwarfs: the evolution of maximal body size. Proceedings of the National Academy of Sciences, 98, 14518–14523. Cohen, J. E. (1991). Food webs as a focus for unifying ecological theory. Ecology International (International Association for Ecology Bulletin), 19, 1–13. Cohen, J. E. & Carpenter, S. R. (2005). Species’ average body mass and numerical abundance in a community food web: statistical questions in estimating the relationship. In Dynamic Food Webs: Multispecies Assemblages, Ecosystem Development and Environmental Change – A Volume of Theoretical Ecology, ed. P. C. de Ruiter, V. Wolters & J. C. Moore. Amsterdam: Elsevier, pp. 137–156. Cohen, J. E., Briand, F. & Newman, C. M. (1990). Community Food Webs: Data and Theory. Biomathematics Vol. 20. Heidelberg, Berlin, New York: Springer-Verlag. Cohen, J. E., Pimm, S. L., Yodzis, P. & Saldan˜a, J. (1993). Body sizes of animal predators and animal prey in food webs. Journal of Animal Ecology, 62, 67–78. Cohen, J. E., Jonsson, T. & Carpenter, S. R. (2003). Ecological community description using the food web, species abundance, and body size. Proceedings of the National Academy of Sciences, 100, 1781–1786. Cohen, J. E., Jonsson, T., Mu¨ller, C. B., Godfray, H. C. J. & Savage, V. M. (2005). Body sizes of hosts and parasitoids in individual feeding relationships. Proceedings of the National Academy of Sciences, 102, 684–689.
Elton, C. (1927). Animal Ecology. (New impression with additional notes 1935.) New York: Macmillan. Gomer, R. H. (2001). Not being the wrong size. Nature Reviews Molecular Cell Biology, 2, 48–54. Hansen, B., Bjı´rnsen, P. K. & Hansen, P. J. (1994). The size ratio between planktonic predators and their prey. Limnology and Oceanography, 39, 395–403. Horning, M. & Trillmich F. (1997). Development of hemoglobin, hematocrit and erythrocyte values in Gala´pagos fur seals. Marine Mammal Science, 13, 100–113. Hutchinson, G. E. (1959). Homage to Santa Rosalia or why are there so many kinds of animals? American Naturalist, 93, 145–159. Jonsson, T. & Ebenman, B. (1998a). Trophic links and the relationship between predator and prey body sizes in food webs. Chapter 2 in T. Jonsson, Food Webs and the Distribution of Body Sizes. Linko¨ping Studies in Science and Technology, Dissertation 535, Linko¨ping, Sweden, pp. 63–81. Jonsson, T. & Ebenman, B. (1998b). Effects of predator-prey body size ratios on the stability of food chains. Journal of Theoretical Biology, 193, 407–417. Jonsson, T., Cohen, J. E. & Carpenter, S. R. (2005). Food webs, body size and species abundance in ecological community description. In Food Webs: From Connectivity to Energetics, Advances in Ecological Research Vol. 36, ed. H. Caswell. San Diego: Elsevier, pp. 1–84. Kleiber, M. (1961). The Fire of Life: An Introduction to Animal Energetics. New York: John Wiley. Leaper, R. & Huxham, M. (2002). Size constraints in a real food web: predator, parasite and prey body-size relationships. Oikos, 99, 443–456. Memmott, J., Martinez, N. D. & Cohen, J. E. (2000). Predators, parasitoids and pathogens: species richness, trophic generality and body sizes in a natural food web. Journal of Animal Ecology, 69, 1–15.
BODY SIZES IN FOOD CHAINS
Menge, B. A., Lubchenco, J., Gaines, S. D. & Ashkenas, L. R. (1986). A test of the Menge-Sutherland model of community organization in a tropical rocky intertidal food web. Oecologia (Berlin), 71, 75–89. Pauly, D. & Christensen, V. (1995). Primary production required to sustain global fisheries. Nature, 374, 255–257. Peters, R. H. (1983). The Ecological Implications of Body Size. Cambridge: Cambridge University Press. Reuman, D. C. & Cohen, J. E. (2004). Trophic links’ length and slope in the Tuesday Lake food web with species’ body mass and numerical abundance. Journal of Animal Ecology, 73, 852–866. Reuman, D. C. & Cohen, J. E. (2005). Estimating relative energy fluxes using the food web, species abundance, and body size. In Food Webs: From Connectivity to Energetics, Advances in Ecological Research Vol. 36, ed. H. Caswell. San Diego: Elsevier, pp. 137–182. Rott, A. S. & Godfray, H. C. J. (2000). The structure of a leafminer-parasitoid community. Journal of Animal Ecology, 69, 274–289. Schoener, T. W. (1968). Size of feeding territories among birds. Ecology, 49, 123–141.
Sheldon, R. W., Prakash, A. & Sutcliffe, W. H., Jr. (1972). The size distribution of particles in the ocean. Limnology and Oceanography, 17, 327–340. Ve´zina, A. F. (1985). Empirical relationships between predator and prey size among terrestrial vertebrate predators. Oecologia (Berlin), 67, 555–565. Wallace, A. R. (1858). On the tendency of varieties to depart indefinitely from the original type. In C. R. Darwin & A. R. Wallace, On the tendency of species to form varieties; and on the perpetuation of varieties and species by natural means of selection. Journal of the Proceedings of the Linnean Society, Zoology, 20 Aug. 1858, 3, 45–62. Online: http://pages.britishlibrary.net/charles. darwin3/jpls.html#natsel Warren, P. H. & Lawton, J. H. (1987). Invertebrate predator-prey body size relationships: an explanation for upper triangular food webs and patterns in food web structure? Oecologia (Berlin), 74, 231–235. Yoshimura, J. & Shields, W. M. (1995). Probabilistic optimization of body size: a discrepancy between genetic and phenotypic optima. Evolution, 49, 375–378.
325
CHAPTER SEVENTEEN
Body size in aquatic ecology: important, but not the whole story ALAN G . HILDREW School of Biological and Chemical Sciences, Queen Mary, University of London
DAVI D G . RAFFAELLI Environment Department, University of York
RONNI EDMONDS - BROWN Division of Geography and Environmental Sciences, University of Hertfordshire
Introduction Ecologists have long been aware of regularities and patterns in the body size of organisms in populations and communities, observations that go back at least to Alfred Wallace and continue through the works of Elton, Thienemann, Hutchinson, MacArthur and many others. The classical contribution of R. H. Peters (1983) codified such patterns through the concept of body-size allometry, of metabolic rate and other features, and led on to many of the phenomena now included under macroecology (Blackburn & Gaston, 2003). Brown and colleagues (Brown et al., 2004; Brown, Allen & Gillooly, this volume), in particular, added new advances in scaling theory and, incorporating the exponential effect of temperature on metabolic rate, sought to explain a wide variety of patterns and processes in ecology at levels of organization from individuals to ecosystems. The focus on aquatic systems at the Hatfield symposium, and in this resultant volume, is justified because body-size patterns may be more important, or at least more obvious, in aquatic ecosystems. Woodward and Warren (this volume) offer three possible reasons. First, the most important primary producers in water are small and, along with small heterotrophic micro-organisms and small detritus particles, are gathered from suspension by larger consumers. Second, they point out that conventional predators, larger in turn than their prey, seem particularly prominent in aquatic systems where there may be fewer parasite food chains (see Cohen, this volume). A third, methodological reason for a particular emphasis on body size by aquatic ecologists, may be the relative ease of automated counting of particles and organisms in suspension across a wide range of scales. Perhaps with particular reference to methodology, it is noteworthy that terrestrial and aquatic ecologists have also generally taken different perspectives on body size. Thus, in describing assemblages, aquatic Body Size: The Structure and Function of Aquatic Ecosystems, eds. Alan G. Hildrew, David G. Raffaelli and Ronni Edmonds-Brown. Published by Cambridge University Press. # British Ecological Society 2007.
BODY SIZE: IMPORTANT, BUT NOT THE WHOLE STORY
ecologists often present their data as a continuous size spectrum of individuals, irrespective of taxonomic affinities, whereas terrestrial ecologists generally view the species as the individual datum, and usually as a fixed entity – mean adult body mass. The ways these two research communities present and subsequently interpret their data affect their understanding of how aquatic and terrestrial systems work (Raffaelli, Solan & Webb, 2005). To these speculations we might add that aquatic systems show one feature for which there is no obvious parallel in the terrestrial world: a viscous medium that greatly shapes how organisms function and behave. This unique attribute of aquatic systems has understandably been a major focus for aquatic ecology and has accordingly produced a perspective that is quite different from that of terrestrial ecologists. As the chapters in this book testify, body size has clearly remained a major research focus for the ecological community, from ecophysiology to the ecosystem. In part this is because body size sets real mechanical limits on what organisms can do (e.g. limits to the dimensions of prey that can be physically ingested), and in part because body size is a super-parameter, which does well at capturing a range of associated physiological and ecological traits, but is much more easily measured (or found in books) than those traits themselves. A few of the authors herein have set out explicitly to test such data against the predictions of the metabolic theory of ecology (Brown et al., 2004), with varying degrees of success, while others chose ostensibly different theoretical backgrounds against which to consider patterns and dynamics related to body size.
Body size and metabolic theory We consider first the few chapters that explicitly mention explorations of metabolic theory, or at least consider metabolic theory against patterns in the different fields of literature. With respect to life history, Atkinson and Hirst (this volume) consider whether selection can alter the scaling exponent (b) relating metabolic rate to body size, the core of metabolic theory. They ask to what extent is b fixed at 3/4? They cite Glazier’s (2005) survey of 642 metabolic rates during ontogeny, around half of which deviated from 3/4 and, interestingly, with higher values (mean 0.95) in pelagic compared to benthic (mean 0.74) species, perhaps due to the increased costs of buoyancy or avoiding predation in pelagic species. Huryn and Benke (this volume) consider biomass turnover and body size in stream benthic invertebrates, analyzing the relationships between body size (M) and population density, biomass (B), production (P) and P/B. They found an encouraging degree of fit with theory for three temperate streams, with the relationship between P/B and body size (as mass) having a scaling exponent varying between 0.24 and 0.27, bracketing the predicted value of 0.25. Huryn and Benke (this volume) consider this precise fit ‘extraordinary when considering the large ecological differences between the streams’, forested or grassland, with or without fish, etc. The ‘snag’ community
327
328
A. G. HILDREW ET AL.
(organisms sedentary on woody debris in the channel) of the warm water Ogeechee River (Georgia, USA) was inconsistent with the theory (scaling exponent of M and P/B, 0.50), though this discrepancy could be accounted for in various ways. Thus, for Huryn and Benke’s data on stream benthos, metabolic theory does fairly well. Several other authors found encouraging, though patchy, fits of pattern with metabolic theory in different contexts (Petchey, Long & Morin, this volume; Humphries, this volume; Hall et al., this volume; Jennings & Reynolds, this volume), leaving us with the intellectually awkward task of judging whether the theory survives because we can explain the discrepancies, or if the theory falls in general. The problems of testing theory in this way are several and were vigorously debated at the meeting, and there seem to be three major issues involved. First, testing has essentially been a post-hoc exercise in which data gathered for completely different purposes are used to challenge the theory, always a tricky process. Second, different measures of body size were used in the present collection of chapters: mass or weight (dry, ash-free, wet, whole-body or elemental), volume or equivalent spherical diameter, individual, average or adult body size. Different measures will be appropriate for different research questions and in many cases a clear rationale for selecting a particular measure of body size is presented. For other studies, however, including many macroecological investigations, the data used are simply those at hand, often adult body sizes, which can be extracted from field guides or species survey lists. If patterns that emerge from the analyses of such data are interpreted in the context of theory that assumes other expressions of body size, such as individual biomass, then the value of the research is compromised. One solution to this potential problem is to make available a much greater variety of body-size measurements for a broad range of systems within a public domain database, so that analyses can be tailored more closely to theory. Third, body-size data and relationships are presented in a variety of forms within the present volume. Several chapters reported bi- or tri-variate relationships, with body size as the explanatory variable, often seeking to allow comparisons with theory (mainly as judged by slopes of regression lines). However, testing for deviation from theoretical relationships opens up a can of worms. For instance, the choice of regression models (least squares, geometric mean, bisector regressions and others) is important because each estimates a different value for the slope. Similarly, it is not good enough to assume that, because a slope through the observed data is not significantly different from the theoretical prediction, the data are consistent with theory. An alternative explanation is that there is not enough power (essentially the number of data points) in the analysis to falsify the model, and formal power analyses should be carried out to check this, although this is rarely done. More suitable statistical procedures to demonstrate negligible trends in data are now available (Dixon & Pechmann,
BODY SIZE: IMPORTANT, BUT NOT THE WHOLE STORY
2005), and these authors also provide a good discussion of the difference between a failure to reject a null hypothesis and verifying that null hypothesis. In addition, for claimed invariant relationships, where a slope of zero is predicted by theory, it is incredibly difficult to distinguish such invariance in a data set from a non- (and therefore meaningless) relationship. Finally, different representations of the data could well generate quite different perspectives on the significance of body size in ecology. Within taxonomic collections (e.g. a family of water beetles or harpacticoid copepods), bivariate plots of body size against abundance provide reasonable representations of relationships. As the taxonomic grouping becomes coarser (e.g. freshwater benthos or marine benthic invertebrates), regression techniques begin to fail to provide an adequate model of the data. At the whole system level, a constraint space (e.g. Brown, 1995) or biomass spectrum (e.g. Stead et al., 2005) might be more appropriate, although formally quantifying such pattern and exploring deviations from theoretical prediction is fraught with difficulty (Leaper & Raffaelli, 1999).
Other empirical, theoretical and conceptual backgrounds Most chapters in this book do refer to metabolic theory, but perhaps wisely do not attempt explicit tests, instead setting their findings in the context of a variety of other areas of theory and concept. For instance, Humphries (this volume) considers suspension feeding, a mode of feeding virtually restricted to aquatic systems, in relation to body size and the physics of water and flow. Body size is crucial for aquatic organisms because it determines flow regime and thus, for suspension feeders, food supply. We are particularly struck by Humphries’ remark that body size is particularly ‘nebulous and difficult to specify’ for suspension feeding organisms, because they stretch to the very limits the relationship between the various measures of size that ecologists can use. Consider for instance gelatinous pelagic filter feeders, of large physical dimensions but very low body mass; benthic filter feeders with large and heavy, but metabolically inactive, shells; and organisms that build extensive nets and mucus sheets to capture food at a rate beyond that expected from their body size alone. Perhaps for this reason there is still no clear relationship between body size and food-particle size for suspension feeders, in contrast to the situation for conventional predators. Townsend and Thompson, Atkinson and Hirst, and Huryn and Benke (all this volume) all deal with body size as a species trait in relation, respectively, to the habitat templet hypothesis, life-history theory and metabolic theory. Townsend and Thompson (this volume) consider body size against the habitat templet hypothesis of Southwood (1977), and allied predictions specifically developed for streams by Townsend and Hildrew (1994), and referring to traits in relation to disturbance, productivity, land use, predation and others. They conclude that
329
330
A. G. HILDREW ET AL.
clear patterns of macroinvertebrate body size are ‘sometimes apparent but rarely account for a large proportion of variation along the (environmental) dimension of interest’. They attribute this result to the complexity of multiple stressors in natural systems and to the inclusion of mixtures of organisms with different ‘solutions’ (including body size) to particular environmental conditions. We also wonder whether in such systems, population and community dynamics will often be transient and thus the ‘snapshot’ patterns often weak. Disturbance is one obvious cause of fluctuations, though others are feasible, and we return to this notion below. Jennings and Reynolds (this volume) relate body size to the pressures of commercial fishing and to fisheries management (see also Atkinson & Hirst, this volume). The effects of fishing on the population size structure of exploited species have been stark and, apart from the well-known phenomenon of ‘fishing down the food chain’ in overexploited systems, Jennings and Reynolds (this volume) have managed to use theory and data to predict that the abundance of large fish (16–66 kg body weight) in the North Sea may now be a remarkable 99.2% lower than before exploitation. Jennings and Reynolds’ paper is wide ranging, and also deals with community-level phenomena such as trophic structure and biomass size spectra, and they point out the practical applications of food-web theory in the adoption of ecosystem and community-level management of fisheries and the oceans. In terms of trophic structure, they highlight the absence of clear trophic steps (trophic levels) in most aquatic (as opposed to terrestrial) ecosystems, and the prevalence of life-cycle omnivory and size-based predation, all arguments for a size-based analysis of the ecosystem impacts of exploitation. Two other chapters focus on body size in relation to species interactions, food-webs and food web theory. Woodward and Warren (this volume) provide much evidence for the importance of body size in determining feeding links, food-web structure and life-cycle omnivory. Communities of freshwater benthic invertebrates evidently provide some of the clearest evidence of all of the primacy of body size in natural food webs, thus chiming with Jennings and Reynolds’ (this volume) conclusions. They analyze explicitly the limits, both upper and lower, to the size disparity observed between predator and prey – i.e. both upper and lower size refugia exist – and show the effect of basing estimates of size disparity on individuals and on species averages. If species averages are taken (i.e. the average size of the predator and the average size of the prey) it turns out that the predator is about ten times greater than the prey, whereas if individual links are considered (i.e. the size of the particular predator and its actual prey) the mean size disparity is about 100. Jones and Jeppesen (this volume) consider further the role of body size in relation to trophic structure, and in their case to the occurrence of alternative regimes and in particular of trophic cascades, many examples of which seem to be ‘wet’ (Strong, 1992). In the
BODY SIZE: IMPORTANT, BUT NOT THE WHOLE STORY
sense that trophic cascades are characterized by clear trophic levels, there seems a disparity between the view that cascades are prevalent and the rejection by Jennings and Reynolds (this volume) of integer trophic levels for most aquatic ecosystems, as well as Woodward and Warren’s (this volume) findings for simple freshwater benthic communities. Jones and Jeppesen (this volume) argue that cascades are common where there is a large size disparity between predator and prey, such that ‘prey’ are so small that they are fed on unselectively by the consumers, and where small productive producers turn over very rapidly and can sustain a large biomass of long-lived predators. These features may be more decisive than body size per se in determining the occurrence of cascades. Several chapters took a biogeographical or macroecological approach. These include Finlay and Esteban; Rundle, Bilton and Foggo; Warwick; and Schmid and Schmid-Araya (all this volume). Finlay and Esteban (this volume) consider the biogeography of organisms in relation to their notion of a ‘biogeographical divide’, a body size below which species, by consequence of their enormous population abundances (thus linking this concept to macroecological theory), are cosmopolitan, but above which they are not. This body-size transition seems to be in the range 1–10 mm. For macro-organisms, Rundle et al. provide evidence of a positive relationship between range size and body size in actively dispersing dragonflies. For passively dispersing freshwater organisms, they find that those within the body-size transition of Finlay and Esteban, such as small crustaceans, do indeed have a biogeography and are not cosmopolitan. For passively dispersing marine and freshwater organisms, range size increases with body size, perhaps related to the covariance between body size and propagule output, and hence dispersal probability. It might be that microbial species below about 1 mm tend to be cosmopolitan, but that above this body size, geographical range again increases, as the ability to disperse actively goes up with body size in active dispersers, and the likelihood of dispersal goes up with propagule output (and therefore with body size) in passive dispersers. It seems clear that data on range size for aquatic organisms lag well behind those for terrestrial organisms, which may explain the general paucity of (biogeographical) macroecological relationships for aquatic organisms. Warwick (this volume) deals with a macroecological pattern for which there is a long history of research in aquatic ecosystems, the species–body size spectrum. He reports a regularly bimodal spectrum for the marine benthos, apparently contrasting with a unimodal one for freshwater. In the marine benthos the small mode consists of meiofauna and the large mode the macrofauna, with few intermediate species. Warwick attributes this pattern not to local ecological interactions or habitat architecture, as originally advocated by Schwinghamer (1981), but to the evolutionary history of the marine benthos. The trough of species in the spectrum equates to the size of the planktonic, larval forms of the larger macroinvertebrate benthos. These cannot be benthic because the smaller
331
332
A. G. HILDREW ET AL.
meiofaunal species would predate on or compete with them (many meiofaunal predators take relatively large prey, often larger than themselves). According to Warwick, this general regional pattern of species–body size then produces the local patterns observed. In contrast, Schmid and Schmid-Araya (this volume) analyze such community patterns in relation to fractal geometry, potentially the most fundamental of all the approaches in this volume since fractals may underly the scaling of metabolic rate with body size itself. Schmid and SchmidAraya did not find a self-similar pattern of body size across all scales, however, referring to the changes in scale as ‘multifractal’. Interestingly, Warwick (this volume) also rejected a single fractal scaling for the marine benthos. Other chapters took experimental or modelling approaches. In their experiments using model ecosystems of protists, Petchey, Long and Morin (this volume) demonstrated the effects of body size at three levels of organization: the population, community and ecosystem. The results were encouraging, but there were discrepancies with allometric theory and we can conclude that the effects of body size on ecological processes are modified by trophic complexity (connectedness) and species richness. Persson and De Roos (this volume) also looked at model systems, but focused on individual variation using physiologically structured population models. They stress the within-species variation, much of it ontogenetic, in food intake, growth and body size that is so characteristic of most metazoans. This variation generates intrinsic dynamics and divergent body-size distributions, not wholly predictable by metabolic theory for instance. Systems with strong cohorts that are constantly changing are non-equilibrial systems that can approach alternate states. In short, they emphasize population and community dynamics rather than the structure and pattern that was addressed by many of the other authors. These models are remarkably powerful, and explain effects apparent in many fish populations, but we can only join Persson and De Roos (this volume) in concluding that ‘a major challenge . . . is to develop approaches that allow the analysis of more complex configurations in terms of the numbers of species present’. When even simple real food webs may contain tens or hundreds of species, this is a challenge indeed. Several authors referred to the ecosystem consequences of body size, including CO2 production (Petchey et al., this volume) and nutrient cycling, though this aspect was most explicitly addressed by Hall et al. (this volume). The latter address the extent to which animals affect nutrient cycling directly, through ingestion, egestion, production and excretion, and how body size controls such effects. Partially as expected, nutrient excretion rates scaled allometrically, with an exponent less than 1. Ammonium excretion of stream invertebrates from 18 taxonomic orders scaled at 0.85 with body mass, though there is an interesting degree of variation, apparently related to taxonomy and thus also potentially to the measure of body size used for different taxa. For instance, why are there differences in the scaling of the excretion of P, but not N, between two species of
BODY SIZE: IMPORTANT, BUT NOT THE WHOLE STORY
zebra mussel (Dreissena), and why does excretion of N and P in Dreissena scale so disproportionately with body size (b ¼ 1.38)? There are some fascinating insights in this chapter into how body size could feed back onto ecosystem processes. For instance, they note that the removal of large, migratory salmon has reduced nutrient subsidies to many rivers and their catchments, and that the harvesting of large fish will similarly feed back onto nutrient regeneration rates. Joel Cohen (this volume) closed our symposium with a consideration of the consequences, in both predator and parasite food chains, of the average body mass of a consumer being related to the average mass of the species consumed by a power law with an exponent less than 1. We were struck by Cohen’s finding that ‘one kilogram of resource supports a predator of larger body mass in a terrestrial community than in a (marine) coastal community’, though it is not at all obvious why this should be so. This and other intriguing outcomes of his analyses only serve to highlight the need for more data on body size in natural systems.
Concluding remarks We finish with some speculations. Body-size allometry, and its extension to the metabolic theory of ecology, does pretty well at predicting some, but not all, of the patterns and processes dealt with in this book, and hence provides our chapter title. Whilst appealing to issues of scale may these days seem like the resort of the scoundrel, it is obvious that patterns appear and disappear as spatial and temporal resolution changes and analyses include more and more heterogeneous groups of species at different trophic levels. Further, the appropriate measure of body size is difficult to specify a priori, and we are struck by nature’s ingenuity in creating ways of cheating the scaling and limitations of body size, allowing organisms to ‘punch above their weight’ in obtaining food, and in the process making the ecologist’s task more difficult. Examples include filterers that build large external feeding structures and Cohen’s (this volume) cases of social hunting that enables predators to feed on prey much larger than themselves. Perhaps even the very large marine filter feeders are only possible because they can exploit whole aggregations of prey, effectively a ‘superindividual’, with a combined body size much larger than the single krill ‘particle’. The greatest challenge to a size-based approach to food webs lies with the enormous terrestrial primary producers, trees, the largest of all organisms but which lie at the base of land-based food webs, though their biomass is mainly metabolically inactive (Cousins et al., 2005). Transient dynamics and temporal fluctuation are referred to in several papers, and clearly point to the notion that patterns based on energetic rules and allometry may be obscured at the scale at which it is possible to observe them. It seems likely that the processes underlying metabolic theory provide an envelope or a constraint outside which natural systems may not sustainably lie,
333
334
A. G. HILDREW ET AL.
but that heterogeneity, fluctuations and non-equilibrial processes mean that systems obey such rules only in an approximate way. Energy subsidies among neighbouring systems can be considered in a similar way to transient dynamics. Thus, subsidies to top predators can produce ‘apparent’ cascades in shallow lakes (Jones & Jeppesen, this volume) and might have produced the discrepancies with metabolic theory in the highly productive snag communities of filter feeding organisms in the Ogeechee River observed by Huryn and Benke (this volume). In short, our title captures the discussions and conclusions at the symposium about the role of body size – it is important, but not the whole story.
References Blackburn, T. M. & Gaston, K. J. (2003). Macroecology: Concepts and Consequences. Oxford: Blackwell Science. Brown, J. H. (1995). Macroecology. Chicago:University of Chicago Press. Brown, J. H., Gillooly, J. F., Allen, A. P., Savage. V. M. & West, G. B. (2004). Towards a metabolic theory of ecology. Ecology, 85, 1771–1789. Cousins, S. H., Bracewell, K. V. & Attree, K. (2005). Measuring the ability of food to fuel work in ecosystems. In Dynamic Food Webs, ed. Ruiter et al. London: Academic Press, pp. 248–257. Dixon, P. M. & Pechmann, J. H. K. (2005). A statistical test to show neglible trend. Ecology, 86, 1751–1756. Leaper, R. & Raffaelli, D. G. (1999). Defining the body size-abundance constraint space: data from a real web. Ecology Letters, 2, 191–199. Peters, R. H. (1983). The Ecological Implications of Body Size. New York: Cambridge University Press.
Raffaelli, D., Solan, M. & Webb, T. J. (2005). Do marine and terrestrial ecologists do it differently? Marine Ecology Progress Series, 304, 271–307. Schwinghamer, P. (1981). Characteristic size distributions of integral benthic communities. Canadian Journal of Fisheries and Aquatic Sciences, 38, 1255–1263. Stead, T. K., Schmid-Araya, J. M., Schmid, P. & Hildrew, A. G. (2005). The distribution of body size in a stream community: one system, many patterns. Journal of Animal Ecology, 74, 475–487. Strong, D. R. (1992). Are trophic cascades all wet? – differentiation and donor-control in speciose ecosytems. Ecology, 73, 747–754. Southwood, T. R. E. (1977). The habitat, the templet for ecological strategies. Journal of Animal Ecology, 46, 337–365. Townsend, C. R. & Hildrew, A. G. (1994). Species traits in relation to a habitat templet for river systems. Freshwater Biology, 31, 265–275.
Index
absolute abundance, microbial species 169–171 abundance and metabolic rate 8–9 consumer:producer ratios 11–12 distribution across species 2 Acartia tonsa (copepod) 24–25 acidification effects in freshwaters 112–113 adaptive dynamics evolutionarily unbeatable strategy 36 fitness definition 36 frequency and density dependence 36–37 relationship of traits to fitness 35–36 trade-offs among traits 35–36 adaptive dynamics life history model 35–36 Allee effect 236–237 allometric equations, form of 1–2 allometric scaling strategies 227–228 alternative states and body size distributions 235–239 aquatic ecosystems body size as a structuring force 98–99 individual biomass production rates 5 individual metabolic rate and body size 4–5 temperature corrections 3–4 aquatic invertebrates (dispersal and body size) 188–193, 195, 196, 198, 200, 201 active dispersers in freshwaters 188–189 active dispersers in freshwaters (case studies) 188, 193–196 dispersal in marine systems 191–193 dispersal in marine systems (case studies) 199–203 passive dispersers in freshwaters 188, 190–191 passive dispersers in freshwaters (case studies) 188, 196–199 Arctic char 236, 237 Atlantic silverside (Menidia menidia) 41 Avelia martinicensis 175 B (mass-specific metabolic rate) 2 b (scaling exponent) 1–44, 45 Bacillus cereus 247 Bacillus subtilus 247 bacteria, body size 1 barndoor skate (Dipturus laevis) 267–268
biodiversity characteristics macroscopic organisms 169 microbial species 169 biogeography and size see microbial species biological scaling relations 1–4 biomass and body size (stream communities) 56, 57, 60, 61, 62, 68–71 and diversity 246–247 consumer:producer ratios 11–12 distribution across species 2 estimates 42 biomass invariance mechanisms 146, 148–152 biomass per unit volume, consistency of 245 biomass production rates, individual organisms 5 biomass turnover (P/B) factors affecting 000 range of rates 000 biomass turnover (P/B) and body size (stream communities) 000 application of the MTE 60, 62–68 biomass 56, 57, 60, 61, 62, 68–71 biomass turnover rate (P/B) predictions and results 56, 57, 61, 62, 63, 66, 68–71 differences in temperate and subtropical streams 60, 62–68 discussion 60, 62–68 metabolic theory of ecology 55–56 population density 56, 60, 61–62, 65, 68–71 predictions 56–57, 68–71 production (P) 56, 57, 61, 63, 64, 68–71 results 60–62, 63, 64, 65, 66 secondary production rate 55–56, 68–71 study streams 57–59 tests of predictions 59–60 bluegill sunfish (Lepomis macrochirus) 39, 40 body size adaptive plasticity in Cladocera 37–39 correlation with characteristics of organisms 1–2 early writings and studies ix–x link to patterns in natural systems 98 range of variation among organisms 1 relation to metabolic rate 1–2
336
INDEX
body size (cont.) relationship with temperature 42 variation within species 226 body-size distribution (BSD) in communities 140 effects of shifts in 112–113 multifractal properties 155–156, 157, 158–159 body-size increase, costs and benefits in suspension feeders 19–24 body-size measurement, difficulties in suspension feeders 17 Boltzmann-Arrhenius factor 2 Boltzmann’s constant (k) 2 brown trout 236, 237 BSD see body-size distribution cannibalistic system 233–234, 235 capelin 237 carbon cycle, and metabolism of organisms 9–11 carbon:phosphorus:nitrogen ratio 9–10 carbon:phosphorus ratio influence on growth rate 6 relation to body size 6 carbon turnover and plant size 10–11 Cladocera, adaptive plasticity in body size 37–39 common skate (Dipturus batis) 267–268 community consequences of body size (microbial systems) 249–250, 252–256, 257, 258, 259, 260, 261–262 community size distributions 110–112 and food web construction 110–112 distinct multiple nodes 111–112 factors affecting size structure 112 invertebrate macrofauna 111–112 megafauna 111–112 meiofauna 111–112 vertebrate macrofauna 111–112 Condylostoma reichi 175 consumer:producer ratios abundance 11–12 biomass 11–12 metabolic energy flux 11–12 copepods body size in relation to prey size 24–25 estimation of global patterns of mortality 46–47, 48 Coregonus albula 229 cosmopolitan-biogeography transition size 171–172 general validity 180–181 cosmopolitan distribution theory (microbial species) 168, 172–180 ‘biogeography’ of microbes 180 cosmopolitan genotypes (rDNA) 178–180 diversity of cryptic protists exceeds active species 172–174 endemic species 168, 175–178, 179 habitat influence on species composition 168, 175–176 habitat selects for cryptic or active phase 172–174 local:global species ratios 176, 177 random dispersal of free-living protists 168, 174
similar species inventories irrespective of geographical distance 174 species concepts 178 undersampling problem 176–178, 179 Cyclidium glaucoma 178–179, 181 density and biomass scaling 146–148 density–body-mass scaling 144–145 with sample area 145–146, 147 density dependence (strength of compensation), response to additional mortality 270–273 determinate growth 228 diatom communities, influences on species composition 175–176 Dipturus batis (common skate) 267–268 Dipturus laevis (barndoor skate) 267–268 dispersal challenges for aquatic organisms 187–188 microbial species 169–171 dispersal ability macroecological significance 187 traits which influence 187 dispersal and body size (aquatic invertebrates) 188–193, 195, 196, 198, 200, 201 active dispersers in freshwaters 188–189 active dispersers in freshwaters (case studies) 188, 193–196 dispersal in marine systems 191–193 dispersal in marine systems (case studies) 199–203 passive dispersers in freshwaters 188, 190–191 passive dispersers in freshwaters (case studies) 188, 196–199 distribution, cosmopolitan-biogeography transition size 171–172 disturbance, effects on marine species diversity 216, 218–220, 221 diversity see species diversity Dreissena spp. (mussels) 289 Dyar’s constant 211, 212 E-state (environmental state) 231 ecological stoichiometry and nutrient relations 2 ecosystems application of MTE 9–12 biomass invariance mechanisms 146, 148–152 body-size distribution (BSD) 140 body-size scaling relationships 98 carbon turnover and plant size 10–11 consequences of body size (microbial systems) 250–251, 256–259, 260, 262–263 effects of community size spectrum 43 element turnover rate and metabolic rate 9–10 energy and materials flow in food webs 11–12 flux and storage of elements 9–12 flux and storage of energy 9–12 fractal geometry framework for patterns and relationships 141–143 influence of body size 225–226 influence of body-size attributes 140 particle-size distribution (PSD) in benthic habitats 142 power-law behaviour 140–143
INDEX
role of metabolism in trophic relationships 11–12 scale-invariance and fractal properties of SADs 141–142 scale-invariance and fractal properties of SARs 141–142 scale-invariance properties 140–143 species-abundance distribution (SAD) mechanisms 146, 148–152 storage of elements in living biomass 9–10 stressor-induced size-spectrum changes 112–113 stressor-induced species loss 112–113 egg hatch times and egg protection strategy 48–49 model 47–49 egg-to-adult times, model 47–49 endothermy and individual growth rate 228 energy availability models 273–274 energy equivalence rule 8–9 energy flux, consumer:producer ratios 11–12 energy transfer in predator–prey interactions 274 energy use across species 2 energy use of populations, and body size 8–9 environmental gradients and body size 78 biomonitoring tools 94 body size patterns 90–94 in stream ecosystems 78 environmental gradients and body size (stream studies) 78–85 agricultural and mining pollution 84–85 altitude 82–83 catchment land use 82 chemical contaminants 84–85 descriptive studies of abiotic gradients 79–81 descriptive studies of biotic gradients 81–82 descriptive studies of complex gradients 82–83 disturbance regime 79–80 experimental studies 83–85 floods 81 hydraulic stress 81 New Zealand streams database 85–88, 89, 90, 91, 92 phosphate availability 81–82 predation 82, 83–84 primary productivity 82 substrate particle size 80, 83–84 evolution fisheries-induced 40–41 of body size see life history analysis; life history theory feeding and size factors affecting variation in prey choice 101–105 individual level processes 100–105 ontogenetic dietary shifts in predators 104–105 scaling to species 105–110 size constraints on predation 100–105 fish eggs, ontogenetic development rate and body size 5–6 fish species, mortality selection for smaller size at maturity 40–41, 42
fisheries describing and predicting community responses 274–280 genetic shifts in mature size of fish 40–41 modelling effects on community size structure 274–280 shifts in fish community size spectrum 43 size-based responses to fishing 274–280 fisheries-induced evolution 40–41 fisheries management changes in size-spectra slopes 269 extinction of marine species 267–268 ‘fishing down the food web’ 268 indirect effects of fishing 269 magnitude of fishing effects 267–268 population and community effects of exploitation 266, 267–268 seabed damage by fishing gear 268, 269 significance of body size 266 size-related effects of fishing 268–269 size-related responses to exploitation 268–269 state of the world’s fisheries 267–268 flow regime (Reynolds number, Re), and body size 17–18 food availability, and body size (suspension feeders) 21, 26–28 food-chain body sizes lower limit to mass of host and parasite 307 metaphoetesis 306–307 parasite chains 306–308 parasitoid chains 306, 307 power-law relationships 306, 307 predator chains 306–308 predator–prey size relationships 306–308 social hunters 306–307 upper limit to mass of predators and prey 307 food-chain body sizes data differences between terrestrial and coastal communities 310, 314, 321–322 empirical values of the exponent 310, 314, 321–322 examination of Hutchinson’s (1959) arguments 319–321 limits to mass of host and parasite 310, 314, 321–323 limits to mass of predator and prey 310, 314, 321–323 power-law relationships 310, 314, 321–323 food-chain body sizes theory 308–313 maximal and minimal body masses 308–311 predicted value of the exponent 311–312 ratios and differences of consumer mass and resource mass 312–313 see also food webs data food-dependent development 235–239 food-dependent growth 228 food webs 2 application of MTE conceptual framework 11–12 body size structure 11–12 construction of size-based models 113–114 construction related to community size distributions 110–112 feeding loops 110
337
338
INDEX
food webs (cont.) ‘life history’ omnivory 110 masking of size-related effects 107, 108, 109–110 rates of energy and material flow 11–12 role of metabolism 11–12 species-averaging effects 105–110 structure by size rather than species 110 synthesis of approaches to 12 food webs data 313–319 studies of a well-defined community 313–317 studies that pool multiple communities 314, 317–319 fractal geometry framework for patterns in ecosystems 141–143 fractal properties of size-structured communities 152–159 Gadus morhua (North Atlantic cod) 237 Gadus morhua (northern cod) 40–41 global patterns of life history, prediction 47–49 global patterns of mortality in copepods, estimation 46–47, 48 growth and development patterns 226–229 allometric scaling strategies 227–228 classification of growth types 228 determinate growth 228 endothermy and individual growth rate 228 food-dependent growth 228 indeterminate/plastic growth 228 link between individual traits and community attributes 229 population-level feedback and dynamics 229 scaling constraints 227–228 see also ontogenetic development growth-rate hypothesis 6 habitat complexity and predator–prey interactions 129–132 and species complexity 129–132 and trophic cascades 129–132 hatching rate and body mass 5–6 Hubbell’s unified theory of biodiversity (UTB) 149–152 Hutchinson’s (1959) arguments, food chain body size and diversity 319–321 Hutchinsonian ratio 211 i-state (individual) distribution models 230–231 indeterminate/plastic growth 228 individual body size and population dynamics 229–240 alternative states and body size distributions 235–239 cannibalistic system 233–234, 235 E-state (environmental state) 231 emergent Allee effect 236–237 extensions to more complex configurations 239–240 food-dependent development 235–239 i-state (individual) distribution models 230–231 modelling framework 230–231
ontogenetic development and community structure 235–239 ontogenetic development dynamics 231–234, 235 p-state (population state) 230–231 physiologically structured population models (PSPMs) 230–231 size-dependent predation 235–239 size-structured consumer-resource system 231–233, 234 structured biomass community model 239–240 tritrophic food chains 235–239 individual organism biomass production 5 influence of body size on performance 225 metabolic rate and body size 4–5 intrinsic rate of increase (rmax), response to additional mortality 270–271 intrinsic rate of increase in a habitat (r) 34–35, 38–39 invariant quantities 44 invertebrate macrofauna 111–112 Leeuwenhoek, Antonie van 167 Lepomis gibbosus (pumpkinseed sunfish) 39, 40 Lepomis macrochirus (bluegill sunfish) 39, 40 life histories and body size 269–271 and population dynamics 269–271 and responses to exploitation 270–271 trade-offs 270–271 see also growth and development patterns life history analysis 33 biomass estimates 42 body size relationship with temperature 42 ecological applications 40–43 ecosystem function and community size spectrum 43 fisheries-induced evolution 40–41 prediction of body size variation within species 40–41, 42 shifts in community size spectrum 43 temperature-size rule (TSR) 42 life history analysis and scaling relationships ecological applications 46–49 estimation of global patterns of mortality 46–47, 48 intra- and interspecific allometries 45–46 invariant quantities 44 key elements 43–44 methodological issues 44–46 prediction of global patterns of life history 47–49 scaling exponents 43–44, 45 selection effects on scaling exponents 45 life history theory adaptive dynamics model 35–37 adaptive plasticity in life histories 37–39 fitness definition 34–35, 36 frequency and density dependence 36–37 interpretation of phenotypic variation 39 intrinsic rate of increase in a habitat (r) 34–35, 38–39 key elements 34–36 lifetime reproductive success (LRS) 34–35, 38–39
INDEX
methodological issues 36–39 optimality (optimization) model 34–35, 36–37 phenotypic plasticity 37–39 prediction of life history plasticity 37–39 reaction norms 37–39 relationship of traits to fitness 35–36 trade-offs among traits 35–36 life history traits, stream habitats 77–78 lifespan and body size 6–7 and temperature 6–7 lifetime reproductive success (LRS) 34–35, 38–39 Lindeman efficiency 11–12 local:global species ratios, size dependence 181–182 Lontra canadensis (river otter) 292–293 Loxodes rex 175 macroecological patterns acidification effects in freshwaters 112–113 aquatic ecosystems 186–187 effects of individual level processes 99–100 effects of shifts in body-size distribution 112–113 effects of species dispersal ability 187 influence of body-size relationships 186 size-spectrum changes in stressed ecosystems 112–113 stressor-induced species loss 112–113 macroscopic organisms, characteristics of biodiversity 169 marine communities describing and predicting responses to mortality 274–280 energy availability models 273–274 energy transfer in predator–prey interactions 274 modelling fishing effects on size structure 274–280 PPMR (ratio of mean predator size to mean prey size) 274 size-based features of aquatic food webs 273 size-based responses to fishing 274–280 size-based structuring 273–274 size-spectra emergence models 273–274 slope of size spectra 273–280 see also fisheries marine systems, similarities to terrestrial systems 210 marine systems species diversity benthic species adult body size distributions 212–216, 217, 218, 219 benthic species all-animal body size distributions 216–217, 219, 220 Dyar’s constant 211, 212 effects of disturbance 216, 218–220, 221 effects of pollution 216, 218–220, 221 evolutionary influences 214–216, 217, 218, 219 Hutchinsonian ratio 211 influence of feeding traits 215–216, 217, 218, 219 influence of habitat architecture 213–214, 215 influence of life history characteristics 214–216, 217, 218, 219 number of co-occurring species in a guild 211–212, 213
pelagic species size distributions 218, 220 relationship with body size 210–211 size difference between competing species 211, 212 species guilds 211–212, 213 mass-specific metabolic rate (B), equation for 2 maximal population growth rate (rmax) 7, 8 megafauna 111–112 meiofauna 111–112 Menidia menidia (Atlantic silverside) 41 metabolic energy flux, consumer:producer ratios 11–12 metabolic rate and body size in aquatic ecosystems 4–5 effect of temperature 2 factors affecting 3 relation to body size 1–2 metabolic theory of ecology (MTE) 98, 226 and body size (summary) 327–329 application to biomass turnover (P/B) and body size 60, 62–68 applications 2–3, 13 conceptual framework 3 development of 1–2 effects of body size 2–3 flow of energy and materials between trophic levels 11–12 food web structure and body size 11–12 in aquatic ecology 2–3 predictions 3 quantitative links from individual to ecosystem levels 10–11 synthesis of approaches to food webs 12 metaphoetesis 306–307 microbial species absolute abundance and dispersal 169–171 cosmopolitan-biogeography transition size 171–172 cosmopolitan distribution hypothesis 167–169 endemism question 169 general validity of geographical distribution data 180–181 geographical restriction hypothesis 167–169 neutral dispersal hypothesis 167–169 neutral theory of community structure 181–182 size dependence of local:global species ratios 181–182 size effects on biodiversity 169 size effects on dispersal and distribution 167–169 size range 170 see also cosmopolitan distribution theory microbial systems (experimental) 246–247 community consequences of body size 249–250, 252–256, 257, 258, 259, 260, 261–262 description of the experiments and data sets 248–249 discussion 259–263 ecosystem consequences of body size 250–251, 256–259, 260, 262–263 methods 247–251
339
340
INDEX
microbial systems (experimental) (cont.) population consequences of body size 249, 251–252, 253, 254, 255, 259–261 results 251–259, 260 species choice 247–248 mortality rate and body mass 7–8 mortality selection in fish species, evolutionary responses 40–41, 42 MTE see metabolic theory of ecology multifractal behaviour in ecosystems 142–143 empirical applications 142 multifractal properties of PSDs in river ecosystems 156–159 multifractal SARs 159–163 multifractality of BSDs in river ecosystems 155–156, 157, 158–159 mussels (Dreissena spp.), nutrient excretion 289 Mycoplasma, body size 1 Neobursaridium gigas 175, 181 neutral theory of community structure 181–182 New Zealand streams database analysis 85–88, 89, 90, 91, 92 algal productivity effects 88, 90, 92 biomonitoring tools 94 body size patterns 90–94 disturbance effects 88, 90 factors affecting body size 88, 89, 90, 91, 92 fish 85 fish-predation effects 88, 92 land-use effects 88, 91 macroinvertebrates 85–86 methods 85–86 physicochemical conditions 86, 87 Principal Components Analysis 88, 92 results 88, 89, 90, 91, 92 statistical analyses 86 study sites 85 substrate effects 88, 90, 92 normalization constant 1–2 North Atlantic cod (Gadus morhua) 237 northern cod (Gadus morhua) 40–41 nutrient cycling in aquatic systems consequences of variation by size 293–300 direct and indirect regulation 286–287 effects of harvesting-induced changes in size structure 298–300 influence of body size 287 nutrient flux estimation from biomass size distributions 294–296 predator impact on prey size structure 296–297, 298 regulation by aquatic animals 286–287 variation in body-size distributions 293–294 nutrient excretion in aquatic systems chemical form of excreted N and P 287 fed and unfed animals 288–289, 290 influence of body size 287–292 influence of temperature 288–289, 290 rates of N and P excretion 287–290, 291 ratios of N and P excretion 291–292 taxonomic differences 289–290, 291
nutrient relations and ecological stoichiometry 2 nutrient translocation in aquatic systems and body size 292–293 and speed of movement 292–293 behavioural constraints on home range 292–293 Onchorhynchus spp. (Pacific salmon) 292–293 ontogenetic development and community structure 235–239 dynamics 231–234, 235 rate and body size 5–6 see also growth and development ontogenetic dietary shifts in predators 104–105 ontogenetic size refugia for prey species 107, 108, 109–110 ontogenetic variation in body size 226 optimality (optimization) model (of life history) 34–35 fitness definition 34–35 frequency and density dependence 36–37 intrinsic rate of natural increase (r) 34–35, 38–39 lifetime reproductive success (LRS) 34–35, 38–39 relationship of traits to fitness 35 trade-offs among traits 35, 36 P see production p-state (population state) 230–231 Pacific salmon (Onchorhynchus spp.) 292–293 Paraphysomonas 172, 181 parasite and host, lower limit to mass of 307, 310, 314, 321–323 parasite chains 306–308 parasitoid chains 306, 307 particle-size distributions (PSDs) in benthic habitats 142 multifractal properties in river ecosystems 156–159 perch (Perca fluviatilis) 234, 235, 238 phenotypic plasticity in life histories 37–39 phenotypic variation co-gradient 88 counter gradient 39, 40 interpretation 38, 39 phosphorus, as a growth-limiting nutrient 6 physiologically structured population models (PSPMs) 230–231 Plagiopyla frontata 174 Poecilia reticulata (Trinidadian guppy) 41, 42 Polarella glacialis 181 pollution, effects on marine species diversity 216, 218–220, 221 Polycentropus flavomaculata 24 population consequences of body size (microbial systems) 249, 251–252, 253, 254, 255, 259–261 population density and body size (stream communities) 56, 60, 61–62, 65, 68–71 population density at steady state, and metabolic rate 8–9 population dynamics and body size 269–271 and life histories 269–271
INDEX
and species interactions 2 intrinsic rate of increase (rmax) 270–271 responses to additional mortality 270–273 strength of compensation (density dependence) 270–273 population dynamics and individual body size 229–240 alternative states and body size distributions 235–239 cannibalistic system 233–234, 235 E-state (environmental state) 231 emergent Allee effect 236–237 extensions to more complex configurations 239–240 food-dependent development 235–239 i-state (individual) distribution models 230–231 modelling framework 230–231 ontogenetic development and community structure 235–239 ontogenetic development dynamics 231–234, 235 p-state (population state) 230–231 physiologically structured population models (PSPMs) 230–231 size-dependent predation 235–239 size-structured consumer-resource system 231–233, 234 structured biomass community model 239–240 tritrophic food chains 235–239 see also adaptive dynamics life history model population-level feedback and dynamics 229 population maximal growth rate (rmax) 7, 8 population turnover rate, and metabolic rate 7–8 Power Fraction (PF), niche-assembly model 148–152 power-law and scaling relationships in ecosystems 144–145 power-law behaviour in ecosystem features 140–143 power-law relationships food-chain body sizes 306, 307 food-chain body sizes data 310, 314, 321–323 PPMR (ratio of mean predator size to mean prey size) 274 predator chains 306–308 predator–prey interactions and trophic cascades 129–132 energy transfer 274 factors influencing variation in prey choice 101–105 feeding loops 110 individual level processes 100–105 ‘life history’ omnivory 110 networks structured by size 110 ontogenetic dietary shifts in predators 104–105 ontogenetic size refugia for prey species 107, 108, 109–110 particle capture by suspension feeders 24–26 PPMR (ratio of mean predator size to mean prey size) 274 predator impact on prey size structure 296–297, 298 scaling to species 105–110 size-dependent foraging 100–105 size-dependent predation 235–239
size relationships 24–26, 306–308 species-averaging effects in food webs 105–110 upper limits to mass of predator and prey 307, 310, 314, 321–323 production (P) (whole organism metabolic rate) and body size (stream communities) 56, 57, 61, 63, 64, 68–71 equation for 2 individual organisms 5 Proteus vulgaris 247 PSDs see particle-size distributions PSPMs (physiologically structured population models) 230–231 pumpkinseed sunfish (Lepomis gibbosus) 39, 40 quarter-power exponents (b), explanation for 1–2 r (intrinsic rate of increase in a habitat) 34–35, 38–39 rmax (intrinsic rate of increase), response to additional mortality 270–271 rmax (maximal population growth rate) 7, 8 Re see Reynolds number reaction norms 37–39 Redfield ratio 9–10 resource supply and abundance 8–9 Reynolds number (Re) 17–18, 293 collecting elements of suspension feeders 19, 20, 21 river ecosystems data collection 143–144 density and biomass scaling 146–148 density-body mass scaling 144–145 density-body mass scaling with sample area 145–146, 147 fractal properties of size-structured communities 152–159 multifractal behaviour 142–143 mulitfractal properties of BSDs 155–156, 157, 158–159 multifractal properties of PSDs 156–159 multifractal SARs 159–163 power-law and scaling relationships 144–145 scale-related patterns 142–143 river otter (Lontra canadensis) 292–293 RNA, influence on growth rate 6 roach (Rutilus rutilus) 238 SADs see species-abundance distributions sapuara (Semaprochilodus kneri) 293 SARs see species-area relationships scale-invariance of power laws 144–145 scale invariance properties of ecosystems 140–143 fractal properties of size-structured communities 152–159 scaling constraints 227–228 scaling exponent (b) 1–2, 43–44, 45 secondary production rate and biomass turnover rate 55–56, 68–71 self-similarity, fractal properties of size-structured communities 152–159 Sequoia trees, size of 1 Serratia marcescens 247
341
342
INDEX
Simulium (blackfly) larvae, feeding strategy 28 size-spectrum changes in stressed ecosystems 112–113 size-structured consumer-resource system 231–233, 234 social hunters 306–307 solid–fluid interfaces, and body size 21, 26–28 species-abundance distributions (SADs) mechanisms 146, 148–152 scale-invariance and fractal properties 141–142 species-area curves 141–142 species-area relationships (SARs) multifractal properties 159–163 scale-invariance and fractal properties 141–142 species-averaging in food webs 105–110 species complexity and habitat complexity 129–132 and trophic cascades 129–132 species diversity, and total biomass 246–247 species diversity (marine systems) benthic species adult body size distributions 212–216, 217, 218, 219 benthic species all-animal body size distributions 216–217, 219, 220 Dyar’s constants 211, 212 effects of disturbance 216, 218–220, 221 effects of pollution 216, 218–220, 221 evolutionary influences 214–216, 217, 218, 219 Hutchinsonian ratio 211 influence of feeding traits 215–216, 217, 218, 219 influence of habitat architecture 213–214, 215 influence of life history characteristics 214–216, 217, 218, 219 number of co-occurring species in a guild 211–212, 213 pelagic species size distributions 218, 220 relationship with body size 210–211 size difference between competing species 211, 212 species guilds 211–212, 213 species interactions and population dynamics 2 stoichiometry, whole body carbon:phosphorus ratio 6 stream habitats and life history traits 77–78 community size structure 78 species traits related to disturbance regimes 78 trophic roles 77 see also biomass turnover; environmental gradients streams database analysis (New Zealand) 85–88, 89, 90, 91, 92 algal productivity effects 88, 90, 92 biomonitoring tools 94 body-size patterns 90–94 disturbance effects 88, 90 factors affecting body size 88, 89, 90, 91, 92 fish 85 fish-predation effects 88, 92 land-use effects 88, 91 macroinvertebrates 85–86 methods 85–86
physicochemical conditions 86, 87 Principal Components Analysis 88, 92 results 88, 89, 90, 91, 92 statistical analyses 86 study sites 85 substrate effects 88, 90, 92 strength of compensation (density dependence), response to additional mortality 270–273 stressor-induced size-spectrum changes 112–113 stressor-induced species loss 112–113 structured biomass community model 239–240 Sugihara Fraction (SF), niche-assembly model 148–152 suspension feeders active 16–17 body size and flow regime (Reynolds number, Re) 17–18 body size and food availability 21, 26–28 body size and solid-fluid interfaces 21, 26–28 body-size measurement difficulties 17 body size relation to food particle size 24–26 collecting elements 16 colonial or clonal animals 22–23 costs and benefits of feeding structure size increase 19–24 dependence on flow characteristics 17–18 deposit-suspension feeders 16–17 development of gelatinous bodies 24 feeding structure 16 hydrodynamic implications of body size 17–18 influence on local sediment deposition 26–27 limits to maximum body size 19–24 modification of body parts into feeding structures 23–24 ontological shifts between flow regimes 19 overcoming velocity gradients 27–28 particle capture as predator–prey relationship 24–26 particle encounter mechanisms 18 passive 16–17 range of feeding adaptations 18 Re of collecting elements 19, 20, 21 role in ecosystem energy transfer 16 roles in the food web 16 transport of particles to feeding structure (particle flux) 16–17 use of external capture apparatus 24 vortex shedding rate and body size 26–27 temperature and lifespan 6–7 effect on metabolic rate 2 temperature correction 3–4 temperature-size rule (TSR) 42 terrestrial ecosystems role of body size 98–99 trophic cascades 134 Tracheloraphis caudata 174 Trichoptera (caseless caddis flies) larvae, use of external feeding structures 24 Trinidadian guppy (Poecilia reticulata) 41, 42 tritrophic food chains 235–239
INDEX
trophic cascades alternative equilibria 132–133 and biomass distribution 122 and habitat complexity 129–132 and predator–prey interactions 129–132 and species complexity 129–132 body-size disparity between predators and prey 119–122 body-size effects 133–134 body size of primary producers 127–128 conditions which lead to 119 definition 118–119 effects of herbivory 127–128 in terrestrial ecosystems 134 predation effects 118 shape of the trophic pyramid 128–131
system productivity and food web structure 122–124, 125–126 turnover rate of primary producers 127–131 trophic roles, in stream ecosystems 77 unified theory of biodiversity (UTB) 149–152 vertebrate macrofauna 111–112 vortex shedding rate, and body size 26–27 whales, body size 1 within-species variation in body size 226 Zero-Sum Multinomial (ZSM) distribution, dispersal-assembly model 148, 149–152 zooplankton eggs, ontogenetic development rate and body size 5–6
343