Cage Aquaculture, 3rd Edition

Author: Malcolm Beveridge

602 downloads 4218 Views 5MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Cage Aquaculture Third Edition

Malcolm C. M. Beveridge

Cage Aquaculture

© 1996, 2004 by Blackwell Publishing Ltd Editorial Ofﬁces: Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK Tel: +44 (0)1865 776868 Blackwell Publishing Professional, 2121 State Avenue, Ames, Iowa 50014-8300, USA Tel: +1 515 292 0140 Blackwell Publishing Asia Pty Ltd, 550 Swanston Street, Carlton, Victoria 3053, Australia Tel: +61 (0)3 8359 1011 The right of the Author to be identiﬁed as the Author of this Work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. First edition published 1987 Second edition published 1996 by Fishing News Books, a division of Blackwell Science Third edition published 2004 by Blackwell Publishing Library of Congress Cataloging-in-Publication Data Beveridge, Malcolm C. M. Cage aquaculture/Malcolm C. M. Beveridge. – 3rd ed. p. cm. Includes bibliographical references (p. ). ISBN 1-4051-0842-8 (pbk. : alk. paper) 1. Cage aquaculture. 2. Cage aquaculture – Environmental aspects.

I. Title.

SH137.3.B48 2004 639.8 – dc22 2004000842 ISBN 1-4051-0842-8 A catalogue record for this title is available from the British Library Set in 10/12 pt Sabon by SNP Best-set Typesetter Ltd., Hong Kong Printed and bound in India by Replika Press Pvt Ltd, Kundli The publisher’s policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp processed using acid-free and elementary chlorine-free practices. Furthermore, the publisher ensures that the text paper and cover board used have met acceptable environmental accreditation standards. For further information on Blackwell Publishing, visit our website: www.blackwellpublishing.com

Contents

Preface Acknowledgements 1 Cage Aquaculture – Origins and Principles 1.1 1.2 1.3

Principles of aquaculture Rearing facilities The origins of cage culture

2 Cage Aquaculture – An Overview 2.1 2.2 2.3 2.4

Diversity of cage types Cages and cage aquaculture Cage culture and aquaculture Advantages and disadvantages of cage culture

3 Cage Design and Construction 3.1 Shape, size and materials 3.2 Traditional designs 3.3 Modern designs Appendix 3.1 Current force on a single panel of a net cage (from Løland 1993a) Appendix 3.2 Example of cage ﬂotation computation Appendix 3.3 Calculation of the buoyancy of a 3 ¥ 3 ¥ 3 m bamboo cage (see section 3.3.2) 4 Site Selection 4.1 4.2 4.3 4.4

Environmental criteria for farmed aquatic species Environmental criteria for cages Site facilities and management Concluding remarks

5 Environmental Impacts and Environmental Capacity 5.1 5.2 5.3 5.4

Resource consumption The cage aquaculture process Wastes Modelling environmental capacity

vii vii 1 2 4 6 9 9 14 22 24 32 33 37 40 107 109 110 111 111 134 151 155 159 159 163 164 183 v

vi

Contents

Appendix 5.1 Example of intensive cage rainbow trout production assessment for a temperate natural lake (see section 5.4.1) (modiﬁed from Beveridge 1984a) Appendix 5.2 Example of extensive cage tilapia production for a tropical reservoir (see section 5.4.1) (modiﬁed from Beveridge 1984a) Appendix 5.3 Example of semi-intensive cage tilapia production assessment for a tropical lake (see section 5.4.1) (modiﬁed from Beveridge 1984a) 6 Management 6.1 6.2 6.3

Transport and stocking Feeds and feeding Routine management

7 Problems 7.1 Currents 7.2 Disease 7.3 Drifting objects 7.4 Fouling 7.5 Oxygen 7.6 Security 7.7 Predators and scavengers 7.8 Wastes 7.9 Weather and climate Appendix 7.1 Example of calculation for a aeration system design for a freshwater rainbow trout cage, assuming airlift pumps are employed References Index

198

199

199 201 201 209 226 240 241 243 250 251 256 265 265 275 281

307 308 361

Preface

Since the ﬁrst edition of this book seventeen years ago, aquaculture has consolidated its position as an important means of producing food and a contributor to global food security. Cage aquaculture too has continued to expand. While undoubtedly there is more caged ﬁsh production in fresh waters than in marine environments, there has been much expansion in the intensive rearing of species such as Atlantic salmon – a ﬁfteen-fold increase in as many years – sea bass and sea bream in coastal environments. The third edition tries to maintain the original aim of producing a synthesis of information on cages and cage aquaculture practices. The past ten years have seen tremendous advances in the body of knowledge pertaining to aquaculture. For example, studies of the behaviour of farmed aquatic animals have resulted in improved welfare, growth and survival of stock and reductions in wastes. However, if cage aquaculture is to continue to develop and contribute to global food supplies, its reliance on environmental goods and services must be fully considered. Context is important and judgements on resource use, economic and social impacts must be made in the widest possible context, including alternative means of food production. With expansion and intensiﬁcation of production methods, integration with other users of coastal and freshwater environments, too long ignored, is now crucial. As in previous editions, this book is intended as a source or reference book rather than as a practical manual and the new edition contains many new references. Its format is little altered, although the balance between sections has been changed to accommodate new information and to reﬂect redundancy in certain practices. I have included little information on cage or equipment suppliers but refer readers to the internet, to trade papers such as Fish Farmer, Fish Farming International, Northern Aquaculture, and to the trade directories published by the European Aquaculture Society and Aquaculture Magazine.

ACKNOWLEDGEMENTS While I have recently moved on to pastures new, this book is very much the product of my long association with the Institute of Aquaculture, University of Stirling – I could not have written it if I had not been a member of staff there for more than twenty years. I am particularly indebted to my former colleagues, especially Donald Baird, Paul Bulcock, Arturo Chacon Torres, Yrong Song Chen, Roy Clarke, Sylvain Huchette, Kim Jauncey, Sunil Kadri, Liam Kelly, Dave Little, James Muir, Kenny McAndrew, Anne Nimmo, Oscar Pérez, Mike Phillips, Lindsay Ross, Fernando Starling, Alan Stewart, Billy Struthers, Trevor Telfer and Md. Abdul Wahab. vii

viii

Preface

I thank the organizations that have supported me over the years in my work on cages, especially the Overseas Development Administration of the UK Government (now the Department for International Development, DFID) and the Highlands and Islands Development Board (now Highlands and Islands Enterprise, HIE). The Food and Agriculture Organization of the United Nations awarded me an Andre Mayer Fellowship to work at the College of Fisheries, University of the Philippines, and the short period spent in that wonderful country greatly inﬂuenced my views of aquatic environments, their conservation and management. Many people have been involved in the evolution and development of this edition of the book and I would particularly like to thank the individuals and organizations who provided information and photographs: Mr Ismael Awang Kechik, Mr Håkan Berg, Dr Asbjorn Bergheim, Mr Alastair Blair, Dr Peter Blyth, Dr Giles Boeuf, Dr Alastair Bullock, Dr Andre Coche, Mr Richard Collins, Dr James Deverill, Dr David Edwards, Dr Magnus Enell, Fusion Marine (One-steel and Mr Coulsen), Dr John Hambrey, Dr John Hargreaves, Dr Steve Hodson, Professor John Huguenin, Dr Kim Jauncey, Professor J. Katoh, Professor Nils Kautsky, Dr Liam Kelly, W&J Knox, Dr M. Kuwa, Dr Geir Løland, Mr Ian Macrae, Mr Ken McAndrew, Professor James Muir, Dr Mike Phillips, Dr Roger Pullin, Professor Ron Roberts, Dr Derek Robertson, Sadco-shelf, Professor T. Sano, Professor Christina Sommerville, Dr Alan Stewart, Stirling Environmental Services, Dr Trevor Telfer, Dr Max Troell, Mr Barney Whelan and Dr F Willumsen. Many people also helped with the production of illustrations and graphs: Graham Brown, Rachel Delaney, Brian Howie, Liam Kelly and Denise Macrae at the University of Stirling and David Hay at Fisheries Research Services, Scotland. To all, I owe a debt of thanks. Last, but not least, I would like to thank my family – Maggie, Sandy and Charlotte – for their patience and constant support.

Cage Aquaculture, Third Edition Malcolm C. M. Beveridge Copyright © 1996, 2004 by Blackwell Publishing Ltd

Chapter 1

Cage Aquaculture – Origins and Principles

Aquaculture is the aquatic counterpart of agriculture and its origins extend back some 4000 years (Beveridge & Little 2002). However, unlike agriculture, which has been the most important way of obtaining food on land for several thousand years, aquaculture has until recently contributed little in real terms to world ﬁsh or shellﬁsh production. Instead of evolving towards cultivation, hunter–gatherer methods of procuring food from the aquatic environment developed along a different path: by improvements in ﬁnding prey and by increases in killing power. There are several reasons why agriculture and aquaculture did not develop in the same way. First, food in the lakes and seas has, until recently, been abundant. Increases in ﬁshing pressure and development of ﬁsheries technology were sufﬁcient to meet growing demands and there was, therefore, little need to learn to farm. Moreover, the aquatic environment was hostile and something to be feared. It must have seemed impossible that a structure that could hold ﬁsh securely and withstand the forces of the tides and currents, waves and storms, could be built in the sea. There were other technical problems, too, to overcome. While the breeding and husbandry of animals and the harvesting and planting of seeds was readily achieved on land, it has proved difﬁcult to breed many aquatic species, to hatch the eggs and to successfully rear the offspring. The problems in part stemmed from the fact that people were dealing with organisms that were very different from themselves and with an environment about which they were largely ignorant. It was not until the rise of the biological sciences in the 19th century that the mysteries surrounding the physiology and reproduction of aquatic animals, and the role the environment played in controlling these processes, began to be solved. World demand for ﬁsh, both as a source of food for human consumption and for reduction to ﬁshmeal, has grown at a steady pace since the end of World War II. Until recently demands were met by the expansion of capture ﬁsheries. Growth was around 5% during the 1950s and 1960s, increasing to 8% during the 1980s, and to 10% per annum during the past decade, production peaking at 95 million tonnes in 2000 (FAO 2003) (Fig. 1.1). However, when production from China is excluded, the supplies of ﬁsh for human food have changed little since the mid-1980s. There is a dwindling number of conventional stocks that can sustain further increases in exploitation and the situation has been exacerbated by steep increases in fuel oil prices, the development of economic exclusion zones (EEZs), the over-capitalization of many ﬁshing ﬂeets and profound anthropogenic changes to the very ecosystems upon which ﬁsheries depend. 1

2

Chapter 1

Fig. 1.1 Growth, real and projected, in world capture ﬁsheries production, world ﬁsh culture and population (data from various sources).

Over the next 25 years or so capture ﬁsheries landings might remain stable, providing appropriate management of stocks and development of new ﬁsheries can be achieved and providing that novel ﬁsh products can be successfully marketed. All the indications suggest that by the end of the ﬁrst quarter of the 21st century farmed ﬁsh production will approximate that from capture ﬁsheries production and be the most important means of providing ﬁsh for food. This scenario, however, takes no account of likely future shortages in some of the raw materials required for intensive aquaculture and ignores growing constraints on land and water availability (Beveridge et al. 1994b, 1997b; Pauly et al. 1998; Naylor et al. 1998, 2000), assuming instead that human ingenuity will rise to the challenges.

1.1

PRINCIPLES OF AQUACULTURE

Fisheries and aquaculture share the same aim: to maximize the yield of useful organisms from the aquatic environment. The classical theories of Russell (1931) and Beverton & Holt (1957) have determined that the size of exploitable stocks is determined by four factors: recruitment rate, growth rate, natural mortality rate and ﬁshing mortality rate (Fig. 1.2). Capture ﬁsheries try to maximize yields by increasing ﬁshing mortality rate, partly at the expense of natural mortality, although if too many ﬁsh are killed recruitment and growth are unable to compensate and stocks dwindle. Aquaculture, on the other hand, seeks to increase yields by manipulation of all four population-regulating factors: growth, reproduction, recruitment and natural mortality rates.

Cage Aquaculture – Origins and Principles

3

Fig. 1.2 Factors governing exploitable stock biomass (redrawn from Pitcher & Hart 1982).

Aquaculture began independently in different societies, both agriculture- and ﬁshing-based, and followed a pattern of development in many respects similar to that of agriculture. The control of natural mortality through the capture and holding of ﬁsh and shellﬁsh while they increased in biomass or value was probably an early achievement (Beveridge & Little 2002). The simplest facilities to construct would have been earth ponds, possibly little more than mud walls built to temporarily hold water and ﬁsh following the seasonal ﬂooding of a river. Manipulation of growth through feeding with household scraps or agricultural wastes would have been a logical next step. However, with one or two notable exceptions, such as that of carp in China (Li 1994), control of spawning and recruitment is comparatively recent as it is difﬁcult to induce many species to breed in captivity. There are also many technical problems involved in the hatching of eggs and the maintenance and feeding of larval and juvenile stages (Bardach et al. 1972). Aquaculture has gradually gained control over all four of these population-determining processes. Recent decades, in particular, have seen great advances in the ﬁelds of nutrition, genetics, engineering, physiology and biochemistry, resulting in hugely improved yields. In summary, aquaculture, or the farming of aquatic organisms, is achieved through the manipulation of an organism’s life cycle and control of the environmental variables that inﬂuence it. Three main factors are involved: control of reproduction, control of growth and elimination of natural mortality agents. Control of reproduction is essential otherwise farmers must rely on naturally spawning stocks. The supply of fry from the wild may be restricted to a particular season and a particular area, and there may also be shortages due to over-exploitation of wild stocks. This step remains to be realized in the culture of many, particularly marine, species. Growth can be increased through selection of broodstock (control of breeding is, therefore, a prerequisite), and through

4

Chapter 1

feeding. While the culture of carnivorous species is dependent upon the supply of largely ﬁshmeal-based diets, there is considerable scope for minimizing feed costs providing that the appropriate omnivorous/detrivorous/planktivorous species and systems are used. Finally, rearing systems are essential to all types of aquaculture. They are designed to hold organisms securely while they increase in biomass by minimizing losses through predation and disease and by excluding competitors (Reay 1979). Rearing systems must also facilitate management.

1.2

REARING FACILITIES

Rearing facilities for ﬁsh can either be land-based or water-based, the former including ponds, raceways, tanks and silos and the latter comprising enclosures, pens and cages. In dictionaries, the terms ‘enclosure’, ‘pen’ and ‘cage’ appear synonymous and may be used interchangeably. However, in aquaculture this has given rise to a degree of confusion, the term ‘enclosure’ often being used to describe something which could either be a cage or a pen and the word ‘pen’ being used in North America to denote a large sea cage. The terms are used here in a more restricted sense. ‘Enclosure’ is used to denote an enclosed natural bay, where the shoreline forms all but one side, which is typically closed off by a solid, net or mesh barrier (Fig. 1.3a). In pen culture, all sides of the structure, except for the bottom, are man-made, often being constructed from wooden

(a) Fig. 1.3 Water-based aquaculture systems. (a) Enclosure (Lake Buhi, Philippines); (b) ﬁshpen (Laguna de Bay, Philippines); (c) cages (Ireland – courtesy M. J. Phillips).

Cage Aquaculture – Origins and Principles

5

(b)

(c) Fig. 1.3 Continued.

poles and netting (Fig. 1.3b). The bottom of the pen, however, is formed by the sea bed. Cages, by contrast, are enclosed on the bottom as well as the sides by wooden, mesh or net screens (Fig. 1.3c). There are other differences among water-based rearing facilities. Pens and enclosures tend to be larger, ranging in size from around 0.1 ha to some that exceed 1000 ha in area. Cages, however, typically have a surface area somewhere between 1 m2 and 1000 m2. Moreover, because of their small size, cages are better suited to intensive culture methods than pens.

6

1.3

Chapter 1

THE ORIGINS OF CAGE CULTURE

Cages were probably ﬁrst used by ﬁshermen as a convenient holding facility for ﬁsh until ready for sale (Beveridge & Little 2002). The earliest cages may have been little more than modiﬁed ﬁsh traps or baskets, and such traditional types of holding facility have been in use in many parts of the world for generations. True cage culture, in which ﬁsh or other organisms were held for long periods of time while they increased in weight, was until recently thought to be a comparatively modern development. According to Li (1994), however, cage culture was established in China during the Han Dynasty 2200–2100 years ago. In the ﬁrst apparent written account Hu (1994) relates how Zhou Mi described fry sales in the ancient Jiujiang River in an appendix (‘Beiji’) to a book entitled Kuixinzhashi written in 1243 during the Sung Dynasty (AD 960–1280). The fry ‘. . . reach home and are placed in cloth cages in open water with bamboo sticks supporting the four corners. The fry actively move about in the cages with the waves as if they enjoy playing. One or half a month later, the fry grow bigger for marketing. The cloth cages are ﬁne-meshed cages. The bamboo sticks serve to frame the cages in which the caught fry are nurtured.’ The culture of fry in what were probably small ﬁxed cages must have been for grow-out purposes in ponds. Hu (1994) gives a further account of fry catching in the Jiujiang River in the late 1840s and describes how in 1876 foreign visitors to Jiangxi described fry catching and temporary holding in cages. In the Great Lake region of Cambodia ﬂoating cages have been used since the end of the 19th century (Lafont & Savoeun 1951; Hickling 1962; Ling 1977; Pantulu 1979). Snakeheads (Channa spp.), catﬁsh (Pangasius spp., Clarias spp.) and marble-headed gobies (Oxyeleotris marmorata) were held in wood or bamboo cages, fed on a mixture of kitchen scraps and trash ﬁsh and transported by river to the markets of Phnom Penh. Cages were either towed behind the boats or occasionally incorporated into the vessel to form a well-boat (Fig. 1.4a, b). During the 20th century, this type of cage culture spread to most parts of the lower Mekong delta and into Vietnam (Pantulu 1979; Tuan et al. 2000). In Mungdung Lake, Sulawesi, Indonesia, ﬂoating bamboo cages have been in use since the early 1920s (Reksalegora 1979) to rear Leptobarbus hoeveni fry captured from the lake. A different form of cage culture appeared in Bandung, Indonesia, around 1940. Small bamboo and ‘bulian’ wood cages were anchored to the bottom of organically polluted rivers and canals and stocked with common carp, Cyprinus carpio, which fed on wastes and invertebrates carried in the current (Vass & Sachlan 1957; Costa-Pierce & Effendi 1988). Traditional cage culture, distinguished by its reliance on natural construction materials and on natural or waste feeds, is still practised in parts of Indonesia and Indo-China (Fig. 1.5). However, these traditional cage-rearing practices had a localized inﬂuence and did not directly give rise to modern cage ﬁsh farming methods (see also Hu 1994). Modern cages utilize synthetic mesh or netting materials and have collars usually fabricated from synthetic polymers and metals, although wood is still widely used in many designs. It is difﬁcult to be precise about the origins of modern cage ﬁsh farming although Japan was undoubtedly a key inﬂuence. According to Milne (1974), Professor Harada, Director of the

Cage Aquaculture – Origins and Principles

7

(a)

(b)

(c) Fig. 1.4 Traditional ﬁsh cage designs, Indo-China. (a) Southern Vietnam; (b) boat-shaped cage from Cambodia; (c) battery of small cages, Cambodia (from Pantulu 1979).

Fisheries Laboratory at Kinki University, ﬁrst started experimenting with cage ﬁsh culture in 1954 and commercial culture of yellowtail Seriola quinqueradiata followed three years later. In Norway, cages were being used to culture Atlantic salmon (Salmo salar) in the early 1960s and in Scotland the White Fish Authority commenced salmon cage rearing trials around 1965. Surprisingly, tilapia (Oreochromis spp.) culture in cages is of even more recent origin and

8

Chapter 1

Fig. 1.5 Cage aquaculture of cobia (Rachycentron canadum) in household cages, Ha Long Bay, Vietnam (courtesy M. J. Phillips).

owes its beginnings to work carried out at Auburn University in the late 1960s (Schmittou 1969). Modern cage farming is thus very much a phenomenon of recent decades. In the following chapter the relative importance of aquaculture in cages and the factors determining its current status are explored.

Cage Aquaculture, Third Edition Malcolm C. M. Beveridge Copyright © 1996, 2004 by Blackwell Publishing Ltd

Chapter 2

Cage Aquaculture – An Overview

Cages are highly versatile and lend themselves to being used in many different ways. This section is intended as an overview of the subject and many of the themes touched on will be explored in detail in later chapters.

2.1 DIVERSITY OF CAGE TYPES Cages have developed a great deal from their humble origins and today there is an enormous diversity of types and designs. They may be classiﬁed as shown in Fig. 2.1 (see also Huguenin 1997). There are four basic types: ﬁxed, ﬂoating, submersible and submerged. Fixed cages consist of a net bag supported by posts driven into the bottom of a lake or river (Fig. 2.2a). They are in common use in some tropical countries such as the Philippines. Fixed cages are comparatively inexpensive and simple to build, although they are limited in size and shape and their use is restricted to sheltered shallow sites with suitable substrates. The bag of a ﬂoating cage is supported by a buoyant collar or, in some cases, a frame. This type is by far the most widely used and can be designed in an enormous variety of shapes and sizes to suit the purposes of the farmer. Floating cages are also less limited than most other designs in terms of site speciﬁcations. Some ﬂoating types are designed to rotate in order to control fouling (see Chapter 7). The design in Fig. 2.2b rotates about a central axis incorporated into the collar while other designs are rotated by means of moving the ﬂotation elements or by adjusting the buoyancy of the frame members (Fig. 2.2c). The much more widely used non-rotating ﬂoating types can be constructed with wide or narrow collars. The former are common on larger cages and serve as work platforms, facilitating many of the routine farm tasks (Fig. 2.2d). Most wide collars are designed to be rigid although some are ﬂexible so that they may be used at more exposed sites (Fig. 2.2e; see also Chapters 3 and 7). Simple and inexpensive ﬂexible-collar narrow cages can be fabricated using rope and buoys (Fig. 2.2f), but in practice are difﬁcult to manage. Rigid narrow collars, constructed from glass ﬁbre or steel section and buoys (Fig. 2.2g), are popular in Western Europe despite the fact that routine operations must be performed from a boat or pontoon. Rigid mesh designs must, of course, utilize a rigid collar. Some ﬂoating net bag designs, including early designs for ﬂatﬁsh culture, have a solid bottom (Hull & Edwards 1979). Neither net nor rigid mesh bag submersible cages have a collar, but instead rely on a frame or rigging to maintain shape. The advantage over other designs 9

10

Chapter 2

FIXED

FLOATING

Nonrotating

Rotating

Wide collar

Narrow collar

Rigid collar

Flexible collar

Rigid bag

Flexible bag Net floor

SUBMERSIBLE

With central axle

Suspended from surface

Without central axle

Rotation by adjustment of float buoyancy

Rigid

SUBMERGED

Adjustable buoyancy

Flexible

Rotation by adjustment of float position

Solid floor

Fig. 2.1 A classiﬁcation system for cages (developed from Kerr et al. 1980).

(a) Fig. 2.2 Different types of cages. (a) Fixed cage (SEAFDEC, Laguna de Bay, Philippines); (b) rotating cage with central axle (Kiel, Germany); (c) submersible, rotating cage without central axle (Dunstaffnage, Scotland – courtesy A. Blair); (d) ﬂoating, wide collar milkﬁsh broodstock cage (SEAFDEC, Philippines – courtesy R. S. V. Pullin); (e) ﬂexible wide collar design (Ireland – courtesy B. Whelan); (f) ﬂexible, narrow collar cage (Lake Titicaca, Bolivia); (g) rigid, narrow collar frame (Scotland – courtesy I. H. Macrae); (h) submersible cage (Germany); (i) submerged cage (Java, Indonesia – courtesy C. Sommerville).

Cage Aquaculture – An Overview

11

(b)

(c) Fig. 2.2 Continued.

is that the position in the water column can be changed to exploit prevailing environmental conditions. The cages are typically kept at the surface during calm weather and are submerged during adverse weather or during a harmful algal event. Some submersible designs rely on the bag being suspended from buoys or

12

Chapter 2

(d)

(e) Fig. 2.2 Continued.

a ﬂoating frame on the water surface (Fig. 2.2h) while others have variable buoyancy (see later in Fig. 7.31). While a number of submerged cage designs have been proposed (see Huguenin & Rothwell 1979; Huguenin 1997), far fewer have gone beyond the design

Cage Aquaculture – An Overview

(f)

(g) Fig. 2.2 Continued.

13

14

Chapter 2

(h)

(i)

Fig. 2.2 Continued.

concept stage or indeed have been built or widely used. Simple submerged cages, which are little more than wooden boxes with gaps between the slats to facilitate water ﬂow and are anchored to the substrate by stones or posts (Fig. 2.2i), have been used to culture common carp in ﬂowing waters in Indonesia (Vass & Sachlan 1957; Costa-Pierce & Effendi 1988) and to culture lobster in coastal waters in Vietnam (Tuan et al. 2000). Submerged net mesh bag designs have been used in lakes and reservoirs in the former USSR and in China (Martyshev 1983; Li 1994). However, it remains a moot point whether all species readily adapt to rearing in submersible or submerged cages (see sections 7.9.1, 7.9.2). Despite the fact that cage designers and manufacturers have produced all sorts of designs in the past half-century or so, the range of cage types today is, if anything, smaller than it was a decade ago. Cost, always important, has now become the overriding design criterion, particularly in the industrial-scale farming industries (e.g. salmon farming) and this has led to uniformity in terms of shape, size and materials (see section 3.1).

2.2

CAGES AND CAGE AQUACULTURE

Because they are a relatively inexpensive and convenient way to keep captive aquatic organisms, cages have been used for a variety of purposes, some unre-

Cage Aquaculture – An Overview

15

lated to aquaculture. For centuries, cages were used to hold and transport bait ﬁsh for tuna pole and line ﬁshing (Ben Yami 1978; Takashima & Arimoto 2000) although today their use has been largely superseded by live-bait holds in boats. They have also been used in Norway and Canada, on a trial basis at least, to move ﬁsh such as herring (Clupea harengus harengus) and pollack (Pollarchius pollarchius) to impoundments (see Kreiberg & Solmie 1987 for a review) and in Japan for keeping ﬁsh caught in traps until ready for market (Takashima & Arimoto 2000). Caged ﬁsh have been used to monitor the water quality of power station efﬂuents (Holt 1977; Chamberlain 1978) and to monitor environmental quality (Grizzle et al. 1988; Meng et al. 2000). Cage aquaculture has been used to treat the symptoms of eutrophication (Yang 1982; Little & Muir 1987; CostaPierce & Effendi 1988; Chang 1989; Starling et al. 1998; Costa-Pierce 2002) and has also been used in conservation initiatives for species as disparate as seahorses (Hippocampus spp.) (Vincent & Pajaro 1997), frogs (Vines et al. 1996) and giant clams (Tridacna spp.) (Bell et al. 1997), and to produce ﬁsh for putand-take recreational ﬁsheries. Cages may be useful in experimental work where it is important to exclude environmental effects (see Kulikovsky et al. 1994) or as an alternative to replicate ponds (Struve & Bayne 1991), although particular care must be given to experimental design to ensure sufﬁcient replication and to avoid pseudoreplication issues (Hurlbert 1984). But cage culture is primarily targeted at producing food. Cages can be used at different stages of the life cycle: for breeding, fry and ﬁngerling rearing, and/or production of ﬁsh for the table. Cage-based tilapia hatcheries (‘hapas’) were developed in the Philippines in the 1970s (Beveridge 1984b; Smith et al. 1985) and a decade or so later the technology had become widespread in Southeast Asia. Little & Hulata (2000) cite low capital costs and intensive management of broodstock and seed as their primary advantage over other systems. Cage production of fry and ﬁngerlings of freshwater ﬁsh species, such as tilapia and carp, is now widespread (Smith et al. 1985; Li 1994; Jayaraj et al. 1998; Anh & Son 2001; Ariyaratne 2001; see also section 2.2.1), as is the use of cages for production of Atlantic salmon smolts in Scotland (Williamson & Beveridge 1994) and elsewhere. The use of cages in growing ﬁsh and crustaceans for the table is also widely discussed in this book. Like other types of aquaculture, cage ﬁsh farming may be classiﬁed on the basis of feed use as extensive, semi-intensive and intensive. In extensive culture ﬁsh rely solely on available natural foods such as plankton, detritus and seston. Semi-intensive culture involves the use of low protein (<10%) feedstuffs, usually compounded from locally available plants or agricultural by-products, to supplement the intake of natural food, while in intensive culture operations ﬁsh rely almost exclusively on an external supply of high protein (>20%) food, usually based on ﬁshmeal (see also Coche 1983).

2.2.1

Extensive cage aquaculture

Various authors have observed caged ﬁsh grazing on the fouling community on cage nets (see Norberg 1999; Huchette et al. 2000). In a cage-based trial in which

16

Chapter 2

Fig. 2.3 Primary production of fresh waters (from Beveridge 1984a).

the surface area available for colonization by fouling organisms was greatly increased, Huchette & Beveridge (2003) attempted to grow tilapias with no additional food. Although production of up to 0.94 g ﬁsh m-2 per day was recorded in the upper 0.5 m of the water column, the system proved not to be economically viable. Further discussion is thus limited to the culture of ﬁsh in cages that are reliant on external supplies of food. Extensive cage culture is restricted to fresh waters and may be practised in two types of environment: highly productive lakes and reservoirs (see Shenoda & Naguib 2000) and water bodies that receive sewage or domestic wastes (see Kibria et al. 1999). Primary production, which fuels all successive energy transactions in aquatic food webs other than in waste-fed systems and systems with high allochthonous inputs (i.e. externally-produced materials, such as leaves), is dependent upon the availability of essential nutrients (for example, phosphorus and nitrogen compounds) as well as light and temperature (Le Cren & LoweMcConnell 1980; OECD 1982). Systems with high nutrient loadings are likely to be highly productive. However, productivity is also strongly correlated with latitude (Brylinsky 1980), and between temperate zones (23–67°N and S) and tropical zones (23°N–23°S) there is a considerable increase in the range of annual primary production values (Fig. 2.3). Tropical water bodies with high nutrient loadings offer the best opportunities for extensive cage culture (Table 2.1). Extensive cage culture on any scale seems to have only been practised in the Philippines and China (Beveridge 1984a; Li 1994; Beveridge & Stewart 1998). Caged bighead carp (Aristichthys nobilis) were used at Selatar Reservoir, Singapore, during the late 1970s and early 1980s to help control eutrophication problems in municipal water supplies (Yang 1982), and similar approaches were employed in reservoirs in Brazil in the 1990s using silver carp

Cage Aquaculture – An Overview

17

Table 2.1 Correlation between the plankton biomass and ﬁngerling production of silver carp and bighead carp in reservoirs (from Li 1994). Phytoplankton

Eutrophic Mesotrophic Oligotrophic a

Zooplankton

¥103 cells l-1

mg l-1

ind.a l-1

mg l-1

Stocking density ind.a m-2

>100 30–100 <30

>5 2–5 <2

>2000 1000–2000 <1000

>3 1–3 <1

300–600 100–300 —

Transfer size cm

Fish yield kg m-2

13 13 —

7–15 2–7 —

Individuals.

Fig. 2.4 Inspecting submersible, illuminated cage used to rear pike fry, Germany.

(Hypophthalmichthys molitrix) (Starling et al. 1998). In Saguling Reservoir, Indonesia, cages of tilapia were sited alongside, or even inside, intensive carp cages, in order to exploit the uneaten food and excess plankton (Costa-Pierce 2002). Lights are also sometimes used to attract zooplankton in the extensive cage rearing of juvenile planktivorous stages of both freshwater and marine ﬁsh (Figs 2.2h, 2.4) (Bronisz 1979; Uryn 1979; Jäger & Kiwus 1980; Holm & Møller 1984; Jäger 1986; Mamcarz & Nowak 1987; Mamcarz & Kozlowski 1989; Willinsky et al. 1995b; Fermin & Seronay 1997; Watson et al. 2001). Sewage-fed ponds and streams and rivers subject to high loadings of domestic or agricultural waste have proved suitable for extensive cage culture (Vass & Sachlan 1957; Noble 1975; Gaigher & Krause 1983; Gaigher & Toerien 1985;

18

Chapter 2

Getabu 1987; Edwards 1992; Kibria et al. 1999; Gooley et al. 2000), although there is concern about the public acceptability and health risks associated with ﬁsh grown in such systems (Buras 1993; Howgate et al. 2002). It is possible to depurate such ﬁsh (Buras 1993), however, or incorporate them into diets for other species (Edwards 1992). Planktivores, detrivores and omnivores, including many of the carp and tilapias, are suitable candidates for extensive culture, although little research has been carried out to determine the best species for different conditions. This can be crucial; for example, there are usually few planktonic species present in ﬂowing water (lotic) systems other than in the slower-ﬂowing, lowland reaches. Extensive cage culture of plankton-feeding species in rivers and streams is likely to be impractical. Trials conducted in the Tengi River, Malaysia, with bighead carp illustrate this. Cages were stocked with 25-g ﬁsh at a stocking rate of 15 ﬁsh m-3. During the two-month trial, 95% of ﬁsh died and the average weight of survivors was 20 g (I. A. Kechik, pers. comm.). Similarly, attempts to grow bighead carp in cages without supplementary feeding in the United States proved unsuccessful because of insufﬁcient food (Engle 1982). In the Philippines, very high production of tilapia by extensive cage culture – up to 1.90 kg m-3 per month – has been achieved (Beveridge 1984a) although it is a moot point whether this magnitude of production is sustainable in the long term (see Aquino & Nielsen 1983; Santiago 1994). Nevertheless, extensive cage culture can be an alternative to the management of a resource as a ﬁshery (Beveridge 1984a; Beveridge & Phillips 1988; Beveridge & Stewart 1998; Beveridge & Muir 1999). Extensive cage culture of ﬁsh in conjunction with intensive lake-based ﬁsh rearing is a means of reducing environmental impacts of wastes while increasing the proﬁtability of the venture (Beveridge & Phillips 1988; Costa-Pierce 2002). It may also be desirable to switch between semiintensive and extensive cage culture at different times of the year in order to take advantage of periods of high phytoplankton densities.

2.2.2

Semi-intensive cage aquaculture

In tropical fresh waters, semi-intensive rearing of ﬁsh is the most common method of cage culture and species which feed low in the food chain, such as Oreochromis niloticus, O. mossambicus, O. aureus, and bighead, silver and common carp, are fed on a variety of materials including rice bran, wheat middlings, brewery and domestic wastes (Jangkaru & Djajadiredja 1979; Pantulu 1979; Dela Cruz 1980; Coche 1982; Beveridge & Phillips 1988; Costa-Pierce & Soemarwoto 1990; Beveridge & Stewart 1998; Beveridge & Muir 1999; McAndrew et al. 2002). Semi-intensive cage ﬁsh rearing is also practised to a limited extent in lakes and reservoirs in Eastern Europe (Muller & Varadi 1980; Martyshev 1983). However, other than some experimental work with herbivorous species such as siganids and mullets (Pitt et al. 1977; Tahil 1978) semiintensive cage culture is not practised in marine environments (Fig. 2.5). Further expansion of semi-intensive cage culture is likely to be restricted to tropical fresh waters since few commercially important marine species and temperate fresh-

Cage Aquaculture – An Overview

19

Fig. 2.5 Siganids feeding on ﬁlamentous algae (courtesy R. S. V. Pullin).

water species can utilize the natural foods available to caged species or low protein, plant-based supplemental feeds. Supplementary feeding practices are largely determined by availability and feeds tend to be delivered on an ad hoc basis rather than according to predetermined rules. Indeed, there is little known about the role of supplementary feeding in semi-intensive aquaculture (De Silva & Davy 1992; De Silva 1993) and so it is difﬁcult to provide any ﬁrm guidelines. Feeding methods are discussed in section 6.2. Most types of freshwater site have proved suitable for semiintensive cage culture.

2.2.3

Intensive cage aquaculture

Intensive cage culture is largely conﬁned to the rearing of high-value carnivorous species. In fresh water, salmonids and channel catﬁsh (Ictalurus punctatus) are reared intensively while in the marine environment Atlantic salmon, yellowtail, sea bass (Lates calcarifer, Dicentrarchus labrax) and milkﬁsh (Chanos chanos) are the principal intensively farmed species. Some intensive rearing of caged tilapias and carp – omnivorous ﬁsh which typically have a low market value – is practised in parts of the world where they fetch high market prices (e.g. Europe, North America, Singapore, Taiwan). A limited growing season in these countries also encourages the use of intensive systems. Trash ﬁsh is still the principal type of feed used in yellowtail, grouper and tuna culture. Formulated, pelleted dry diets have been developed and are widely available for salmonid and channel catﬁsh (Ictalurus punctatus) culture and, in the

20

Chapter 2

Philippines, for milkﬁsh and tilapias (Marte et al. 2000). Intensive culture is not recommended in fast-ﬂowing riverine sites as feed losses can be excessive.

2.2.4

Monoculture versus polyculture

The terms ‘monoculture’ and ‘polyculture’ are used here, respectively, to describe the culture of one or several species within a cage. The term ‘integrated aquaculture’ is applied to situations where caged ﬁsh are grown in ponds or in association with other aquaculture systems (ponds, longlines, etc.). In cage culture, monoculture is the rule and polyculture the exception. There are several reasons for this. There are fewer exploitable feeding niches in cages than in pens or ponds. Natural foods are unavailable to caged benthic or macrophyte-feeding species and even for plankton-feeding species food supplies may be limited. Experimental polyculture of Chinese carp in cages was tried in Hungary (Muller & Varadi 1980) but never became a commercial proposition. Sea urchins (Psammechinus miliaris) have been grown in pearl nets suspended in Atlantic salmon cages in Scotland with some success, the sea urchins utilizing waste feed and growing faster during the winter months than their wild counterparts (Kelly et al. 1998). In Brazil, various seaweeds have been grown in commercial shrimp cages, the caged seaweed beneﬁting from the shrimp feed and faecal-derived nutrients and from exclusion of grazers, while the seaweed provides shelter and shade for the shrimp (Lombardi et al. 2001). Polyculture of silver and bighead carp in cages, however, is quite widely practised in China, with yields of up to 7.5 kg m-3 per year from extensive systems and up to 13.5 kg m-3 from semi-intensively managed cages (FAO 1983; Li 1994). Stocking ratios are determined by trial and error, based on plankton biomass and species composition. At ﬁrst sight there would appear to be few advantages associated with growing several top predatory species together in intensive cage aquaculture, and indeed results have been mixed. Trials carried out in Scotland to evaluate the feasibility of growing two high-value species together, turbot (Scopthalmus maximus) and Atlantic salmon, had little success. However, polyculture of channel catﬁsh with rainbow trout (Oncorhynchus mykiss) resulted in signiﬁcantly greater weight gains in the former than when grown in monoculture (Beem et al. 1988). Although no conclusions were drawn, the catﬁsh probably beneﬁted from polyculture conditions through increased availability of food. In another experiment in which channel catﬁsh were cultured with tilapia, growth and production of channel catﬁsh was enhanced, possibly because the vigorous feeding activity of the tilapias stimulated the feeding response of the catﬁsh (Williams et al. 1987). Culture of snapper (Lutjanus johni) and grunt (Pomadasys kaakan) proved more proﬁtable than the culture of either species alone (Hussain & Khatoon 2000), although why has not been fully explained. Not all beneﬁts are food related. Holm (1992) explored various salmon-based polycultures and found improved production when Atlantic salmon parr and Arctic charr (Salvelinus alpinus) were grown together. Salmon dispersed themselves among charr, avoiding conspeciﬁcs, thus reducing the incidence of intraspeciﬁc aggression. With growing pressures to improve proﬁtability and to

Cage Aquaculture – An Overview

21

Fig. 2.6 Poultry facilities mounted above ﬂoating cages, Sri Lanka.

reduce mortalities and waste loadings, there are also increasing incentives to introduce herbivorous ﬁsh to control fouling (see Huy et al. 2001), to use ﬁsh which feed on ectoparasites to control disease and to stock benthos/detritus feeders to utilize uneaten food (see Chapters 5 and 7).

2.2.5

Integrated cage aquaculture

The integration of cage aquaculture with other rural activities was reviewed by Little & Muir (1987). Cage culture may be more difﬁcult to integrate than pond farming; by using existing water bodies the ﬁsh-rearing operations may be at some distance from the crop and animal rearing. Attempts to house poultry or other livestock over cages have rarely proved satisfactory (Fig. 2.6). Integration on any scale is apparent in only a few countries (Chang 1989; Li 1994; Zoran et al. 1994; Gooley et al. 2000; Chopin et al. 2001; Costa-Pierce 2002). In some parts of Asia cages are incorporated into the management of multi-use water bodies or in ponds, while in Israel, Australia and the United States attempts are being made to integrate cage culture with irrigation or wastewater treatment systems in order to improve efﬁciency of water use (Budhabhatti & Maughan 1993, 1994; Haylor 1993; Zoran et al. 1994; Costa-Pierce 1996; Troell & Norberg 1998; Beveridge & Muir 1999; Kibria et al. 1999; Gooley et al. 2000; Yang & Lin 2000). Although the potential for exploiting aquaculture wastes is increasingly recognized (Folke & Kautsky 1989, 1992; Troell et al. 1999; Chopin et al. 2001; McVey et al. 2002) – indeed, some would say necessary – attempts to integrate

22

Chapter 2

intensive cage ﬁsh rearing with other activities in the marine environment remain largely conﬁned to pilot-scale trials. Trials integrating cage aquaculture with seaweed and sea cucumber rearing have been conducted in North America (Mazhari & Petrell 1991; Petrell & Alie 1996; Ahlgren 1998; Chopin et al. 1999; Nelson et al. 2001b), Japan (Hirata et al. 1994, in Chopin et al. 2001) and the Philippines (Hurtado-Ponce 1992). Culture of scallops and mussels, both freshwater and marine, have been widely employed in trials in order to try to address eutrophication problems caused by cage ﬁsh culture (Jones & Iwama 1991; Taylor et al. 1992; Stirling & Okomus 1995; Parsons et al. 1999; Soto & Mena 1999; Mazzola & Sarà 2001). Troell & Berg (1997) modelled the effects of particulate cage wastes on the growth of mussels, but found that only under exceptional circumstances (high biomass of caged ﬁsh, slack tide) was there likely to be any measurable impact on mussel production. Mazzola & Sarà (2001) used carbon isotopic analysis to establish that under Mediterranean conditions (i.e. oligo-mesotrophic conditions, relatively low water movement) waste uneaten feed and faecal material from caged sea bass (Dicentrarchus labrax) and sea bream (Sparus aurata) is taken up in measurable amounts by adjacent mussel (Mytilus galloprovincialis) and clam (Tapes spp.) populations, especially the deepwater clams, both cleaning the environment and enhancing bivalve production. Industrial-waste heated water has also been exploited for aquaculture (Marcello & Strawn 1973; Chamberlain & Strawn 1977; Langford 1983; Schneider et al. 1990; Zmyslowska et al. 2001) (see also section 4.1.1). In Europe, the EC-funded BIOFAQ project was established to look at integrating bioﬁltration with intensive mariculture (http://www.sams.ac.uk/biofaqs). Despite the few tangible examples of successful integrated cage aquaculture, there is evidence of a change in mindset with regard to aquaculture development. In the United States, which still has a relatively small marine aquaculture sector in relation to its coastline, the need to integrate aquaculture with other sectors and to balance intensive ﬁsh production with extensive mollusc or seaweed culture has been recognized (McVey et al. 2002).

2.3

CAGE CULTURE AND AQUACULTURE

More than 150 ﬁsh species and nearly a dozen species of prawn, lobster and crab have been grown in cages, on an experimental basis at least (Coche 1983; Chua & Tech 2002). Some of the most unlikely candidates for domestication and culture in small enclosures are included in this list. Fast-swimming pelagic species, such as the tunas (Thunnus thynnus, Euthynnus pelamis), have been found to adapt readily to the conﬁnes of a ﬂoating cage, albeit that the cages used are large (Piccinetti et al. 1980; Butler 1982; Anon 1992; Doumenge 1996; Petrusevics & Clarke 1998; Gooley et al. 2000). Bottom-dwelling ﬂatﬁsh (e.g. Scopthalmus maximus, Hippoglossus hippoglossus), too, have been successfully cultured (Martinez Cordero et al. 1994) although it may be necessary to install a solid bottom in the cages (Hull & Edwards 1979; Linfoot et al. 1990). Some ﬁsh, such as the estuarine groupers, are crepuscular predators and prefer to hide among rocks and corals and ambush their prey. However, even this life-style

Cage Aquaculture – An Overview

23

Fig. 2.7 Small (4 ¥ 1 ¥ 1 m) ﬁxed cages stocked with juvenile Penaeus monodon, Laguna de Bay, Philippines.

seems to be of little disadvantage when crowded together in cages (Chua & Teng 1980), although the provision of shelter in the form of used car tyres can increase production (Teng & Chua 1979). Prawns are another group of animals that are difﬁcult to culture at high densities. However, if the bottom of the cage is buried in mud, then prawns can be successfully reared, even in small units (Fig. 2.7). Indeed, few commercially important cultured species are absent from Coche’s list, the most notable exception being eels (Anguilla spp.). However, cage culture of eels (A. japonica) has been reported from China using ﬁne mesh net bags (0.6–1.5 cm) (Anon 1988; Li 1994).

2.3.1

Cages and world aquaculture production

In a recent review of world ﬁnﬁsh, crustacean and mollusc culture, the FAO estimated that the annual production in 1998 was nearly 40 million t (FAO 2003). Around 25% of production was accounted for by molluscs and just over

24

Chapter 2

50% by ﬁnﬁsh, with less than 5% of production being of prawns and other crustaceans. Production in fresh water accounts for something like 95% of world ﬁnﬁsh culture and around 90% is from ponds. Thus cage culture at most accounts for 10% or 1.8 million t of freshwater ﬁnﬁsh per annum. In certain sectors of the industry, however, cage culture assumes a particularly important role. Cage production of rainbow trout in Scotland, for example, amounts to some 40% of total production (Williamson & Beveridge 1994; SEERAD 2002). Culture of marine and diadromous ﬁsh is currently somewhere in the region of 1 million t per annum. An estimated 800 000 t is derived from cages, most of the balance being produced in tropical coastal ponds. The yellowtail, tuna and salmon farming industries are almost exclusively cage-based. The culture of prawns and other crustaceans in cages is still in its infancy, commercial production being limited to a few hundred tonnes of penaeid prawn species in Singapore, Thailand and China, and of lobster in Vietnam.

2.4

ADVANTAGES AND DISADVANTAGES OF CAGE CULTURE

In evaluating cages they must be judged against other systems such as pens and ﬂoating enclosures, and ponds, tanks and raceways in terms of: • • • • • • •

resources required for construction and operation; level of technology required for construction; ease of management; adaptability; quality of the ﬁsh reared; social and environmental implications; economic performance.

2.4.1

Technology, management and the farmed product

Although designs for rearing ﬁsh in offshore locations tend to be technically sophisticated, many cages are simple to construct and can be assembled in a day or so using local, unskilled labour. Cages are also readily managed. Observation of stock is easy and, unlike pens, ﬁsh may be harvested relatively easily using little more than a scoop net. Once installed land-based culture systems can be difﬁcult or expensive to change. However, cage farms can be expanded simply by adding a few more cages as experience grows and circumstances allow and, regulations permitting, cages may be readily redeployed to facilitate management. Although cage-grown ﬁsh can succumb to ﬁn and skin damage through abrasion (Moring 1982; Anh & Son 2001) some of the problems are related to aggressive interactions among ﬁsh (Boydstun & Hopelain 1977; Turnbull et al. 1995, 1996), and may be minimized through modifying rearing densities. Li (1994) believes that caged carp show less signs of physiological stress than freeswimming ﬁsh. In terms of quality, cage-reared ﬁsh are either the same as, or superior to, ﬁsh reared in other systems and even wild ﬁsh in terms of condition factor, appearance and taste (NORDA 1984; Li 1994; Kelly & Kohler 1996).

Cage Aquaculture – An Overview

2.4.2

25

Social and environmental issues

Cages have proved appropriate for both small-scale artisanal and large-scale commercial aquaculture production, the materials used to fabricate the cages and the size for the rearing units being chosen accordingly. The past decade has seen a tremendous shake-up in the Atlantic salmon farming industry with fewer, but larger, companies involved and a much greater degree of vertical integration. The larger multinational companies now produce in excess of 10 000 t of Atlantic salmon in cages, operating many sites in several countries and have annual turnovers in the tens of millions of dollars. These changes have been accompanied by substantial increases in efﬁciency of labour use. However, the impacts of cage aquaculture development, let alone the effects of the radical restructuring of the cage farming industry witnessed in major producing countries such as Scotland, Norway or Chile, on local employment has been little examined (see http://www.abdn.ac.uk/aqcess/). One of the overriding advantages of cages is that they make use of existing water bodies, giving non-land-owning sectors of the community access to ﬁsh farming – in theory at least. This can be important in areas where ﬁsheries are in decline, capitalizing on skills and augmenting the income of ﬁshermen’s families (see, for example, Gonzales 1984; Hambrey et al. 2001b). By virtue of their morphometry, lack of shallow inshore spawning areas and of suitable ﬁsh species to exploit the extensive pelagic areas, large reservoirs can be unproductive. Here cages can be especially useful. Demonstrating that cage culture can be highly proﬁtable in an open access, multi-purpose water body, however, can result in an accumulation and consolidation of exploitation rights by the richer sections of the community, often outsiders, to the detriment of the indigenous poor (Zerner 1992; Beveridge & Stewart 1998; Costa-Pierce 2002; Hambrey & Roy 2002). Cage aquaculture has recently been used to great effect in Bangladesh by the NGO (non-governmental organization) CARE-Bangladesh, where many thousands of landless people and poor farmers have successfully adopted cage culture of local carp and catﬁsh (McAndrew et al. 2000, 2002; Brugere et al. 2001; Hambrey et al. 2001a). However, despite the fact that by mid-2002 there were some 10 000 cages in the country, a review of the CARE-Bangladesh project highlighted some emerging issues that needed careful consideration (Hambrey & Roy 2002). There remained no clear written agreements on access rights, leaving poor farmers highly vulnerable. The gradual withdrawal of subsidies to buy nets, and the withdrawal of support and learning assistance, were also judged likely to increase the failure rate among new entrants. Provided due attention is paid to social issues, cages can be used to generate additional ﬁsh production and create jobs (Beveridge & Stewart 1998; Beveridge & Muir 1999; Costa-Pierce 2002). Cage aquaculture was successfully used in the resettlement of people displaced by reservoir construction in Malaysia and Indonesia. During 1985–88, in excess of 7500 jobs were created at Saguling Reservoir, Indonesia (Costa-Pierce 2002), although since the mid-1990s both social and environmental problems have arisen. Because cage farms have a much more direct and intimate association with the aquatic environment than ponds or tanks, they can have a pronounced impact on the natural environment and on local social issues (Zerner 1992;

26

Chapter 2

Beveridge & Stewart 1998; Kim 2000; Rawson et al. 2002). Cages occupy space that can disrupt access and make navigation difﬁcult. They can reduce landscape value and also alter current ﬂows and increase local sedimentation rates. Like other aquaculture systems, cage aquaculture uses resources and produces wastes. Irrespective of the method of rearing, cage aquaculture can be responsible for the introduction of pathogens or the disruption of disease and parasite cycles, changing the aquatic ﬂora and fauna, and altering the behaviour and distribution of the wild ﬁsh community. Ho (2000) has pointed out that sea lice (Copepoda: Caligidae) are likely to become a particular constraint to the expansion of cage aquaculture in Asia. Furthermore, associated with intensive cage culture is the release of uneaten food and faeces into the environment, stimulating primary production and adversely affecting water quality, while in freshwater lakes where extensive cage rearing of ﬁshes is practised, algal populations can be over-grazed, effectively reducing primary productivity. The impacts of cage ﬁsh farming on the aquatic environment can not only bring about conﬂict with other interests, such as ﬁshing or recreation, but can also exert a negative feedback effect on the cage operations themselves. It is thus important to understand and be able to quantify the impacts of cage aquaculture and so develop guidelines for the rational exploitation of aquatic resources. Impacts, their prediction, prevention and amelioration, are thoroughly discussed in Chapters 5 and 7.

2.4.3

Economics

Capital costs vary with cage size and materials. Huguenin & Ansuini (1978) demonstrated a three-fold difference in costs per unit volume between two cages constructed from identical materials but which differed in size. Similarly, a cage constructed from rigid copper–nickel (Cu–Ni) mesh costs around three times as much as a cage of similar size that utilizes a synthetic ﬁbre net bag (Huguenin et al. 1981). Cages designed to withstand exposed conditions can cost between 25 and 100% more per unit volume than those designed for sheltered, inshore sites (Gunnarsson 1988; Huguenin 1997; Lisac & Muir 2000). Operational costs are also highly variable, being determined by species, site, method of culture, management and scale of operations. Seed costs vary between 10 and 40% of operating costs for tilapias (Escover & Claveria 1985), 17% for intensive milkﬁsh farming (Marte et al. 2000), between 7 and 15% of grow-out costs for salmon (Huguenin & Ansuini 1978; Heen 1993; Higgs 1997) and 31% for grouper (Epinephelus spp.) (Marte et al. 2000) depending on size and source of stocked material. In intensive cage farming operations, feed often accounts for more than 50% of operating costs (Collins & Delmendo 1979; Brown 1983; Chong 1993; Marte et al. 2000) compared with perhaps only 3–4% in semiintensive cage tilapia operations (Table 2.2). Marine cage farms often experience problems with fouling and are generally more labour intensive than freshwater operations, as are farms that have problems with disease, poachers or predators (Escover & Claveria 1985). Depreciation costs can be as high as 45% of operating costs for wooden cages (Aragon et al. 1985) (Table 2.2). Metal mesh cages and those designed for exposed locations, however, have considerably longer

Cage Aquaculture – An Overview

27

Table 2.2 Operating costs, expressed as percentages, of small (mean cage area = 420 m2), medium (mean cage area = 848 m2) and large (mean cage area = 2499 m2) cage ﬁsh farms in the San Pablo City region, Laguna, Philippines, 1982 (data from Aragon et al. 1985).

Seed Feed Laboura Depreciation Interest Miscellaneous

a

Small

Medium

Large

34 3 19 30 10 5 100

42 4 11 31 9 3 100

38 3 8 45 5 1 100

Labour costs include estimates for unpaid labour which can account for 50% of total labour.

Table 2.3 Cost structure (US$ kg-1) in different types of Atlantic salmon farms (modiﬁed from Heen 1993). Sea-cage farming, Norway Expert evaluation

Smolt Feed Labour Pumping costs Miscellaneous costs Capital Total costs kg-1

Land-based, Iceland

15 best farms

US$

%

US$

%

US$

%

0.55 1.90 0.55 — 0.55 0.70 4.25

13 45 13 — 13 16 100

0.55 1.30 0.30 — 0.25 0.60 3.00

18 44 10 — 8 20 100

0.60 1.20 0.55 0.25 0.55 1.10 4.25

14 28 13 6 13 26 100

useful lives. There are also differences in operating costs between small and large operations, as illustrated by data for semi-intensive tilapia farms in the Philippines (Table 2.2) and salmon farms in Norway, and between ‘average’ farms and those which are particularly well-run (see Table 2.3). Gasca-Leyva et al. (2002) found that although input costs for sea bream (Sparus aurata) production were higher in the Canary Islands than in the Mediterranean, environmental conditions in the Canary Islands result in more rapid growth and higher economic returns. Is it possible then to reach any conclusions about the economics of cage ﬁsh farming compared with pen, pond, raceway or tank culture? Both the capital investment and the operational costs of intensive cage salmon culture are considerably lower than alternative methods of production which must rely on pumped sea water and either land-based tanks or raceways or ﬂoating enclosed systems (Shaw & Muir 1986; Blakstad 1993; Heen 1993; Lisac & Muir 2000). Reductions in feed use and in mortality caused by predation, algal blooms and ectoparasites in land-based production systems have resulted in similar production costs in some analyses (Table 2.3), although in others the savings have been insufﬁcient to signiﬁcantly reduce overall production costs (Table 2.4).

Ratio to total annual operational costs Ratio to gross income

Capital costs Operational costs Gross income Net income

11 500

Initial costs 2 764 22 765 45 000 19 471

Annual costs

Cages

76.0 43.0

10.8 89.2

% 14 500

Initial costs 3 232 24 136 45 000 17 632

Annual costs

Pens

64.0 38.1

11.8 88.2

%

47 000

Initial costs

3 908 21 866 45 000 19 226

Annual costs

Raceways

43.0

75.0

15.2 84.8

%

Table 2.4 Production costs (US$) of 50 000 channel catﬁsh in cages, pens and raceways. Data based on a 160-day growth period and a mean weight at harvest of 800 g. No cost for water is included (data from Collins & Delmendo 1979).

28 Chapter 2

Cage Aquaculture – An Overview

29

Table 2.5 Production costs (US$ kg-1) associated with landbased and offshore sea bass/sea bream culture in the Mediterranean (from Lisac & Muir 2000). Land-based

Offshore

Juvenile costs Feed cost Harvest and packing Insurance Medication Oxygen cost Electricity cost Fuel costs Net ﬁnancial costs

1.85 1.81 0.24 0.24 0.11 0.31 0.94 0.00 0.69

1.89 1.81 0.31 0.28 0.06 0.00 0.00 0.08 0.26

Total variable costs

6.19

4.70

Labour Maintenance Other ﬁxed costs Depreciation

0.54 0.15 0.31 0.58

0.55 0.09 0.39 0.59

Sum ﬁxed costs

1.58

1.63

Production costs (per kg) Price per kg Proﬁt per kg

7.77 8.02 0.25

6.33 8.02 1.69

Comparisons of intensive channel catﬁsh production in the United States in cages, raceways and pens conﬁrm that initial capital costs are lower for cages although operating costs are slightly higher than for raceways (Table 2.4). Nevertheless, the proﬁtability of the cage-based farm in this instance was marginally higher. Other studies conﬁrm that channel catﬁsh production in cages can be at least as proﬁtable as other methods of rearing (Lowell et al. 1982). In an extensive analysis of land-based versus offshore cage production systems for production of sea bass and sea bream in the Mediterranean, Lisac & Muir (2000) clearly demonstrate the returns on investment and the proﬁt are signiﬁcantly better for the offshore option (Table 2.5), while pointing up the enormous variability in per unit volume costs for offshore cage options. None of these studies, however, incorporate the economic costs associated with the work the environment must do (ecological services) in dispersing and assimilating the wastes, costs that are increasingly being passed on to the aquaculture producer through environmental taxes. Comparative data for less intensively managed systems are difﬁcult to ﬁnd. Using information from Guerrero (1981) for tilapia farming in the Philippines, pond farming appears the more attractive option (Table 2.6). However, costs do not include land purchase or rental charges, the latter accounting for up to 30% of operational expenditure (Rosario 1985). Moreover, capital costs are based on 1-m3 cages, which are considerably more expensive to construct than a single

30

Chapter 2

Table 2.6 Costs (pesos) of producing 3 t of tilapia per annum in cages and ponds in Nueva Ecija, Philippines. Data based on a 1ha pond and 421-m3 cages (data from Guerrero 1981).

Capital costs Fixed costs Variable costs Fingerlings Fertilizer Feed Labour Total costs Gross income Net income Return

Cages

Ponds

4 200 1 400

27 000 2 880

6 300 7 728 1 950 17 378 24 000 6 622

4 000 150 600 845 8 475 24 000 15 525

38%

183%

large cage, and feed costs quoted for farms in Nueva Ecija were around ten times greater than those incurred in Laguna Province at the time. Hambrey & Roy (2002) reviewed the CARE-CAGES project run by CAREBangladesh in Bangladesh, which had successfully helped thousands of resource poor farmers establish cage farms over the course of some 5 years. However, they also found some worrying economic trends. Proﬁtability seemed to be falling, perhaps due to the many novice participants in the programme who left stocking until late in the season, resulting in poor growth. They also found that over the course of the project, cage ownership per household had fallen slightly. Farmers often chose to reinvest proﬁts not in cages but in more traditional poultry and livestock activities, perhaps due to a lack of conﬁdence or avoidance of risk. In summary, cages are still the most economically feasible technology for intensively rearing ﬁsh such as salmonids, yellowtail, sea bass and grouper in marine waters. Cages can be a proﬁtable means of production for other species and at other types of site although this depends very much upon local circumstances. Introduction of costs for discharging wastes, however, would change the economic basis of production (see Chapter 5).

2.4.4

Security

From the farmer’s point of view, the disadvantages of water-based culture systems are summed up by the word ‘vulnerable’. Cages are usually sited in public or multi-user water bodies where farmers are powerless to stop or control pollution. Pollution has long been a problem for cage ﬁsh farmers in Japan causing millions of dollars of damage each year (Tabira 1980; Nose 1985; Doi 1991; Takashima & Arimoto 2000). Cages and pens are more susceptible to storm damage than ponds, tanks or raceways and this type of damage is cited

Cage Aquaculture – An Overview

31

as a major reason why a number of lake-based tilapia hatchery owners in the Philippines moved operations back to the land (Beveridge 1984b; Smith et al. 1985). Cages are also highly vulnerable to poaching and vandalism (see Gooley et al. 2000), which are especially prevalent in areas where social injustice has occurred as a result of unregulated cage farming development or where access rights are not guaranteed.

Cage Aquaculture, Third Edition Malcolm C. M. Beveridge Copyright © 1996, 2004 by Blackwell Publishing Ltd

Chapter 3

Cage Design and Construction

Good, practical cage designs should resolve engineering and cost considerations with the requirements of the species being farmed and those of the farmer who must operate the system. Cage structures must withstand the forces of winds and waves while holding stock securely. The design is important to the ﬁsh in that it both circumscribes living space and, within the site chosen, inﬂuences environmental quality inside the cage. From the farmer’s point of view the cage must be safe, secure and easy to manage. All of this must be achieved cost-effectively. In practice, ignorance of the effects of cage size or shape on ﬁsh production, and lack of research into the interaction between impinging environmental forces and cage structures, has meant that cage designs have largely evolved empirically. Until recently cage designs, traditional or otherwise, have reﬂected more the limitations of the materials used, the production economics and the available skills than the requirements of the ﬁsh or the environmental conditions in which the cages were to be sited. This is not surprising. The links between system design and ﬁsh behaviour, and between behaviour and production, have largely been regarded as irrelevant in the context of production economics. The forces acting on a cage and mooring system are complex and difﬁcult to quantify, and analysis of the responses of a cage structure to these forces requires the development and use of sophisticated test rigs and computer models. Research and development work of this nature is expensive and, as Kerr (1981) pointed out, was unlikely to be carried out while the cage ﬁsh farming industry remained small and fragmented. Moreover, there was little incentive to carry out this type of investigative work since most designs worked reasonably well in the conditions of comparative shelter afforded by most coastal and inland sites. However, the situation is changing. Ultrasonic telemetry, hydroacoustics and video technology are increasingly being used to investigate ﬁsh behaviour in a cage environment and to explore the inﬂuences of design and management on production (Phillips 1985; Kadri et al. 1991; Srivastava et al. 1991; Blyth 1992; Blyth et al. 1993; Dunn & Dalland 1993; Huse & Holm 1993; Jobling et al. 1993; Juell & Westerberg 1993; Smith et al. 1993; Martinez Cordero et al. 1994; Juell 1995; Oppedal et al. 2001). The cage farming industry, particularly in northern Europe, North America, Chile and Japan, expanded dramatically during the 1980s and 1990s and attracted the interests of a growing number of large multinational companies seeking to diversify into a new and growing market and with resources to carry out research and development (see Gunnarsson 1988, 1993; Fearn 1990; Slaattelid 1990). In the early 1990s, organizations such as SINTEF in Norway diverted part of their attention from oil 32

Cage Design and Construction

33

and shipbuilding to aquaculture systems. Moreover, there is growing pressure to culture ﬁsh in more exposed, offshore locations and increasing concern over security of stock and staff safety (Kuo & Beveridge 1990; Polk 1996; Muir & Basurco 2000). Fish welfare considerations aside, in the future cages will have to be more stringently designed, tested and cost-effectively manufactured and they will have to satisfy not only the ﬁsh farmers but also the insurance companies and government safety standards. Plans to introduce a system of licensing for cages, similar to that governing shipping, and to extend structural engineering codes of practice have been discussed in several countries (see Barker 1990; Cairns & Linfoot 1990). In Norway, it has been announced that a Norsk Standard for technical requirements for cage ﬁsh farming will be introduced for all new farms by 2004 and that all farms will have to comply by 2008. This chapter is intended as a brief review of cage and mooring system design and construction, and focuses on the most widely used ﬂoating types. Offshore designs are considered further in Chapter 7.

3.1

SHAPE, SIZE AND MATERIALS

At the outset of any cage design process, designers should give some consideration to the shape and size of the cage bag, which as stated above, should resolve engineering and cost considerations with the requirements of the species being farmed and of the farmer who must operate the cages. Present knowledge remains sufﬁcient to provide only general guidelines.

3.1.1

Shape

While circular cage bags make the most efﬁcient use of materials (Fig. 3.1), and thus have the lowest costs per unit volume, the surface area : volume ratio is comparatively small and water exchange may be constrained (Howard & Kingwell 1975). Cage shape can also be important to ﬁsh. Observations made of shoaling, actively swimming species such as salmonids or milkﬁsh (Chanos chanos) (Sutterlin et al. 1979; Wilton 1979; Yu et al. 1979; Srivastava et al. 1991; Juell & Westerberg 1993) suggest that structures that are circular or near circular (i.e. polygonal or square) in plan area are best in terms of utilization of space, corners of rectangular structures being little utilized (Fig. 3.2). Shape is likely to inﬂuence effective stocking densities and swimming behaviour, which in turn can inﬂuence production, although it is likely to be less important for less active species such as tilapias and carp (see Kelly & Kohler 1996).

3.1.2

Size

Few sites, other than perhaps irrigation canals, impose size limitations on cages (e.g. Budhabhatti & Maughan 1994); size is usually determined by other factors. One advantage of increasing bag size is that costs per unit volume fall. This is one reason for the trend towards the use of increasingly large rearing units in intensive marine ﬁsh culture. In western Europe the 100–150-m3 cages used to

34

Chapter 3

Fig. 3.1 The perimeter lengths of various cage shapes of the same surface area.

rear Atlantic salmon in the 1960s and 1970s have gradually been replaced by cages of up to ﬁfty times this volume. However, large cages require more sophisticated materials and technology and they can be more difﬁcult to manage, requiring several staff to perform even the most simple of routine operations. Inspection of nets and removal of mortalities may have to be carried out by divers or by using expensive, sophisticated technology. Monitoring of ﬁsh health and grading become difﬁcult. In other words, advantages that cages might have over other rearing systems in terms of ﬂexibility in production planning may be offset when large units are used. The new generation of large cages may prove to be best suited to the culture of species such as yellowtail, tunas or of sterile (triploid) salmon that do not require regular grading or sequential harvesting. Water exchange in very large cages may also be poorer as the surface area : volume ratio decreases with increasing size. Other doubts concerning the trend

Cage Design and Construction

35

Fig. 3.2 Circular swimming pattern of ﬁsh in cages of different shape. Assuming ﬁsh maintain this pattern irrespective of cage shape, note the amounts of unused space in cages of different shape.

towards large rearing units have been voiced by insurers worried about the risks involved in holding tens of tonnes of ﬁsh in a single cage. There has been little systematic study of the relationship between the size of aquaculture systems and growth and production. Some general principles seem intuitively obvious: certain species, by virtue of their life-style, may require greater living space than others while large animals are also likely to require more space than small animals. Through trial and error it has been found that larger shoaling and pelagic ﬁsh, such as salmon, tunas and yellowtail, grow faster in cages with larger surface areas (Nose 1985). Trials conducted in Norway, comparing production of salmon in 50-m and 90-m circumference cages, showed that ﬁsh in the larger structure grew faster, that there was improved food conversion and less feed loss, lower mortality and a reduction in the incidence of sexual maturation (Guldberg et al. 1993). There is little evidence, however, that

36

Chapter 3

Fig. 3.3 Mean number of halibut observed at different depths in 4-m deep cages. Records based on hourly observations made with video recording equipment (from Martinez Cordero et al. 1994).

cage area affects less active species to the same degree; the ﬁnding that tilapias grow faster in larger cages seems to be because feed losses are reduced (McGinty 1991). Little is known about depth either. In the extensive culture of planktivorous ﬁsh, Li (1994) recommends that the cage should be held within the planktonrich surface (<2 m) waters. However, very shallow cages (<1.5 m) have been shown to affect body shape and retard growth in carp and tilapia for reasons not yet understood (Maruyama & Ishida 1976, 1977). By contrast, ﬂatﬁsh such as turbot can be grown in cages of 0.9–1.6 m in depth (Kerr et al. 1980) while halibut, another ﬂatﬁsh, spends most of the time on the cage bottom (Fig. 3.3), little utilizing the rest of the caged volume (Martinez Cordero et al. 1994). Catﬁsh, too, being bottom feeders, tend to utilize the deeper areas of the cage except when feeding. Studies of air-breathing ﬁsh maintained in tanks have shown that there are also depth-related energy and time costs associated with travel for aerial respiration (Kramer 1983; Bevan & Kramer 1986). The swimming depth of salmon is inﬂuenced by a number of factors, the relative importance of which varies with season and time of day (Juell 1995). Trade-offs between feeding opportunity and predator avoidance, mediated by differences in maturity or competitive ability, are important. While it was once assumed that depths much greater than 10–12 m would be poorly utilized by ﬁsh and that cage depths of 3–10 m would be acceptable for most species under most

Cage Design and Construction

37

circumstances, commercial trials in Norway showed that Atlantic salmon will utilize cage depths of up to 35 m (Huse & Holm 1993). While few studies have examined the inﬂuence of system dimensions on behaviour and production, even fewer have considered the importance of the ratio of one parameter, such as area, to another (e.g. depth). It is likely that cage depth and area interact in some, as yet, undetermined way in deﬁning the environment for caged stock.

3.1.3

Materials

Ideally, the materials used to construct the cage bag should be: • • • • • • • •

strong; light; rot, corrosion and weather resistant; fouling resistant; easily worked and repairable; drag free; smooth textured and thus non-abrasive to ﬁsh; inexpensive.

Some would also argue that the materials should be rigid and thus resistant to deformation in strong currents: the converse argument is that a degree of ﬂexibility is desirable since this helps reduce drag, absorbs current energy (see later) and facilitates harvesting. To a certain degree, the materials used to fabricate the cage frame and collar should have similar properties. No single material is ideal; all to some extent are a compromise. However, undoubtedly some materials are better suited to certain species, sites and purposes.

3.2

TRADITIONAL DESIGNS

As traditional cages developed from ﬁsh traps and ﬁsh holding facilities, so they borrowed heavily from them in design and were fabricated using existing skills from whatever materials were to hand and familiar to the builders. Natural materials, such as grasses or wood, were used not only to form the collar or frame but also to enclose the ﬁsh. Reeds (e.g. Phragmites, Cypsus spp.), jungle vines and creepers such as rattan (Calamus spp.) and ridan (Salacca glabescens), bamboos, hardwoods and softwoods have all been used in this way (Huet 1956; Hickling 1962; Watson & Raja 1979; Pantulu 1979; Murugesan & Parameswaran 1984; Fedoruk & Srisuwantach 1984; Costa-Pierce & Effendi 1988). Reeds and vines, however, are unsatisfactory; being largely composed of cellulose they decompose within months in tropical waters, although life expectancy may be extended slightly by application of tar-based compounds (Klust 1982). They are also easily damaged by the caged ﬁsh and by predators. Bamboo, on the other hand, is a much more appropriate material being strong, cheap, widely available, and easily worked with simple tools. Typically, slats are

38

Chapter 3

Fig. 3.4 Traditional basket-shaped bamboo sand goby cage from central Thailand (redrawn from Fedoruk & Srisuwantach 1984).

cut and planed using a knife or machete and either ﬁxed directly to the cage frame or woven into matting, sometimes in combination with synthetic or natural ﬁbres, and subsequently ﬁxed to the frame. In parts of Thailand and Vietnam traditional cages are still woven entirely from bamboo slats to form basket-shaped structures (Fig. 3.4). Provided the bamboos have been properly selected and treated the resultant mesh will last for up to 2 years. A number of trials have compared growth, survival and production of various species of ﬁsh in similar-sized bamboo and nylon net cages. All-bamboo constructions are sometimes claimed to be as good as, and in some cases superior to, those ﬁtted with nylon mesh and they cost considerably less (Lipton 1984; Natarjan et al. 1984). However, other studies have shown they were inferior: they had poor water exchange and caused damage to the caged stock (Anh & Son 2001). It can be difﬁcult to ﬁnd straight bamboos, resulting in a non-uniform mesh that may allow fry to escape. Bamboo slatted cages are also labour intensive to

Fig. 3.5 Slatted hardwood sand goby cage from central Thailand (redrawn from Fedoruk & Srisuwantach 1984).

Fig. 3.6 Bamboo cage installed in canal bottom, West Java, Indonesia. Overhead wires are to deter predators (courtesy C. Sommerville).

40

Chapter 3

construct, heavy and cumbersome to manage – particularly if more than a few cubic metres in volume – and unsuitable for use in the more corrosive brackish and marine environments (see Anh & Son 2001). Hardwoods are generally denser than bamboo or softwoods and thus are stronger and more resistant to impact damage and to infestation by boring organisms, such as the mayﬂy Povilla adusta, which are prevalent in many African waters (Coche 1979) (see Fig. 4.8), and to fungal attack. A variety of species has been used throughout the tropics. Planks are simply lashed or nailed together (zinc nails are recommended) to form box-like constructions that are either ﬂoated using supplementary ﬂotation materials or anchored to the bottom with stones and posts (Figs 3.5, 3.6). Economic analyses suggest that although hardwood cages typically cost around 10–15 times more than those fabricated from bamboo, they have a useful working life of 9–12 years in fresh waters (Sodikin 1978; Fedoruk & Srisuwantach 1984). Traditional ﬂoating cages utilize a variety of materials such as bundles of bamboos or hardwood logs and oil drums that are lashed to the sides of the structure for supplementary ﬂotation (Fig. 3.5). Simple anchoring systems – ropes and block weights, or posts driven into the substrate – are most commonly used.

3.3 3.3.1

MODERN DESIGNS Cage bag

The International Organization for Standardization (ISO) deﬁnes netting as ‘a meshed structure of indeﬁnite shape and size’, but the term is used here in a more restricted sense to differentiate ﬂexible netting composed of natural and synthetic ﬁbres, such as cotton and nylon, from rigid and semi-rigid materials, such as extruded plastics and metals. The latter materials are referred to as rigid meshes. Netting and rigid meshes can be made from a range of man-made and natural materials and are available in a variety of different forms. Rigid mesh materials have been developed for many different purposes with the result that there is no single accepted system for their characterization. They are usually described in terms of density (weight m-2) and the diameter of the mesh bars, mesh size generally being a measure of the distance between parallel bars (Fig. 3.7). For netting, the ISO has laid down stringent regulations governing terminology (http://www.iso.ch/iso/en). Mesh size is illustrated in Fig. 3.7. The yarn used for net fabrication is usually described in terms of tex, which is a measure of linear density (1 tex = 1 g 1000 m-1) (Klust 1982). Flexible mesh materials (netting) Natural ﬁbres are rarely used today for cage netting. They suffer from a number of disadvantages, not least of which is their susceptibility to rotting and consequent loss of strength (Klust 1982).

Cage Design and Construction

41

Fig. 3.7 Net and mesh terminology.

Modern netting is composed of synthetic ﬁbres, the most common being polyamide (PA), polyester (PES), polyethylene (PE) and polypropylene (PP), which unfortunately, for commercial reasons, trade under a plethora of names. These vary not only from country to country but also within a country from manufacturer to manufacturer. Klust’s somewhat dated list (Klust 1982) cites 274 brand names for PA netting yarns (both PA6 and PA6.6), 100 for PES, 78 for PE and 136 for PP! The use of one of the generic terms for PA ﬁbres, nylon, has become so widespread that it is applied to all polyamide materials and henceforth will be used here. Not only do netting materials have different properties, the textiles used in their fabrication are available in a number of different forms that partly determine the ultimate characteristics of the netting. Nylon and PES netting are usually manufactured from continuous ﬁlament (also known as multiﬁlament) yarn, composed of extremely long, ﬁne ﬁbres, generally less than 50 mm in diameter. All ﬁlaments run the entire length of the yarn and thus there are exactly the same numbers of ﬁlaments present throughout the yarn. Although some PP netting yarn is manufactured from continuous ﬁlaments, some is also manufactured from split ﬁbre yarn which is produced either by twisting plastic ﬁlm under tension or by mechanically ﬁbrillating the ﬁlm immediately after extrusion. The

42

Chapter 3

Fig. 3.8 Construction of twisted and folded netting yarns (from Klust 1982).

resultant ﬁbres are considerably coarser than continuous ﬁlaments and vary a great deal in their thickness. Split ﬁbre PE netting yarns are also occasionally used for netting although most PE netting is still manufactured from folded monoﬁlaments. Monoﬁlaments are similar to the continuous ﬁlaments described above, but are considerably thicker (100–500 mm in diameter) and are sufﬁciently strong to function on their own as a yarn. However, the PE netting used for cages is usually manufactured from a twisted monoﬁlament yarn. The yarns described above are known as single yarns and form the basic components of netting yarns. There are two principal types of netting yarn used to make knotted netting: braided and cabled. Most knotted netting used in cage ﬁsh farming is fabricated from cabled yarns, which are produced in three stages: (1) the ﬁbres are combined as described above to form single yarns; (2) several – usually three – single yarns are twisted together to form a folded yarn or netting twine; (3) several – usually three – of these folded yarns or netting twines are twisted together by a second twisting operation, carried out in alternative directions at each stage of the manufacture, to form a cabled netting twine or cabled yarn (Fig. 3.8). Single or double weaver’s knots are used to construct most knotted netting (Fig. 3.9), although in Japan reef knots and square knots are

Cage Design and Construction

43

Fig. 3.9 (a) Types of knot used in netmaking: (i) single weaver’s or English knot; (ii) reef or Japanese knot; (iii) double weaver’s knot. (b) Examples of various methods for the construction of bars and joints in knotless netting: (i) Japanese twisting; (ii) Raschel; (iii) braiding (from Klust 1982).

also used. In factory-produced nets, this process is entirely mechanized (Fig. 3.10). The yarns used for the fabrication of knotless netting are either single or folded PES, PE or nylon yarns. Two types of knotless netting are commonly used for cages; the Japanese twisted and the Raschel (Fig. 3.9). According to Klust (1982) there are many more variations possible in the construction of bars and joints in the Raschel than in the Japanese twisted netting and it is also easier to repair. The Raschel type is widely available throughout the world whereas the Japanese twisted knotless netting is mainly manufactured and used in Japan. Braided knotless PE netting (Fig. 3.9b (iii)), with less extensibility and fewer inter-ﬁbre crossings, is claimed by some manufacturers to be superior for use in offshore cages where cyclical loads can induce fatigue failure. The above section is intended as an introduction to netting materials and readers who wish to discover more are recommended to consult Klust (1982) and Garner (1989). For most, however, it is sufﬁcient to know only the general properties of different netting materials. Ideally, cage nets should be slightly denser than water, thus facilitating hanging, but not so dense as to make handling excessively difﬁcult or labour intensive, or to signiﬁcantly affect the ﬂotation or mooring systems. The netting should be strong enough to resist ripping

44

Chapter 3

Fig. 3.10 Net-making factory assembling Raschel knotless netting cage bags.

by ﬂoating objects and predators, and be capable of supporting a portion of the biomass of enclosed ﬁsh when the nets are lifted for harvesting. High extensibility, so that the netting can endure stretching without breaking, and resistance to abrasion are also desirable. Flexural stiffness, or resistance of the netting to lateral and bending deformation, is important since a stiff net will resist collapsing in currents. On the other hand, the netting should be soft and smooth textured so as to minimize abrasion damage to the skin, ﬁns and eyes of the caged ﬁsh. However, softness and ﬂexural stiffness are to some extent contradictory in synthetic ﬁbre netting. Other desirable properties are that the netting should have as low a resistance as possible to water currents, so as to minimize deformation and dynamic loadings, and that it should be resistant to fouling, minimizing increases in loading to the cage mooring system while facilitating the best possible water ﬂow through the cage. Finally, durability, availability, cost and maintenance requirements are extremely important considerations. The characteristics of netting are determined not only by the properties of the yarn (which are a product of the material used), the type of ﬁbre (e.g. monoﬁlament or multiﬁlament), the diameter of yarn and the degree of twisting, but also by the method of net fabrication (e.g. knotted or knotless, Raschel or Japanese twisted) and the mesh size. The addition or application of chemical stabilizers, dyestuffs, antifoulants and stiffening compounds can also modify these properties. The general characteristics of various netting ﬁbres are summarized in Table 3.1. Direct comparison between knotted and knotless netting is more difﬁcult

d

c

b

Polyamide. Polyester type Terylene–Dacron–Diolen–Tergal–Tetoron–Trevira. Polyethylene type high density, polymerized at low pressure. Continuous ﬁlament form.

Moderate

Moderate

a

High Flexible Soft Medium

High Flexible Soft Medium

Extensibility, wet Stiffness Softness Resistance against weathering without treatment or colouring Resistance to fouling

1.14 Very high 85–95d 12

1.14 Very high 85–95d 12

PA 6

Fibre density (g cm-3) Breaking strength Breaking strength, wet, (% dry breaking strength) Weight in water (% air-dry weight)

PAa 66

—

Low Moderately stiff Moderate High

1.38 High 100 28

PESb

Low

0.96 High 110 0 (Buoyant) Intermediate between PA and PES Stiff Hard Medium

PEc

0.91 High 100 0 (Buoyant) Low Stiff Hard Low– medium Moderate

PP

Table 3.1 Characteristics of synthetic ﬁbres. Note that the two types of nylon are referred to, although both have near identical properties (developed from Milne 1972; Klust 1982).

Cage Design and Construction 45

46

Chapter 3

because many of the mechanical tests commonly carried out on knotted netting cannot be directly applied to knotless materials. Nevertheless, it is possible to make a number of qualitative comparisons. The principal problem with knotted netting is the knots; they are costly to make, both in terms of time and materials, result in signiﬁcant increases in weight (up to 100% increase in weight per unit area for some types) and bulk, and are the weakest part of the net. Furthermore, they considerably increase drag, and knotted netting fouls and abrades more readily than knotless materials and has been widely reported as damaging to caged ﬁsh. By contrast, knotless netting is often cheaper and because it is knotless, panels can be machine stitched to form extremely strong enclosures (see later). It is usually the more suitable material for fabricating cage bags and is readily available in most parts of the world. It suffers from the disadvantage that it seems to be less resistant to fatigue wear than knotted materials, making it perhaps less appropriate for offshore cages. Knotless netting is generally only fabricated in nylon, PE or PES. The latter is extremely expensive and considerably heavier than nylon (Table 3.1), although some believe that it may be somewhat superior. However, the economic advantages offered by nylon far outweigh any disadvantages that it might have in terms of stiffness or resistance to weathering. PE is a suitable alternative to nylon, being much cheaper and more readily available in many parts of the world. In most other respects, however, it is inferior. Although knotless netting can be manufactured in a variety of shapes, square mesh remains more open when subjected to currents. Rigid mesh materials Modern rigid mesh cages consist of mesh panels attached to a frame and supported by a ﬂoating collar. Two basic types of material have been used: rigid polymer and metals. Rigid polymer mesh materials such as Netlon® or Durethene® are formed from PE by an extrusion process. They are usually square or diamond mesh, are readily available in a range of mesh sizes and are semi-rigid rather than rigid and thus may be conveniently stored in rolls. They have a density intermediate to that of ﬂexible synthetic ﬁbre netting and rigid metal meshes – somewhere in the order of 0.35–0.60 kg m-2, although some companies also market a range of heavy gauge plastic meshes (0.70–1.1 kg m-2). The durability and performance characteristics of rigid plastic meshes have not been evaluated fully although there is some published information on Netlon®. Although it does not corrode and is highly rot resistant, like all other synthetic polymers it is prone to weathering. It is more resistant to fouling than untreated ﬂexible netting but less so than galvanized steel or copper alloy meshes (see Table 3.5). Metal meshes have been extensively tested for use in the construction of rigid cages. Metals suffer from one major disadvantage in that they are subject to chemical attack or corrosion when immersed in water. The ease with which a metal corrodes depends upon its electronegativity (its capacity for attracting electrons) and its standard electrode potential. When a metal is immersed in water a potential difference, or electrode potential, is established between the

Cage Design and Construction

47

Table 3.2 Standard electrode potentials (E0) of some common metals measured relative to the hydrogen electrode at 25°C. Metal

Electrode reaction

E0

Metal

Electrode reaction

E0

Copper Hydrogen Lead Tin Nickel Cadmium

Cu = 2Cu2++ 2eH = H+ + ePb = Pb2+ + 2eSn = Sn2+ + 2eNi = Ni2+ + 2eCd = Cd2+ + 2e-

+0.34 0 -0.13 -0.14 -0.25 -0.40

Iron Chromium Zinc Aluminium Magnesium Potassium

Fe = Fe2+ + 2eCr = Cr3+ + 3eZn = Zn2+ + 2eAl = Al3+ + 3eMg = Mg2+ + 2eK = K+ + e-

-0.44 -0.71 -0.76 -1.66 -2.37 -2.93

metal and the liquid and metal ions pass into solution leaving the metal with an excess of free electrons: M

∫M

+

+ e-

The electrode potential is dependent upon the electronegativity of the metal and the temperature and ionic concentration of the solution in which it is immersed. Electrode potentials are measured by immersing the metal in a solution of its own ions and using a standard hydrogen or calomel electrode as reference. The electrode potentials of some common metals are given in Table 3.2. In Table 3.2 the metals are arranged in order of increasing negativity and reactivity, with copper at the top and potassium at the bottom. If two different metals are connected in a conducting medium, current ﬂows from the metal with the more negative potential (i.e. anode) to the metal with less negative potential (the cathode) resulting in the corrosion of the more anodic material. For example, by coupling zinc (-0.76) to copper (+0.34) in an electrolytic medium, zinc will form the anode and thus suffer corrosion while copper will become the cathode and thus be protected. The electromotive force (EMF) between the metals will be 0.34 - (-0.76) = 1.1 V. If iron (-0.44) replaces zinc as the anode, then the EMF will be smaller (0.78 V) and the resultant rate of corrosion will be reduced. Because the rate of corrosion is dependent upon the nature of the electrolytic medium – and in particular its ionic composition, temperature and oxygen concentration – and the structure and chemical uniformity of the metals, a range of standard electrode potential (E0) values (in volts) exists for each metal and its alloys. The galvanic series, as it is known, is empirically derived. Fig. 3.11 shows the galvanic series for metals and their alloys in ﬂowing seawater conditions, measured against a calomel electrode. Although the E0 values given in the ﬁgure may change outside the given temperature, current and salinity regime, the relative galvanic positions will remain the same. Because galvanic cells can be established by connecting together two dissimilar metals in a conducting medium, such as sea or brackish water, care should be taken when choosing metal components for a cage system (see below). However, galvanic cells can also be caused by the following:

48

• •

•

•

Chapter 3

Differences in the microstructure of a metal or alloy component. Differences in the ionic composition of the conducting medium. In sea water, areas in contact with less saline water form the anode, while those in contact with higher salinities act as cathodes. Differences in the dissolved oxygen (DO) concentration of the conducting medium. Low DO areas act as anodes while high DO areas become cathodes. The existence of stress cells within the metal or alloy. Differences in grain size (ﬁne grain – anode; coarse grain – cathode), and the presence of imperfections (defect – anode; perfect – cathode), strains (cold-worked – anode; annealed – cathode) and stresses (loaded areas – anode; non-loaded areas – cathode) can all lead to corrosion (for reviews see ICE 1979; Hey & Pagan 1983).

Some metals are particularly unreactive and are thus relatively resistant to corrosion (e.g. copper, tin), while others, although subject to corrosion by virtue of their high electrode potentials, are protected by the formation of a surface oxide coating when immersed in water (e.g. aluminium, zinc). The corrosion resistance of a metal may also be improved by alloying it with a more corrosion resistant material (e.g. the addition of 12–20% chromium in stainless steels or copper in ferritic steel). Cathodic protection by sacriﬁcial corrosion, in which a base metal is connected to the part to be protected and thus corrodes in its place, is a technique commonly employed to protect marine installations such as pipelines and ships. Magnesium alloy anodes, containing 6% aluminium, 3% zinc and 0.2% manganese are widely used for this purpose (Chong 1977). Another technique to suppress corrosion is the use of protective coatings. Metallic coatings may be applied by dipping, electroplating, spraying, cladding or cementation. Inorganic oxide, phosphate or chromate coatings are formed by chemical treatment. A variety of paints, varnishes, plastics, lacquers and rubber compounds may also be used to isolate the metal from the corrosive environment, although few of the above protective coating methods are appropriate for metal meshes and even fewer are appropriate for use in cages designed to produce ﬁsh for food. In summary, the most suitable metals for construction of rigid mesh cages, from a corrosion-resistance point of view, are those with low electrode potentials. Alternatively, a coating or sacriﬁcial corrosion technique must be employed to protect the mesh. Unfortunately, many of the most eligible materials are unsuitable because of their cost, weight, strength or brittleness, and as a result only three types of metal mesh have been widely employed in the fabrication of cages: 90 : 10 copper–nickel wire and expanded metal mesh (alloy CA-706): Although a range of copper–nickel (Cu–Ni) alloys has been tested (Ansuini & Huguenin 1978; Ravindran 1983; Woods Hole Engineering Associates 1984), 90 : 10 Cu–Ni has proved the most appropriate for use in aquaculture, combining structural strength, lightness and corrosion resistance with antifouling properties. Although there have been no controlled comparative tests of durability of wire and expanded Cu–Ni mesh materials, the latter is probably superior since there

Cage Design and Construction

49

Fig. 3.11 Galvanic relationships in ﬂowing sea water (2.5–4.0 m s-1; 10–27°C). Both active and passive potential ranges are given for stainless steels (redrawn from Woods Hole Engineering Associates 1984).

are no crevices in which corrosion can occur. Expanded Cu–Ni mesh is also widely available in a variety of mesh sizes (3 mm to 5 cm) and gauges. However, almost all of the research and development work has been carried out on hexagonal mesh material (gauge 0.89 mm, strand diameter 1.27 mm), with a nominal mesh size of 1 cm. This material weighs 1.86 kg m-2 and has an ultimate yield strength of 3.3–4.3 ¥ 105 N m-2 (Woods Hole Engineering Associates 1984). Most experimental and commercial designs to date have used material with the above speciﬁcations. The long-term stabilized corrosion rate for 90 : 10 Cu–Ni is around 2.5 ¥ 10-3 mm per year or less in ﬂowing seawater conditions in a temperate climate (Efﬁrd 1975) and 22.5 ¥ 10-3 mm per year as determined from

50

Chapter 3

120-day ﬁeld trials in Cochin, India (current speed = 0.6 m s-1), in coastal tropical conditions (Ravindran 1983). Quantitative data on rates of fouling have been published by Huguenin & Ansuini (1981) and Ravindran (1983) and suggest that in typical ﬂowing seawater conditions fouling will rarely cause more than a 10% blockage of the meshes. Galvanized steel weldmesh or chain link netting: The raw material for these types of meshes is cold-drawn mild steel wire, galvanized with zinc, preferably by a hot-dipping process. The wire can then be woven to form diamond mesh chain link or welded together to form square mesh (galvanizing of the latter is usually carried out after welding) (Fig. 3.12). For equivalent mesh sizes, chain link is generally made of ﬁner gauge wire and is consequently lighter but less strong (Table 3.3). The service life of galvanized steel mesh in a particular environment is dependent upon the type of mesh and the quality of galvanizing. Milne (1970, 1972) found that chain link had a shorter service life than weldmesh due to abrasion at the links, which accelerated corrosion, and that chain link retained only 50% of its initial strength after immersion for one year in the sea. By contrast, he found that weldmesh made of the same material retained 80% of its initial strength after immersion for the same period. However, modern chain link is superior to the old products. Kuwa (1983) has shown that the effectiveness of galvanization depends upon its thickness as well as the method of application, hot-dipping is superior to electroplating or spraying. In Japanese waters it was found that at 400 g m-2 the service life of galvanized steel mesh was about 15 months, with 260 g m-2 about 10 months, and

(a) Fig. 3.12 Galvanized steel mesh cage technology, Australia. (a) Tassie cage; (b) caged barramundi and other species; (c) close up of MarineMesh galvanized steel wire mesh (courtesy Fusion Marine, One-Steel and Mr Coulsen).

Cage Design and Construction

51

(b)

(c) Fig. 3.12 Continued.

with 250 g m-2 galvanic protection lasted 9 months. After consumption of the galvanized coating, corrosion of the steel wire was rapid (0.4–0.6 mm per year). However, Kuwa (1983) also showed that cathodic protection with aluminium alloy anodes can effectively double the service life of the material in sea water (Fig. 3.13; Table 3.4). The fouling rates for galvanized steel mesh materials are shown in Table 3.5. Commercially fabricated galvanized steel cages have begun to be used in Queensland, Australia for barramundi (Lates calcarifer) culture where they have been found to reduce fouling, problems with predators, loss in volume due to

52

Chapter 3

Table 3.3 The physical properties of various rigid mesh materials used in cage fabrication. Material

Netlon Durethene

90 : 10 Cu–Ni expanded metal Galvanized steel weld mesh Galvanized steel chain link

Plastabond

Mesh size (mm)

Gauge (mm)

Density (kg m-2)

Strength

Source

50 6–44 6 29 37 10

3.3 — — — — 1.3

0.34 0.61 0.79 0.68 0.71–1.10 1.86

— — — — — 3.3–4.3 ¥ 108 N m-2c

25

2.5

3.40

205 kgd

Woods Hole Engineering Associates 1984 Milne 1970

25 32 45 55 76

2.0 2.6 3.2 4.0 2.5

2.03 — — — 3.25

127 kgd — — — 127 kgd

Milne Kuwa Kuwa Kuwa Milne

Milne 1970 a b

1970 1983 1983 1983 1970

a

Values refer to Durethene polyethylene marine netting; ADPI Enterprises, Philadelphia. Values refer to Durethene polyethylene marine netting, heavy duty series; ADPI Enterprises, Philadelphia. c Yield strength of panel. d Tensile wet break strength. b

currents and are argued to provide a better rearing environment and savings in labour (Figs 3.14, 3.15). Plastic-coated galvanized chain link or weldmesh: Chain link is typically fabricated from polyvinyl chloride (PVC) coated galvanized steel wire which is woven into diamond mesh netting (Milne, 1972), whereas the plastic-coated weldmesh is manufactured in a similar method to galvanized steel weldmesh and then coated in PVC. Comparisons of 25 mm rigid mesh netting show that PVCcoated chain link is similar in weight to galvanized weldmesh and has similar strength characteristics to galvanized chain link (Table 3.3). It is, however, considerably less resistant to fouling (see Table 3.5).

Design criteria The function of the bag component of a cage is to hold the ﬁsh securely while permitting sufﬁcient water exchange to replenish oxygen (and, in some cases, food) and remove potentially harmful toxic metabolites. It is also important that the cage volume remains relatively resistant to deformation by external forces that would not only cause crowding and stress but may also injure or cause mortalities among stock.

Cage Design and Construction

53

Fig. 3.13 Installation of aluminium alloy anode (P) and test piece (C) for cathodic protection of wire netting (from Kuwa 1983).

Table 3.4 Details of anodic protection for two ﬁsh cages (modiﬁed from Kuwa 1983; see text). Submerged metal mesh bag and frame

Protective area Current density Current required Anode material Anode speciﬁcations Anode installation Date of installation a b

Cage 1a

Cage 2b

119 m2 130 mA m-2 15.5 A Aluminium alloy 8 ¥ 9.30 kg pieces Suspension type 4 months after cage installed

135 m2 110 mA m-2 14.9 A Aluminium alloy 8 ¥ 5.67-kg pieces Suspension type 10 months after cage installed

Season 1 stocked with yellowtail, Season 2 with Japanese striped knifejaws. Season 1, Season 2, stocked with Japanese striped knifejaws.

(a) Clean

(b) Fouled

Fig. 3.14 Current forces (kg m-2) on (a) clean and (b) fouled netting (from Milne 1972).

Fig. 3.15 Current deﬂection around cages (from Løland 1993a, b).

Cage Design and Construction

55

Table 3.5 Seasonal weight increases for various fabrics due to marine fouling (from Milne 1972). Weight (kg m-2)

Fabric

Nylon Ulstron Courlene Polythene Standard Cupra-proofed Netlon Plastabond Galvanized Chain-link Weld mesh

July

0.23 0.34 0.20

2 2 2

0.18 0.18 0.34 3.25

2 1 1 0.75

2.03 3.40

0.3 0.3

Multiplication factor for weight increase September

November

85 64 85

108 110 126

112 44 36 10

200 94 48 13

0.5 0.5

2.75 2.50

Two principal types of force act on the cage bag: static and dynamic. Static loads act vertically and are imposed by the weight of the bag plus fouling and, in the case of small rigid mesh structures which are usually lifted during harvesting, by the biomass of ﬁsh. Static loads can be estimated from the area and density of netting or rigid mesh materials used, the weight of fouling, the ﬁsh biomass, the quantities of ropes and the weights of frame components used to strengthen and stiffen the structure. However, while it is important to quantify static loads for design of the ﬂotation and mooring systems, these are less critical in determining the design of the bag. Although waves acting on the collar induce a degree of dynamic vertical loading the most important dynamic forces impinging on the cage bag are usually caused by currents, which act horizontally. Quantiﬁcation of currents and the response of the materials can help predict how a particular design will perform in terms of water exchange and deformation of the bag, and can aid in the design of frame members for rigid mesh materials and in rigging systems for netting. Data generated can also be used in the computation of moorings. Water ﬂowing through a mesh or netting panel imposes a load which is transmitted to the supporting frame, collar and mooring system. The load on a panel has been shown to be dependent upon the nature of the material used, the shape and size of the mesh, current velocity and the density of water such that: Fc =

Cd rV 2 A 2

(Kawakami 1964; Rudi et al. 1988)

where Fc = the force applied to the panel by the current (kg); Cd = coefﬁcient of drag of the material (dimensionless); r = density of water; V = current velocity (m s-1); and A = projected area of the mesh members (m2). The latter parameter

56

Chapter 3

is calculated from the panel area, number of meshes, and length and diameter of the mesh members. Cd values have been experimentally determined using panels fabricated from a variety of clean and fouled materials (Table 3.6). For materials other than expanded metal meshes, Cd values, which approximate those determined experimentally, can be calculated using the following formulae: Cd = 1 + 3.77

Ê dˆ Ê dˆ + 9.37 Ë a¯ Ë a¯

Ê dˆ Ê dˆ Cd = 1 + 2.73 + 3.12 Ë a¯ Ë a¯

2

for knotted materials, and 2

for knotless materials

(Milne 1970)

where d = yarn diameter (mm) and a = bar length (mm). Similar equations have been derived by Kumazawa et al. (1994). Note that Cd values increase with bar width and consequently are greater for fouled than unfouled materials. The equations given above also indicate that Cd values are considerably greater for knotted than knotless materials. The lowest observed Cd value (0.41) was determined from tests carried out by Woods Hole Engineering Associates (1984) on 90 : 10 Cu–Ni mesh, although for design purposes it is suggested that a more conservative value of 0.56, based on fouling trials, is used. From experimental work, the above relationships have been found to be valid for ﬁxed panels of materials where mesh members do not occlude more than 40% of the panel area (this is known as the solidity ratio), and for currents which ﬂow perpendicular to the panel (Woods Hole Engineering Associates 1984). There is also evidence that Cd values are independent of current velocity over the range typically encountered at ﬁsh farm sites (0.33–1.87 m s-1). However, because expanded metal meshes are asymmetrical, the angle of attack of currents impinging on the mesh is important. In Table 3.7, the effect of current angle on Cd is shown, and it can be seen that there is a greater than three-fold variation. Interestingly, although mesh blockage is least at an angle of -20° and greatest at -70°, these angles did not correspond with the minimum and maximum Cd values, possibly because of the creation of complex ﬂow regimes around the mesh members (Woods Hole Engineering Associates 1984). Attack angle is unlikely to inﬂuence Cd values for other materials to the same extent. Using the equations presented above it is possible to derive current velocity versus loading curves of the type shown in Fig. 3.14 for different materials. However, while the relationships illustrate the effects of current velocity on nets of different mesh sizes, they are only valid for frame-mounted panels of material. Cage bags change aspect due to lift forces and undergo deformation in currents with the result that current forces on ﬂexible net bags have often been overestimated. In order to compute current forces on a cage net bag the following steps, described by Aarsnes et al. (1990) and Løland (1993a,b), should be followed:

e

d

c

b

50 50 50 50 50 50 10 25 25 25 76

Mesh sizec (mm)

Knotted netting. Fouling accumulated after two months of immersion in Scottish sea conditions. Mesh size for netting assumed to be twice the bar length (see Fig. 3.7). Polypropylene. Polyethylene.

Diamond Diamond Diamond Square Square Square Hexagonal Square Diamond Diamond Diamond

Nylon PPd (Ulstron) PEe (Courlene) PE PE (cupra-proofed) Netlon 90 : 10 Cu–Ni expanded metal Galvanized steel weldmesh Galvanized steel chainlink Plastabond (PVC-coated chainlink)

a

Mesh type

Material

2.3 2.5 1.9 1.5 1.5 3.3 1.3 2.5 2.0 2.5 2.5

Diameter (mm)

Unfouled yarn

1.42 1.47 1.33 1.26 1.26 1.19 0.41 1.30 1.24 1.30 1.09

Cd

10.2 10.2 8.9 7.6 5.1 7.6 — 3.3 3.8 5.1 6.4

Diameter (mm)

Fouled yarn

3.99 3.99 3.46 2.95 2.13 1.48 — 1.41 1.48 1.67 1.25

Cd

Table 3.6 Cd values determined for perpendicular current forces acting on various nettinga and rigid mesh materials before and after foulingb (data from Milne 1970 and Woods Hole Engineering Associates 1984.)

Cage Design and Construction 57

58

Chapter 3

Table 3.7 The effects of current angle on the drag coefﬁcient (Cd) of 10-mm 90 : 10 Cu–Ni mesh. Current velocity = 0.267 m s-1 (modiﬁed from Woods Hole Engineering Associates 1984). Angle -50 -40 -30 -20 -10 0 +10 +20 +30 +40 +50 Cd 0.23 0.41 0.29 0.48 0.51 0.47 0.73 0.62 0.63 0.58 0.59

(1)

(2) (3) (4) (5)

The cage bag or system of cages is considered as a series of net panels that lie parallel to the current ﬂow, normal to the current ﬂow or at some speciﬁed angle to the ﬂow. The number of panels depends upon cage shape, a square cage consisting of four side panels and one bottom panel, for example, while a hexagonal cage consists of six side panels and one bottom panel. Lift and drag coefﬁcients are established for the net type as a function of solidity and current attack angle. Local current velocities at each net panel are then calculated as a function of the cage system. The current force on each net is calculated as a function of the weights used to maintain cage bag shape and local current velocities. The total force on the net cage system can then be calculated as the sum of the forces acting on each net panel (Løland 1993a,b).

The method is summarized in Appendix 3.1. According to Løland (1993a,b) drag forces calculated for a group of cages lie in the range 0.9–1.2 times the measured forces, which is sufﬁciently accurate for most purposes. Currents are deﬂected around cages, distorting the shape of the bag (Fig. 3.15) and it is not uncommon for deformation to result in a loss of 25% of volume (see section 7.1). Tomi et al. (1979), in a study of the effects of currents on a cage bag, showed that the forces acting on the bag increased with current velocity, as shown in Fig. 3.16, resulting in a marked decrease in volume. However, while the effect of fastening weights to the bottom four corners of the bag resulted in a signiﬁcant amelioration of the problem, horizontal forces acting on the cage bag increased two- to six-fold. Similar ﬁndings have been made by Løland (1993b). The type of material, mesh size and degree of fouling also inﬂuence water exchange through the cage and are thus important determinants of stocking density. Studies carried out in Japan and Scotland during the 1960s and 1970s demonstrated that the transmission ratio, T (ratio of internal to external currents, Vi : Ve), falls with decreasing mesh size and with increased fouling (Hisoaka et al. 1966; Milne 1970; Inoue 1972; Edwards & Edelsten 1976; Wee 1979). In heavily fouled cages, fabricated from 13 mm nylon mesh, Wee (1979) found that T could be as low as 0.13. However, much of the work was carried out using empty cages, over a limited range of Ve values, sometimes using current meters that were operating outside their design limits, and recent research has suggested that these studies oversimplify the situation. For example, Chacon Torres et al. (1988a) have demonstrated that caged ﬁsh can greatly modify the rate and

Cage Design and Construction

59

Fig. 3.16 Relationship between volume of cage bag and current velocity with no weights (0), with 4 ¥ 52-kg weights (1) and with 4 ¥ 102-kg weights hanging from the bottom corners of the bag (modiﬁed from Tomi et al. 1979).

direction of water ﬂow through a cage by their behaviour. From a timed sequence of photographs of the dispersion of dyes in empty and stocked 1-m3 cages, they found that the exchange pattern changed from unidirectional to multidirectional and estimated that the exchange rate increased by a factor of approximately four when rainbow trout were present. During feeding, the change in ﬂow pattern was even more pronounced (Fig. 3.17). The behaviour of ﬁsh in a cage will vary with species and size and also be dependent upon environmental cues such as time of day, weather, current speed, salinity and temperature. While T may be largely independent of ﬁsh behaviour in fast-ﬂowing conditions and in extremely large cages, behaviour is likely to be an important inﬂuence on water exchange in many situations. In summary, the values given in the literature for forces exerted on netting and for T may be of limited use, but emphasize the importance of using as large a mesh size as possible and minimizing the degree of fouling. The effects of weights on net deformation and horizontal forces acting on the netting are also well illustrated. Detailed calculations are unnecessary for the construction of small cages, while most large net cage bags, which are fabricated by commercial net makers, have evolved through a continual process of feedback from ﬁsh farmers and technological innovations rather than from information generated by experimenters.

Fig. 3.17 Water movement through small (1-m3) net cages, as shown by dye dispersal over time. (a) Empty cage; (b) cage stocked with 65-g rainbow trout (13.5 kg m-3); (c) dye dispersal during feeding (from Chacon Torres et al. 1988a).

60 Chapter 3

Cage Design and Construction

61

Experimental data have proved invaluable in the design of large rigid mesh cages. Huguenin and co-workers used data derived from trials to establish the forces likely to act on various expanded metal mesh materials. These were employed in estimating the deformation of the panels and the forces that would be transferred to the frame members (Woods Hole Engineering Associates 1984). Thus the optimum materials and dimensions for panels and frames were determined. Work by Løland (1993a,b) and colleagues show that it is possible to apply a similar approach to the design of net cage bags. Construction As rigid mesh materials are usually only available in a restricted range of panel sizes or roll widths, this can dictate the size, shape and design of the cage unless sophisticated engineering techniques are employed. Rigid mesh cages usually require some sort of supportive framework, otherwise they may buckle or fracture. Small designs can be assembled by nailing or binding sheets to a frame. A top panel, usually hinged, is often included in rigid mesh designs. One deﬁciency of plastic-coated or galvanized materials is that the cut edges are prone to corrosion and must be treated with some sort of coating for protection. Similarly, bending sheets at right angles can accelerate corrosion not only by damaging any surface coating that is present, but also by inducing stress. Small, simple frames can be constructed from a variety of woods, metals or synthetic materials (Fig. 3.18). The properties of many of these materials are discussed above. Care must be taken over the choice of fasteners, not only to ensure that they are sufﬁciently strong but also to avoid creation of galvanic cells. Commercially available plastic or nylon ties and synthetic twines are inert but may be less appropriate than metal fasteners for large, metal mesh cages. Metal fasteners with similar electrode potentials to the metal mesh should be used. Huguenin & Ansuini (1978) employed nickel aluminium bronze fasteners (E0 = -0.16 to -0.22 V) for 90 : 10 Cu–Ni sheets (E0 = -0.15 to -0.30 V), as the electrode potentials were very similar (Woods Hole Engineering Associates 1984). Stainless steel fasteners, electrically isolated by plastic bushes, proved unsatisfactory due to inadvertent electrical contact. If such a system must be employed, each joint should at least be checked with a current meter prior to the cage being installed. Although some larger rigid mesh cages rely on readily available galvanized steel or 90 : 10 Cu–Ni pipe for the framework, commercial development of modular systems has resulted in specially designed and fabricated frame members. Huguenin and co-workers used a pultruded form of PES and glass composite (ﬁbreglass) which was light, strong, highly resistant to weathering, and could be produced in a wide variety of shapes and sizes (Woods Hole Engineering Associates 1984). Following extensive testing, C-sectioned 5.1 ¥ 1.4 ¥ 3.2 cm pultruded ﬁbreglass was chosen as most appropriate. Panels are constructed by stretching the 90 : 10 Cu–Ni sheeting around the frame members, as shown in Fig. 3.19, to form a panel 30 cm wide of variable length. Panels are joined together to form a rectangular or circular enclosure of variable size.

62

Chapter 3

(b)

(a)

(c) Fig. 3.18 Examples of rigid cage frames. (a) 6 ¥ 6 ¥ 3.6 m galvanized scaffolding cage for Atlantic salmon (from Milne 1979); (b) 1 ¥ 1 ¥ 1.3 m mahogany cage frame with reinforced corners, for experimental culture of tilapias, Lake Kossou, Ivory Coast (from Coche 1979); (c) 3-m3 experimental cage for use in marine waters, fabricated from ﬁbreglass hoopnet rings and vinyl-coated steel mesh (from Swingle 1971); (d) production cage, Bangladesh, with frame fabricated from 2.5-cm plastic pipe (courtesy K. I. McAndrew).

Cage Design and Construction

63

(d) Fig. 3.18 Continued.

Electrically isolated galvanized steel reinforcements are used at corners, joints, and at points where loads become concentrated. The last few years have seen companies in Australia, South Korea and Japan experiment with galvanized steel cages. Australian designs have used heavily galvanized steel chain mesh netting, ﬁtted with an aluminium sacriﬁcial anode for extra protection (Anon 1999; http://www.wwi.es/ukbienve.htm). Synthetic ﬁbre net bags are usually designed with an area of freeboard (an area which protrudes above the water surface) to prevent ﬁsh jumping out. The height of freeboard is determined by the species, some, such as the silver carp, being able to jump considerable heights. However, the inﬂuence of wind forces on exposed netting should be borne in mind (see later). A compromise is to incorporate a top net, which also serves to help exclude predators. Unless prefabricated bags are bought from a commercial company, cage bags must be fabricated from rolls of synthetic ﬁbre netting. Details of procedures are given in IDRC/SEAFDEC (1979), Bautista (1984), Christensen (1989), CostaPierce et al. (1989), Beveridge et al. (1994a) among others. A large, reasonably level area clear of debris is best. The fabrication of small square or rectangular cage bags from ﬁne-mesh (2–5 mm) mosquito netting (hapa), suitable for a lake-based

64

Chapter 3

(a)

(b) Fig. 3.19 (a) Construction details of mesh-wrapped 90 : 10 Cu–Ni panels and examples of ways in which they can be connected together to form units of various shapes and sizes (from Woods Hole Engineering Associates 1984). (b) 6.3 ¥ 3.2 ¥ 3.2 m cage fabricated from 90 : 10 Cu–Ni mesh panels (courtesy J. E. Huguenin).

Cage Design and Construction

65

Fig. 3.20 Assembly of a simple hapa net cage. The sides are cut as one piece and are formed from two lengths of material stitched along the middle. The base is made separately (not to scale).

tilapia hatchery or nursery, is comparatively simple, as shown in Fig. 3.20. The dimensions of the bag must be decided beforehand and it is simplest and most costeffective if the total depth of the bag, including freeboard, and width of the base, is some multiple of the diameter of the roll of material. For example, in Fig. 3.20, the depth and width of the bag are twice the width of the roll. Double stitching of the edges in this case is recommended, and can be done using an ordinary domestic sewing machine ﬁtted with strong nylon thread. For larger mesh netting, cutting and assembly is a little more difﬁcult. Most cage bags are fabricated from knotless netting, which can be hung on the square

66

Chapter 3

or diamond. The former is reportedly easier and more economical for small mesh (~6 mm bar length) sizes. However, if the netting is to be hung so that the meshes adopt a diamond shape, it is necessary to ﬁrst decide on the hanging ratio. Because of its shape, diamond mesh netting can be hung in a variety of ways that dictate the horizontal and vertical mesh openings. For ﬁsh cages, however, the netting is often hung with a hanging ratio (r) of 0.3, i.e. the horizontal opening of the mesh is equivalent to 70% of the stretched mesh value. For practical purposes, the following formula is used to calculate the length of netting: Ls =

Ld 1- r

where Ls = length of stretched netting (m) and Ld = desired length (m). Thus, for a 4 ¥ 3 m cage bag, a total length of 14 m of netting is desired. Assuming r = 0.3, Ls = 14/0.7 = 20 m. For the depth of netting: Ds =

Dd 2r - r 2

where Dd = desired depth (m) and Ds = depth of stretched netting (m). An alternative to measuring the stretched netting is to count the number of meshes (nm). The number of meshes can be computed using: nm =

Dd

(2r - r 2 )l

where l = stretched mesh length (m). Details of how to measure and cut netting for a circular cage bag are given in IDRC/SEAFDEC (1979). After calculating the dimensions of the side and bottom panels, cutting the roll should be planned so as to minimize waste. If possible, the four sides should be cut in one piece, as in Fig. 3.20. An additional advantage of using a knotless material is that seams can be machine stitched (Fig. 3.10). However, machine stitching is impossible without specialized machinery and so, unless the net bag is commercially fabricated, stitching of the sides to the bottom panel must be done by hand using a small net needle and twine that is similar to that of the netting yarn. If carried out by skilled craftsmen, hand sewing is arguably superior to machine stitching, the binding being tighter and the knots better. This is important in reducing friction between panels which can be a cause of net failure. A technique, often referred to as lacing, is used with knotless materials, in which twine is looped around each mesh of the two materials, a knot (e.g. clove hitch) being made every fourth loop or so. Flexible frame cages are often supplied with purpose-built net bags that are designed to cope with harsh, offshore conditions and the frame distortion that occurs during heavy seas. The net bag of the Bridgestone HI-SEAS cage, for example, is designed to absorb wave action in the upper diamond mesh section,

Cage Design and Construction

67

which distorts in response to waves and consequently reduces wave induced motion in the square hung mesh below (Gunnarsson 1996). The extent and nature of the rigging required depends primarily upon the size of the cage bag, whether it is to be used in a ﬁxed or ﬂoating type, and the design of the collar. The key principles are that ropes should be deployed to take the heaviest loads, that the density and tightness of the ropes should be sufﬁcient to secure an even transfer of the loads from the netting and vice versa, and that loads should be transferred over as many meshes as possible to reduce the risk of net failure (Christensen 2000). For small bags less than 50 m3 in volume, ropes should be attached to the outside of all edges, while for larger designs rigging may also have to be attached at intervals along the panels for strengthening and stiffening, and to facilitate lifting during harvesting and net changing. PE or PP (6–10 mm) rope is used. There are many different ways to attach netting to rope (Garner 1989). Usually the rope is simply laced to the net using a twine that is similar in diameter to that of the netting yarn. Each mesh is picked up by the needle and bound to the rope with a blanket stitch. Every 5–10 cm a clove hitch is tied to secure the rope ﬁrmly. For hapa netting, a twine of around 23 tex is used and a clove hitch tied every few centimetres. Some recommend threading the rope directly through the meshes of the netting and securing it at each mesh with a clove hitch while others suggest running a balch line through the netting in the same way and attaching the rigging to the balch line (Alferez 1982; PPD 1986). A balch line is an intermediate rope, typically 2–5 mm in diameter, of low breaking strain, which reduces tension between the main rope and the netting, minimizing the risk of tearing. On larger cage bags, ropes are stitched to the outside of the walls and ﬂoor of the bag every few metres, as described above. For ﬂoating cages, loops should be included not only at all corners but also at metre intervals on all ropes, to facilitate lifting and net changing. The loops should either be knotted and whipped, as shown in Fig. 3.21, or spliced. Weights can be hung from the rigging that runs along the ﬂoor seam in order to help the bag hang properly in the water. Alternatively, a metal frame can be secured to the exterior of the ﬂoor seam. Because of shock vertical loadings it is important that the areas where weights are attached to the cage bag are sufﬁciently strong and are reinforced if necessary. The number and size of weights required is generally determined through trial and error. Alternative methods of rigging are given in the Appendix in IDRC/SEAFDEC (1979). Many commercial submersible/semi-submersible cages, such as the Ocean Spar®, have purpose-built rigging (Loverich & Croker 1993). Bags fabricated for ﬁxed cages should have at least 0.6 m of free rope on each rigging line to facilitate ﬁxing to the frame. Top nets, usually fabricated from large mesh knotted PE or nylon monoﬁlament, are cut to size and ﬁtted to deter predators (see section 7.7).

3.3.2

Cage collars and support systems

The function of the cage collar or support system is to support the bag securely in the water column and, particularly in the case of net bag designs, to help

68

Chapter 3

Fig. 3.21 Detail of a corner of a net bag illustrating how the rope is stitched to the netting along the top and corner seams of the bag, one stitch per mesh. Note that a loop has been provided to facilitate attachment to the collar.

maintain shape. Collars and support systems may also serve as work platforms. The overall robustness of a cage (i.e. its ability to function under adverse weather and sea conditions) is governed by: • • • •

the properties of the construction material, and size and proﬁle of the structural members; the strength of joints between structural members, especially when they occur at points of stress concentration; the degree and distribution of ﬂexibility in joints and members; the design and location of mooring points to distribute resultant stresses (Kerr et al. l980).

Cairns & Linfoot (1990) review cage design from a structural engineering perspective and introduce the concept of ‘limit states’, a term that refers to ‘the conditions in which a structure would be considered to have failed to fulﬁl the purposes for which it was built’. Two types of limit state are identiﬁed: ultimate and serviceability limit states (Table 3.8). The former are catastrophic states that require a substantial safety margin to minimize risks while the latter deﬁne acceptable behaviour during routine operation. (Note that brittle fracture and durability are considered functions of the materials used rather than of design.) In order that each limit state be thoroughly considered, Cairns & Linfoot (1990) suggest the following relationship be applied:

Cage Design and Construction

69

Table 3.8 Structual engineering limit states, as applicable to cage designs (from Cairns & Linfoot 1990). Ultimate limit states

Serviceability limit states

Strength (yield, buckling) Stability against overturning and sway Fracture due to fatigue Brittle fracture

Deﬂection Vibration Repairable damage due to fatigue Corrosion

Applied force actions ¥ partial safety on loads Æ resistance of section ¥ partial safety factor on materials Both static and dynamic forces act on a cage collar or support system. They are considered in the discussion of ﬂoating cage structures. Fixed cage design is considered principally from the point of view of the materials and construction methods used. Fixed cages With ﬁxed cages the bag is supported by posts driven into the substrate. Bamboo is commonly employed for support. Bamboos are grasses that are widely distributed throughout the tropics, subtropics and mild temperate areas of the world, and are particularly common in Southeast Asia, China, India and the Paciﬁc region. The woody, hollow stems (culms) grow in branching clusters from thick underground rhizomes. They are fast growing, the culms of many species reaching 15 m or more in 2–3 years. However, of the 1000 or so species, only around 20–30 are suitable for construction purposes. Favoured species include Dendrocalamus spinosa, D. hamiltonii, D. merillianus, Bambusa blumeana, B. vulgaris and Giganthocloa gigas. The anatomy of a bamboo culm is shown in Fig. 3.22 and mechanical properties are summarized in Table 3.9. When compared with other materials on a weight-for-weight basis, bamboo is exceptionally resistant to bending and tension forces but is much weaker in terms of resistance to shear forces. Once the ultimate bending force is exceeded failure occurs because of a loss of cohesion between ﬁbres (Janssen 1981). Thus, under excessive bending forces, bamboos tend to split rather than snap. While they may have certain disadvantages for use as supports in a ﬁxed cage, bamboos are easily split and readily fabricated into a cage mesh. The properties of bamboo vary with species, moisture content, age and position along culm. Culms of 3–5 years old have the best mechanical properties and dry (12% moisture content) specimens are more resistant to bending forces than green ones. Seasoning is thus important. Position along the culm has also been shown to inﬂuence resistance to shear and bending, top sections being signiﬁcantly weaker but stiffer. Despite their many excellent properties, bamboos suffer from one important disadvantage: they have a comparatively short useful working life. The material

70

Chapter 3

Fig. 3.22 Anatomy of a bamboo culm (redrawn from Janssen 1981).

rots comparatively quickly, is susceptible to destruction by boring insects and eventually becomes waterlogged, losing ﬂotation properties. Sections at the water line degrade most rapidly through the combined action of weathering and continual immersion and drying out. Thus in ﬁxed cages it is the section exposed to the air that tends to fail ﬁrst. Bamboo tends to have a useful working life of 18–24 months in fresh water and 12–18 months in sea water (IDRC/SEAFDEC 1979). Hardwoods and softwoods are considerably more resistant to weathering and rotting but are much more costly. In the Philippines, a common

Cage Design and Construction

71

Table 3.9 Summary of mechanical properties of various cage collar construction materials (data from Chong 1977; Janssen 1981; Bjerke 1990; Rossbach 1993).

Density (kg m-3) Ultimate tensile strength (N mm-2) E-valueb (tensile strength) (N mm-2) Ultimate short-term bending strength (N mm-2) E-value (bending stress) (N mm-2) Ultimate shear stress (N mm-2) a b

Bamboo

Softwood

Hardwood

Concrete

HDPMa

Steel

600–800 200–300

450–650

400–750 20–110

2 400 4

— —

7 800 270–700

17 400–10 300

5000–9000

8000–18 000

25 000–35 000

900

210 000

84

—

—

—

18.5

—

20 500

—

—

—

—

—

2.25

11

14

—

—

—

VESTOLEN® A5041 R (high-density polyethylene material). E-value = modulus of elasticity = stress/strain ∫ ratio of load per unit area to deformation per unit length.

alternative to bamboo is the anahaw palm (Livingstonia rotundifolia). Although it costs at least ﬁve times as much as bamboo, it has a lifespan of 3 years and, since it is considerably stronger, only a fraction of the number are required as supports, an added advantage at sites where storms and water hyacinth problems occur (see section 7.3). The posts in a ﬁxed cage fulﬁl the same functions as the collar in a ﬂoating cage, i.e. they must resist both the static vertical forces imposed by the weight of the net and working staff and the horizontal dynamic forces exerted by wind, waves and currents. Unfortunately, there has been insufﬁcient experimental work carried out to provide any but the most general guidelines. In view of the high compressive strength of bamboo (Janssen 1981) static forces are likely to be negligible and, of the dynamic forces exerted on the netting, Milne (1970) has shown that currents are the most important. Currents induce deformation of netting and so supporting posts must be used at regular intervals. The methods developed for computing spacings of supports for net barriers (Milne 1970) cannot be readily applied to cages due to their threedimensional conﬁguration. Fish farmers have found that bamboo culms spaced 1–2 m apart are adequate for most conditions. The posts supporting the net bag must be buried sufﬁciently deep to resist moving as a result of wind, wave and current forces acting on the structure. Milne’s method (Milne 1970, 1972) for net barrier pile calculations, applied to a typical ﬁxed cage design using bamboo support posts at 1-m intervals, suggests that a depth of 2 m is sufﬁcient to support the structure securely, even in soft puddle clay. This is likely to be conservative, considering the support given by horizontal cage frame members directly opposing the prevailing currents. Depths of 1–2 m have been found to be adequate, depending on the nature of the netting, the prevailing forces, the dimensions, spacing and type of posts used, and the cohesive properties of the soil.

72

Chapter 3

Fig. 3.23 Framework, constructed to facilitate the cleaning and preparation of bamboos.

Assuming a buried depth of 2 m, a water depth of 6 m and that the bamboos extend for 3 m or more above the water line, culms of at least 11 m are required. Ideally the bamboo should be well seasoned, aged >2 years, and should have a diameter of around 10–12 cm at the base and 3 cm at the tip. Culms should ﬁrst be cleaned of any sharp projections at the nodes which could damage the netting, and the node diaphragms or internodes of the section to be submerged should be perforated to help sinking. Simple tools, such as machetes and chisels, and a purpose-built work base by the water’s edge are all that are required (Fig. 3.23). The bamboos are ﬂoated to the site on a raft that also serves as a construction platform (Fig. 3.24). Two or three people can construct a small ﬁxed cage in a few hours. Posts are driven into the mud by a twisting motion, often with someone in the water to aid accurate placement. The corner posts of one side are ﬁxed ﬁrst and a rope stretched between the two to ensure that the posts lie along a reasonably straight line. After one side has been completed, the others are ﬁnished in a similar manner. Horizontal supports, which utilize the top and middle sections of the culms, are lashed to the vertical posts – usually one on either side of a row – about 1.5–2.0 m above the water line and serve not only to strengthen the structure but also as a narrow walkway. The completed framework should be around 0.5 m longer on each side than the net bag dimensions to minimize the risk of abrasion. The cage net is then tied to the posts, the bottom rigging being secured by divers. A commercial ﬁxed cage, as used for tilapias in the Philippines, is shown in Fig. 3.24. Where a shallow surface mud stratum is encountered, anchorage of the posts can be improved by driving bamboo pegs, 25 mm diameter ¥ 450 mm long,

Cage Design and Construction

(a)

(b)

(c)

(d)

73

Fig. 3.24 Construction of a ﬁxed cage, training course, SEAFDEC, Binangonan, Philippines. (a) Corner posts are established ﬁrst and a line is stretched between posts to aid positioning of intermediates; (b) intermediate posts being driven into the substrate; (c) laterals are lashed in place to reinforce the structure; (d) completed cages.

through the internodes at the base of the culm to form two cross-pieces, about 30 cm apart (Alferez 1982). Bracings to strengthen the structure may also be used at the corners and on the exposed side of the cage. Additional horizontals will stiffen the cage framework. If softwood or hardwood posts are used instead of bamboo, then notches can be cut into the trunk and temporary rungs nailed in place so that a person can use his or her weight to help drive the posts into the substrate. Under difﬁcult conditions, a raft equipped with scaffolding may be useful (Fig. 3.25).

74

Chapter 3

Fig. 3.25 Construction raft to help drive posts into difﬁcult substrates.

Floating cages Wave, wind and current forces may be considered as principally horizontal loads imposed on the cage collar while the mooring forces are reactions. Similarly, gravity can be considered as an imposed vertical load and buoyancy forces as reactions. The horizontal loads are dynamic in nature while the vertical loads tend to be static although the inertia of the cage system and damping forces impose some dynamic loadings in the vertical plane, particularly with respect to waves (Cairns & Linfoot 1990) (Fig. 3.26). In an analysis of ﬂoating steel cage designs Cairns & Linfoot (1990) state that the resistance of most cage collar components to horizontal bending forces is in the order of a thousand times greater than resistance to vertical bending, and that stresses in main structural members exerted by horizontal loadings should be fairly low, provided mooring systems are adequate. They support their argument by stating that most cage structural failures in Scotland have arisen from excessive vertical loadings. Gravitational static loads comprise the weight of the bag, superstructure (including stanchions, freeboard netting and top nets, handrails, etc.) and additional loads that may arise during routine operations (farm staff, feedbags, graders, pumps, aerators). The static vertical loads imposed by ice may also have to be taken into account. Such loads must be quantiﬁed prior to designing the ﬂotation system (Appendix 3.2 gives an example).

Cage Design and Construction

75

Fig. 3.26 Principle forces acting on a cage system (dashed lines = dynamic forces; solid lines = static forces).

In order to compute the weight of the cage in water, the densities of the materials used must be established. For the cage to ﬂoat, the static loads acting on the structure (i.e. the weight in water) must be counter-balanced by buoyancy forces. The buoyancy of the collar is dependent upon the upward force acting on those components wholly or partially immersed in the water and is equal to the weight of water displaced. The buoyancy force can be calculated from data on the density of materials used to construct the collar: FB = VWQW - VMQM where FB = buoyant force (kg); VW and VM = volumes of water and the ﬂotation material respectively (m3); and QW and QM = the densities of water and the ﬂotation materials (kg m-3), respectively. For example, 1 m3 of water at 20°C weighs l000 kg, and 1 m3 of bamboo weighs approximately 600 kg (from Table 3.9). Thus, the buoyancy force acting on 1 m3 of bamboo in water = (1 ¥ 1000) - (1 ¥ 600) = 400 kg, i.e. 1 m3 bamboo can support a load of 400 kg. In some designs the ﬂotation system is an integral part of the cage collar. A simple collar may be constructed almost entirely from bamboo, the bamboo serving not only to maintain the shape of the bag but also to support the weight of the structure (see later). Other such designs use gas and expanded polyurethane and polystyrene foam ﬁlled plastic and synthetic rubber pipes. The alternative is to choose dense materials such as steel and aluminium alloys for strength and to add supplementary ﬂotation such as steel drums or blocks of expanded synthetic polymers. While it would thus seem possible to specify the ﬂotation capabilities of a cage from the densities of materials alone, in practice it is not quite so

76

Chapter 3

straightforward. The above equation assumes an equal weight distribution around the collar, taking no account of vertical loadings that may be acting on one side or indeed at one point (e.g. at a corner), which occurs when cage collars also serve as work platforms. The effect of cages being linked together is also ignored. One way to estimate the necessary buoyancy is to assume that each side of the cage must support not only its share of the loadings imposed by the net and superstructure but also the appropriate maximum proportion of the mass of staff and equipment likely to be on the cage. For most materials, the data for quantifying the amounts required for ﬂotation are readily available (see Table 3.9). However, for natural materials such as bamboo, density is variable and depends upon species, age and seasoning. Because of its irregular and hollow shape, it is also difﬁcult to translate data on density per unit volume into ﬂotation properties per culm, although tables estimating culm weight are available (see also Appendix 3.3). The buoyancy of steel or plastic drums can be calculated from their weight and volume using the formula given above. A further difﬁculty with wood or bamboo is the loss of buoyancy with time as a result of rotting and weathering. Some otherwise ideal materials, such as plastic or steel drums, rapidly accumulate fouling organisms, reducing ﬂotation capabilities and increasing the drag on moorings. As a result maintenance (cleaning) costs can be high. Antifouling compounds retard the rate of fouling (see section 7.4). An alternative is to cover plastic and steel drums in tough polythene bags, which are changed as necessary. Steel and plastic drums may also be rotated every few months, exposing fouling organisms to the air. Some ﬂotation materials are brittle or prone to weathering. Expanded polystyrene, for example, suffers from both deﬁciencies; it is not only brittle but also becomes increasingly susceptible to damage on impact with exposure to ultraviolet (UV) light (see Fig. 7.37). These materials may be protected by coating in glass reinforced concrete or ﬁbreglass. In view of the difﬁculties in quantifying the materials required to provide adequate buoyancy, it may seem tempting to over-engineer the ﬂotation capabilities of the collar. However, while cage builders include a substantial margin for safety, over-speciﬁcation can be costly and can adversely affect the behaviour of the cage, exacerbating shock loadings on the collar and mooring system. A body can be said to be in a state of stable equilibrium if, when displaced, forces are established that tend to restore the body to equilibrium. By virtue of its shape and because most of the mass lies close to the water line, a ﬂoating cage has a low centre of gravity. When a wave passes, the collar inclines through a small angle. Although the position of the centre of gravity of the cage (G) remains the same, the centre of buoyancy, B (the centre of gravity of the displaced water), moves to a new position, B1 (Fig. 3.27). The intersection of the vertical line through B1 with the line that passed through the original centre of buoyancy is known as the metacentre (M). The magnitude of the restoring couple that restores the body to equilibrium is WGM sin q, where W is the mass of the cage (kg). If the buoyancy of the cage collar is increased without greatly increasing the total mass, W, which is possible if low-density materials such as expanded polyurethane or polystyrene are used, then the centre of gravity is raised. The

Cage Design and Construction

77

Fig. 3.27 Centres of buoyancy (B) and gravity (G) in a ﬁsh cage. When the cage tilts B moves to a new position B1 (see text).

resultant decrease in draft also causes the distance between the metacentre and the centre of gravity (the metacentric height) to increase, increasing the stability of the cage (for review see Stokoe 1973). However, the magnitude of the restoring couple also increases so that the period of roll is reduced. From studies of the behaviour of ships, it is known that increasing the metacentric height results in increased stresses on the structure while the resultant motion causes those on board to feel increasingly uncomfortable. Excess surplus buoyancy in cages is not only likely to induce a great deal of motion, making it unpleasant or even dangerous for staff, but also excessive movements of the bag may adversely affect stock (see Linfoot et al. 1990 for discussion of effects of cage motion on ﬂatﬁsh). Moreover, any increase in motion of the collar will unnecessarily increase the forces acting on the structure. Because wave and current forces acting on the collar are proportional to surface area (see later), reductions in the dimensions of the ﬂotation system are desirable, particularly if cages are to be located at an exposed site. It follows, therefore, that the use of low-density materials for supplementary ﬂotation is desirable, although resistance to fouling and weathering are also important considerations. As stated above, the dynamic horizontal forces acting on a ﬂoating collar are principally due to wind, currents and waves. The loadings on a horizontal member of a cage collar can be deﬁned by six independent force actions – three linear forces in three orthogonal directions and three moments about the same three axes, resulting in a torque about the axis Tx and bending in the horizontal and vertical planes (Fig. 3.28). Wind forces act on the superstructure – stanchions, handrails and freeboard netting, ﬂoats, etc. – that lie above the water line. For design purposes, gust speeds are used to calculate wind forces, and the necessary long-term values can often be obtained from meteorological ofﬁces. Shellard (1965), in an analysis of data for the British Isles, reported that 49–54 m s-1 gust speeds occur on average once every 50 years in coastal sea lochs while 54–58 m s-1 gust speeds can be expected in offshore areas and the Outer Isles. Milne (1970) proposed the following equations be used to estimate wind forces:

78

Chapter 3

Fig. 3.28 Torsion and bending forces in a sea cage component (modiﬁed from Cairns & Linfoot 1990).

FWIND = 0.0965 A V 2 FWIND =

1 r aV 2 10 -2 2

(Boven 1968) (Pankhurst & Holder 1952)

where FW = wind force (kg); A = area (m2); V = wind velocity (m s-1); k = resistance coefﬁcient = I - b/2; b = blockage coefﬁcient = (1 - d/l)2; d = mesh diameter (m); I = nominal mesh size (m); and ra = density of air = 1.227 kg m-3. The former equation is used for solid components and the latter is used to calculate wind forces on freeboard netting and mesh. Pankhurst and Holder’s equation is likely to overestimate loadings since it was based on knotted materials. In Fig. 3.29, the relationship between gust speed and wind force on various types of netting and meshes is shown. Assuming that 25-mm nylon is used and that the freeboard height is around 1 m, then wind forces at coastal marine sites are unlikely to exceed 42 kg m-2. However, the forces produced by equivalent wind velocities on solid structures are around 280 kg m-2, emphasizing the importance of designing buoyancy carefully and reducing cage superstructure to a minimum. In order to illustrate the impact of wind forces on cage superstructure, Turner (2000a) uses a typical 60-m plastic circle cage as an example. Some 18.5 m2 of material is likely to be exposed to prevailing winds and, at gust speeds of 54 m s-1, the total load on the cage is in the order of 34 kN, or 3.4 t. To maintain such a cage on station during such an event, a towing vessel of at least 500 hp would be required. The horizontal drag forces exerted by currents on the collar can be described using an equation identical to that given above. The current velocity at most coastal marine sites is unlikely to exceed l.5 m s-1 (see Chapter 4) and, again, k must be determined experimentally. In trials with a commercial high-density polyethylene cage design, Slaattelid (1990) determined that forces imposed by current drag on the collar (as measured by tension on mooring lines) were as much as 4.5 kN at currents of up to 0.58 m s-1, equivalent to approximately 25%

Cage Design and Construction

79

Fig. 3.29 Wind forces on netting for various wind speeds (from Milne 1972).

of the total (collar plus net bag) drag force imposed by currents. However, if deformation of the circular structure is permitted, forces transmitted by the current to the moorings fall dramatically. The long fetches and relatively stable wind conditions that prevail in the Mediterranean can produce extremely long wavelength seas and swells, which can result in very severe inshore conditions for cage farmers (Turner 2000b). The Mediterranean apart, breaking waves are usually conﬁned to waters considerably shallower than half the incident wavelengths. It is, therefore, the impact of the kinetic rather than the potential energy of waves that is of concern. To calculate the wave forces acting on a collar, the horizontal and vertical orbital velocities of the water particles must be known. These can be derived from information on prevailing wave periods, signiﬁcant wave heights and water depth at the site (see Chapter 4) and by employing Stoke’s theorem for waves of ﬁnite amplitude. The horizontal (m) and vertical (w) components of wave particle orbital velocity can be calculated from: m=

pH cos h[2p(z + h) L] cos q t sin h(2ph L)

w=

pH sin h[2p(z + h) L] sinq t sin h(2ph L)

(Muir Wood & Fleming 1981)

where H = wave height (m); t = wave period (s); L = wavelength (m); h = depth of water (m); z = variation from mean water level (m); and q = angle of wave relative to the structure. The equations apply to waves where 0.04 < h/L < 0.5 (i.e. conditions prevailing at most sites). In Fig. 3.30 the maximum horizontal component of orbital velocity, m, is plotted against wave height for waves of restricted period (2.5 艋 t 艋 8 s) in 6

80

Chapter 3

Fig. 3.30 Horizontal orbital velocities in 6 m (20 ft) and 15 m (50 ft) of water at various wave heights and periods (from Milne 1972).

and 15 m water depth. According to Milne (1970), m is unlikely to exceed 2 m s-1 in most ﬁsh farming conditions, and thus this value can be used for design purposes for marine cages. The forces generated by m on exposed netting have been discussed by Milne (1970) who proposed that the equation given above for calculating the drag forces exerted by currents on netting and mesh materials is applicable. From observations of fouling on exposed netting he suggested that for design purposes approximately 20% of the surface can be considered as fouled and that in effect this results in a two-fold increase in horizontal wave forces. However, most of the horizontal wave energy will be expended on the collar rather than on freeboard netting. The force exerted on the collar can be described by an equation similar to that derived for the effect of currents on ﬁxed net and mesh panels: FWAVE = kr m 2 A where FWAVE = force (N); r = density of water (kg m-3); and A = area of the cage collar perpendicular to the wave train (m2); k is a dimensionless constant, similar to Cd for netting and meshes, whose value will depend upon the nature of the collar (material, shape, construction, etc.) and wave characteristics. Unfortunately k must be determined empirically. However, assuming a design m value for marine cages of 2 m s-1, F will be considerably less than 400 kg m-2 which occurs when k = 1. The importance of minimizing the surface area of the collar perpendicular to the water surface is also apparent from the above equation. The vertical component of wave particle maximum orbital velocity, w, is approximately 83% of the maximum horizontal orbital velocity (Wiegel 1964). Cairns & Linfoot (1990) and Linfoot et al. (1990) conducted extensive experi-

Cage Design and Construction

81

mental tests to determine torsional, shear and bending forces that would be exerted on cage collar members by vertical wave motion. Using a conventional square collar design they concluded that when the wave train was incident to the cages (i.e. at an angle of 0°) torsional and shear forces were comparatively unimportant and that bending forces on cage collar members lying parallel to the wave train were of greatest signiﬁcance. Bending moment maxima and minima were recorded at the mid-point of the cage member and forces reached a maximum when wavelengths were slightly greater than the dimensions of the cage. Torsional forces became signiﬁcant when the waves impacted on cages at an angle, maximum torsional forces being generated during the mid-point of the wave period. However, greatest forces occurred at linkage points rather than in cage collar structural members. The authors concluded that maximum vertical bending stresses on cage members are unlikely to exceed 100 N mm-2; by comparison, the design strength of Grade 43 steel used in commercial designs is 275 N mm-2. Of greater relevance is the effect of the cyclical nature of the loadings imposed by waves, which result in continuous stressing and de-stressing. Wave period (the time taken for a wave crest to travel a distance equal to one wavelength) deﬁnes loading cycles. At inshore sites with small effective fetch lengths (see Chapter 4) wave periods are typically shorter than in offshore sites and are around 2.5–6 s. In one month, a component at an inshore site can be exposed to one million such cycles. Prolonged exposure to cyclical loading forces in the order of only 10% of those required to cause failure under a single application may result in failure through fatigue, particularly of welded components. In assessing structural strength of various types of steel construction, Cairns & Linfoot (1990) concluded that channel section cage collar components have lower torsional and bending moments than identically scaled hollow section components (Fig. 3.31).

Fig. 3.31 Torsion in sea cage components: inﬂuence of section type. Dashed line = channel section; solid line = hollow section (redrawn from Cairns & Linfoot 1990).

82

Chapter 3

In summary, although much is still to be learned about the complex forces impinging on cage collars, it is clear that vertical dynamic forces imposed by waves are the most important. These exert bending forces on cage structural members when waves are incident to the cages, but also exert signiﬁcant torsional forces at linkages when impinging at an angle. Bending forces are at a maximum at wavelengths similar to the dimensions of the cage; hence, inshore sites may be as vulnerable to such forces as offshore sites. Cyclical stresses imposed by the periodicity of waves may cause fatigue in cage members, which may be of greater signiﬁcance than occasional very large waves. In this respect, inshore sites with short fetches and small wave periods may be worse than offshore sites. In terms of construction, Cairns & Linfoot (1990) and Slaattelid (1990) have shown that certain construction types are better able to resist bending and torsional forces than others and recommend avoiding placing joints at the most highly stressed locations. A proportion of the wave energy absorbed by a cage will be translated into kinetic and potential energy causing the cage to move, the residual forces being either dissipated within the structure or transmitted to the adjoining cages, moorings and the water. A ﬂoating body, such as a ﬁsh cage, can have three rectilinear and three rotational motions. The rectilinear motions are: • • •

heave: vertical motion; surge: horizontal motion along the longitudinal axis; sway: horizontal motion along the transverse axis.

The rotational motions are: • • •

yaw: rotation about the vertical axis; roll: rotation about the longitudinal axis; pitch: rotation about the transverse axis.

When a body is freely ﬂoating, heave, roll and pitch occur as a result of wave action and the displacement from equilibrium is corrected by gravity acting on the mass (see earlier). When moored, the body has three additional natural periods, surge, sway and yaw, with the elasticity and line weight of the mooring system and the cage net being the restoring force (Wiegel 1964). Studies of heave and surge in cages with and without nets show that the correction effect of the net bag on heave is only of signiﬁcance in conditions with very short wave periods, although the net bag signiﬁcantly damps surge under a range of wave periods (Ormberg & Slaattelid 1993). While mooring has little effect on heave, roll or pitch of a cage or cage group moored from a single point, multiple-point moorings will greatly reduce and/or modify these motions. Energy dissipated within the collar is translated into deformation of structural members and joints. While horizontal forces acting perpendicular to a moored square or rectangular cage will tend to induce bending, most materials and construction types are highly resistant to such forces (Cairns & Linfoot 1990). All structures are ﬂexible to some degree and for most cage types some ﬂexibility in the collar is desirable, not only for the reasons given above but also since it

Cage Design and Construction

83

allows the cages to ‘ride’ the waves to a certain extent, reducing both the incident forces acting on the collar and the motion of the bag. Collars constructed from high-density polyethylene materials (HDPE/HDPM), such as VESTOLEN®, for example, can deform (Slaattelid 1990; Ormberg & Slaattelid 1993), reducing incident wave forces on the structure and dissipating energy. Kerr et al. (1980) argue that since it is difﬁcult and costly to design effective moveable joints, joints should be kept as rigid as possible and that the design should instead exploit ﬂexibility inherent in structural members. Flexibility is, of course, undesirable in the collar of rigid mesh designs and can be minimized by mounting the mesh panels or framework at the points of least distortion (Huguenin & Ansuini 1978). To quantify the energy absorbed by the collar and the resultant deﬂection or deformation of structural members and joints, complex and detailed analyses are required (Cairns & Linfoot 1990; Slaattelid 1990; Ormberg & Slaattelid 1993). However, an indication of the degree of the deformation that may occur in a structural member can be derived from the modulus of elasticity (E-value; ratio of stress : strain). Mild steel has an E-value of around 210 ¥ 103 N mm-2, while wood and bamboo have values of 5–9 ¥ 103 and 10–17 ¥ 103 N mm-2, respectively. By comparison, HDPM has a modulus of elasticity of only 900 N mm-2 (see Table 3.9). In sheltered inland waters where currents are low, simple cages constructed from materials of comparatively low structural strength are adequate. One of the simplest, widely used in the Philippines, is fabricated from bamboo. For a 3 ¥ 3 ¥ 3 m cage, suitable for rearing tilapias or carp, 20 seasoned bamboo culms and 100 m of 4-mm PE rope are required. The bamboos are prepared as described above for ﬁxed cages and cut into lengths as shown in Fig. 3.32. Using tyres to support each corner, ﬁve lengths cut from the culm bases are laid out to form each side of the collar. Corners are secured using bamboo pegs and rope. The pegs, approximately 25 cm long ¥ 12 mm in diameter and tapered to a point, should be cut from an inter-node section as near to the base of the culm as possible and driven through the bamboos (see Fig. 3.33). The rigidity of each side is improved by lashing sections of split bamboo, outer face upwards, across the walkways. Stanchions, approximately 1 m long, are secured at each corner and handrails run between them. Two struts are required at each corner to secure the stanchions ﬁrmly enough to support the cage bag. The construction process is shown in Fig. 3.33. Approximately 3 man-days are required to build a cage. The buoyancy of the collar is around 110 kg – sufﬁcient to support the weight of one person and the cage bag (see Appendix 3.3).

Fig. 3.32 Use of a bamboo culm for ﬂoating cage construction.

84

Chapter 3

(a)

(b) Fig. 3.33 Construction of a ﬂoating bamboo cage collar. (a) Detail of corner; (b) uprights, sides and support struts being lashed in place; (c) the cage near completion; (d) the cage ready to be towed to its site.

Cage Design and Construction

85

(c)

(d) Fig. 3.33 Continued.

For larger cages plastic or steel drums can be used for supplementary ﬂotation and where cages are to be used in more exposed locations (e.g. sheltered coastal sites) wood and steel bolts should be substituted for bamboo. Hardwoods such as Hopea odorata (chengai pasir) and Dryobalanops aromatica (kapur), which are

86

Chapter 3

Fig. 3.34 Two sketches of simple ﬂoating cage designs used for cage culture of ‘mojarra casta rica’, Yucatan, Mexico. (a) Side view of a wood/oil drum construction; (b) plan view of a larger, wood, galvanized steel pipe and polystyrene construction (from Beveridge et al. 1994a).

used for cage construction in Southeast Asia, and mangrove woods (Rhizophora mangle, Avicennia spp.), which have been used in Mexico and elsewhere (Beveridge et al. 1994a), are considerably denser and more resistant to bending and shear forces and are less susceptible to rotting and damage by boring organisms. If regularly treated with preservative chemicals, a hardwood collar should

Cage Design and Construction

87

last up to 4 years (Chua 1979). Cage collars constructed from mangrove wood, galvanized steel pipe and oil drums/polystyrene are shown in Fig. 3.34, variations on which can be found in FAO/UNDP South China Sea Programme publications, in Bautista (1984), Christensen (1989) and Costa-Pierce et al. (1989). Among the ﬁrst commercial designs was the Kames cage, developed by Stuart Cannon, Kames Fish Farming Ltd, Scotland, in the late 1960s. Over the past 35 years, it has been extensively used in both inland and sheltered coastal areas throughout the world (Fig. 3.35). Although there are now numerous variations, it typically consists of walkways that run along two sides of the cage joined together at either end by timber cross-beams. The walkways are fabricated from larch (Larix spp.) – a comparatively dense (590 kg m-3) and strong softwood, which should be pressure treated with wood preservative compounds to improve rot-resistance – and blocks of expanded polystyrene. Galvanized steel brackets secure the corners and galvanized steel stanchions bolted to the walkways and cross-beams support the wooden hand and net rails. Construction details are given in Berry (1981). As discussed at the beginning of this section, there is a growing tendency to use larger cages and to rear ﬁsh in less sheltered locations. Both trends have exposed weaknesses inherent in the types of cage design discussed above and have resulted in a number of solutions being employed. Some have involved no more than a few modiﬁcations. Materials with higher speciﬁcations have been introduced: galvanized steel or high-density polyethylene (HDPE) has replaced much of the timber, and larger and heavier brackets and bolts have been used at joints between structural members. Such materials are, of course, considerably more expensive, but arguably have a much longer life-span – 10–12 years for a steel-framed design. In other cages corner joints have been re-designed to increase resistance to twisting, compression and extension forces (see Fig. 3.36), and to withstand the additional strain of supporting feeding tubes from feed barges. Steel tubing should be joined by welding or with properly designed clamps, care being taken not to place joints between sections of structural members at the mid-point (Cairns & Linfoot 1990). Instead of bolting stanchions to the walkways, some designs have them incorporated into the walkways and have handrails fabricated from galvanized steel so that the cage superstructure performs a greater structural role, increasing resistance to bending and twisting. Other designs have abandoned conventional shapes and materials but have retained the principle of maintaining rigid joints and using structural members with some inherent elasticity (see earlier). By virtue of their shape, octagonal and hexagonal designs may absorb less energy than rectilinear collars (they present a smaller surface to prevailing forces) and distribute the loads over a larger number of joints so that the load per joint is reduced. However, 135° joints are weaker than right-angles and must be carefully designed. By eschewing corners, circular collars are inherently more robust although they can also be more difﬁcult to fabricate. For example, the circular ﬂoating cages developed by SEAFDEC for culturing milkﬁsh (see Fig. 2.2d) are constructed from epoxy-resin coated galvanized iron pipe and require a purpose-built frame for bending pipes to speciﬁcations (Yu et al. 1979). HDPE tubing which is light, strong and highly resistant to rotting and weathering, ﬁlled with high-density expanded

88

Chapter 3

(a)

(b) Fig. 3.35 (a) Kames-type cage. (b) Kames collar under construction.

Cage Design and Construction

89

Fig. 3.36 Corner joint for a ﬁsh cage fabricated from 6 mm galvanized steel plate and tubing. Dimensions in mm (redrawn from Berry 1981).

polyurethane, can be easily formed into circular collars and is widely used for marine ﬁsh culture throughout the world. The base is typically fabricated from 16–25-cm diameter pipe, smaller diameter, hollow sections serving as stanchions and handrails. Lengths of pipe are welded or bolted together on site to form circular or octagonal collars of various sizes. Collars fabricated from HDPE have some ﬂexibility. However, collars may incorporate two or three concentric rings connected by hollow-sectioned struts and saddle joints (Fig. 3.37) which strengthen and stiffen the structure, increase buoyancy and improve the working environment (see also Fearn 1990; Slaattelid 1990; Ormberg & Slaattelid 1993; Brittain 1996; Gunnarson 1996). Square cages, fabricated from HDPE, are also common. A number of commercial companies have approached the problems of farming in large cages at exposed sites by designing ﬂexible collars that undergo pronounced deformation under the inﬂuence of external forces. This both reduces the forces impinging on the collar, since it tends to adopt the shape least resistant to external forces, and dissipates absorbed energy through distortion of the structural members. The collars may be square or six- to eight-sided and the structural members fabricated from synthetic rubber ﬁlled with air under pressure. Joints are rigid and made of steel or aluminium (see section 7.9 for further discussion).

90

Chapter 3

(a)

(b) Fig. 3.37 HDPE ﬂexible circular cage. (a) Sea bream cage, Djerba, Tunisia; (b) detail showing saddle joint connecting inner and outer rings.

Cage Design and Construction

91

Discussion of the design and construction of ﬂoating collars has so far ignored the working environment. In many small cages the collar provides little more than ﬂotation and so must be serviced from a pontoon or pier or by boat. Many of those used for culture of freshwater species in the United States have been designed in this way (see McLarney 1984). Although the large, ﬂexible cages provide a narrow collar for staff to carry out simple tasks, they are difﬁcult to work from in rough weather. Moreover, on account of their size all heavy routine jobs such as grading and net changing must be carried out from pontoons or boats, often using power-assisted gear. The majority of ﬁsh cages have collars that serve as work platforms. Not only must they be sufﬁciently buoyant, they must also provide a base that is safe to work from. Most working collars have walkways that are at least 1 m wide. Others, which have narrow walkways on several sides, may be connected to a wide central pontoon that facilitates the transport of feed, etc. In some designs the pontoons are sufﬁciently wide and buoyant to support small trucks. Wooden slats, such as those used in Kames-type designs, give an excellent non-slip surface providing they are regularly cleaned and kept free from algae and other debris. Covering wooden decking with chicken wire or strips of scrap rubber conveyor belt material can help. The decking of steel or aluminium designs either has a non-slip ﬁnish, or is fabricated from an expanded metal mesh which not only provides good grip but is also self-draining. Handrails and stanchions are always ﬁtted on the inside of walkways and, in some designs, on the outside as well. In some designs, such as the Dunlop ‘Tempest’, the handrails have been developed and strengthened to provide additional support for staff during routine operations such as harvesting ﬁsh or lifting nets (Brittain 1996). Although strong, metal handrails can be unpleasant in climatic extremes (see also http://www.plasticfab.com.au for details of commercial designs).

3.3.3

Cage groupings and linkages

Groupings Although ﬂexible collar and some small rigid collar cages may be moored individually, it is common practice to group cages together, not only because it simpliﬁes and minimizes the costs of mooring arrangements but also for management reasons. The number and arrangement of cages grouped together depends on: • • • • • •

the size of the farm; the size and nature of the site; the shape and design of the cage and linking system; mooring constraints; environmental considerations; disease control and other management policies.

For a small production or artisanal unit, the ﬁrst of the above considerations may be of little consequence. However, for larger farms, the farmer may wish to organize production into convenient groupings of cages. Salmon farmers increas-

92

Chapter 3

Fig. 3.38 DO concentrations (mg l-1) as a function of the position of the net cage, for different current velocities (U) and net solidity ratios (Sn). The model assumes 10 ¥ 10 ¥ 10 m cages stocked at a rate of 20 kg m-3, and a ﬁsh oxygen consumption rate of 2.3 mg kg-1. a: Sn = 0.15, U = 0.25 m s-1; b: Sn = 0.15, U = 0.10 m s-1; c: Sn = 0.25, U = 0.25 m s-1; d: Sn = 0.25, U = 0.10 m s-1. (From Løland 1993a.)

ingly wish to separate cohorts and indeed stocks of ﬁsh from different origins as part of a disease management strategy. The size, shape, depth and physical characteristics (e.g. degree of exposure) of the site may restrict the positioning of the cages and dictate the number that can be grouped together. While square and rectangular cages can be readily assembled in a variety of conﬁgurations, hexagonal, octagonal and circular designs may be more limiting. Similarly, the linkage system on certain commercial designs is such that the cages can only be grouped together in a limited number of ways. Limitations may also be imposed by the availability of suitable anchorages at the site. Environmental considerations are also important. In an experimental study of ﬂow rates through a series of empty cages aligned parallel to the principal current direction, Inoue (1972) showed that transmission rates fell dramatically from cage to cage. Løland (1993a) has also modelled the effect of grouping cages together on oxygen concentrations (Fig. 3.38). Although net deformation and effects of caged stock on water currents (see Chacon Torres et al. 1988a) were ignored in the analysis, marked differences in water quality between cages on the outside and in the centre of a group can be expected. Effects on currents and waste dispersal will occur. It is recommended that although cage groups may have eight or ten cages lying across the current there should be no more than two or three cages parallel to the principal currents. Linkages As shown above, grouping cages together markedly reduces the forces acting on individual cages and modiﬁes cage behaviour, dampening rectilinear and rota-

Cage Design and Construction

93

Fig. 3.39 Old car tyres being used as fenders between cages.

tional motions. Linkages between cages should be designed so that the pitching motion is only moderately damped while rolling and yawing, and surge and sway are kept to a minimum. Not only is it difﬁcult and expensive to construct a joint that permits universal movement, but it is also undesirable to facilitate free movement between cages from either the perspective of the staff or the ﬁsh. On the other hand, the more rigid the linkages the greater the forces that are concentrated at these points and the stronger the designs must be, particularly at exposed sites. The simplest type of linkages use rope or chain, secured sufﬁciently tightly to reduce all motion, except in the vertical plane, to a minimum. Rope tends to abrade and chain is preferable although it may be necessary to protect timber collar members from abrasion. Shackles of appropriate size, or C-links, may be used to secure chain. Rubber tyres may be lashed or bolted between cages to act as fenders (Fig. 3.39). More sophisticated designs of linkages are required for larger and heavier cages and for more exposed coastal sites. When a wave train impinges on a group of cages at an angle, torsional forces at the linkages can be signiﬁcant (Cairns & Linfoot 1990). Linkages designed for coastal sites tend to be larger, made of stronger materials (able to withstand loads as high as 10–20 t) and are designed to distribute loads over a greater surface area or number of elements. The linkage shown in Fig. 3.40 consists of brackets welded to the walkways which dovetail together and which are secured by a galvanized steel pin, secured at either end by pins or bolts. Nylon or synthetic rubber bushes help reduce noise and wear, and dampen vertical motion. In the design and construction of linkages due care

94

Chapter 3

Fig. 3.40 Example of simple linkage between cages.

should be given to the choice of materials not only to satisfy loading criteria but also to avoid galvanic action (see earlier). Linkages that facilitate the rapid removal of a cage from the group during harvesting or for relocation or repair may also be desirable.

3.3.4

Access and mooring systems

Access Access to cages is usually by boat. However, where cages are moored in a sheltered site close to shore as, for example, is common in the steeply shelving fjords of Norway, pontoons may sufﬁce. Pontoons must be able to accommodate tidal rise and fall and, where the shoreline has a steep gradient or where access is via a jetty or pier, a ramp may be necessary (Fig. 3.41). Mooring systems Most mooring systems consist of lines and anchors that secure cages in a particular location. However, the moorings also inﬂuence the stresses acting on cage structural members and the behaviour of the cages in rough weather, and can affect production, proﬁtability and staff safety. They are, therefore, an important – indeed, integral – part of the cage system and should be carefully designed. Thus the collar, net and mooring components of a cage system should be

Cage Design and Construction

95

Fig. 3.41 Ramp from jetty to ﬂoating pontoon.

designed together, although in practice the cages are usually chosen or built ﬁrst with the mooring system being designed as an afterthought. However, commercial ﬁsh cage manufacturers are increasingly willing to, or even insistent that, they design and install the mooring system. Mooring requirements should be determined by the design and type of cages and the characteristics of the site. It would ﬁrst be necessary to quantify the incident forces that are likely to act on the cages under the worst possible weather conditions, and then to evaluate the proportion of energy transferred to the mooring lines and anchors. Two types of analysis can be used: quasi-static and dynamic response (Turner 2000a). Both have their relative advantages and disadvantages, but the more and better the data, even if gathered from model systems, the more accurate the outputs from either approach. As an illustration, Turner identiﬁes the steps necessary in conducting a quasi-static mooring analysis as follows: (1) (2) (3) (4) (5) (6) (7)

The mooring geometry and mooring excursion/force equations are deﬁned. The mean environmental force is applied to the system and the excursion (offset) calculated. The periodic wave forces and response amplitude are now applied to the system. The line tensions resulting from this maximum excursion are now calculated. The line tensions are now compared with the minimum breaking load of the riser components. The maximum peak anchor loads are calculated for each riser and direction. A safety factor (generally 2) is introduced into calculating the riser strengths.

96

Chapter 3

(8)

The maximum peak line loads are re-calculated, assuming one line is broken. (9) If the proposed mooring speciﬁcation fails the safety factor test, then a new speciﬁcation is tried. The present section conﬁnes itself to a discussion of the principles involved, to qualitatively evaluating different mooring systems and to providing general guidelines. Tests using models show that loadings transferred to mooring lines vary enormously depending on current and wave conditions, cage design and number of lines employed (Slaattelid 1990; Whittaker et al. 1990; Gace et al. 1996; Lien et al. 1996; Lien 2000; Frediriksson et al. 2003). Although currents can produce signiﬁcant drag forces, long period waves tend to result in the greatest loads, as much as several hundred kN per mooring line. Such wave loadings are cyclical, peak loads lasting for only a fraction of a second. In practice, mooring requirements are generally greatly over-speciﬁed, although this can lead to over-damping of heave and surge motions (Frediriksson et al. 2003). For particularly exposed locations it is best to consult other similar operations or a specialist mooring installation company. Two types of mooring system may be used: multiple and single point. The former is more common and involves securing the cages in one particular orientation while with the latter the cage(s) are moored from one point only, allowing them to move in a complete circle. Single-point moorings tend to be used with rigid collar designs in sheltered sites. They use less cable and chain than multiple-point moorings and, because they adopt a position of least resistance to the prevailing wind, wave and current forces, both inter-cage forces and torsional forces at linkages are reduced (Linfoot & Hall 1987; Thoms 1989). Lien (2000) states that single-point mooring systems also reduce the enormous net deformation seen in conventional mooring systems and have been used with success in Norway to moor large offshore cages. Cages moored from a single point also distribute wastes over a considerably larger area than those secured by a multiple-point system (Fig. 3.42). In a ﬁeld study carried out in a freshwater loch, Beveridge (unpublished) estimated that if the areas covered by identical groups of cages moored by the two systems are compared, then there is a 20–40-fold increase in the area over which the single-point moored cages discharges waste, although wastes will not, of course, be evenly distributed over the entire area. Goudey et al. (2001) proposed that the reduction in impact could be determined from: -2 rp w qf = qspm

where qf = the accumulation of solids under a ﬁxed (multiple-point mooring) system; qspm = the estimated accumulation of solids under a single-point mooring system; w = the width of the cage; and r = the distance from the anchor. Using this formula, together with various assumptions on the proportion of time the cage spends ‘wandering’, the authors estimated that there was a 2–70-fold reduction in deposition of waste per unit area of sea bed. Recent research has

Cage Design and Construction

97

90 m PEH cage Buoy 1000–2000 l

Bottom ring (170 N/m weight)

64 mm braided rope

30 mm chains

Fig. 3.42 Frøya Ringen single-point mooring system, Norway (modiﬁed from Lien 2000).

indicated that waste loadings to the sediments may play a crucial role in determining water quality and ﬁsh health, particularly at marine sites (see Chapter 5). However, while costs may be reduced by up to 50% (see Goudey et al. 2001) the use of single-point moorings also reduces the fraction of a site used for ﬁsh production. As much as 20–30% of a site may be occupied by multiple-point moored cages, the remaining space being occupied by moorings, whereas with single-point moorings the cages only occupy 2–4% of the entire site area. Lien (2000) states that research is underway in Norway to address this issue. The orientation of cages with multiple moorings depends upon the nature of the site and upon the type and group conﬁguration of the cages (see also Løland 1993a,b). If particularly exposed or if currents are strong, then it may be best to secure cages in the position of least resistance to the prevailing wind and current forces. Where a site is sheltered and water circulation is poor it may be better to moor cages so that water exchange is maximized. However, there may be restrictions on mooring orientation imposed by site size or by suitability of mooring grounds. The number of mooring lines used determines the distribution of forces to the anchors. An additional consideration for ﬂexible collar designs is that the number and deployment of mooring lines inﬂuences cage shape (see Slaattelid 1990). For conventional rigid cage designs, there remain arguments about whether to use a few judiciously placed anchors or to rely on many mooring lines. Chettleburgh (1991) argues that if an anchor slips with the former, loadings are likely to be maintained along the length of the cage group, while with the latter an imbalance of forces may be created when an anchor slips, leading to failure at joints or linkages (Fig. 3.43). However, Turner (2000a) argues that although use of multiple, orthogonal moorings may use more materials, there

98

Chapter 3

(a)

(b) Fig. 3.43 Mooring line layout. (a) A typical single-point mooring line layout; (b) typical multiplepoint mooring line layout. a = marker, b = connector, c = chain (equal to depth of water), d = polypropylene rope to buoy (4 ¥ depth of water), e = mooring buoy, f = PP rope connected to cage, g = chain lifting ﬂoats, h = PP rope (1.5 ¥ water depth), i = marker ﬂoat. (Redrawn from Thoms 1989.)

are low riser loads, small anchors are required and there is a high degree of redundancy (i.e. if one riser fails, then the other should hold). Moreover, the additional moorings provide additional support and reduce internal loadings on hinges, links, etc., for groups of cages. Most methods of mooring involve the use of rope and chain to link the cage or cage group to anchors or pegs secured to the sea bed. The mooring line is often termed a ‘riser’. Although this is the most common system, there are alternatives. Some cages – e.g. tension-leg and PolarCirkel designs – may use a sub-

Cage Design and Construction

99

merged rope, or cable-based mooring grid, to which cages may be attached temporarily using near-horizontal lines (for discussion see Lien et al. 1996; Turner 2000a). One further alternative is to drive long posts into the substrate and to attach the cages directly, either with ropes or with metal hoops or tyres that permit some vertical tidal and wave-induced movement. In theory the number and dimensions of the posts required, and the depth to which they must be buried, could be computed from estimates of forces acting on the cage system and data on soil characteristics (see Milne 1970); in practice, however, it is determined by experience. Stout hardwood posts, driven at least 2 m into the substrate are recommended for small cages. Although sometimes employed in sheltered and shallow inland and coastal sites with suitable substrates (Sodikin 1978; Chua 1979; Henderson 1980), this method of mooring is not widely used. There are a variety of methods of using single- and multiple-point moorings. One or two heavy ground chains can be laid which connect the cages to the anchors via mooring lines. Alternatively, mooring lines can be run directly from the cages to the anchors. Points of stress are formed where mooring lines are secured to the cages and so it is important that they are secured at a number of places. Joints, where stresses accumulate or are transferred from one structural member to another, are frequently used. Anchors are usually deployed to resist the principal directions of force and sometimes may be installed on shore as well as at sea. Examples are shown in Figs 3.43 and 3.44. Note that for single-point

Fig. 3.44 Examples of various cage conﬁgurations and mooring systems.

100

Chapter 3

Table 3.10 Speciﬁcations of seven-strand galvanized steel cable (sources: various). Diameter (mm) 3.2 4.8 6.4 7.9 9.5 12.7 25.4b a b

Weight unit length-1 (kg 100 m-1)

Safe loada (kg)

4.8 11.2 18.6 31.3 43.9 75.9 296.8b

227 635 1 043 1 724 2 268 3 856 14 300b

Safe load is approximately 0.25 times breaking load. Estimated.

moorings, at least three anchors are recommended (see also Goudey et al. 2001 for discussion). Mooring lines must, of course, be secured to cage collars via attachment points able to withstand the forces generated. Structural members should be used and, where abrasion is expected, the line should be protected by encasing in plastic pipe. A short length of chain may be preferable if damage from boat propellers is likely. Mooring lines must perform two functions: they must withstand and transmit forces. The loads imposed on a cage mooring system are principally dynamic. It is important that mooring lines have a high breaking strength and can absorb much of the kinetic energy of rapidly changing forces, such as cyclic loads imposed by waves and shock loads imposed by sudden gusts of wind, otherwise these forces will be transmitted directly to the anchors. Natural ﬁbre rope is not particularly suitable as it is easily abraded and prone to rotting (Klust 1983). Steel cable, although immensely strong, is expensive and heavy (Table 3.10) and has little elasticity, although it is sometimes used for securing cages to land-based ﬁxtures. Chain is extremely strong but again is heavy and is usually used in conjunction with synthetic ﬁbre rope. Synthetic ﬁbre ropes are composed of nylon (PA), PE, PES, PP or one of the new synthetic polymers such as Kevlar® or Vectran® (see Kery 1996 for review). In the production of laid ropes, ﬁbres are spun and twisted into rope yarn, several of which are then twisted into a strand. Three or four strands are twisted – the technical term is ‘laid’ – into a rope and for a cable laid rope three or more ropes are twisted together (see Klust 1983 for details). At each stage of production the direction of twist is reversed. Braided or plaited ropes differ from laid ropes in that the yarns are tied together by a process of interlacing, as opposed to twisting. There are a number of different types of braided rope described in detail by Klust (1983). For laid ropes of the same diameter, nylon and PES are considerably heavier than PP or PES (Table 3.11). However, nylon is much stronger on a per unit weight or equivalent diameter basis than ropes fabricated from the other materials. Braided ropes are lighter than laid ropes but, for equivalent weight or diameter, are generally weaker. They also cost more and have few advantages other than they are easier and more pleasant to handle and do not kink. Although

f

e

d

c

b

kp

320 750 1 350 2 080 3 000 4 100 5 300 6 700 8 300 10 000 12 000 — 15 800 20 000 30 000

kp = kilopond = 9.81 N. Polyester. Polypropylene. Polyethylene. Round braided rope. Solid braided rope.

1.1 2.4 4.2 6.5 9.4 12.8 16.6 21.0 26.0 31.5 37.5 — 51.0 66.5 104.0

4 6 8 10 12 14 16 18 20 22 24 26 28 32 40

a

mass

Nylon (PA)

Nominal diameter (mm)

1.5 3.0 5.1 8.1 11.6 15.7 20.5 26.0 32.0 38.4 46.0 — 63.0 82.0 128.0

mass

kp 295 565 1 020 1 590 2 270 3 180 4 060 5 080 6 350 7 620 9 140 — 12 200 15 700 23 900

PESb

— 1.7 2.0 4.5 6.5 9.0 11.5 14.8 18.0 22.0 26.0 — 35.5 46.0 72.0

mass

kp — 550 960 1 425 2 030 2 790 3 500 4 450 5 370 6 500 7 600 — 10 100 12 800 19 400

PPc

0.8 1.8 3.3 4.9 7.2 9.5 12.8 16.1 20.0 24.3 29.5 32.8 39.3 52.5 78.5

mass

kp 200 400 700 1 090 1 540 2 090 2 800 3 460 4 270 5 080 6 100 6 910 8 030 10 400 15 600

PEd

0.9 2.0 3.6 5.6 8.1 10.5 14.3 18.1 22.3 27.0 32.2 — — — —

Mass 280 620 1 110 1 700 2 475 3 200 4 350 5 550 6 700 8 100 9 650 — — — —

kpe

kpf 225 500 900 1 400 2 025 2 325 3 500 4 450 5 350 6 500 7 750 — — — —

Braided nylon (PA)

Table 3.11 Mass and breaking strength (kpa) of common braided and three-strand laid ﬁbre ropes (modiﬁed from Klust 1983).

1.1 2.4 4.4 6.8 9.8 13.3 17.4 22.0 27.2 32.8 39.0 — — — —

mass

260 575 1 000 1 540 2 160 2 860 3 650 4 500 5 300 5 800 6 250 — — — —

kpe

Braided PES

215 440 760 1 160 1 620 2 130 2 700 3 300 3 950 4 600 4 450 — — — —

kpf

Cage Design and Construction 101

102

Chapter 3

it can cost twice as much as PE or PP rope of equivalent strength, nylon has high extensibility and thus energy absorbing properties (Table 3.12), an important factor in designing cage moorings. In a comparison of three-strand laid ropes of equal diameter, Klust (1983) found that at 20% of breaking load nylon extended by about 20%, PE by 7–11%, PP by 6–9% and PES by around 3%. Recovery once the load is removed is extremely rapid for nylon rope. There is no marked difference between elongation characteristics of laid or braided ropes. Ropes should not be attached directly to either shore or sea anchors, but instead should be connected via a section of chain (Fig. 3.43). The chain serves to increase the effectiveness of the mooring system, partly owing to its mass and shape, which directly act as an efﬁcient type of anchor (Wiegel 1964), and partly owing to its accentuation of the catenary shape of the mooring line which improves the holding power of the existing anchor by both reducing the angle between the mooring line and anchor (see below), and by increasing the energy absorbing properties of the mooring line. Moreover, a section of chain is necessary at the anchor since it is much more resistant than synthetic ﬁbre rope to the prevailing high abrasion forces. There are several types of chain available, including wrought iron and various types of carbon and alloy steels. Wrought iron is very variable in quality; the best has excellent corrosion resistance while the poorer grades are inferior in all respects. Some stainless steel chain is suitable for marine use, but is prohibitively expensive. Mild steel chain, with low carbon and manganese contents, has been widely recommended for cage anchorages. Speciﬁcations of one type are given in Table 3.13. The length of chain is usually no more than one-third of the total length of the mooring line, otherwise vertical loadings can be excessive. A fairly heavy grade of chain is recommended. The total length of the mooring line should be at least three times the maximum depth of water at the site (see later) and where the rope joins the chain a galvanized, heavy-duty thimble should be spliced into the rope and a galvanized shackle of the appropriate size – preferably a bow shackle one size larger than the diameter of the rope – used to connect the chain and to the rope. Thoms (1989) recommends welding or brazing shackles closed, as they are prone to becoming undone, or splicing the rope and chain together. An alternative mooring line, composed largely of chain, is occasionally employed. Typically, 18–25-mm chain, two to three times the maximum depth of water in length, is connected from the anchor to a ﬂoat positioned 10 m or so from the cage, and a section of rope – PES or nylon – used to link the ﬂoat to the cages. The section of chain between the ﬂoat and the anchor adopts a pronounced catenary shape, the vertical and horizontal holding power depending on the mass of the chain and the angle between the chain and the anchor. The buoy minimizes the vertical loading on the cages and must be sufﬁciently large to support the mass of the chain in the water and to resist the vertical forces imposed by the cages on the mooring system. Various buoy types can be used, the best probably being expanded foam ﬁlled. A single ﬂoat per mooring line tends to be used, although Lien et al. (1996) clearly show reductions in line tension from using a series of ﬂoats with the same ﬂotation capacity as a single ﬂoat. Under shock loads, the chain/buoy acts as a spring absorbing much of the energy that would otherwise be transmitted to the anchor. In an analysis of the

small, varying medium

Group d Shrinkage in water Resistance to sunlight

b

Continuous ﬁlament. Monoﬁlament.

very very very very

Group c Resistance to sustained load Resistance to repeated load Resistance to shock loading Flexing endurance

a

23.3 high 4.9 very high

Group b Elongation (%), dry, at 30% breaking strength Elasticity Elongation ratio to manila at 30% breaking strength Toughness high high high high

30.6 80–90 2.83

Group a Breaking length, dry, km Wet strength (percentage dry strength) Strength ratio to manila

PA cont. ﬁla

none high

very high very high high very high

5 high 1.1 high

19.2 100 1.78

PES cont. ﬁla

none medium

high high high medium

12.7 medium 2.7 high

28.8 100 2.67

PP monoﬁl.b

none medium

high high high low

8.5 medium 1.8 high

28.8 100 2.67

PP split ﬁbre

none medium

medium low medium low

9.9–15.1 low, creep 2.1–3.2 medium

21.2 100–115 1.96

PE monoﬁl.b

great medium

low low low low

4.7 low 1 low

10.8 105–120 1

Manila

Table 3.12 Comparative properties of three-strand ﬁbre ropes. The resistance to sunlight refers to synthetic ﬁbres that are stabilized but not coloured (from Klust 1983).

Cage Design and Construction 103

104

Chapter 3

Table 3.13 Speciﬁcations of open link proof steel chain (sources: various). da (mm) 4.8 6.4 7.9 9.5 12.7 15.9 19.1 22.2 25.4 28.6 30.2 a b

Weight per unit length (kg m-1)

Safe loadb (kg)

0.6 1.1 1.6 2.4 4.1 6.1 8.6 11.6 14.9 18.9 23.2

263 408 617 844 1497 2268 3209 4355 5635 7076 8709

Diameter of link. Safe load is approximately 0.25 ¥ working load.

effects of varying the distance between the anchor and buoy and the buoy and cage, Lien et al. (1996) showed that the restoring force increased when the former is shortened and the latter is lengthened. However, the importance of maintaining a strong catenary shape to the line to ensure good anchoring must also be borne in mind. Land anchors are sometimes used where cages are sited inshore, close to land only marginally above sea level. They are convenient to install and maintain. Galvanized steel ring bolts, with 20–30 cm long shanks, are either driven into rock or a hole is drilled and the bolt grouted in with cement. Wooden pegs, 15 cm ¥ 15 cm in cross-section by 200 cm long, are also sometimes used for marine anchors (Chua 1979). No chain is used and instead the rope is passed through a hole drilled in the peg and spliced to form a loop. The peg is then driven into the substrate using a guide pole. Where the substrate is rocky, an iron pin with an eye at one end may be used and the chain at the end of the mooring line secured to the eye by a shackle. The simplest and cheapest type of marine anchor is the deadweight or block anchor, which typically consists of a bag of sand or stones or a block of concrete or scrap metal. Consider a block anchor, sitting on the sea bed and connected to a ﬁsh cage by a length of mooring line. The block will begin to move when the horizontal component of the force exerted by the cage and mooring line equals the frictional force between the block and substrate. The holding coefﬁcient, k, of the anchor is deﬁned as R, the horizontal force divided by the mass of the anchor. It can be shown that k depends upon the angle between the anchor and the cage and thus the ratio between water depth and line length – and the nature of the substrate. In Table 3.14 data on the k values of sandbag anchors are summarized. Block anchors are inefﬁcient, i.e. have low holding power per unit installed weight. For example, a sandbag anchor weighing 100 kg in water and installed on a sandy substrate would have a holding power of between 19 and 27 kg, depending upon the angle between the mooring line and the sandbag, compared

Cage Design and Construction

105

Table 3.14 The ﬁxing coefﬁcient, k, of sandbag anchors on different substrates and varying mooring cable length : water depth ratios (l : d) (sources: various). Nature of substrate

Sand (vertically pulled) Sand (horizontally pulled) Mean value Sandy mud (vertically pulled) Sandy mud (horizontally pulled) Mean value Mud (vertically pulled) Mud (horizontally pulled) Mean value

l:d 1

2

3

4

5

0.16 0.21 0.19 — — 0.10 0.05 0.05 0.05

0.47 0.59 0.53 0.33 0.31 0.32 0.13 0.33 0.23

0.60 0.65 0.63 0.41 0.31 0.36 0.20 0.34 0.27

0.71 0.68 0.70 0.49 0.43 0.46 0.23 0.48 0.35

0.73 0.74 0.74 0.61 0.62 0.62 0.30 0.52 0.41

with a modern lightweight anchor which will have a holding power 50–70 times that of its mass. The performance of the sandbag anchor is much poorer in mud. Note that there is a curvilinear relationship between the cable length : water depth ratio and the horizontal holding power which begins to level off at a k-value of 3–4. Similar relationships exists for other block anchors. In practice, block anchors are more efﬁcient than this since they tend to bed down into the substrate. Consider, for example, a concrete block 1 ¥ 1 ¥ 0.4 m, with a mass in water of 970 kg, sitting on a sandy substrate and buried to a depth of 10 cm. The cross-sectional area of the buried block opposed to the horizontal forces is 0.1 m2. Using data from Milne (1970) for the cohesive resistance of ﬁrm sandy soils to forces gives a resistance value of 548 kg, increasing the holding power by a factor of almost two. Concrete block anchors may be simply fabricated using wooden shuttering, tyres, or any other convenient object as a mould. Steel rods for strengthening and an eye-bolt for a mooring attachment are usually incorporated. Once fabricated, the blocks can be transported to the water’s edge at low tide and ﬂoats attached, so that they can be ﬂoated to the required location at high tide. Once installed, they are difﬁcult to recover. There are numerous types of embedding anchor (Fig. 3.45; see also Kery 1996). The holding power of an embedding anchor is related to its frictional resistance in soil, and so is dependent upon ﬂuke area, soil penetration and the mechanical properties of the soil rather than simply the mass of the anchor. Anchor penetration is a function of ﬂuke shape and the angle between the ﬂuke and the shank, while the frictional resistance of the soil is dependent upon soil cohesiveness and shear strength. As with block anchors, the angle between the anchor and the mooring cable is critical in determining holding power. In Table 3.15 the holding coefﬁcients for various mooring cable length:water depth ratios are given for a ‘kedge-type’ anchor, illustrating differences between soil types and the effect of mooring line angle. Embedding anchors are very efﬁcient, i.e. they have a high holding power to mass ratio. Under optimum conditions (sandy mud soil, low angle between mooring line and anchor, etc.) they are 10–500 times as efﬁcient as block

106

Chapter 3

Kedge

Navy Stockless

Danforth

Lightweight

Grapnel

Mushroom

Fig. 3.45 Various types of anchor.

Table 3.15 The ﬁxing coefﬁcient, k, of kedge-type anchors on different substrates and varying cable length : water depth ratios (l : d) (sources: various). Nature of substrate

Sand Sandy mud Mud Mean

l:d 1

1.5

2

3

4

5

0.26 0.23 0.11 0.20

1.10 1.90 0.60 1.20

1.90 3.27 1.99 2.39

4.37 4.40 3.29 4.02

— 5.50 5.11 5.31

5.83 5.15 6.46 5.81

anchors. They are, however, more expensive than block anchors in terms of cost per unit holding power and have to be bedded in properly. When dropped, embedding anchors do not immediately grip the substrate but instead have to be dragged several metres before they become secure. Moreover, if they drag, they can travel a considerable distance before bedding in again. ‘Piggybacking’ (i.e. the use of two anchors connected together) gives greater holding power than the sum of two independently moored anchors.

Cage Design and Construction

107

There are numerous other types of anchor, some combining the properties of block and embedding types (e.g. shaped concrete blocks) while others are designed for particular types of substrate (e.g. soft mud anchors). Anchor speciﬁcations can be obtained from manufacturers or from chandlers, many of whom supply cage ﬁsh farmers. Prior to choosing or installing anchors it is advisable to survey the lake or sea bed. Anchors should be positioned ﬁrst. The position of the anchors can be accurately established using a global positioning system (GPS) or by taking bearings with respect to local, easily visible land marks. Small buoys attached via light riser cables to the anchors may also be used (Fig. 3.43). Turner (2000a) gives a plan of operation for installing mooring systems. In summary, this involves: • • • • • • • • •

establishing a work base; sourcing a work boat and appropriate lifting gear; securing access; marking out the site; assembling the moorings; laying the lines; proving the anchors; tensioning the risers; conducting a diver survey of the installed lines.

Mooring systems must be checked at regular intervals and fouling removed from buoys and mooring lines. It is more economical to lift moorings that are more than 50 m deep than to inspect by diving. Turner (2000a) states that it is essential that any mooring inspection assesses component strength to see if it deviates signiﬁcantly from design strength and that it should also assess likely deterioration in the interval to the next inspection.

APPENDIX 3.1 Current force on a single panel of a net cage (from Løland 1993a) The drag force FD on a cage net panel is deﬁned as the force in the direction of the ﬂow, while the lift force FL is deﬁned as the force normal to the ﬂow. Mean drag and lift forces on a cage net panel can be written as dFD =

1 [rCD (a)U 2dA] 2

dFL =

1 [rCL (a)U 2dA] 2

where CD and CL = the drag and lift coefﬁcients as functions of the angle between the normal to the net and ﬂow direction; dA = the area of the portion of the net panel; U = current velocity; r = density of water; a = angle between the ﬂow direction and the normal vector to the net in the direction of ﬂow.

108

Chapter 3

The drag coefﬁcient is a function of the solidity ratio Sn, (i.e. the ratio between the area covered by the meshes and the total area of the net panel), the Reynolds number Rn and the angle of the approach, where Sn = 2.0 Rn = U

t Dt

Dt V

where t = mesh bar diameter and Dt = mesh size (see Fig. 3.7 for terminology); and where v = kinematic viscosity. Although it is not possible to ﬁnd a general expression for CD valid for all combinations of Sn and Rn – this must be derived from tests with models – an estimate can be made by summation of the drag on each mesh bar. The total drag force FD is an integral of the above equation over the entire panel, such that FD =

1 Ú 2 [rC A

D

(a)U 2dA]

where dA = area of a portion of the net panel. The above equation gives the drag force when the net panel has a rigid, plane form, as a would be constant over the entire net panel. However, a cage net panel typically has little bending stiffness. A net panel suspended from a ﬁxed point would tend to rotate as a curved plane about the suspension point. Hence the angle a will vary along the depth of the net panel. The drag force and the shape of the net panel must be found by numerical analysis, and assumes that the panel has no bending stiffness and that it can be considered as a number of sub-elements. Each sub-element then has an equilibrium position at the angle that gives zero moment. Each sub-element is exposed to a drag force, a lift force, the weight and buoyancy of the net itself and the reaction forces from neighbouring elements. The equilibrium position is determined by an iteration process beginning at the lower end of the net panel where the reactive forces are known and are equal to the drag and the weight of the sinkers. When the equilibrium position of the ﬁrst element has been determined the reaction forces on the next sub-element are also known. This procedure is repeated until the suspension point is reached.

Current force on a system of net cages (from Løland 1993a) Consider, as an example, a group of six cages in which the current direction is normal to the downstream incident panel (see Fig. 3.15 for four cages). The system may be considered as consisting of 12 net panels normal to the current direction and 18 net panels oriented parallel to the ﬂow. Of the 12 panels lying normal to the ﬂow, two are unshielded, two are shielded by one panel, two are shielded by two panels, etc. The two most downstream panels are shielded by ﬁve panels.

Cage Design and Construction

109

Current velocity is reduced as a geometric series as it ﬂows through subsequent panels ui = Ur (i -1) where ui = the reduction in velocity as the upstream current ﬂows through a cage net panel and r = the velocity reduction factor (u/U). In extensive model tests Løland and colleagues showed that the incident current velocities on net panels normal to the ﬂow direction is the same as that inside the cage bag, i.e. they are exposed to the same current velocity as the downstream panel of the cage bag. Where the cage bag is constructed from rigid mesh and no deﬂection occurs the total drag can be expressed as 4N c

FD = N N

Ê 1 - r0 ˆ È 1 r2 ˘ rU 2 Á C (a = 0)A N + CD (a = 90)A p 2 ˜Í D 2 Ë 1- r ¯Î 1 + r 2 ˙˚

where AN and AP are the area of the net panel normal to and parallel to the ﬂow direction, respectively. The drag force on a ﬂexible net cage bag system is calculated in a similar manner although, because it is necessary to iterate on the deformation of each net panel, the results cannot be obtained in closed form: FDN = r CD (a 1 )U 2 A1N + r CD (a 1 )U 2 r 2 A2N + . . . + r CD [r CD (a 6 )]U 2 where FDN = the drag force on the net panels normal to the free ﬂow. The calculated drag force lies in the range 0.9–1.2 of the measured drag force, sufﬁcient for most purposes (Løland 1993a).

APPENDIX 3.2 Example of cage ﬂotation computation Cage bag dimensions = 5 ¥ 5 ¥ 5 m = 125 m3 Material = nylon, raschel woven 24-mm stretch mesh Density = 0.24 kg m-2 Total weight of bag in air when clean and dry = 30 kg Assume a 15-fold increase in weight in air owing to fouling Thus, design weight of bag in air = 450 kg Quantity of rope used for rigging = 90 m Density = 10 kg 100 m-1 Total weight of rope in air when dry = 9 kg Assume a two-fold increase in weight in air owing to fouling Thus, design weight of rope = 18 kg Maximum stocking density at time of harvest = 25 kg m-3 = 3125 kg Assume 10% of ﬁsh may be supported by net during harvesting = 313 kg Weight of superstructure and staff, equipment, etc. = 500 kg

110

Chapter 3

Total weight of cage in air = 1272 kg Assuming a factor of safety at 1.25, the vertical design load in air = 1591 kg

APPENDIX 3.3 Calculation of the buoyancy of a 3 ¥ 3 ¥ 3 m bamboo cage (see section 3.3.2) The buoyancy of a bamboo collar can be computed from the volume and density of the bamboo used. Volume can be derived from

[

2

V = pnL (0.5D) - (0.5 - d)

2

]

where V = volume (m3); n = number of bamboos used; L = mean length of bamboos (m); D = mean outer diameter (m); and d = mean wall thickness (m). D and d are means of several measurements. For the 3 ¥ 3 ¥ 3 m cage, 20 sections 4.5 m long are used for the collar base. Assuming D = 0.12 m and d = 0.008 m

[

2

V = 3.14 ¥ 20 ¥ 4.5 (0.6) - (0.52) = 0.253 m

2

]

2

If the density of the bamboo used is 550 kg m-3 then the buoyancy = (1000 550) ¥ 0.253 = 114 kg, which would be sufﬁcient to support the cage bag and a member of staff, although without much margin of safety.

Cage Aquaculture, Third Edition Malcolm C. M. Beveridge Copyright © 1996, 2004 by Blackwell Publishing Ltd

Chapter 4

Site Selection This chapter examines the problems of selecting a suitable cage farm site in marine and freshwater environments. It does not, however, consider the broader aspects of how cage aquaculture should be incorporated into coastal and catchment management plans (see Barg 1992; Pullin et al. 1993; Brzeski & Newkirk 1997; GESAMP 2001). The carrying capacity of sites is considered in Chapter 5. Choice of site in any ﬁsh farming operation is of paramount importance since it greatly inﬂuences economic viability by determining capital outlay and affecting running costs, production and mortality. Cage- and pen-based aquaculture systems suffer in comparison to land-based operations in that there is less room for error in site selection. Poor pond sites, for example, may be improved by drilling boreholes to increase water supply or by introducing ﬁlters and sediment traps to remove suspended material. However, there is little that can be done at a cage or pen farm if the site proves too exposed and water exchange is poor or if water quality deteriorates. One further disadvantage is that in exploiting what is generally a common property resource, for example a lake or area of coastline, the cage farmer has little control over subsequent developments, including establishment of other farms, that may adversely affect the quality of the site. There are three categories of cage site selection criteria that must be addressed (Table 4.1). The ﬁrst is primarily concerned with the physicochemical conditions that dictate whether a species can thrive in an environment (temperature, salinity, oxygen, currents, pollution, algal blooms, exchange). From a welfare perspective, water quality, ﬂow rates and temperature should be appropriate for the species concerned (see FSBI 2002). The second comprises those factors that must be considered in order to site a cage system successfully (weather, shelter, depth, substrate) while the third is concerned with the establishment of a farm and proﬁtability (legal aspects, access, land-based facilities, security, economic and social considerations). Although much of the data required must be collected by survey methods and analysis of water samples, invaluable information can also be gained by talking to local people about prevailing weather conditions and the occurrence of algal blooms or pollution. Consultations with local people prior to establishment of a cage farm may also help minimize poaching and vandalism.

4.1 ENVIRONMENTAL CRITERIA FOR FARMED AQUATIC SPECIES 4.1.1

Water quality

For a comprehensive treatment of water quality criteria relevant to cage aquaculture, reference should be made to Kinne (1976), Alabaster & Lloyd (1980), 111

112

Chapter 4

Table 4.1 Criteria for cage site selection. Note that several factors are common to more than one category (see text). Category 1

Category 2

Category 3

Temperature Salinity Pollution Suspended solids Algal blooms Disease organisms Water exchange Currents Fouling

Depth Shelter Substrate Currents Fouling

Legal aspects Access Security Proximity to markets

Lewis & Morris (1986), Lloyd (1992), Svobodová et al. (1993), Howells (1994), Poxton (1996) and Black (1998). Cage sites must have good water quality. Sites should not only be uncontaminated by toxic industrial pollutants, but also should meet the pH, temperature, oxygen and salinity requirements of the species to be farmed. Advice on water quality assessment and methods for speciﬁc tests are detailed in a number of texts such as Stirling (1985). The use of remotesensing techniques, such as satellites, for evaluating and monitoring water quality of cage aquaculture sites has been discussed by Chacon Torres et al. (1988b), Meaden & Kapetsky (1991) and Willumsen et al. (1993) (Fig. 4.1; http://www.oceanor.no/products/systems/seawatch.). Temperature and salinity Fish and other farmed aquatic organisms have no means of controlling body temperature, which changes with that of the environment. A rise in temperature increases metabolic rate and causes a concomitant increase in oxygen consumption and activity as well as production of ammonia and carbon dioxide. Salinity is a measure of the amount of dissolved solids present in water (Table 4.2) and is usually expressed in parts per thousand (‰). Its relevance to aquaculture lies principally in its control of osmotic pressure, which greatly affects the ionic balance of aquatic animals. When selecting a site for cage culture the optimum temperature and salinity conditions of the species should be met since even immediately outside these optima, feeding, food conversion and growth are adversely affected. Sub-optimal conditions also contribute towards stress, leading to increased susceptibility to parasitic infections and reduced resistance to disease (Alabaster & Lloyd 1980; Pickering 1981, 1993, 1998; Anderson 1990; Schreck 1990). Rapidly ﬂuctuating temperatures and salinities are often more harmful than seasonal changes, although some species are more tolerant than others. Optimal conditions for some of the more important cultured species can be found in Poxton (1996), Black (1998) and Ross (2000).

Site Selection

113

Fig. 4.1 The SEAWATCH marine surveillance and information system (from Willumsen et al. 1993).

Table 4.2 The major elements of sea water (35‰) (modiﬁed from Kalle 1971; Poxton 1996). Elements

g kg-1

Cations Sodium Potassium Magnesium Calcium Strontium

10.752 0.375 1.295 0.426 0.008

467.56 10.10 106.50 20.76 0.18

Anions Chlorine Bromine Fluorine Sulphate Bicarbonateb Boric acid

19.345 0.066 0.001 2.701 0.145 0.027

545.59 0.83 0.07 56.23 — —

a b

Milli-equivalents kg-1 = (g kg-1/atomic weight of ion) ¥ ionic charge ¥ 1. Bicarbonate and carbonate vary according to the pH of sea water.

Milli-equivalents kg-1a

114

Chapter 4

It is desirable to have as much information as possible about the temperature and salinity conditions at a prospective site, especially where conditions for aquaculture are marginal. In the Bay of Fundy on the eastern seaboard of Canada ‘winterkills’ of Atlantic salmon occur when temperatures fall to -0.7 or -0.8°C. A combination of local hydrographic and meteorological conditions results in particular inshore areas being especially prone to lethal temperatures (Page & Robinson 1992; Saunders 1995) (see also section 7.9.2). There is a close correspondence between temperature regime and solar radiation received by a water body, the most important determinant of solar radiation being latitude (Le Cren & Lowe-McConnell 1980). The longer wavelengths of the electromagnetic spectrum (>700 nm) are the most important in the transfer of heat and, because of the associated high extinction coefﬁcients, they are completely absorbed in the top few metres of the water column. If it were not for mixing processes, there would be an exponential fall in temperature with depth. The mixing of surface water with deeper water requires energy and the degree of mixing depends both on the energy inputs to the water body and on the density differences between the surface and underlying water, the greater the difference the greater the energy required. In fresh waters, density differences are principally caused by temperature. Above 4°C the density of water decreases with increasing temperature. In lentic (i.e. still water) freshwater systems, a pattern of thermal stratiﬁcation is often observed in which a layer of warm, less dense surface water (epilimnion) sits on top of a colder, denser deep layer (hypolimnion). Temperature differences within the layers are small. Separating the layers is a zone where temperature changes rapidly with depth, the thermocline (Fig. 4.2). In temperate regions, surface waters begin to warm in the spring causing the onset of stratiﬁcation that may persist until the end of summer. As autumn advances, the surface waters cool and the prevailing winds increase in strength

Temperature °C 0

4

8

12

16

0

Depth m

4

Thermocline

8

12

16 Winter

Summer

Fig. 4.2 Stratiﬁcation patterns in a freshwater body.

Site Selection

115

Table 4.3 Lentic water body types. Type

Description

Distribution

Holomictic

Mix from top to bottom

Universal

Dimictic

Two circulation periods

Continental mid-latitudes and high altitude sub-tropics

Warm monomictic

One circulation period (autumn)

Maritime mid-latitudes

Polmictic

Frequent stratiﬁcation and circulation

Shallow tropical and mid-latitude lakes

Meromictic

Only partially mixed

Extremely deep lakes, universal

and frequency until there is sufﬁcient energy to overturn the whole water body, bringing cooler water up from below. During winter, some lakes may freeze on the surface, resulting in a second period of stratiﬁcation that lasts until spring when the surface waters warm sufﬁciently to reduce density differences and facilitate wind-induced mixing. Stratiﬁed lakes and reservoirs are generally deep and occur throughout temperate regions and in some parts of the tropics. Water bodies that exhibit two periods of circulation per year (spring and autumn) are known as dimictic and tend to be continental in distribution (i.e. situated away from the warming inﬂuence of coastal air currents). Monomictic lakes, which turn over once in September or October, are common in Western Europe. While monomictic and dimictic lakes turn over completely, very deep lakes require such an immense amount of energy to mix completely that they experience only a partial turn over. These are known as meromictic lakes. Shallow and exposed lakes may not stratify at all or stratify for short periods of time (days). Polymictic lakes as they are called, occur throughout tropical and temperate regions. Lake and reservoir types and their distributions are summarized in Table 4.3. Stratiﬁcation in lotic (ﬂowing water) systems is less pronounced than in lentic systems and is restricted to slow-ﬂowing lower reaches. In marine areas, temperature variations are complicated by salinity. The salinity of sea water varies between about 32 and 40‰, and in open waters is determined by evaporation and precipitation. In offshore areas, where the stability of the water column is dependent upon tidal turbulence and water column depth (Pingree et al. 1978), stratiﬁcation typically occurs in deep waters with low tidal velocities. The temperature regime of coastal areas may be greatly inﬂuenced by runoff from the land. In temperate regions, river discharges and surface freshwater runoff are colder in winter and warmer in summer than the receiving coastal seas, whereas in tropical parts of the world freshwater runoff is likely to be warmer than the sea during the dry season and cooler during the wet season. Because water density is determined by both salinity and temperature (Fig. 4.3), the mixing of terrestrial runoff with sea water requires energy. The degree to

116

Chapter 4

Fig. 4.3 Speciﬁc gravity of water at different temperatures and salinities. Lines of equal density are shown (redrawn from Kalle 1971).

which the two factions mix depends principally upon the volume of freshwater runoff and available mixing energy (tidal, wind). An estuary may be deﬁned as ‘a semi-enclosed coastal body of water which has a free connection with the open sea and within which the sea water is measurably diluted with fresh water derived from land drainage’ (Pritchard 1967). Although marked changes in temperature and salinity with depth might be expected in such an environment, there are a number of stratiﬁcation patterns resulting from differences in estuarine geomorphology. At one extreme, is the strongly stratiﬁed estuary where surface salinity differs from bottom salinity by several parts per thousand and where there is a strong salinity gradient (halocline) at mid-depth. There is a net ﬂux of surface water seawards and a net ﬂux

Site Selection

117

of deep water landwards, with some mixing of the two layers along the entire length of the estuary. However, where the river ﬂow is large, undiluted fresh water may ﬂow out to open sea above the underlying layer of sea water and there is no mixing between layers. This type of strongly stratiﬁed estuary is called a salt-wedge estuary. Another type of strongly stratiﬁed estuary is the fjord (see page 131). Here three layers may be apparent: a surface brackish water layer, an underlying layer of increasing salinity with depth, and a deep, stagnant layer where salinity is constant to the bottom. At the other extreme are well-mixed estuaries where there is no change in salinity with depth. In such estuaries, the transport of salt water landwards and fresh water seawards is through horizontal eddy diffusion (Ketchum 1983) and the net ﬂux at all depths is seawards. Intermediate in nature to the well-mixed and strongly stratiﬁed estuaries are weakly stratiﬁed estuaries, in which there is an increase in salinity of a few parts per thousand with depth. At the landward end of such estuaries is a zone of river water, undiluted with sea water, while at the mouth of the estuary we would ﬁnd sea water undiluted by fresh water. The above description is a considerable over-simpliﬁcation, since within a single estuary changes in stratiﬁcation with space and time occur (see Ketchum 1983; Dyer 1997). Some attempt has been made to model the stratiﬁcation process in inshore coastal areas. Since density can be determined from temperature and salinity data, it is possible to calculate the energy required to bring about complete mixing of the water column: 0

q = 1 h Ú (r ¢ - r)gzdz -h

0

r ¢ = 1 h Ú rdz -h

(Simpson 1981)

where q = potential energy anomaly (J m-3); h = water column depth (m), r¢ = density of water at surface, and r = density at depth z. Low values of q (<10) correspond to a low stability or mixed water column, high values of q (>30) are indicative of a stable or stratiﬁed water column, and values between 10 and 30 indicate a transitional or weakly stratiﬁed water column. Jones et al. (1984) applied the model to data from the west coast of Scotland and found a general agreement between water column depth and stability. Exceptions occur during unusually stable weather conditions when there is an increase in heat inputs to surface waters. The aquaculture potential offered by thermal discharges from power stations is widely recognized. Their suitability for cage ﬁsh farming very much depends upon the cooling water source and the design and operation of the power station. Water of varying salinities and quality is used to generate steam and drive turbines, which in turn drive the generators, waste steam being fed to a condenser and discharged as thermal efﬂuent. Some older stations, where water supplies are plentiful, use single-pass cooling systems while others recycle water. Heated cooling water is either discharged directly into the receiving water body or cooled

118

Chapter 4

prior to discharge. Cooling methods involve ponds or towers of various design, heated efﬂuent eventually being discharged via pipes or canals into the water body from which it was abstracted (for review see Langford 1983). Cages may be sited in cooling ponds or canals or in the receiving water body. Water temperature depends upon location, water supply (estuarine river/lake), the design of the system, the site (pond/canal/receiving water body) and the proximity of the cages to the power station. Thermal efﬂuent is usually between 3 and 15°C warmer than the intake (i.e. thermal power stations have a DT of 3–15°C). When operating at full output DT is constant, although DT can ﬂuctuate widely because of changes in operating schedule. Changes in DT of 10–15°C over a few days are not unknown (Langford 1983). It is, therefore, important to determine the DT pattern at a prospective thermally heated site. Different species may have to be cultured at different times of the year. The problems associated with using thermal power sources are illustrated by Schneider et al. (1990). Large and rapid changes in water quality conditions can occur that greatly inﬂuence growth and production. Oxygen Oxygen is required by all higher organisms for the production of energy; energy that is required to fuel essential functions such as digestion and assimilation of food, and maintenance of osmotic balance and activity. Oxygen requirements vary with species, stage of development and size, and are also inﬂuenced by environmental factors such as temperature. If the supply of oxygen deviates from the ideal, feeding, food conversion, growth and health can be adversely affected. It is, therefore, important that good oxygen conditions prevail at a site. Although the oxygen content of water is in a state of dynamic equilibrium with atmospheric oxygen, the gas passing in both directions across the air/water interface by diffusion, it is never more than 5% that of a similar volume of air because of the low partial pressure and poor solubility of the gas. Solubility declines with increasing temperature and salinity; thus a sample of sea water contains less dissolved oxygen (DO) than an equivalent volume of fresh water of the same temperature while a sample of cold water contains more DO than the equivalent volume of warm water, provided salinities are the same. An increase in altitude results in a decrease in partial pressure and a consequent reduction in the quantity of oxygen that the water can hold. The relationships are summarized in Table 4.4. Pressure corrections can be made by multiplying by a factor of P/760, where P is the observed barometric pressure in mmHg, or by P/1010 if the barometric pressure is recorded in mbar (Stirling 1985). At equilibrium, water is described as being 100% saturated. However, DO levels may temporarily deviate from saturation because of changes in oxygen supply and demand. A major inﬂuence on the pool of DO is the planktonic algae community. Phytoplankton require only light and a supply of nutrients in order to convert carbon into plant tissue by photosynthesis, releasing oxygen as a by-product: light

6CO 2 + 6H 2O æ ææÆ C6 H 12O6 + 6O 2

Site Selection

119

Table 4.4 Solubility of oxygen in water (mg l-1) at different temperatures and salinities when exposed to water-saturated air at a total pressure of 760 mmHg (∫ 1.01 bar) (from Stirling 1985). Temperature °C

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40

Salinity ‰ 0

5

10

15

20

25

30

35

14.6 13.8 13.1 12.5 11.8 11.3 10.8 10.3 9.9 9.5 9.1 8.7 8.4 8.1 7.8 7.6 7.3 7.1 6.9 6.7 6.5

14.1 13.3 12.7 12.1 11.5 10.9 10.5 10.0 9.6 9.2 8.8 8.6 8.3 8.0 7.7 7.4 7.2 7.0 6.8 6.6 6.5

13.6 12.9 12.2 11.6 11.1 10.6 10.1 9.7 9.3 8.9 8.6 8.3 8.1 7.7 7.5 7.2 7.0 6.9 6.7 6.5 6.3

13.2 12.5 11.8 11.3 10.7 10.2 9.8 9.4 9.0 8.6 8.3 8.1 7.8 7.5 7.3 7.0 6.9 6.7 6.5 6.4 6.2

12.7 12.1 11.5 10.9 10.4 9.9 9.5 9.1 8.7 8.4 8.1 7.9 7.6 7.3 7.0 6.8 6.6 6.4 6.2 6.1 6.0

12.3 11.6 11.1 10.5 10.1 9.6 9.2 8.8 8.5 8.1 7.8 7.7 7.4 7.1 6.8 6.6 6.3 6.2 6.1 5.9 5.7

11.9 11.3 10.7 10.2 9.7 9.3 8.9 8.6 8.2 7.9 7.6 7.5 7.1 6.8 6.6 6.4 6.1 6.0 5.9 5.7 5.6

11.5 10.9 10.3 9.8 9.4 9.0 8.6 8.2 7.9 7.6 7.3 7.2 6.9 6.6 6.4 6.1 5.9 5.8 5.7 5.6 5.5

During the day there is a net production of oxygen but at night, when photosynthesis stops, the algal community becomes a net oxygen consumer. Where there are large algal communities, supersaturation of DO may occur during the day and subsaturation conditions prevail at night, with late afternoon maxima and pre-dawn minima owing to the inefﬁciencies of air/water diffusion processes. In highly productive ﬁsh ponds, diurnal DO variations can be as great as 7– 8 mg l-1 (Boyd 1979), severely stressing ﬁsh. However, in water-based culture systems, where algal populations tend to be much smaller, diurnal variations in DO generally do not exceed 2–3 mg l-1, although these may still be stressful. The environmental conditions conducive to blooms usually occur during the warmer months in areas subject to high nutrient inﬂuxes. External sources, such as sewage discharges and agricultural runoff, may be important contributors. However, a sudden upwelling of nutrient-rich water from the deeper parts of the water body during the breakdown of stratiﬁcation may also stimulate blooms. Certain types of water body, such as enclosed areas with poor ﬂushing rates (see section 4.1.2), tend to be more subject to blooms. Problems can occur when algal blooms die as a result of sudden changes in climatic conditions, affecting light and temperature or exhaustion of an essential nutrient. During decomposition, microbial respiration may remove much, or even all, of the DO resulting in ﬁsh kills (Grave 1981; Santiago 1994; Bagarinao & Lantin-Olaguer 1998; Yambot 2000; Costa-Pierce 2002).

120

Chapter 4

Another community that makes demands on DO is the benthos, and during stratiﬁcation benthic oxygen demand can cause deoxygenation of the hypolimnion. At the end of stratiﬁcation, the upwelling of deoxygenated, hypolimnetic water can result in ﬁsh kills as has happened at cage sites in reservoirs in Israel (Zoran et al. 1994). The high particulate waste loadings associated with some intensive cage ﬁsh farm sites may greatly increase benthic oxygen demand, which in turn may reduce the DO content in and around cages (see sections 5.3 and 7.8). Supersaturation of dissolved gases (O2, N) in thermal power station efﬂuents can arise from pressurization of water through pumping or from ﬂowing over spillways, raising saturated water from depth to the surface or from heating saturated water (Sylvester 1975). High dissolved gas levels have caused extensive mortalities of ﬁsh held in cages at power stations through gas bubble disease (Marcello & Strawn 1973; Ciesluk 1974). Chamberlain & Strawn (1977) found that submerging cages to reduce exposure to gas supersaturation reduced mortalities, saturation decreasing by about 10% per metre depth because of increased hydrostatic pressure. In summary, sites that are strongly stratiﬁed for much of the year, and/or where algal blooms carry risks of periodically poor oxygen condition, should be avoided if possible. Marine sites which have good bottom currents and which disperse sedimenting wastes are desirable. pH pH is a measure of the hydrogen ion concentration [H+] of a solution and is determined by the equation: pH = -log[ H + ] pH is important to aquaculturists because extreme values can damage gill surfaces, leading to death (McDonald 1983), and because it affects the toxicity of several common pollutants (ammonia, cyanide) and heavy metals (aluminium). pH is expressed on a scale of 0 (acidic) to 14 (alkaline). A pH value of 7 is considered to be neutral. Because it is a logarithmic scale, a change in pH of one unit results in a ten-fold increase in hydrogen ion concentration. The pH of sea water usually lies in the range 7.5–8.5, there being more anions than cations. Sea water is also well buffered, i.e. comparatively resistant to changes in pH through the addition of alkaline or acidic compounds and, unlike fresh waters, the pH of sea water does not show marked seasonal or diurnal changes due to photosynthesis. Because of the great variability in ionic composition, fresh waters may have pH values of between pH 3 and 11. Acidic waters are often associated with highland areas or areas subject to acid rain precipitation. Alkaline waters typically occur where the geology of the watershed is dominated by alkaline sedimentary rocks, e.g. chalk or limestone. Lakes with exceptionally high pH values are often found in desert areas where outﬂows are less than inﬂows and where salts accumulate. Exceptionally high pH conditions (pH 8.5–9.2) have also been recorded

Site Selection

121

in volcanic Taal Lake, Philippines, adversely affecting caged ﬁsh growth and production (Yambot 2000). The ideal pH for most ﬁsh is in the range 6.5–8.5. While pH is, therefore, not a problem at most marine sites, care must be taken in fresh water as there can be marked seasonal and diurnal changes, particularly in low-buffered waters. Intensive cage operations can stimulate phytoplankton production, leading to elevated pH values, particularly in summer, due to increased photosynthetic carbon ﬁxation. Ammonia toxicity may be a problem at such times (see also Black 1998). In areas prone to acid precipitation, a sudden fall in pH often occurs during January–March because of run-off of acidic snow melt waters. Turbidity Turbid conditions arise from organic or inorganic solids suspended in the water column as a result of soil erosion, mining wastes, sewage efﬂuents, pulp and paper mill wastes, and other industrial efﬂuents (Alabaster & Lloyd 1980). Cage ﬁsh farms are themselves a source of suspended solids. Some suspended solids may have toxic properties (e.g. the salts of various metals) while organic solids can cause depletion of DO. These will not be considered here while planktonic algae, although a constituent of suspended organic material, will be considered elsewhere (see section 4.1.2). The quantity and quality of material suspended in the water column at any particular moment is largely determined by water movement, which transports, fractionates and modiﬁes solids. Large, dense particles are more easily sedimented than small, less dense particles. Water currents can also prevent particles from sedimenting and re-suspend sedimented materials. Water chemistry, and salinity in particular, inﬂuences turbidity through its effect on ﬂocculation and sedimentation, and is particularly important in the transport of sediments in estuaries (see Postma 1967; McCave 1979). Although suspended solids cause a number of problems in aquatic systems (Alabaster & Lloyd 1980), it is principally their effects on stock that are of concern to cage ﬁsh farmers. High levels cause gill damage, inducing the gill epithelial tissues to proliferate and thicken (Ellis 1944; Eller 1975). If damage is sufﬁciently severe ﬁsh will die. Ellis (1944) has argued that the larger the particles, the greater their hardness and angularity and the greater the possibility of injury to gill tissues. Mortalities often occur after the period of exposure, complicating diagnosis (Alabaster & Lloyd 1980). Suspended solids have also been implicated in diseases such as ‘ﬁn-rot’ (Myxobacteria) (Herbert & Merkens 1961; Herbert & Richards 1963) and in poor ﬁsh growth. Poor growth may in part be caused by the effect of turbidity on visibility, adversely affecting food intake (Sigler et al. 1984). Turbidity levels less than 100 mg l-1 have little effect on most species. However, duration of exposure is important. Although cage ﬁsh farmers should avoid sites where high turbidity levels occur, this may not be possible in riverine sites where concentrations of several thousand mg l-1 suspended solids can occur during ﬂoods (Alabaster & Lloyd 1980).

122

Chapter 4

Pollution Pollution can be deﬁned as ‘the introduction by Man into the environment of substances or energy liable to cause hazards to human health, harm to living resources and ecological systems, damage to structure or amenity, or interference with legitimate uses of the environment’ (Holdgate 1979). With respect to cage ﬁsh farming, a pollutant is something that damages the cage structure, adversely affects farmed stock or food, or accumulates in the farmed ﬁsh to such a degree that it is toxic if eaten. There are enormous numbers of pollutants entering aquatic ecosystems (e.g. heavy metals, PCBs, oil-based products, pesticides, detergents) and it is beyond the scope of this book to discuss or list even those most relevant to cage ﬁsh farming (see Alabaster 1982; Lloyd 1992; Svobodová et al. 1993; Howells 1994; Poxton 1996). Although extensive sampling and sophisticated analytical techniques are required to detect many of these compounds, risks can be reduced by siting cages as far as possible from industry. Unfortunately, in many parts of the world, ﬁsh farming has had to develop alongside industry or in heavily populated areas. In Asia, pollution – industrial, agricultural, domestic – is one of the most important constraints to cage aquaculture development (Tabira 1980; Ikenoue & Kafuku 1988; Liao & Lin 2000). As much as 80–90% of aquaculture losses in Japan were at one time attributed to oil pollution alone (Sakiyama 1979). Geographically isolated areas can also be at risk. Major oil spills from tankers using routes close to coastal ﬁsh farms have caused problems. For example, the Braer accident in southern Shetland, Scotland in 1993 resulted in 1740 t of farmed Atlantic salmon being killed and a further 3659 t being destroyed the following year owing to hydrocarbon ﬂesh taint (Black 1998). The increasing risks of reductions in pH owing to acid precipitation in some of the remoter lakes of Northern Europe and North America, although these trends are at last being reversed thanks to reductions in SO2 emissions. Thermal efﬂuents from power stations may contain biocides, such as chlorine, to control fouling, corrosion inhibiters and heavy metals (e.g. copper, zinc, cadmium, nickel and iron) (for review see Langford 1983). These should be investigated during site selection. Pollution is dealt with further in Chapter 7.

4.1.2

Harmful algal blooms

Phytoplankton blooms occur whenever appropriate conditions prevail: high light and nutrient levels, and warm temperatures combined with favourable hydrographic conditions. Blooms adversely affect ﬁsh not only through their physical presence, which can damage or clog gills, but also through their inﬂuence on DO. Some phytoplankton species impart a musty ﬂavour to ﬁsh ﬂesh while others produce toxins that kill aquatic organisms or accumulate in their tissues and prove toxic when eaten. While phytoplankton blooms are discussed in general in this section, emphasis will be placed on factors that govern the development of harmful algal blooms (HABs), a term that also covers the effects of toxic organisms when present at low concentrations. Phytoplankton and DO are discussed in section 4.1.1.

Site Selection

123

At any moment the phytoplankton community consists of tens of species of algae. Species differ in their temperature, light and nutrient requirements and, under any given set of environmental conditions, one species will tend to dominate. As conditions ﬂuctuate over time owing to changing light levels, temperature, nutrient availability and not least through the inﬂuence of the algae themselves, so the dominant species may change. This process of succession is often less marked in the tropics where environmental conditions are more stable. Changes in phytoplankton community composition may also occur through the inﬂux of algae from other areas. During bloom conditions the algal community, typically dominated by one species, becomes extremely large, densities in excess of 106 cells ml-1 often occurring. Blooms of certain species occur more frequently than others and in fresh waters the most important groups are the diatoms and cyanobacteria (blue-green algae). Freshwater blooms In temperate regions freshwater diatom blooms tend to occur during the spring and occasionally during the autumn months. In fact, diatom blooms are usually only termed blooms because they are present in relatively large numbers compared to the cell numbers of other species present. Nevertheless, although scarcely discernable to the naked eye they can still cause gill damage because of their silicaceous cell coats (Mamcarz & Wornial´´l o 1986). Blooms of cyanobacteria are more common than diatoms in inland waters. They are sometimes a permanent feature in tropical lakes and reservoirs while in temperate regions blooms typically occur from mid to late summer as the result of a combination of factors. Light and nutrients stimulate growth and cell division, and hence blooms occur most commonly in productive lakes where nutrient loadings are high. However, many species produce gas vacuoles resulting in the accumulation of algae on the water surface (Fig. 4.4). If this occurs during a period of calm weather there may be insufﬁcient turbulence to disperse the algae (for review see Reynolds 1984). The more common genera of bloom-forming cyanobacteria are given in Table 4.5 and include species of Oscillatoria, Anabaena and Microcystis that produce compounds such as geosmin and 2-methylisoborneol which cause musty, earthy off-ﬂavours in farmed ﬁsh (Persson 1980; Tucker & Martin 1991; Millie et al. 1995; Tucker 2000). Also of concern to cage ﬁsh farmers are toxin-forming species of Microcystis, Anabaena, Aphanizomenon and Oscillatoria. Although toxin-producing cyanobacteria are ubiquitous, the factors determining toxicity are complex and poorly understood (Falconer 1993; Carmichael 1994). The toxins are of two types: alkaloid neurotoxins and protein or peptide hepatotoxins. However, not all species within a genus and not all populations of a particular species may be toxin-producing; even within a single water body some bloom patches may be toxin-producing while nearby patches may not. The toxin produced by a particular species may vary from strain to strain, making identiﬁcation of a toxic bloom or toxin impossible without laboratory tests.

124

Chapter 4

Fig. 4.4 Dying bloom of cyanobacteria in a ﬁsh cage (courtesy J. A. Stewart).

Table 4.5 Genera of common bloom-forming cyanobacteria in inland waters (modiﬁed from Reynolds & Walsby 1975). Order

Family

Genus

Chroococcales

Chroococcaceae

Nostocales

Oscillatoriaceae

Coelosphaerium Gomphosphaeria Microcystis Oscillatoria Spirulina Trichodesmium Anabaenopsis Aphanizomenon Gloeotrichia

Nostocaceae Rivulariaceae

While cyanobacteria toxins are among the most potent in nature (Fig. 4.5), their role in ﬁsh kills is not necessarily due to poisoning. While deaths of farmed channel catﬁsh in ponds in the United States have been widely attributed to cyanotoxins (Zimba et al. 2001), deaths at a farm in Spain were shown to be due to asphyxiation rather than cyanobacterial toxins (Toranzo et al. 1990). Immersion of rainbow trout in cultures of toxic Microcystis aeruginosa for 10 days resulted in no mortalities (Phillips et al. 1985b), suggesting that toxic effects only occur when ﬁsh are directly exposed to the toxin. This may happen when algae are ingested or when gills are damaged, increasing permeability to toxins (Phillips et al. 1985b; Rodger et al. 1994; Tencalla et al. 1994). However, while

Site Selection

125

Fig. 4.5 Toxicities of various animal, fungal, bacterial and cyanobacterial toxins. LD50 represents the dose, in g puriﬁed toxin per kg body weight, required to kill 50% of a population of laboratory mice or rats by intraperitoneal injection (redrawn from Skulberg et al. 1984).

phytoplanktivorous species such as silver carp and tilapia reduce or stop feeding when exposed to toxic cells (Beveridge et al. 1993; Keshavanath et al. 1994; Beveridge & Baird 2000), it is believed that accidental ingestion of toxic cells while feeding on ﬂoating pellets has caused the deaths of channel catﬁsh (Zimba et al. 2001). Although many of the claims that mortalities at cage ﬁsh farms have been caused by toxic algae remain unsubstantiated, there may be human healthrelated issues associated with consumption of ﬁsh that have bioaccumulated microcystins (De Magalhaes et al. 2001). It would thus seem wise to avoid lakes that are most likely to suffer cyanobacterial blooms (i.e. water bodies with high nutrient levels and elevated algal biomass; Downing et al. 2001).

126

Chapter 4

Coastal blooms A number of marine algal groups form blooms, including diatoms, cyanobacteria, prymnesiophytes and dinoﬂagellates. Global warming seems to be associated with an increase in frequency and range extension of HABs (De Sylva 1999). There are, however, few data on the costs to the aquaculture industry. In an assessment of losses in the Scottish and Irish aquaculture industries between 1988 and 1993 Kennedy (1994) estimated that 3% of the 450 claims, accounting for 10% of settlements (£2.1 million) by his company, were due to HABs. In the space of two months (March–April 1998), a dinoﬂagellate bloom in Hong Kong killed some US$32 million worth of ﬁsh (Yin et al. 1999). Diatom blooms are particularly common in temperate and sub-tropical coastal waters during the cooler months of the year. The diatoms Chaetoceros convolutus and C. concavicornis have been responsible for heavy losses of caged salmonids in North America (Rensel et al. 1989; Albright et al. 1993; Whyte et al. 1997). The prominent, barbed setae become embedded in the lamellar epithelium of the gills and at sub-lethal levels (<5 cells ml-1) cause production of excess amounts of mucus, limiting oxygen uptake and increasing susceptibility to bacterial pathogens such as Vibrio anguillarum. In extreme cases deaths have occurred as a result of interference with gill function and loss of blood from injury. Atlantic salmon seem to be particularly susceptible (Rensel et al. 1989). However, as only certain sites have suffered losses, problems may be avoided by careful site selection. High losses have also been reported from Scotland by Bruno et al. (1989) where a bloom of diatoms and a silicoﬂagellate caused severe irritation of farmed Atlantic salmon gills leading to necrosis, extensive sloughing of gill lamellae and consequential respiratory failure. Bloom-forming marine species of cyanobacteria are uncommon, but include the genera Trichodesmium and Nodularia (Round 1981) and, although a few tropical and sub-tropical species of Lyngbya and Oscillatoria are toxic (Moore 1982), they seldom form blooms and are rarely associated with ﬁsh kills (see also section 4.1.3 below). Cyanotoxins, however, have been associated with net pen liver disease (Kent et al. 1996; see below). Toxin-producing prymnesiophytes have also caused problems at ﬁsh farms. In Norway in 1989, 750 t of caged trout and salmon in a fjord were killed by a bloom of Prymnesium parvum. Although the mass release of fresh water from a hydro-electric power plant was held primarily responsible, nutrient loadings from the ﬁsh farms are thought to have played a role in the development of the bloom (Kaartvedt et al. 1991). In the late 1980s to early 1990s huge blooms of Chrysochromulina leadbeateri, originating in the polluted Skaggerak, were carried by currents along the Norwegian coast, killing hundreds of tonnes of salmon (Aune et al. 1992; Black 1998; Eikrem & Throndsen 1998). The most important group of HABs is undoubtedly the dinoﬂagellates, responsible for so-called ‘red tides’. However, of the 1200 species, many of which cause red tides, only a very small proportion has been shown to produce toxic substances, and even fewer have been implicated in ﬁsh kills (Okaichi et al. 1989; Graneli et al. 1991). Dinoﬂagellate-related deaths at ﬁsh farms must now be regarded as widespread and common: see reports from Japan (Nishimura 1982;

Site Selection

127

Fig. 4.6 Section of a gill affected by the marine dinoﬂagellate Gyrodinium aureolum, showing distorted secondary gill lamellae. Cytoplasmic and nuclear debris is evident between the lamellae (H & E ¥ 325) (courtesy R. J. Roberts).

Okaichi 1989; Yang et al. 2000), Korea (Kim 2000) and Hong Kong (Wong & Wu 1987; Yin et al. 1999; Yang et al. 2000). In August 2002, a bloom of Cochlodinium polykrikoides reportedly caused the deaths of some 24 000 3-kg yellowtail, sea bream and globeﬁsh (Tetradontidae) in the Yatsuhiro Sea, worth 300 million Japanese yen. In Western Europe dinoﬂagellates have been implicated in mass mortalities of farmed salmonids, eels and gadoids (Tangen 1977; Taylor & Seliger 1979; Parker 1981; Jones et al. 1982; Mortensen 1985; Dahl & Tangen 1990). The toxicological mechanisms, however, remain unclear, being attributed to both neurotoxins and to reactive oxygen species, the latter causing cause oxidative damage to ﬁsh gills (Kim et al. 1991). Roberts et al. (1983) have described the sequence of events leading to death in rainbow trout following exposure to Gyrodinium aureolum: increased opercular beat rate, hyperactivity followed by inactivity, and death within 1 to 24 h. The gills of dead ﬁsh displayed marked histological changes characterized by necrotic degeneration and disintegration of the lamellar epithelium (Fig. 4.6). Toxic dinoﬂagellate blooms tend to occur in warm water and, therefore, are restricted to the summer months in temperate regions. Gowen et al. (1985) propose three ways in which blooms might occur at a coastal site: (1) (2)

the bloom could develop at the site from species already present; the bloom could develop at the site from species transported into the farm area; (3) the bloom could develop outside the site and be transported into the farm area.

128

Chapter 4

The risk of a type (1) or type (2) bloom occurring depends largely on the nature of the site. If water exchange is in days and if nutrient levels, especially dissolved NO3, are low, then risks are low. Risks of type (1) and (2) bloom occurring are much higher in weakly stratiﬁed fjordic bays with long (weeks) exchange periods, and where there are also high levels of dissolved NO3, even in deep water, risks are further increased. For sites at risk from type (3) blooms, environmental conditions are less important than proximity to areas where dinoﬂagellate blooms occur. Studies of northern European waters have demonstrated that dinoﬂagellates thrive in shelf sea-front boundaries (i.e. areas where tidally mixed inshore waters and thermally stratiﬁed offshore waters meet) (Pingree et al. 1978), and that certain locations, because of their proximity to fronts and prevailing currents, are particularly susceptible to blooms (Gowen et al. 1985; Dickman 2000). Sites close to highly polluted coastal areas may also be vulnerable. It is thus apparent that some coastal sites, which are otherwise suitable for cage ﬁsh farming, may be at risk from toxic dinoﬂagellate blooms. Talking to local people and relevant authorities may help identify such sites since toxic blooms are often recurrent events. Coastal shelf sea-fronts may also be detected by satellite. It would be best, therefore, to avoid nutrient-rich marine sites and sites where the exchange period is longer than a few days. Using the above criteria Gowen (1984) identiﬁed high-risk sites on the west coast of Scotland. This approach might also be applicable to fjordic marine sites in other parts of the world (e.g. Canada, New Zealand, Norway, Chile). The risks posed by toxic algal blooms cannot be eliminated entirely by judicious site selection. Cage farms change the nature of sites through the addition of nutrient-rich wastes, and research in Japan and elsewhere has implicated these wastes in the formation of algal blooms (Nishimura 1982; Kaartvedt et al. 1991; Black 1998). Kennedy & Kreiberg (2001) highlight the importance of water quality monitoring in relation to Heterosigma akashiwo blooms. A depression of long-term mean monthly salinity values by 1–3‰ during spring greatly increases the probability of a summer bloom.

4.1.3

Disease

The worst types of site from a disease risk point of view are those in which pathogenic or potentially pathogenic organisms exist prior to establishment of the farm, and those in which disease organisms are likely to thrive following the establishment of a farm. With regard to the former, cage ﬁsh farms are likely to be more prone to disease than the latter as they exploit natural water bodies and are thus more exposed to risk (see also Chapter 2). Organically polluted water bodies harbour more disease agents than unpolluted water bodies. ‘Red-boil disease’, a dermal form of vibriosis occurring in the estuarine grouper (Epinephelus salmoides), for example, is produced by the bacterium Vibrio parahaemolyticus and can be contracted following skin damage through handling (Wong et al. 1979; Chua & Teng 1980; Leong & Colorni 2002). The organism has been found in exceptionally high numbers in sewagepolluted waters such as the Straits of Penang. Vibrio-induced mortalities of up

Site Selection

129

to 90% of caged stocks have been reported here, while almost no deaths were caused at a clean water site on another part of the island (Chua & Teng 1978). Similar differences between the two sites in terms of Escherichia coli numbers were also recorded (Chua & Teng 1980). Haemorrhagic septicaemia is a common disease among cultured freshwater carp. It is caused by the bacterium Aeromonas punctata (Toor et al. 1983) whose presence in fresh waters has been related to trophic state (Rippey & Cabelli 1980). The occurrence of other ﬁsh pathogens also seems to be related to trophic conditions. There is evidence of a link between trophic state and pathogenic fungal infections in ﬁsh (Grimaldi et al. 1973), while Chua (1979) has observed that the ectoparasitic marine isopod Nerocilia sp., which attacks rabbit ﬁsh, is more prevalent in organically enriched marine waters. Facultative pathogenic organisms, therefore, are often associated with water bodies where a source of infection, such as untreated sewage, is present or where there is a supply of organic nutrients that the organisms can utilize. Both wild ﬁsh populations and organisms that are intermediary hosts in a particular ﬁsh parasite’s life cycle can cause disease problems for the ﬁsh farmer. The introduction of large numbers of caged ﬁsh to a system can have a dramatic effect on disease agents, although the mechanisms involved in the transmission of parasites from wild to cage ﬁsh and vice versa are not fully understood (McGuigan & Sommerville 1985; Sommerville 1998). Cages of salmon attract scavenging saithe (Pollachius virens) which often harbour the sea lice Lepeophtheirus salmonis and Caligus elongatus, and laboratory trials have clearly shown that lice can transfer between host species (Bruno & Stone 1990; Kent & Poppe 2002). In several cases in the UK caged ﬁsh have become severely infested with the cestodes Triaenophorus nodulosus and Diphyllobothrium spp. resulting in heavy mortalities and the closure of at least one farm (Wootten 1979; Jarrams et al. 1980). Infections were subsequently traced to the wild ﬁsh populations that were found to be carrying the parasites. Matheson (1979) showed that Atlantic salmon parr raised in cages in a freshwater Scottish loch became heavily infected with D. ditremum and D. dendriticum within 2 months of being introduced to the site. Although parasites were not isolated from wild brown trout (Salmo trutta) in the loch, few specimens had been examined. The metacercariae of the marine digenean Cryptocotyle lingua encyst in the skin of ﬁsh, causing disﬁgurement. The ﬁrst host of the parasite is the common whelk Littorina littorea, which occurs on rocky coasts, and the parasite is commonly encountered by salmon farms sited close to such areas (Sommerville 1998; Kent & Poppe 2002). Disease risks, therefore, can be minimized by avoiding sites where parasites occur or where there are disease agents in the wild ﬁsh, in intermediary hosts, or in the environment that could be transmitted to the caged ﬁsh (Beveridge 2002). Russian workers suggest that diplostomatosis in caged trout and carp may be avoided by siting cages 80–100 m from fringing macrophyte beds where the intermediate host of Diplostomum spathaceum, the snail Lymnaea stagnalis, resides (Martyshev 1983). Unfortunately, once cage ﬁsh farms are established, the conditions that promote disease are enhanced: an increase in host density, a deterioration in

130

Chapter 4

water quality, an increased risk through the introduction of diseased stock to the farm or the attraction of birds and other opportunistic predators to the area (see section 7.2).

4.1.4

Water exchange

Good water exchange, or ﬂushing, at a site intended for intensive cage operations is essential in order to minimize the accumulation of wastes and all the associated problems this can cause. Water exchange is dependent upon currents, although it is complicated by the effects of salinity, temperature and topography. In open coastal sites, such as straits, or those located in the shelter of a group of coastal islands or a coastal indentation (bay or bight), water exchange tends to be rapid, occurring in a period of days rather than weeks. From Edwards & Edelsten (1976), the volume of water in the bay, V, can be calculated from: V = AD where V is in m3; A = surface area (m2); and D = mean depth (m). The ﬂushing time, T, can be estimated by dividing V by the incoming ﬂow of water, F, such that: T =

V F

The hourly mean ﬂow of water, F, is dependent upon tidal currents: F=

AH 12.5

where H = mean tidal height (m). Thus, T =

12.5D H

However, many cage farms in Scotland, Norway and elsewhere are situated in enclosed sea lochs (fjords) or bays where the pattern of water exchange is much more complicated. Fjords are typically rectangular in cross-section, with a small width : depth ratio. They may extend as much as 150 km inland and have one or more inﬂowing rivers. Although the above equation can be used to estimate ﬂushing rate, it assumes that the fjord is well-mixed and ignores the effects of ‘short-circuiting’. The ﬂushing of a fjord is dependent upon tidal movements of water at the sea end and on freshwater inﬂows at the landward end (see Fig. 4.7). In reality, the idealized picture shown in Fig. 4.7 is greatly complicated by the size and topography of the basin, the number, location and depth of sills, the magnitude and location of the freshwater inﬂows and tidal range. In some fjords, part of the water in the incoming ﬂood tide may be re-circulated ebb water,

Site Selection

131

Fig. 4.7 An idealized fjordic circulation. The surface, brackish layer discharges to the sea. A compensation current ﬂows in over the sill to replace water which is entrained (vertical arrows) into the surface outﬂow. Strong mixing may occur at the sill and there may be recirculation of brackish water into the compensation current (modiﬁed from Edwards & Edelsten 1976).

leading to a short-circuiting of the ﬂushing system and an increase in ﬂushing time. Sites in fjords with sills that restrict water movement or with long and complicated topographies tend to suffer from these deﬁciencies. The mouths of some fjords, such as those on the Paciﬁc coast of North America, have sills which extend to 150 m or more, while those on the Norwegian, Scottish and New Zealand coastlines often have sills of 10 m or less in depth (Edwards & Edelsten 1976; Dyer 1997). Such shallow sills not only increase the exchange time of the whole basin through short-circuiting, but also greatly limit the ﬂushing of deep layers of water (Fig. 4.7). Increasing freshwater ﬂow into a fjord system decreases the overall exchange time (Dyer 1997), although because of salinity/density effects (see section 4.1.1) this will have a proportionally greater effect on surface than on deep water exchange. An example is given by Edwards & Edelsten (1976) who found that the ﬂushing time of Loch Etive, a fjord on the west coast of Scotland, was 8 days at a depth of 2 m, whereas at 14 m it was 28 days. The above points concerning the inﬂuence of sills and freshwater inﬂows illustrate that although currents at a site may appear to be ideal (10–60 cm s-1) for cage ﬁsh culture, water exchange can still be low, leading to poorer than expected water quality in surface waters and a build-up of wastes on the bottom. In freshwater lentic sites, the exchange time, or ﬂushing rate (volumes per year, r), depends upon the volume of the water body, V, and the total volume of water ﬂowing out of the water body each year, Qo, such that: r=

Qo V

This formula is appropriate for cages sited in the main body of a lake or reservoir but may give an over-optimistic estimate if applied to a farm sited in a bay

132

Chapter 4

within a lake. Although the ﬂushing rate of bays can be computed using the formula above, substituting the volume of the bay for V, and the total annual outﬂowing water from the bay for Qo, computations can be difﬁcult. In general, however, it is strongly advised that large-scale intensive units are not sited in enclosed bays. It should also be noted that improved exchange at a site (i.e. an overpessimistic prediction of r) can also occur when cages are located near outﬂows (see Kelly 1995). The exchange time or ﬂushing rate of inland waters ranges from a matter of seconds for riverine locations to several years for many lakes. As stated in the introduction to this section, intensive cage ﬁsh farms are best sited in areas with good water exchange. In marine areas, this means at sites where there are both good bottom and surface currents and where the exchange period is in days rather than weeks. In fresh water, it is desirable that the ﬂushing rates of lakes and reservoirs are in weeks and months rather than years. In rivers, currents are generally too strong for intensive cage culture (see Chapter 2). Semi-intensive and extensive cage culture operations, however, depend upon the supply of plankton/detritus, and thus a high rate of primary production is necessary. As is demonstrated in Chapter 5, primary productivity and ﬂushing rate are inversely correlated, and thus sites with poor water exchange may be preferred for these types of culture.

4.1.5

Fouling

Fouling decreases the speciﬁed mesh size of netting while increasing mesh surface area. The reduction in mesh size restricts the ﬂow of water through cages, thus reducing the rate of DO supply and waste metabolite removal. The increased resistance to water ﬂow can cause deformation of the bag and a decrease in cage volume as well as increasing stresses on the cage collar and moorings. The additional weight of fouling can lead to net failure and makes net changing difﬁcult and time consuming. Fouling organisms have also been implicated in ‘netpen liver disease’ (Andersen et al. 1993; Kent et al. 1996) and have been shown to act as a reservoir of Neoparamoeba pemaquidensis, the causative agent of amoebic gill disease in Atlantic salmon in Tasmania (Tan et al. 2002; DouglasHelders et al. 2003). Fouling has some positive attributes; it reduces the risk of abrasion to caged stock and studies in Canada suggest that encrusting ﬁlter-feeding mussels reduce the risks of bacterial kidney disease to caged salmon posed by the pathogenic bacterium Renibacterium salmoninarum (Paclibare et al. 1994). The periphyton community that encrusts tropical freshwater cages has also been shown to enhance growth and production of caged tilapias and can reduce production costs (Norberg 1999; Huchette et al. 2000; Huchette & Beveridge 2003). There are some 200 species of marine fouling plants and animals (Lovegrove 1979). The few studies that have been carried out on fouling of marine cages and pens indicate that the diversity of colonizing organisms can be large and that several species (e.g. the mollusc Martesia striata) can damage wooden structures by boring. Cheah & Chua (1979) identiﬁed more than 34 species of algae

Site Selection

133

(cyanophytes, rhodophytes, chlorophytes), coelenterates, polyzoans, annelids, arthropods, molluscs, and simple chordates clinging to net cages sited in the Straits of Penang after only 2 months of immersion. No increase in diversity was found during the subsequent 2 months of the study although the weight of attached organisms more than doubled. Colonization begins with biochemical conditioning, which occurs immediately after immersion, and which results in an equilibrium between macromolecules on the structure and those in the surrounding medium. Within an hour of immersion, a primary ﬁlm of bacteria is formed; colonization by unicellular algae, then multicellular organisms, follows (Horbund & Freiberger 1970; Milne 1970; Santhanam et al. 1984; Hodson & Burke 1994). However, the size, diversity and development pattern of the fouling community depends upon the cage materials used, species succession, physical disturbance, competition, predation, the prevalence of fouling species in the environment and, of course, immersion time. In Scotland, Milne (1970) investigated the rates of fouling of mesh and netting panels of different material and found that galvanized panels developed much less fouling than synthetic ﬁbre netting panels (see Table 3.5). In Tuticorin, India, Santhanam et al. (1984) observed that different organisms colonized the bamboo, oil drum and PE netting parts of cages. Studies conducted in Malaysia, Bangladesh and Australia observed that the fouling community on net cages changes over time (Lai et al. 1993; Hodson & Burke 1994; Huchette et al. 2000). The size and diversity of the fouling community is also greatly affected by the temperature and productivity of the environment (Dubost et al. 1996). In warm, organically enriched waters such as the Straits of Penang and Hong Kong Harbour, the rates of fouling are high (Cheah & Chua 1979; Tseng & Yuen 1979). The rate of fouling is also higher in cages sited in thermal efﬂuents (Chamberlain & Strawn 1977). There is evidence too, that fouling is most rapid in areas of slow currents. In Singapore, for example, fouling is principally a problem where ﬂood tide currents are less than 25 cm s-1 (PPD 1986). The cage environment, with elevated dissolved nutrient levels, is believed to promote fouling (Costa-Pierce & Bridger 2002). The range of fouling organisms and the rate of fouling decreases with salinity, and so fouling is generally not a problem at freshwater sites, although the boring larval stage of the mayﬂy Povilla adusta may cause considerable damage to wooden cages in some African lakes (Coche 1979) (Fig. 4.8). Santhanam et al. (1984) found many fewer fouling animals on net cages and pens sited in brackish than marine waters. Observations of fouling at a freshwater cage site in Scotland have shown that algae are the principal organisms responsible. In studies of fouling of tilapia cages in Laguna de Bay (Laguna Lake), Pantastico & Baldia (1981) also found algae to be the principal fouling organisms present and that growth was highest in the upper portions of the net cages where light levels were greatest. Greenland et al. (1988) and Dubost et al. (1996) found fouling in fresh waters to be principally composed of bryozoans. Greenland et al. (1988) reported that the fouling adversely affected growth of caged channel catﬁsh maintained in ponds. In summary, fouling is of greatest signiﬁcance at marine sites with low current velocities and is greatly inﬂuenced by temperature and nutrient status. Either

134

Chapter 4

Fig. 4.8 Damage caused by Povilla adusta larvae to a submerged cage structure made of softwood after 13 months of immersion in a freshwater lake (from Coche 1979).

such sites should be avoided (see Dubost et al. 1996) or management practices adapted to take account of the problem (see section 7.4).

4.2 4.2.1

ENVIRONMENTAL CRITERIA FOR CAGES Weather

The weather can determine the suitability of a particular site or area for cage ﬁsh culture through its inﬂuence on cage structures and on caged stock. Of particular concern are violent storms and conditions of extreme cold. The storms of tropical latitudes may be classiﬁed into four types: • tropical cyclones (wind strengths ≥ Force 12, 33 m s-1); • severe tropical storms (wind strengths = Force 10–11, 24–32 m s-1); • moderate tropical storms (wind strengths = Force 8–9, 17–23 m s-1); • tropical depressions (wind speeds < Force 8, 17 m s-1).

Site Selection

135

Fig. 4.9 Average number of hurricanes (wind Force 12) and tropical storms (wind Force 8–11) per year (redrawn from Couper 1983).

With the most violent of these, tropical cyclones, torrential rain accompanies hurricane force winds. Such events occur principally between latitudes 5° and 30° and are known by a variety of regional names, e.g. cyclones (south-west Paciﬁc and Bay of Bengal), hurricanes (the North Atlantic, West Indies and north-east Paciﬁc), typhoons (the China Sea and north-west Paciﬁc). Cyclones develop from disturbances in the Intertropical Convergence Zone (ITCZ; the area near the equator where the south-east trade winds meet the north-east trade winds) and tend to follow a westerly track curving pole-wards in a clockwise manner in the northern hemisphere and anticlockwise in the southern hemisphere. As they travel they intensify and accelerate until they are typically 600–800 km in diameter. The vast amounts of energy that fuel their progress are derived from heat released in the condensation of water held in the moist air sucked upwards from the ocean surface (Couper 1983; http://www.ncdc.noaa.gov/ogp/papers/houze1.html). There is more energy released when water vapour is condensed from warm air, since warm air holds more moisture. Thus, cyclonic development is greatest over the western areas of oceans during late summer and autumn when sea temperatures are highest (Fig. 4.9). As a cyclone travels across a large land mass, it becomes starved of moisture and decays rapidly. Cyclones are thus maritime phenomena.

136

Chapter 4

Fig. 4.10 Approach routes of typhoons, Taiwan (redrawn from Twu et al. 1986).

Most tropical cyclones occur in the West Paciﬁc and Indian Oceans (Fig. 4.9). The development of cage marine ﬁsh culture has been hampered in many typhoon-affected areas, such as Taiwan (Twu et al. 1986; Liu 2000) (Fig. 4.10). In Japan and elsewhere cage ﬁsh farmers accept the risks. Breakwaters and submersible cages may minimize the risk of damage and stock loss (see section 7.9.1). Inland waters adjacent to coastal areas are also sometimes hit by typhoons, causing enormous damage and loss of stock (Fig. 4.11). Ice not only damages cages but also can cause super-chilling of water, leading to ﬁsh deaths. Among the worst affected areas are those inﬂuenced by Arctic currents, including the coastlines of northern and eastern Canada as far south as Nova Scotia, northern and western Alaska, Greenland, northern and eastern Iceland and the Baltic (see section 4.1 above; Page & Robinson 1992; Boghen 1995). Where ice and low surface temperatures are prolonged, cage culture may be impossible without recourse to the use of submersible cage designs, while in more marginally affected areas (e.g. New Brunswick, Baltic) risks can affect insurance premiums. Occasional super-chilling of sea water can occur outside the areas mentioned above. In northern Norway, in 1982, one ﬁsh farmer lost 20 000 Atlantic salmon smolts after temperatures fell to -1.2°C (Anon 1982). Temperate continental inland lakes and reservoirs also often freeze during winter, although in more marginally affected areas the movement of the ﬁsh can help keep the cage environment ice-free (Paetsch 1977) (see also section 7.9).

Site Selection

137

Fig. 4.11 Typhoon damage, ﬁsh pens and cages, Laguna de Bay, Philippines.

4.2.2

Shelter and waves

Few structures can withstand the impact of the open sea: winds tear at structures projecting above the water while waves attack objects on the sea surface. Shelter from these forces, and waves in particular, is therefore a prime consideration in site selection. Also to be considered are general wave conditions at prospective sites since they determine fatigue wear on structural components and operational conditions for staff. A sinusoidal wave pattern is shown in Fig. 4.12. In the aquatic environment there are many kinds of waves that differ in origin, form and velocity (Fig. 4.13), although those that are of paramount importance in siting cages are generated by the wind. Wind-driven waves are propagated by the frictional drag of wind blowing across a stretch of water that transfers energy to the ﬂuid. Wave size is determined by wind velocity, wind duration and the distance of open, unobstructed water across which the wind blows (fetch) (Bascom 1964) and is also inﬂuenced by the waves present when the wind starts to blow. At the windward end of the fetch, waves are poorly developed with small wave heights and short periods of oscillation. However, they develop with distance, reaching maximum size when they attain the same velocity as the wind. In practice, only waves generated by low-velocity winds attain full potential (Table 4.6). As waves leave the generating area they modify and become what is often termed ‘swell’. The shorter waves lose energy quickly and die out, and the swell becomes more uniform and longer crested while wave height gradually decays. Wave height increases with wind velocity (Table 4.6) and wave energy increases proportionally with the square of wave height. When considering the

138

Chapter 4

Fig. 4.12 Wave terminology.

Fig. 4.13 Period and relative energy of different ocean waves (redrawn from Bascom 1964).

suitability of a site from the point of view of shelter, it is the worst possible conditions that are of concern, i.e. the highest waves and most damaging wave periods and frequencies that are likely to occur at the site. Several publications discuss the prediction of wave heights from meteorological and topographical data (Milne 1972; Muir Wood & Fleming 1981; US Army Corps of Engineers 1984). In order to assess likely wave characteristics at a site, information on the long-term frequency and direction of surface wind speeds is needed. This may be obtained from national meteorological ofﬁces, such as the UK Meteorologi-

Site Selection

139

Table 4.6 Conditions in fully developed seas (modiﬁed from Bascom 1964).

knots

m s-1

miles

km

h

Average height m

10 15 20 25 30 35 40

5.1 7.7 10.3 12.9 15.4 20.6 25.7

10 34 75 160 280 710 1420

19 63 139 297 519 1315 2630

2 6 10 16 23 42 69

0.3 0.8 1.5 2.7 4.3 8.5 14.6

Wind velocity

a b

Fetch length

Time

H3a

H10b

m

m

Period where most energy is concentrated s

0.4 1.1 2.4 4.3 6.7 13.4 23.8

0.6 1.5 3.1 5.5 8.5 17.4 30.2

4 6 8 10 12 16 20

Signiﬁcant wave height. Average height of the highest 10% of waves.

Fig. 4.14 A typical windrose. The values in the centre of the rose give the total number of observations and percentage of calms (see text).

cal Ofﬁce or the Norwegian Meteorological Institute. Information may be supplied in the form of a windrose, in which the percentage frequencies with which winds of various strengths occur are plotted as vectors. The total length of all vectors, calm periods included, is 100%. In Fig. 4.14 the winds have been summarized as vectors representing eight principal directions, and it is readily apparent that the prevailing winds are from the south-west and west. Each vector has been subdivided into four segments of unequal lengths, illustrating the proportion of winds of different strengths blowing from each direction. Reviews of wind speed data may also be available. In the review by Shellard (1965), highest mean hourly wind speeds and gust speeds for different areas of the British Isles, which would on average be exceeded only once every 50 years,

140

Chapter 4

are given. Mean hourly speeds rather than gust speeds are used to predict wave characteristics. Electronic databases, such as UKD-map, are potentially a useful source of information although they still contain insufﬁciently detailed information of coastal areas for aquaculture purposes. Much wind data is from ship-board observations and has a considerable degree of bias that may be corrected using the expression: W = 2.16Ws0.777

(Cardone 1969, in US Army Corps of Engineers 1984)

where Ws = the ship reported wind speed (knots) and W = corrected wind speed. If no information relevant to the site being considered can be found then estimates of highest mean hourly wind velocities, based on nearest-to-site information and modiﬁed by local observations, may have to be used. Correction factors for alternative sources of data (e.g. wind data collected at elevations other than 10 m; wind speeds recorded when there are large air–sea temperature differences) are detailed in the Shore Protection Manual (US Army Corps of Engineers 1984). To account for drag, wind speed data should be to converted to wind stress factors (UA), such that: U A = 0.71U 1.23 m s -1

(Note: W ∫ U)

Fetch lengths at the site corresponding to the directions of the strongest prevailing winds must be determined. A fetch is deﬁned as an uninterrupted region in which the wind speed and direction are reasonably constant. For any given wind direction it is recommended that the fetch length be determined as a mean of nine measurements, 3° apart, taken from a map (see Fig. 4.15). Unlike the open expanses of the oceans, fetches in inland waters are limited by surrounding land masses, and fetches that are long in comparison to width occur frequently. Here, fetch width may have a limiting effect on the generation of waves. While Saville (1954, in Milne 1972) has proposed a method for determining the effective fetch, FE, for the relatively uniform fetch widths which occur in inland waters or coastal fjords, it greatly overestimates the restrictive effect of fetch width on wave heights and it is recommended that fetch is estimated using the above procedure. For a given set of fetch and wind data, wave heights are smaller and wave periods shorter if the waves are generated in shallow rather than deep water because of friction exerted by the sea bed and percolation in the permeable sediments (US Army Corps of Engineers 1984). In Figs 4.16 and 4.17, deep water (>15 m) and shallow water (3 m, 7.5 m, 10.5 m, 15 m) wave forecasting curves are plotted, assuming constant depths over the fetch. Milne (1970) has recommended that inshore and sheltered coastal sites use shallow water curves and that exposed coastal sites use deep water values. Wave heights are estimated by tracing a line from the wind stress factor ordinate until it intersects with the appropriate fetch length ordinate. This method of estimating wave height assumes that fetch length and not wind duration is limiting (see above), although Figs 4.16 and 4.17 include plots of wind duration. The predicted wave height actually refers to the signiﬁcant wave

Site Selection

141

Fig. 4.15 Map of a cage site. Nine radials, 3° apart, have been drawn in order to calculate fetch length along the principal wind direction.

height. However, the distances between consecutive waves are highly irregular and wave heights are also variable owing to the presence of a large number of wave components at any particular time. The practical solution is to deal with waves statistically. The signiﬁcant wave height is the mean height of the highest 33% of waves and is around 40–60% higher than the overall mean wave height (see Table 4.6). Note that the maximum wave height at a site may be 1.8–2 times the signiﬁcant wave height (Turner 2000b). A further assumption is that the atmospheric pressure system is stable. However, in rapidly moving pressure systems, such as can occur during tropical cyclones, the situation becomes complex. As an illustration of the effect this can have on waves, Muir Wood & Fleming (1981) cite Bretschneider’s example of a 12 m s-1 advancing cyclone producing a 50% increase in signiﬁcant wave heights. Swell can also cause problems for farms. Coastal swell typically has wavelengths of several hundred metres and periods of 7–15 seconds and the origin may be several thousand kilometres distant. Some Scottish sites exposed to the open Atlantic Ocean experience heavy swell, which has caused de-scaling of ﬁsh and consequent high mortalities through excessive motion of the cage bag.

Fig. 4.16 Nomogram for deep water signiﬁcant wave predictions as a function of wind speed, fetch length and wind duration (from US Army Corps of Engineers 1984).

142 Chapter 4

Fig. 4.17 Nomograms for shallow water signiﬁcant wave predictions as functions of wind speed, fetch length and wind duration (a = 3.0 m; b = 7.5 m; c = 10.5 m; d = 15.0 m) (from US Army Corps of Engineers 1984).

Site Selection 143

Fig. 4.17 Continued.

144 Chapter 4

Site Selection

145

In summary, the above methods can be used to estimate signiﬁcant wave heights at prospective sites although they ignore wave–current interactions which can be signiﬁcant in areas of strong current. More sophisticated forward tracking wave models have also been used by Bell & Barr (1990). Turner (2000b) proposes the following factors must be taken account of in considering wave climate and shelter: • • • • •

What type of cage system will survive, and for how long? What is the best orientation of the cages? Will ﬁsh survive in storm conditions? What type and strength of nets are required to avoid storm damage? What mooring excursion–restoring force limits will be necessary to minimize cage and anchor loads? • What maximum peak mooring loads will occur, both at the top and at the anchor? • Will the cages be workable on sufﬁcient days (see also Rudi et al. 1996)? • What type of boat will be required? • What type of feeding system will be required? • What offshore ‘deep’ counter currents may be generated by offshore storms? Although cages and moorings can be designed and built to withstand a range of weather conditions, as a guide it can be assumed that most ﬂoating cage structures are no stronger than other types of coastal structure and, therefore, unless specially designed will be unable to withstand signiﬁcant wave heights greater than 1–1.5 m (Möller 1979; Rudi & Dragsund 1993). Based on decades of experience, Turner (2000b) has offered a personal view of the appropriateness of different cage types to sites with different degrees of exposure and wave climate (Table 4.7), adding that ‘. . . they should not be used without a manufacturer’s warranty for such conditions’ (!). It must also be borne in mind that exposure can also affect operational activities and it is recommended that estimates of ‘down-time’ be made prior to ﬁnal site selection (see Rudi et al. 1996; Turner 2000b).

4.2.3

Currents

Good water exchange through cages is essential both for replenishment of oxygen and removal of waste metabolites. Currents inﬂuence ﬁsh behaviour, affecting social hierarchies, growth and growth disparities among stock (Phillips 1985; Leon 1986; Jobling et al. 1993; Jobling 1995) and, reportedly, ﬂesh quality. In extensive culture situations, water currents are also essential for the supply of food. However, excessive currents impose additional dynamic loadings to the cage, supporting structures and moorings, may adversely affect ﬁsh behaviour and contribute towards food losses from semi-intensive and intensive operations. High ﬂow rates have also been implicated in skeletal deformities of cage-reared carp (see later). In all but a few coastal regions of the world, such as in the Mediterranean or Baltic, tidal currents are the predominant source of surface water currents.

<4 800 <10 700 <12 500 <10 700

n ¥ 5 000 n ¥ 5 000 <10 000 <2 000

Flexible hose: 20–25 m diameter hexagon

Steel, tubular, hinged: 20 m square Semi-submersible

New generation Rigid multiple cage barges: n ¥ 20 m

Semi-submersible barges: n ¥ 20 m

Tension leg cages: submerged Sea bed bottom structures

<1 000 <2 700 <3 600 <8 500

Volume (m3)

Flexible hose: 15–20 m square

Conventional cage design Timber: 6–12 m square Steel, hinged: 12–15 m square Plastic circles: 12–25 m diameter Offshore plastic circles: 20–30 m diameter

Cage type and size

15.0 15.0

6.0

5.0

6.0 10.0

6.0

5.0

0.8 2.0 3.5 4.5

HS max (50-year) (m)

2.0 2.0

2.0

2.0

1.5 1.8

1.5

1.4

0.2 0.4 1.2 1.2

Mean HS 40% days (m)

<20 <20

<10

<10

<2.0 <5.0

<2.0

<2.0

<0.5 <1.0 <1.0 —

Distance offshore (km)

High capital costs; live aboard; prone to fatigue; easy to operate; ﬁsh abrasion may be a problem High capital costs; live aboard; easy to operate; limited capacity Not enough data; good potential; telemetry required Not enough data; good potential; telemetry required

Low capital costs; limited buoyancy; low cost Low labour; prone to fatigue; easy to operate Low capital costs; difﬁcult husbandry; labour intensive Low capital costs; need automatic feed system; difﬁcult to operate; labour intensive Medium capital costs; need automatic feed system; difﬁcult to operate; labour intensive Medium capital costs; need automatic feed system; difﬁcult to operate; labour intensive High capital and maintenance costs; easy to operate Capital intensive; need automatic feeders; net changing and harvesting difﬁcult

Notes

Table 4.7 Subjective, outline guidance to capabilities of various generic types of cage to withstand different degrees of exposure. Exposure is assessed by the average wave heights prevailing at a site (mean HS) and by the likely highest waves that will occur over a 50-year period (HS max (50-year)). An indicative offshore distance is also given (modiﬁed from Turner 2000b).

146 Chapter 4

Site Selection

147

Attractive forces, exerted by the moon and sun on the Earth produce tidal waves, which have extremely long wavelengths (half the circumference of the Earth) and periods of oscillation (12 h 25 min). The crest and trough of the wave are termed high and low tide respectively while the wave height is referred to as the tidal range. Associated with the rise and fall of the tide are horizontal motions of water, or tidal currents. Maximum current speeds occur at the middle of the rise (ﬂood) and fall (ebb), and minimum current speeds with slack water at the time of high and low tides. The Earth’s motion results in a rotational tidal current, which in the northern hemisphere where semi-diurnal tidal patterns predominate, is principally clockwise. Around the UK coastline, for example, an eastward ﬂowing tide at high water will ﬂow westwards 6 hours later. In the southern hemisphere, where mixed and diurnal tides prevail, the direction of rotation varies much more. However, the pattern of tidal currents is greatly inﬂuenced by local topography and other factors, such as surface runoff from the land and prevailing winds, resulting in marked changes in current over short distances and over depths of only a few metres. It is possible to estimate tidal currents in some types of coastal area. For example, maximum current speeds during ebb and ﬂood tides in a fjord can be calculated from:

Vmax =

Ê AH ˆ 4 10 m s -1 Ë B ¯

(Edwards & Edelsten 1976)

where A = surface area of the fjord landward of the site; B = cross-sectional area of the fjord at the site; and H = tidal range (m). However, in practice the formula is of limited value because of local inﬂuences. Rapid changes in depth or coastal contours can result in eddies with areas of dead current. The presence of a sill at the mouth of a fjord accelerates tidal currents in the region of the sill and greatly complicates the movement of water in the inner part of the inlet while in other types of bay, resonance effects may greatly increase tidal range and current velocities. A signiﬁcant fall in barometric pressure may also increase tidal range and currents. Further details concerning tides and tidal currents can be found in Muir Wood & Fleming (1981) and Pond & Pickard (1983). The methodology for surveying currents at a prospective site is discussed by Edwards & Edelsten (1976), Landless & Edwards (1976, 1977), Turner (2000b) and others. Details of PCbased current prediction models are also given in Bell & Barr (1990). Current velocities in coastal marine areas typically range from 0 cm s-1 at slack water to ﬁgures in excess of 250 cm s-1 at some sites during ﬂood and ebb tides. Good sites should have minimal periods of slack water. It is difﬁcult to recommend speciﬁc maximum current velocities for particular species because of the inﬂuences of cage design and stocking density. Capital costs of cages and moorings must increase with current velocity. While production can be increased at sites with high tidal current velocities (since stocking rates can be increased), beyond a certain point there will be unacceptable reductions in net bag volume

148

Chapter 4

and ﬁsh may have to expend excessive energy to maintain station, adversely affecting production. In practice, ebb and ﬂood tidal currents in the range 0.1–0.6 m s-1 and mean tidal currents of 0.03–0.2 m s-1 have been found to be satisfactory; sites where currents exceed 1 m s-1 are not generally recommended (Braaten & Saetre 1973; Chen 1979; Chua & Teng 1980; Kerr et al. 1980; PPD 1986; Ikenoue & Kafuku 1988; Rudi & Dragsund 1993; Turner 2000b). Inland waters have currents that differ both in origin and pattern. In lotic water bodies (i.e. ﬂowing water systems, such as streams and rivers) water is derived from three principal sources; surface runoff, through-ﬂow and groundwater, and is subject to two major forces, gravity and friction. As water ﬂows along a channel there is a loss of energy through frictional forces exerted by the banks and the stream bed. The simplest, and best known equation which describes this relationship is the Manning equation (Manning 1891): V =

1 23 12 R S n

where V = velocity (m s-1); n = roughness coefﬁcient; R = hydraulic radius (crosssectional area divided by wetted perimeter, m); and S = slope. The radius, R, and slope, S, can be obtained from ﬁeld survey measurements. Roughness, n, is dependent upon the size and shape of the grains forming the wetted perimeter of the stream, and other characteristics such as the quantity and type of vegetation present. Although there are exceptions, such as when sand is formed into dunes or ripples, in general the smaller the grain size, the lower the n value will be. Most streams have values of between 0.025 and 0.075 although tables of roughness values can be obtained from many standard hydraulics textbooks (e.g. Chow 1959; Richards 1982). Lotic systems differ markedly in their characteristics depending on how and where they were formed and on whether or not they have been modiﬁed by man. Upper sections characteristically have high current velocities, low ﬂows and turbulent ﬂow patterns whereas lower sections tend to have low current velocities, high ﬂows and laminar ﬂow patterns. As discussed earlier, cage culture in lotic systems is principally practised in South and Southeast Asian countries where ﬁsh are reared in traditional extensive/semi-intensive situations. In Cambodia, slow-ﬂowing reaches of the Mekong and Tonle Sap Rivers have cages (Lafont & Savoeun 1951; Pantulu 1979) while in Indonesia, rivers and irrigation canals, which for much of the year are also fairly slow ﬂowing, are used (see Fig. 2.2). In a study of traditional cage aquaculture practices in Bandung, Java, Vaas & Sachlan (1957) found that both branches of the River Tjibunut that were being used to culture carp were about 5 m broad by about 0.35 m deep with typical current velocities of around 10 cm s-1, although, as in many other tropical rivers, current velocities greatly exceeded this during monsoon ﬂoods. However, the cages can withstand such conditions because of their construction and method of anchorage. While slow ﬂowing irrigation canals in the United States are increasingly being considered for cage aquaculture, there is also some concern because of effects on ﬂow rates and on risks of ﬂooding (Little & Muir 1987; Costa-Pierce & Effendi 1988; Budhabhatti & Maughan 1993; Haylor 1993) (see also section 4.3.1 below).

Site Selection

149

Current patterns differ between inland and marine sites. Instead of four or eight periods of fast-ﬂowing currents per day associated with peak ebb and ﬂood conditions, currents in inland waters are more constant. High rates of water exchange through cages may be tolerable in marine situations because feeding during peak ﬂow conditions can be suspended. However, the relatively constant ﬂow of water through cages in inland sites means that feeding regimes cannot be adjusted in this way. Feed loss is a problem in lotic sites and is one of the main reasons why commercial intensive cage culture in fast-ﬂowing rivers or streams is rarely viable. Current velocities in thermal efﬂuent canals can also be highly variable. For example, at La Casella oil-ﬁred power station on the Po River, ﬂows in the discharge canal can vary from 1 m s-1 to nearly zero, depending upon the water level in the river (Bronzi & Ghittino 1981). At the Kozienice power station, Poland, carp reared for 1 year in cages in thermal efﬂuent canals suffered severe body ﬂexures which Backiel et al. (1984) attributed to currents of 2.6–12.3 cm s-1. Surface current velocities in lakes and reservoirs are usually much lower (0.2–2.0 cm s-1) than in rivers or marine sites. The currents are principally winddriven, although draw-down from outﬂows and discharges from inﬂowing rivers and streams can have localized effects on velocities while other factors, such as seiches can modify whole lake patterns (Csanady 1975; Smith 1975). Being usually fairly uniform, current velocities in lentic water bodies are rarely of any importance in site selection. The exception is where extensive operations can be sited near outﬂows to take advantage of increased current velocities and supplies of planktonic food (see Fig. 4.18).

Fig. 4.18 Over-crowding of cages at the outﬂow of Lake Buhi, Philippines.

150

4.2.4

Chapter 4

Depth

Some cages, such as the rigid wood and bamboo constructions used in traditional ﬁsh culture in Indonesia and Vietnam, are designed to sit in shallow water in contact with the substrate. Water depths here are not critical provided that cages are covered, or largely covered, by water for most of the culture period. Too great a fall in water level could reduce cage volume, thereby increasing stocking density and adversely affecting water quality. Fixed cages are installed in shallow areas of lakes and reservoirs or rivers where depths do not exceed 8 m or so, since it is difﬁcult to ﬁnd sufﬁciently strong supporting posts which are any greater in length. In theory, depth of water is not important for ﬂoating cages. However, costs and problems associated with mooring increase with depth. For most types of cage culture, cages should be sited in sufﬁcient depth to maximize the exchange of water yet keep the cage bottoms well clear of the substrate. As discussed in Chapter 3, ﬁsh often augment passive, externally generated currents that ﬂow through the cages by drawing in water as a result of swimming and feeding. Water is drawn into the cage not only through the sides but also through the bottom panel (Chacon Torres et al. 1988a), and as the cage bottom approaches the substrate, ﬂows become increasingly impeded. Although the relationships between cage culture, sedimentation of wastes, water quality and disease remain poorly understood, it is best to hold ﬁsh at least 4–5 m above the sediments. This is not always practical, particularly in lotic sites or ponds where cages are being used. However, in freshwater lakes and reservoirs, suitable depths can readily be found using simple survey equipment. A plumb line is ﬁne, although slow, in depths up to 10 m or so (Turner 2000b). Annual water level ﬂuctuations can be considerable, particularly in reservoirs or lakes where water is abstracted for irrigation purposes (Little & Muir 1987; Haylor 1993; Beveridge & Stewart 1998; Beveridge & Muir 1999). In water bodies of this type, ﬁxed cages are inadvisable because although the net bags are lowered to a new position, this is only possible if the drawdown is <4 m per annum or so, given the dimensions of the posts. Floating structures have the advantage in that they can be moved to deeper water if and when necessary. In the marine environment, Turner (2000b) recommends using an echo sounder with a high frequency transducer (~200 Hz), arguing that because it produces a narrow beam it gives more accurate readings, particularly on sloping contours. Tidal ﬂuctuations must be taken into account in any depth calculations. Spring tides of maximum rise and fall occur every two weeks at new and full moons, and can range from little more than 0.5 m in some parts of the world to over 10 m in the Bay of Fundy, Newfoundland. For the ﬁsh farmer’s purposes the tidal range for a particular site can be calculated from nearest-to-site values given in tidal tables augmented by site observations over several tidal cycles. Observations should be made during relatively stable and calm weather as strong offshore or onshore winds can cause correspondingly smaller or larger than expected tides. A difference in barometric pressure of 34 mbar from the average

Site Selection

151

will result in a height difference of 0.3 m (i.e. 10 mm per mbar) (Muir Wood & Fleming 1981).

4.2.5

Substrate

The substrate at sites can range from rocky to soft mud and may inﬂuence the choice of cage design. At freshwater sites, it can prove difﬁcult, if not impossible, to drive supports for ﬁxed cages into hard, rocky ground, even with the assistance of a pile-driving platform (see Fig. 3.25), and ﬂoating cages may have to be used. For marine sites, there are some advantages to choosing sites with rocky substrates since it indicates good current scour with reduced risk of waste build-up, although mooring cages in this type of site can be a problem (see section 3.3).

4.3 4.3.1

SITE FACILITIES AND MANAGEMENT Legal requirements and planning

The law with respect to aquaculture varies enormously from one part of the world to another and, while it is relatively easy to set up a ﬁsh farm in some countries, it is very difﬁcult in others. Most countries have not yet drafted or passed speciﬁc legislation dealing with aquaculture, relying instead on amendments to existing laws. In this section the types of conditions and controls that prospective cage ﬁsh farmers are likely to meet are brieﬂy reviewed. Controls on efﬂuents are discussed in Chapter 5. Comprehensive reviews of aquaculture legislation can be found in Van Houtte et al. (1989), Howarth (1991), Fernandes et al. (2000), McCoy (2000), Uriarte & Basurco (2001) and in various on-line sources of information on the internet (e.g. http://www.intraﬁsh.com/laws-and-regulations/; http://www.oneﬁsh.org/static/index.jsp). Cage ﬁsh farms sit in the water column and are moored to the lake or sea bottom and/or the shore. In some cases, the cages may also be attached to the land above the high water mark via a pontoon. In most countries the land below the low water tidemark is owned by the state and permission to moor a structure must be obtained from the appropriate government department. In the UK, for example, the state owns the sea bed and anyone wishing to build a cage ﬁsh farm must apply for a lease from the District Valuer of the Crown Estate Commissioners. Since most land above the high water mark is privately owned it follows that ﬁsh farmers must buy, lease or obtain permission from the lawful owners if they wish to use the land for access or mooring. Some countries insist that prospective cage ﬁsh farmers obtain a licence and, in so doing, lay down restrictions concerning site, species, size and type of development. Few countries have yet gone so far as to allocate speciﬁc areas for cage culture development, although zoning, in which speciﬁc types of water body are excluded from aquaculture development, does exist in many, especially European, countries (Van Houtte 1993). In Poland, for example, it is forbidden

152

Chapter 4

Fig. 4.19 Map of Laguna de Bay, Philippines, showing legal ﬁsh pen and ﬁsh cage development area and ﬁsh sanctuary (from Beveridge 1984a).

‘to discharge aquaculture efﬂuents into lakes which do not receive efﬂuents for 3 km upstream’. Cage aquaculture is also prohibited in freshwater lakes in South Korea (Kim 2000). By contrast, in Hong Kong, Singapore, the Philippines and, most recently, Australia, speciﬁc areas have been designated for aquaculture development (Tseng 1983; Beveridge 1984a; PPD 1986). In Hong Kong, Singapore and Australia the legislation applies to marine cage farms while in the Philippines laws have been passed in an attempt to control cage and pen farming developments in Laguna de Bay (Beveridge 1984a) (Fig. 4.19). In Norway, the outcome of the LENKA project is a series of production guidelines for planners in which areas of the coastline have been identiﬁed as having high (60 t km-2), medium (30 t km-2) or low (10 t km-2) capacity for salmon farming, based upon a combination of environmental and socioeconomic criteria (Kryvi et al. 1991; Ibrekk et al. 1993). However, with the rapid development of intensive cage culture in many parts of the world, it has become increasingly apparent that certain types of site or speciﬁc areas are particularly sensitive or vulnerable to cage aquaculture developments. Controls on the use of irrigation canals for cage culture, for example, have long been in force in several countries, including Egypt and Indonesia (IDRC/SEAFDEC 1979). In West Java cage ﬁsh farmers using irrigation canals must maintain them free from detritus (IDRC/SEAFDEC 1979). Cages must also

Site Selection

153

Table 4.8 Separation distances for a new salmon cage farm and existing developments in Scotland (modiﬁed from Telfer & Beveridge 2001). Activity

Salmon farm Shellﬁsh farm Public viewpoint Hotel/tourist centre Houses Wildlife colony Anchorage Fishing ground

Distance (km) Fish farm

Shellﬁsh farm

8.1 3.2 1.6 1.6 0.8 0.8 0.4 0.4

3.7 0.8 0.8 0.8 0.4 0.4 0.4 0.4

be sited at least 3 m from any bridge and are not permitted in canals where the gradient is less than 2.5%. Fish farms too are seen as posing a risk to other ﬁsh farms and many countries have set a mandatory minimum distance not only between ﬁsh farms but also between ﬁsh farms and other aquaculture and nonaquaculture developments (Rudi & Dragsund 1993; Telfer & Beveridge 2001; Table 4.8). There may even be a presumption against cage farm development in particularly vulnerable types of site. Production restrictions operate in most of Scandinavia. In Finland at one time even small units producing less than 3 t per annum had to notify the authorities while farms producing more than 40 t required a government concession. Farms larger than 350 t were discouraged. Restrictions also operate with regard to the use of food (Mäkinen 1993). Size regulations operate in Norway, farms being restricted by production volume. Although farm volume has increased over the years, the restrictions encouraged farmers to maximize production per unit volume, exacerbating disease and environmental problems (Rudi & Dragsund 1993). In Germany, growth in the aquaculture industry is presently very restricted as regulations are stringent and new licences in both fresh and marine waters are extremely difﬁcult to obtain (Rosenthal & Hilge 2000; see Chapter 5). Because cage ﬁsh farms are often located in coastal areas or lakes frequented by shipping, local navigation regulations may have to be complied with. In the UK, for example, marine site licences are issued by the Crown Estate Commissioners and applications are forwarded to relevant groups from an indicative list of groups representing stakeholders. Individuals are also invited to respond to public advertisements. In some countries the procedures for setting up a cage ﬁsh farm are straightforward. In Thailand, all that is required is permission from the provincial ﬁsheries ofﬁce and the payment of a nominal licence fee (IDRC/SEAFDEC 1979). In Scotland the procedures remain somewhat more involved. Applications are processed through local planning authorities. In addition to obtaining permission from the legal owners of the sea bed (i.e. the Crown Estate

154

Chapter 4

Fig. 4.20 Regulatory procedure for marine farms in Scotland (modiﬁed from Telfer & Beveridge 2001). CEC = Crown Estates Commissioners; LPA = Local Planning Authority; SEPA = Scottish Environmental Protection Agency.

Commissioners) and from the Department of Transport, coastal cage farms must register with the statutory disease control body (Fisheries Research Services) and obtain a consent to discharge efﬂuents from the environmental protection agency (Scottish Environmental Protection Agency) (Henderson & Davies 2000; Telfer & Beveridge 2001) (see Fig. 4.20). While there are detailed locational guidelines for the authorization of marine ﬁsh farms in Scottish coastal waters (FRS 2002a; Gillibrand et al. 2002), the entire regulatory framework for aquaculture in the country is under review.

4.3.2

Situation, services and shore facilities

For large and intensive cage ﬁsh culture developments the availability of sufﬁcient land to construct an ofﬁce, feed store, laboratory, manager’s house, etc., close to ﬁsh cages may be an important consideration in site selection. However, if the desired land site is in a conservation area or if there are common grazings or rights of way, then there may be problems in obtaining planning permission. Services are also important. Fresh water, three-phase electricity, telephone and postal services, scavenging and sewerage, road and rail services, and veterinary

Site Selection

155

assistance must all be assessed prior to leasing or purchasing a site. Availability of local labour and housing should also be a consideration. Proximity to markets and food supplies may affect production costs and proﬁtability.

4.3.3

Security

Security is a problem for cage ﬁsh farmers in many parts of the world (Edwards 1978; Secretan 1980; Beveridge & Muir 1982; McAndrew et al. 2000) since cages are often sited in water bodies that are publicly owned or have unrestricted access and are thus vulnerable to poachers and vandals. Cage aquaculture is also often seen as a new venture which conﬂicts with traditional industries of tourism, ﬁshing and angling, and is detrimental to infrastructure such as roads. Local people can feel further disenfranchised if the company and staff come from outside the area and speak a different language or hold different values. For these reasons, public opinion in some areas of Ireland regarding salmon cage farming, for example, is very much divided. Sensitive development and an awareness of local feelings can greatly help dissipate local concerns (see Garvey & Bennett 1991; Barnaby & Adams 2002). Although there are a number of security measures that can be taken to protect installations (see section 7.6), farmers may prefer to site cages where they can be readily observed. A survey of cage ﬁsh farmers in the Philippines has shown that this is a prime consideration in site selection (Escover & Claveria 1985), and that farmers are unwilling to locate cages where they cannot be checked daily.

4.4

CONCLUDING REMARKS

Site selection criteria can be categorized as those that are important for the species under consideration for culture to thrive, those that determine the suitability of the site with respect to the type of cage being considered while the third category is concerned with the factors that determine the possibility of establishing a proﬁtable business (see Table 4.1). A site selection scheme is shown in Fig. 4.21. In their study on the inﬂuence of site selection criteria on capital costs for brackish water ponds Muir & Kapetsky (1988) observed that ‘there have been few attempts to deﬁne the relationships between site factors, development, capital and operating cost’. This certainly applies to cage aquaculture. Is ideal depth more important than proximity to a shore base, for example, and what effects might currents of 0.3 m s-1 have on production costs compared with currents of 0.5 m s-1? Gasca-Leyva et al. (2002) used a bioeconomic modelling approach to investigate the proﬁtability of sea bream cage culture in the Mediterranean and the Canary Islands. Kite-Powell & Hoagland (2001) further developed this approach to site selection for marine ﬁnﬁsh in New England. Nevertheless, while checklists of parameters and scoring systems, such as those of Black & Truscott (1994) (Table 4.9) and Huguenin (1997), can help guide

Fig. 4.21 A decision-making scheme for cage site selection (modiﬁed from Rudi & Dragsund 1993).

Table 4.9 Marine ﬁsh farm area evaluation on the basis of score values assigned for 12 environmental factors (from Black & Truscott 1994). Area description: Grappler Sound, Hopetoun & Carridean Bay. Analysts: Richler, McDonald, de Lange Boom. Ranking = medium for Grappler Sound with caution for low oxygen levels. Ranking = not available at Hopetoun Indian Village, Watson Island. Ranking = not available at shallow Carridean Bay. Factor

Possible score

Actual score

Temperature Salinity

10 10

5 5

Dissolved oxygen

10

4

Plankton Currents

10 8

5 4

Plants/fouling organisms/ herring spawn

8

2

Hydrology/slope stability/river debris Depth/substrate geology Wind and waves Climate/rain/temperature/ freezing/icing Predators Pollution/tidal debris

8

7

5

3

5 5

5 3

2 10

0 7

Total score

50

Comments

No data; warm summer No data; some dilution expected during high runoff No data; autumn low (?) in Queen Charlotte Strait Found in Carridean Bay (species?) No data; 50–200 cm s-1 at narrows, lower ﬂows in most of area Sometimes important to herring spawn; kelp to east of Hopetown in narrow, shallow channel Some bays in area of steep topography Shallow bays not available; rest variable depths; mud bottom Calm >30 cm max snowfall (24 h); no freeze ups reported Raptors; two sea lion haul-outs Small anchorages; two Indian villages camp; possible bark

Site Selection

157

prospective cage farmers, for all but a handful of situations the relationships between site characteristics and proﬁtability remain poorly understood. A useful tool both for site selection and with which to explore these relationships is a geographical information system (GIS). A GIS is a database management system that allows users to store, retrieve and manipulate data, integrated with a series of routines that allow sophisticated spatial analysis and display

Fig. 4.22 Layers used in GIS site assessment for coastal aquaculture. (a) General GIS approach for siting salmon cages showing map layers; (b) GIS approach used to site salmonid cages at Camas Bruaich Bay, Scotland (from Beveridge et al. 1994c).

(a) Fig. 4.23 (a) Optimum area selected for cage culture in Camas Bruaich Bay, Scotland, based on scoring of data in source layers in Fig. 4.22b. (b) Optimum area when dispersion of solid wastes from site is taken into account. (From Ross et al. 1993; Beveridge et al. 1994c.)

158

Chapter 4

(b) Fig. 4.23 Continued.

(Burroughs 1986). GISs have begun to be used in aquaculture to determine siting at national, regional and local scales (Meaden & Kapetsky 1991; Beveridge et al. 1994c; Nath et al. 2000; Pérez et al. 2002, 2003a,b). A GIS permits rapid manipulation of scoring to evaluate effects on outcome and can incorporate modelling of factors such as waste dispersion (see Pérez et al. 2002, 2003a). Ross et al. (1993) used a simple GIS package to determine siting for salmon cages within a 1-km bay on the west coast of Scotland. Detailed maps, derived from ﬁeldwork, were produced for bathymetry, wave height, water quality (salinity variation) and currents (Fig. 4.22). Scoring systems were then incorporated to establish the best area within the bay for siting cages (Fig 4.23).

Cage Aquaculture, Third Edition Malcolm C. M. Beveridge Copyright © 1996, 2004 by Blackwell Publishing Ltd

Chapter 5

Environmental Impacts and Environmental Capacity

Aquaculture, like other economic activities, uses and transforms resources into commodities valued by society, in this instance farmed ﬁsh, and, in so doing, produces wastes. Environmental services are also required to remove and assimilate wastes. Impacts stem from these three processes – the consumption of resources, the transformation process itself and the production and assimilation of wastes (Fig. 5.1) – which not only impose costs on society but have implications for the sustainability of the aquaculture venture itself (Beveridge et al. 1994b, 1997a, b; Naylor et al. 1998, 2000; Costa-Pierce 2002). The present chapter explores resource use, the transformation process (cage aquaculture) and the production of wastes and how these impact on the environment, and concludes by considering environmental capacity. Reducing and managing environmental impacts is considered in section 7.9.

5.1 RESOURCE CONSUMPTION Cage aquaculture requires space in the sea or in a lake, materials to construct cages, seed to stock the system and feed to encourage growth. Water is also required, not only to physically support the animals but also to provide oxygen and remove wastes. Labour and energy are considered later.

5.1.1

Space

Cages are usually sited in the busy coastal zone or in multi-purpose lakes and reservoirs. The area they occupy, although not large by comparison with ponds (cages are deeper and employ higher stocking densities) or in terms of a country’s coastline or freshwater resources, can nevertheless be important. By the mid-1980s, for example, pen and cage ﬁsh farming in Laguna de Bay in the Philippines had grown to such an extent that structures occupied 38–45% of the surface area of the lake (Beveridge 1984a) while in nearby Sampaloc Lake tilapia cages at one time covered more than 30% of the surface area (Santiago 1994). Space occupied by cages may otherwise serve other purposes. If the allocation of space has not been fair, or is perceived as such, social tensions can develop (Beveridge & Stewart 1998; McAndrew et al. 2000, 2002; Hambrey et al. 2001a, 159

160

Chapter 5

Capital Labour Energy Resources • water • construction materials • seed • feed

aquaculture process

Environmental services

Environmental impacts • resource depletion • habitat loss • food web changes • reduced biodiversity • disease

Goods • fish • shellfish

Consumer

Wastes • uneaten food • faecal / urinary products • chemicals • pathogens • escaped animals

Fig. 5.1 Diagram summarizing the relationships between the aquaculture process, inputs of resources and labour, and outputs (farmed products and wastes) (modiﬁed from Beveridge et al. 1994b).

Hambrey 2002). This is apparent in the UK where conﬂicts between anglers and cage ﬁsh farmers sometimes arise. Unregulated development of the ﬁsh pen and cage industry in Laguna de Bay, Philippines, in the 1980s impeded access to ﬁshing grounds and boat trafﬁc to such an extent that it led to violence (Beveridge 1984a). It is not just the space occupied by cages that must be considered. An insensitively sited and poorly managed cage farm can adversely affect landscape values; indeed, visual impact is one of the most important causes of public concern about cage farm developments in Europe, Chile and North America (EDAW & CH2M/Hill 1986; Cobham Resource Consultants 1987; SWCL 1988; Uriarte & Basurco 2001) and has been an important factor in the development of aquaculture and tourist facilities in new reservoirs in Malaysia. Such problems should, of course, be identiﬁed and resolved at the planning stage; in practice, however, this is often not the case.

5.1.2

Construction materials

Materials are required to construct the cage collar and cage bag. In developing countries cages tend to be made from cheap, locally available materials such as oil drums and scrap wood while softwoods, galvanized steel, plastic and aluminum are used in the intensive cage aquaculture industries of Europe and elsewhere. Some materials are derived from non-renewable resources, or have implicit high energy costs. In Vietnam, demands for hardwood for construction of ﬂoating cages for catﬁsh rearing on the Mekong River have been blamed for fuelling illegal felling of trees in neighbouring Cambodia. However, cage farming is not usually of any signiﬁcance in this regard.

Environmental Impacts and Environmental Capacity

5.1.3

161

Seed

Seed is required for all aquaculture operations including cage ﬁsh farming. By contrast with freshwater species, the culture of marine, especially tropical, ﬁsh often requires wild seed, spawning and larval rearing of groupers, siganids, etc., proving either technically difﬁcult or uneconomic (Beveridge et al. 1997a; Beveridge & Haylor 1998; Beveridge 2001; Hair et al. 2002). Although spawning of milkﬁsh in captivity was ﬁrst carried out successfully 25 years ago, the Indonesian and Philippine industries are still overwhelmingly dependent upon some 2 billion wild fry each year. Constraints to development of a hatchery industry include the long (5–9 years) and expensive broodstock rearing period and the poor quality of hatchery reared seed. In the Mediterranean and elsewhere the culture of yellowtail (Seriola spp.) and tuna (e.g. Thunnus thynnus) remains dependent upon the capture of wild juveniles (Beveridge 2001; Hair et al. 2002), largely for technical reasons. Although the numbers captured to support a few hundred tonnes of tropical cage ﬁsh production are small, the use of wild seed in aquaculture is undesirable. Not only is the method of seed collection and transport highly wasteful, large numbers of animals often dying or being discarded in the process, it also results in the farming of strains that have undergone no selection for aquaculture conditions (Beveridge et al. 1997a).

5.1.4

Feed

The bulk of world ﬁsh culture production is of carp and tilapias grown in the tropics and sub-tropics. A small component of ﬁsh production, an estimated 6%, comes from intensive cage culture. Nevertheless, Atlantic salmon farming in particular is becoming an important consumer of ﬁshmeal. By the turn of the millennium, world production of Atlantic salmon was around 1 million t. If an average food conversion ratio (FCR) of 1.3 : 1 is assumed, then 1.3 million t of salmon food are needed to sustain the industry (see also Pike & Barlow 2003). Since the ﬁshmeal component of salmon diets is presently around 35% and 3.5 t of ﬁsh are required to produce 1 t of ﬁshmeal, then 1.6 million t of industrial ﬁsh were being used to support the industry, equivalent to some 25% of global ﬁshmeal supplies (6 million t) and around 10% of total capture ﬁsheries production. Assuming a 28% dietary inclusion rate at present, a further 365 000 t of ﬁsh oil is required, equivalent to ~30% of the 1.2 million t produced annually from capture ﬁsheries. These ﬁgures relate to one sector only: cage farming of Atlantic salmon. Other aquaculture sectors, especially marine ﬁnﬁsh farming, are increasingly reliant on ﬁshmeal and ﬁsh oil, and despite endeavours to ﬁnd replacements there remain not only many technical, but also farmed product quality and marketing, issues to resolve (Pike & Barlow 2003). In the meantime, the dependency on capture ﬁshery products is, if anything, increasing. Growth in production of ﬁsh at the top of the food chain is rapid, and while ﬁsh that feed lower in the food web, such as carp and tilapias, can be produced by less intensive methods, the increasing pressures on land and water in countries such as China are forcing farmers

Fig. 5.2 Input of natural capital and energy ﬂows to intensive cage farming (top) and semi-intensive pond farming (bottom) in Lake Kariba, Zimbabwe. Estimates of energy ﬂows (MJ) and material ﬂows (kg) are based on the production of 1 kg of ﬁsh (from Berg et al. 1996).

162 Chapter 5

Environmental Impacts and Environmental Capacity

163

to produce more ﬁsh per unit of land or water (see also Folke 1988; Folke & Kautsky 1992; Chamberlain 1993; Beveridge et al. 1994b, 1997a; Baird et al. 1996; Naylor et al. 1998, 2000; Hardy & Tacon 2002). Pike & Barlow (2003) argue that the research into dietary feed formulations will result in some reduction in ﬁshmeal and ﬁsh oil dependency. They also argue that despite the growth and intensiﬁcation of aquaculture, industrial ﬁsheries landings have remained constant and that the increase in supply has come from substitutions in other sectors that have traditionally used these products. They further argue that industrial ﬁsheries are very effectively managed and policed. Hence, despite any increase in demands from intensive cage aquaculture, there is likely to be little adverse impact on stocks. Extensive cage culture of tilapias in Philippine lakes has caused over-grazing with the result that poorer farmers have been forced out of production while those remaining have become increasingly reliant on supplemental food (Beveridge 1984a; Santiago 1994).

5.1.5

Conclusions

From one perspective cage aquaculture is not a major drain on ﬁnite resources except, perhaps, with regard to ﬁsheries products. Intensive cage aquaculture is an important consumer of ﬁshmeal and ﬁsh oil. If demands from the aquaculture industry continue to increase – and they look like doing so, in the short term at least – the industry must accept its share of the blame for any resultant over-exploitation of industrial ﬁsh stocks. The ecological footprint (i.e. the area of land and water required to provide resources and services) or energy ﬂow concepts can be helpful in assessing impacts of resource use (Kautsky et al. 1997). Berg et al. (1996) compared inputs of natural capital and energy ﬂow for intensive cage culture of tilapias on Lake Kariba and semi-intensive pond culture carried out nearby. Although the former used less energy, the latter proved much more efﬁcient in terms of energy transfer (Fig. 5.2). However, it is difﬁcult to say whether cage aquaculture in general is more resource hungry than other forms of aquaculture without a comprehensive analysis of land-based and water-based aquaculture production.

5.2

THE CAGE AQUACULTURE PROCESS

The establishment of a cage ﬁsh farm increases activity and noise levels, not only at the site but also along sea and land routes that service the farm. As a result, adverse impacts on wildlife, especially in remote areas, can occur. Breeding or foraging activities may suffer and animals may be displaced (NCC 1989, 1990; Beveridge 2001). High densities of caged ﬁsh and food attract predators and opportunistic species, which may displace residents (Phillips et al. 1985a; Beveridge 1988; NCC 1989; Carss 1990; Furness 1996; Beveridge 2001). Increased conﬂicts, especially with piscivores, are also likely to occur and the end result may be the death of wildlife, either deliberate or accidental (Beveridge 1988, 2001; Ross 1988;

164

Chapter 5

Rueggeberg & Booth 1989; Carss 1993a, b, 1994; Pemberton & Shaunhnessey 1993; Furness 1996; Quick et al. 2004). Codes of practice have been introduced in a number of cage farming industries in order to minimize conﬂicts. Few quantitative studies of impacts of cage aquaculture on wildlife have been carried out. In studies conducted during the late 1980s it was estimated that 200–800 herons (Ardea cinerea), 1650–2050 cormorants (Phalacrocorax carbo) and 940–2880 shags (P. aristotelis) were being killed each year at Scottish ﬁsh farms (Ross 1988; Carss 1994; Furness 1996). Although persecution may have been largely of birds ﬂedged that year and that were surplus to carrying capacity (NCC 1990), effects at the population level are unknown. While impacts on wildlife are unlikely to be important on a global scale, they may be of local signiﬁcance. Predator-related problems are considered further in Chapter 7.

5.3

WASTES

Aquaculture wastes include uneaten food, faecal and urinary products as well as chemicals, microorganisms, parasites and feral animals.

5.3.1

Uneaten food, faecal and urinary products

The origin of wastes A proportion of food thrown into a cage of ﬁsh is not eaten. Ingestion is dependent upon a sequence of events in which ﬁsh must ﬁrst recognize that there is food present (Fig. 5.3). They must then be able to reach the food (strong currents, for example, may wash pellets out of the cage before they can be ingested) and be motivated (appetite, appearance) to take it into their mouths. Even at this stage, a food pellet may be rejected rather than swallowed if it feels wrong or is contaminated (for reviews see Thorpe & Huntingford 1992; Smith et al. 1995; Beveridge & Kadri 2000). As ingested material passes through the gut it is attacked by enzymes, the products of digestion being absorbed into the bloodstream and the undigested fraction being voided as faeces. Metabolic breakdown products such as CO2 and NH4 and excess nutrients are passed out across the gills and in the urine. In addition, mucus and sloughed scales from caged ﬁsh, fouling organisms that have either become dislodged or have been discarded as a result of in situ net cleaning, mortalities and blood from harvesting operation, may be released into the environment. Although not considered further, the amounts can be substantial. For example, it has been estimated that 1.8 t of fouling material is produced per tonne of ﬁsh production in Hong Kong, equivalent to 31 kg biological oxygen demand (BOD), 7.5 kg N and 70 g P (Wu et al. 1994). Quantifying wastes There are two methods for estimating material lost to the environment: direct, through sampling and analysis of the water column and of sedimenting partic-

Environmental Impacts and Environmental Capacity

165

Fig. 5.3 Sequence of events associated with feeding in caged ﬁsh.

ulate material; and indirect, using a mass balance approach. The former is difﬁcult; it is only practical for estimating uneaten food and faecal solids, and usually involves the suspension of traps below cages. Hydroacoustic and video techniques have been used to determine food losses (Juell 1991; Blyth et al. 1993; Mayer & McLean 1995). Losses associated with use of pelleted feeds in the range 1–30% have been reported, but on caged salmon farms today are estimated to be typically around 3–5% (Gowen & Bradbury 1987; Beveridge et al. 1991; Findlay & Watling 1994; Brooks et al. 2002). Losses associated with use of trash ﬁsh, however, can be as high as 40% food fed (Wu et al. 1994; Leung et al. 1999). A mass balance approach is useful, particularly when used in conjunction with ﬁeld and laboratory data. Uneaten food, faecal and excretory losses to the environment can be estimated using data on feed quantities and quality, food conversion ratios (FCR), digestibilities and faecal composition. Mass balance equations for various wastes, such as nitrogen, carbon or phosphorus, can then be derived. Consider the following as an example. The nitrogen content of tilapia is around 3% (Meske & Manthey 1983) while that of a particular food is 8%. Assuming an FCR of 1.6 : 1, approximately 98 kg of nitrogen is released into the environment for every tonne of tilapia produced, i.e.

(1600 ¥ 0.08) - (1000 ¥ 0.03) By incorporating other data and a number of assumptions, a more detailed picture emerges. If it is assumed that 20% of food is uneaten then 102.4 kg of

166

Chapter 5

Fig. 5.4 Mass balance of nitrogen for 1 t of intensive cage tilapia production. Figures are in kg (see text) (from Beveridge & Phillips 1993).

nitrogen is ingested. According to Beveridge et al. (1991) 360 g of faeces with a nitrogen content of 4% is produced per kg food consumed. Thus, 18.4 kg of nitrogen is voided as faeces, 30 kg is harvested as ﬁsh ﬂesh and the remainder, 54 kg, is excreted as urinary ammonia and urea (see Fig. 5.4). Mass balance approaches to waste estimation have been widely used by Beveridge (1984a), Gowen & Bradbury (1987), NCC (1989, 1990), Beveridge et al. (1991), Barg (1992), Beveridge & Phillips (1993), Angel et al. (1992), Leung et al. (1999), Pitta et al. (1999) and Pearson & Black (2001) among others. While estimates for any particular species tend to vary among studies, this is largely because of differences in feed formulations and FCR values (Gavine et al. 1995; see also below). The mass balance approach provides insights into why and where wastes occur. However, it also has several deﬁciencies. It provides no information on the material that enters the environment, its fate or impact: how much of the solids, for example, remain as solids and how much dissolves? There is little information on the form of the wastes. For example, how much of the nitrogenous wastes is organic and how much is inorganic, and what proportions of the latter are in the form of nitrates, nitrites or ammonia? Equally important, little is known about the bioavailability of the various waste fractions. In addition to waste food, faecal and excretory organic matter, carbon, nitrogen and phosphorus that can stimulate aquatic production, there is also some concern about the fate of zinc and contaminants such as PCBs and dioxins that can be present in farmed ﬁsh foods. These are considered in section 5.3.2 below.

Environmental Impacts and Environmental Capacity

167

Impacts on the water column, plankton and nekton In many countries there are restrictions on the discharge of ﬁsh farming wastes into the environment. Some countries, however, go further, prohibiting the use of cages on certain types of water body (e.g. drinking water supplies, Germany; all lakes, South Korea; see also section 4.3). Excretory products are dispersed in the water column by currents while solids (uneaten food, faeces) tend to settle towards the sea or lake bottom. During sedimentation, some of the uneaten food is consumed by ﬁsh (Phillips et al. 1985a; Carss 1990; Johansson et al. 1998) while some breaks down into ﬁne particles (Chen et al. 1999a, 2000; Sutherland et al. 2001; Stewart & Grant 2002). Nutrients are solubilized, the quantities released depending upon the composition of faeces and uneaten food, physical properties, temperature, depth of water and turbulence (Phillips et al. 1993; Chen et al. 1999a, 2003). Nutrients are also released from sedimented solids (Enell & Lof 1985; Hall et al. 1990, 1992; Holby & Hall 1992, 1994; Kelly 1992; Chen et al. 1999b; Heilskov & Holmer 2001) and it has been estimated that as much as 60% of total phosphorus and 80% of total nitrogen wastes end up in the water column (Pettersson 1988; Wallin & Håkanson 1991a, b; Hall et al. 1992; Holby & Hall 1992). Hypernutriﬁcation (increases in dissolved nutrients levels) is often apparent at freshwater cage sites where currents are low and where dilution is limited (Eley et al. 1972; Kilambi et al. 1976; Loyacano & Smith 1976; Korycka & Zdanowski 1980; Enell 1982; Penczak et al. 1982; Beveridge 1984a; Costa-Pierce & Soemarwoto 1990; Cornel & Whoriskey 1993; Costa-Pierce 1996; Yokom et al. 1997; Diaz et al. 2001; Kelly & Elberizon 2001). Changes also may be apparent in DO, BOD, chemical oxygen demand (COD), turbidity and Secchi disc depth. At marine sites, where dilution is much more rapid, effects are often transitory and only apparent during slack tides when temporary elevations in ammonia levels or decreases in DO levels can be detected (Fig. 5.5) (Kadowaki et al. 1978; Gowen et al. 1989; Aure & Stigebrandt 1990; Gowen 1990; Weston 1991; Pitta et al. 1999; Karakassis 2001; Pearson & Black 2001; Brooks et al. 2002). Nevertheless, Wallin & Håkanson (1991a, b), studying impacts of Swedish and Finnish coastal cage ﬁsh farming, found strong correlations between ﬁsh farm loadings and dissolved nutrient levels, especially between total nitrogen and total phosphorus loadings from farms and inorganic nitrogen and phosphorus concentrations in surface waters. Eutrophication, as indicated by increases in plankton and ﬁsh standing crop or productivity, is readily apparent in many fresh waters used for cage aquaculture (Eley et al. 1972; Kilambi et al. 1976; Loyacano & Smith 1976; Korycka & Zdanowski 1980; Enell 1982; Penczak et al. 1982; Beveridge 1984a; Costa-Pierce & Soemarwoto 1990; NCC 1990; Cornel & Whoriskey 1993; Costa-Pierce 1996; Kelly & Elberizon 2001). Changes in plankton and nekton community structure and function may also result, for example, in shifts in phytoplankton community structure towards cyanobacteria and domination of the zooplankton community by Daphnia spp. (Stirling & Dey 1990; Cornel & Whoriskey 1993; Yokom et al. 1997; Kelly & Elberizon 2001). The degree of

168

Chapter 5

Fig. 5.5 The spatial distribution of (a) ammonium (mg-atoms m-3) and (b) chlorophyll (mg m-3) in Loch Spelve, a fjordic Scottish sea loch (from Gowen et al. 1989).

eutrophication, as will be discussed below, is dependent upon the characteristics of the water body and the size, nature and management of the cage operation. Many studies have failed to ﬁnd any inﬂuence on productivity in marine waters (Müller-Haekel 1986; NCC 1989; Aure & Stigebrandt 1990; Gowen 1990; Weston 1991; Pitta et al. 1999; Arzul et al. 2001; Karakassis 2001; Pearson & Black 2001; Brooks et al. 2002) while others have found only weak relationships between nutrient loadings and chlorophyll a (e.g. Wallin & Håkanson 1991a, b). Given the degree of water movement and ﬂushing at most sites, this is not surprising. Highly enclosed, poorly managed sites can show signs of eutrophication. In Hong Kong, a measurable DO sag is apparent up to 1 km

Environmental Impacts and Environmental Capacity

169

from the cages, although changes in suspended solids and chlorophyll a levels are insigniﬁcant (Wu et al. 1994). Studies conducted in the Baltic, a highly brackish area with limited currents and water exchange, show enhanced growth and production of macroalgae and periphyton, and changes in ﬁsh community structure and function in the vicinity of the cages, depending upon the size of the farm and prevailing environmental conditions (Koivisto & Blomqvist 1988; Ruokolahti 1988; Henriksson 1991; Rönnberg et al. 1992; Nordvarg & Johansson 2002). Impacts on sediments and sediment communities Cage ﬁsh farming does not always result in changes in sediment chemistry or in macrobenthic community ecology (e.g. Cornel & Whoriskey 1993; Johannessen et al. 1994), the degree of nutrient enrichment depending upon species being farmed, food, management, currents and depth. For organic carbon, for example, Hargrave (1994) cites a 500-fold range in sedimentation rates under salmon cages. Effects of solids loadings, however, are apparent at many marine and freshwater sites (see Fig. 5.7). Faeces and waste food, especially from intensively managed operations, have much higher levels of carbon, nitrogen and phosphorus than sediments. The result is that sediments below and in the immediate vicinity of cages have elevated levels of organic matter and nutrients (Eley et al. 1972; Kilambi et al. 1976; Korzeniewski & Korzeniewska, 1982; Trojanowski et al. 1982; Tsutsumi & Kikuchi 1983; Merican & Phillips 1985; Gowen 1990; Hall et al. 1990; Kupka-Hansen et al. 1991; Laurén-Määttä et al. 1991; Ye et al. 1991; Angel et al. 1992; Kelly 1992; Cornel & Whoriskey 1993; Johnsen et al. 1993; Wu et al. 1994; Berg et al. 1996; Karakassis et al. 1998, 1999; Pearson & Black 2001; Brooks et al. 2002). Sedimented food and faeces stimulates microbial production, changing sediment chemistry, and structure and function (Enell & Lof 1985; Kaspar et al. 1988; Hall et al. 1990, 1992; Kupka-Hansen et al. 1991; Holby & Hall 1992, 1994; Kelly 1992; Sowles et al. 1994; Berg et al. 1996; Findlay & Watling 1997; McCaig et al. 1999; Pearson & Black 2001; Brooks et al. 2002). Changes are positively correlated with waste loadings and accumulation: oxygen demand increases and sediments become increasingly anaerobic and reduced (Fig. 5.6), there is an increase in the release of nitrogen and phosphorus compounds into overlying waters and in marine environments methanogenesis and evolution of hydrogen sulphide increase. Effects of cage farming wastes on macrobenthic communities in fresh waters have been discussed by Kilambi et al. (1976), Phillips (1985), Dobrowolski (1987), NCC (1990), Cornel & Whoriskey (1993), Costa-Pierce (1996), Troell & Berg (1997) and Kelly & Elberizon (2001). Again, changes are highly correlated with accumulation of sedimented waste food and faeces. There tends to be a positive correlation between sedimented material and biomass of macrobenthos, and a negative correlation between organic matter and diversity. Heavily impacted sediments are dominated by pollution-tolerant species such as oligochaetes and certain species of chironomid larvae, while less tolerant taxa, such as Ephemeroptera, disappear.

170

Chapter 5

Fig. 5.6 Reduction–oxidation (redox) proﬁles (mV) of sediments: (a) at a heavily fouled site – note rapid fall in redox potential with depth and strongly negative values, although sample sites 25 m from the cages are identical to control; (b) at a less heavily fouled site – note similarity between sample and control sites (courtesy Stirling Environmental Services).

Marine benthic communities show similar patterns of response, again at most but not all sites. Immediately under the cages at heavily impacted sites an azoic zone, devoid of oxygen and macrobenthos, may be apparent. Here, surface mats of the sulphur bacteria Beggiatoa can occur (Fig. 5.7a). Beyond this is an area of organic enrichment where exceptionally high numbers of opportunistic species (e.g. polychaetes Capitella, Scolelepis) occur (Fig. 5.7b). Biodiversity in this zone is characteristically low. The area of benthos affected by cage farming typically extends for 20–50 m beyond the cages, although in some sites, because of poor management or unusual hydrographic conditions, effects are evident up to 150 m from the cages (Brown et al. 1987; Nakao et al. 1989; NCC 1989, 1990;

Environmental Impacts and Environmental Capacity

171

(a)

(b) Fig. 5.7 Appearance of sandy sediments (a) immediately under Atlantic salmon cages – note uneaten food pellets and presence of Beggiatoa; (b) 15 m away from cages; (c) control site 150 m from cages. (Courtesy Stirling Environmental Services.)

172

Chapter 5

(c) Fig. 5.7 Continued.

Ritz et al. 1989; Gowen et al. 1991; Kupka-Hansen 1991; Laurén-Määttä et al. 1991; Tsutsumi et al. 1991; Weston 1991; Angel et al. 1992; Hargrave 1994; Johannessen et al. 1994; Wu et al. 1994; Beveridge et al. 1997b; Karakassis et al. 1998, 1999; Karakassis 2001; Pearson & Black 2001; Brooks et al. 2002) (see also below). It is not only the benthic infaunal communities that may be impacted, of course. In the Mediterranean, there has been particular concern about seagrass (Posidonia) meadows, already in decline in many areas due to pollution. Cage ﬁsh farm wastes have been shown to adversely affect distribution and production, both during and after farming (Delgado et al. 1999; Ruiz 2001).

5.3.2

Chemical wastes

Through the use of compound feeds, it is possible for chemicals such as pesticides and dioxins to be inadvertently introduced into the farmed aquatic food web, not only accumulating in the farmed product but also entering the environment in waste food, faeces or excreta (Karl et al. 2002). Minerals and pigments are incorporated into feeds and a wide range of chemicals is also employed in, and applied to, cage construction materials (stabilizers, pigments, plasticizers, ultraviolet (uv) absorbents, antifoulants). Disinfectants are used to control the spread of pathogens and chemotherapeutants are used to tackle disease (for reviews see Beveridge 1984a; Beveridge et al. 1991, 1997b; Alderman et al. 1994; Weston 1996; GESAMP 1997; Costello et al. 2001; Haya et al. 2001; Zitko 2001; Brooks et al. 2002; Easton et al. 2002).

Environmental Impacts and Environmental Capacity

173

While most are used in extremely small quantities, traces nevertheless have been found in ﬁsh farm sediments and occasionally in farmed ﬁsh (Easton et al. 2002). Some compounds have no known toxic effects while others have, and, as a result, depending upon country, use and classiﬁcation (e.g. biocide or medicine), may be subject to controls to protect both environmental and human health. A number of reviews, however, have concluded that there remains insufﬁcient information on the use and impact of some aquaculture chemicals (Alderman et al. 1994; Costello et al. 2001; Zitko 2001). Zinc Zinc is essential for ﬁsh growth and, in addition to being present in relatively high quantities in ﬁsh meal, may be added to salmonid diets in trace amounts equivalent to 30–100 mg kg-1 of feed (Brooks et al. 2002). Various sediment threshold limits, beyond which zinc is known to have important adverse effects on the biological structure and function of the sediment community, have been established. Brooks and colleagues examined nearly 200 sediment samples from 27 farm sites in British Columbia, Canada, and found 9% of samples from eight farms that exceeded the overall apparent effects threshold of 260 mg Zn g-1 dry sediment (Brooks et al. 2002). However, all samples were associated with sediments containing high sulphide concentrations, binding the zinc and rendering it largely biologically unavailable. Feed manufacturers have also since modiﬁed the quantities and form of dietary zinc. In Europe, zinc is listed under the EU Dangerous Substances legislation and its release into the environment requires control (Costello et al. 2001). Chemotherapeutants Chemotherapeutants comprise a range of antibacterial, antifungal and antiparasite compounds (Table 5.1). Use is largely determined by species, pathogens and the part of the world in which the farm is operating. Few antimicrobials are permitted for use in North America, for example, compared to Japan (Schnick 1991;

Table 5.1 Chemotherapeutants used in mariculture. Antibacterial agents

Fungicides

Parasiticides

Natural antibiotics Tetracyclines Macrolide antibiotics b-lactams Aminoglycosides Phenicols Synthetic antibiotics Sulphonamides Potentiated sulphonamides Quinolones Nitrofurans

Malachite green Formalin

Hydrogen peroxide Organophosphates Avermectins (incl. semi-synthetic compounds) Pyrethroids (incl. synthetic compounds) Teﬂubenzuron Benzoylphenylureas

174

Chapter 5

Fig. 5.8 Changes in antibacterial use in relation to various disease outbreaks and vaccine developments in Norway. Data from Norsk Medisinaldepot (from Beveridge et al. 1997; Costa-Pierce & Bridger 2002).

Weston 1996; Jahncke & Schwarz 2002). Quantities also vary with intensity of production, the bulk of those used in ﬁsh farming undoubtedly being at intensive marine cage farming operations in Western Europe, Japan, North America and Chile. The amounts used can be enormous. Norway is the only country that has kept records and, in 1987, the year in which antimicrobial use peaked, some 49 t were introduced into the coastal marine environment by the industry, although by the mid-1990s use had fallen dramatically (Grave et al. 1990; Bangen et al. 1994; Costa-Pierce & Bridger 2002) (Fig. 5.8). Comparatively little use is made of chemotherapeutants in freshwater cage culture, especially in the tropics. Antimicrobials Antimicrobials are administered in the diet and most end up in the environment in association with uneaten food and faeces, the small amounts taken up by the target organisms eventually being metabolized and excreted. Some drugimpregnated food is ingested by scavengers while some chemotherapeutants also diffuse from sedimenting food into the water column. However, most of the chemotherapeutants ﬁnd their way to the sediments. Water-borne antimicrobials are rapidly diluted and some, such as oxytetracycline and furazolidone, are also highly susceptible to photodegradation (Samuelsen 1989; Weston 1996). The

Environmental Impacts and Environmental Capacity

175

sediments, however, can act as a long-term reservoir of drugs and their residues, compounds such as oxytetracycline and the quinolones oxolinic acid and ﬂumequine persisting for many months after chemotherapy (Lunestad 1992; Samuelsen 1992; Coyne et al. 1994; Weston 1996). The bioavailability of many sediment-entrapped chemotherapeutants remains, however, a matter of debate. Impacts of antimicrobial compounds can be summarized as: • • •

effects on non-target organisms; effects on sediment chemistry and processes; the development of resistance.

Many studies have examined the non-target organisms that take up antimicrobials through foraging for food around cages, ﬁlter-feeding on current-borne feed and faecal particles, or through direct absorption from the water. High levels of drugs such as oxolinic acid are detected in wild ﬁsh, crabs and mussels several hundred metres from salmon farms for up to 2 weeks following treatment (Lunestad 1992; Samuelsen et al. 1992; Ervik et al. 1994b) and there is concern that such levels pose risks – toxic, allergic, development of resistance – to humans catching and eating these foods (Yndestad 1992; Jahncke & Schwarz 2002). While laboratory trials involving exposure to high concentrations of compounds have demonstrated antimicrobial inhibition of sulphate reduction and nitriﬁcation (Hansen et al. 1992; Klaver & Matthews 1994) there is little evidence for quantitative and qualitative changes in sediment ﬂora or in the rates of organic matter degradation under ﬁsh cages (Weston 1996). Many studies have reported increases in resistance – and even multiple resistance – in pathogens as a result of the widespread use of antimicrobials by the industry in some parts of the world (Hastings & McKay 1987; Aoki 1992; Richards et al. 1992; Ervik et al. 1994b; Kerry et al. 1994, 1996; Herwig et al. 1997; WHO 1999; Miranda & Zemelman 2002). The signiﬁcance of the ﬁndings remains difﬁcult to assess as little is known about the persistence of resistance. The use of antimicrobials in intensive cage aquaculture in Europe has decreased greatly since the early 1990s through a combination of the use of single generation sites, the adoption of fallowing regimes and the development of vaccines (Wheatley et al. 1995). Chloramphenicol, furazolidone and dimetridazole are now speciﬁcally banned for use on all food-producing animals in Europe and effective vaccines against Aeromonas salmonicida, Vibrio anguillarum and Yersinia ruckeri are now widely available (Costello et al. 2001). The EU has also implemented a programme of testing and monitoring to ensure no harmful residues remain in edible tissues of aquaculture products (WHO 1999). Parasiticides There are concerns about other chemotherapeutants, particularly those used to control parasites, in terms of the quantities used, their toxicity, release into the environment, persistence and impact on both humans and the environment (GESAMP 1997; WHO 1999). The following section focuses on the control of sea lice Lepeophtheirus salmonis and Caligus elongatus on farmed salmonids;

176

Chapter 5

the chemicals used to control ectoparasites in warmwater species is similar, although little has been done to study the impacts. Chemotherapy has been the key strategy in the control of infestations of sea lice on farmed salmon in Scotland and Norway. The range of chemotheraputants currently used is summarized in Table 5.2. Much of what is known about their toxicity and the environmental risks has been discovered by scientists in Scotland, following procedures laid down by the Veterinary Medicines Directorate (VMD) and the Scottish Environmental Protection Agency (SEPA). Until recently, one of the most widely used chemicals has been a proprietary veterinary medicine containing the organophosphorus insecticide dichlorvos, which inhibits acetylcholinesterase activity, leading eventually to paralysis (see Davies et al. 2001). The chemical is used as a bath and is released into the environment following treatment (see section 7.2.4). It is potentially hazardous to economically important non-target organisms such as crab and lobster larvae (Egidius & Moester 1987). Although it has been argued to cause only limited environmental damage due to rapid dilution and breakdown (e.g. Kent & Poppe 2002), use was prohibited in the UK in 2002. Azamethiphos (Salmosan®), another pesticide administered in bath form, is very soluble in water and is not known to accumulate in sediments. It is, however, more toxic to crustaceans (Brooks et al. 2002). In Norway, the synthetic pyrethroid cypermethrin is now the most widely used bath treatment (Boxaspen & Holm 2001). Again, while highly toxic to crustaceans, present evidence is that rapid dilution results in no deleterious effects on non-target animals, resulting in some form of temporary registration in many countries. Many of the remaining parasiticides used to combat sea lice are administered orally, impacts arising primarily from uneaten food and faecal material deposited on sediments adjacent to the cages. Ivermectin, a semi-synthetic avermectin widely used in agriculture to control parasites, is more toxic to crustaceans and polychaetes – especially the pollution tolerant opportunist Capitella capitata – than to molluscs. The compound has been detected in sediments close to and at distances of up to 50 m from farms although, in general, not in biologically signiﬁcant concentrations beyond 20–30 m (Davies et al. 1998; Brooks et al. 2002). Slice®, the proprietory name of emamectin benzoate (Table 5.2), is currently being used at many sites in Scotland, Chile, Ireland and Norway. Evidence to date suggests that it is only around one-tenth as toxic to marine life as ivermectin, and that despite the long half-life (175 days), levels in sediments remain below detection (Brooks et al. 2002). Predicted environmental concentrations in the vicinity of treated farms are also below no effect concentrations for planktonic copepods (Willis & Ling 2003). Calicide® (teﬂubenzuron) is a pesticide licensed for use of sea lice control in Scotland, Chile, Ireland and Norway. It is a chitinase inhibitor, effective on larval stages only. While residues have been measured in sediments up to 1000 m from cage farms, it has been concluded that it is not bioavailable (Brooks et al. 2002). For reasons of costs and efﬁcacy, hydrogen peroxide, reputedly a less environmentally harmful alternative, has not found wide favour (see Kent & Poppe 2002), other than at the few cage farms run on organic farming principles.

Slicea Lepsidon Ektobann, Calicidea

a

Those with marketing authorization in the UK.

Cypermethrin Ivermectin

Emamectin benzoate Diﬂubenzuron Teﬂubenzuron

Synthetic pyrethroid Semi-synthetic avermectin Benzoylphenylureas

Hydrogen peroxide

Aquagard Salmosana Salartecta, Paramovea Excisa Ivomec

Dichlorvos Azamethiphos

Organophosphate

Product name

Chemotheraputant

Compound type

Properties/environmental fate Diluted and dispersed by currents Diluted and dispersed by currents Degrades rapidly to O2 and H2O Strongly adsorbed to soil and sediments Low solubility; strong afﬁnity to lipid, soil and organics; slow degradation Strong afﬁnity to soil and probably sediment Low solubility; sediment Low solubility; sediment

Treatment Bath; 1 mg l-1, 30–60 min Bath; 0.1–0.2 mg l-1, 30–60 min Bath; 1500 mg l-1, 20 min Bath; 0.005 mg l-1, 30–60 min Oral; 16 ¥ 0.025 mg kg-1 per day (twice weekly) Oral; 7 ¥ 0.05 mg kg-1 per day Oral; 14 ¥ 2.7 mg kg-1 per day (smolts) Oral 7 ¥ 10 mg kg-1 per day

Table 5.2 Characteristics of chemotheraputants used in the control of sea lice infestations of farmed salmon (modiﬁed from Davies et al. 2001).

Environmental Impacts and Environmental Capacity 177

178

Chapter 5

Assessing the likely environmental impacts of chemicals is complex and biomarkers are increasingly widely advocated as, judiciously used, they can determine exposure and pathology and degree of damage. Davies et al. (2001) recommend a suite of such techniques in combination with scope-for-growth measurements in wild and deployed mussels and lugworm (Arenicola marina) bioassays to monitor the environmental effects of sea lice medicines.

5.3.3

Microorganisms and parasites

Comparisons between water supplies to farms and ﬁsh farm efﬂuents show quantitative and qualitative changes in microorganisms (Austin 1993). The intestinal tracts of farmed ﬁsh act as sites of proliferation and microorganisms tend to thrive in the farm environment where high levels of particulate and dissolved organic material occur. In general, diversity and density increase with efﬂuent organic content and decrease during chemotherapy (Austin 1993). The impacts of microorganisms or parasites from aquaculture operations on the environment at large have been little studied (Korzeniewski & Korzeniewska 1982; Iwama 1991; Austin 1993; McVicar 1997; Howgate et al. 2002). While wild ﬁsh can certainly act as sources of infection for cage farm populations (e.g. Rivas et al. 1993; McVicar 1997; Sommerville 1998) there is growing concern about the role played by cage ﬁsh farming in the introduction and spread of infection among wild ﬁsh stocks (NCC 1990; Riddell 1993; Beveridge 2001). Unregulated movements of ﬁsh have been responsible for the spread of a number of diseases and parasites including the introduction of the monogenean Gyrodactylus salaris to Norway from Sweden which resulted in a catastrophic decline of a number of wild Atlantic salmon populations (Halvasen & Hartvigsen 1989; Riddell 1993) (for reviews see also Baltz 1991; Holcik 1991; Weston 1991; Shariff et al. 1992; Sindermann 1993; McVicar 1997). The spread of furunculosis from farmed to wild salmon in Norway is similarly well documented (Johnssen & Jensen 1994). The introduction of cage aquaculture can also upset the host–parasite balance by greatly increasing infection pressure. The impacts of the rapid growth of cage farming of Atlantic salmon in Western Europe on wild salmonid populations has been a contentious issue for more than a decade. There is growing evidence that when sea lice epidemics occur on ﬁsh farms, infective larval stages can spread to inshore areas where trout and Atlantic salmon smolts congregate when they ﬁrst enter the sea (McKibben & Hay, in press). Early returning sea trout smolts, heavily infected with sea lice, are recorded at such times. However, the impacts of lice infestation on wild salmonids at the population level remain to be quantiﬁed. Norway, Ireland and Scotland have introduced limits to gravid female lice levels on farmed ﬁsh. However, epidemiological modelling is also required to assess the effectiveness of the measures.

5.3.4

Feral animals

Aquaculture is heavily reliant on the production of a handful of species that have been extensively moved around in order to capitalize on production

Environmental Impacts and Environmental Capacity

179

Fig. 5.9 Harvesting rainbow trout. Fish are occasionally lost during such operations.

technology and markets. In an analysis of records held on Fishbase (http://www.ﬁshbase.org/home.htm), Bartley & Casal (1998) found that aquaculture was the single most important cause of the 2600 recorded introductions. Among the ﬁsh, the Chinese and Indian major carp, a few species of tilapia, and the rainbow trout and Atlantic salmon dominate world production. It is widely argued that if such farmed stock escape and become established, it is to the detriment of wild stocks. Escapes are almost inevitable and cages are more vulnerable than other systems. There is an inevitable slow trickle of escapes that go largely unnoticed and unrecorded, and which occur during routine farm operations such as stocking, grading and harvesting (Fig. 5.9). Some ﬁsh species, such as sea bream and puffer ﬁsh (Takifugu rubripes) are also known to chew cage nets (Moon et al. 1993). Occasional mass releases, involving tens of thousands of animals, can also happen because of storm damage or vandalism (e.g. Billington & Herbert 1991; Gausen & Moen 1991; Webb & Youngson 1992; Youngson et al. 2001) (Fig. 5.10). There is a growing body of information on the numbers of animals that escape from cage aquaculture operations. Phillips et al. (1985a) estimated that annual angling catches of feral (escaped) rainbow trout in a Scottish loch were equivalent to 2.5% of caged production, while Penczak et al. (1982) estimated that about 5% of caged rainbow trout escaped each year. In Scotland and Norway reporting of escape incidences is mandatory. The numbers of salmon that escaped from Norwegian ﬁsh farms during 1993–2002 are shown in Fig. 5.11. During the same period production increased by 150% and strenuous efforts were made to reduce the numbers of escapes. While the proportion of farmed salmon that

180

Chapter 5

Fig. 5.10 Cages lying broken at the edge of a lake after a storm during which several thousand ﬁsh escaped.

Fig. 5.11 Numbers of escaped farmed salmon, Norway, 1993–2002 (data source: Norwegian Fisheries Directorate). Dotted line indicates an upper limit of 400 000 escaped ﬁsh set by the Directorate in 1999.

escape has shown a consistent decline from around 3.5% of the farmed ﬁsh population in 1997 to around half that in 2001, numbers remain high. According to the Norwegian Directorate of Fisheries, escapes during the period 1994–99 can be attributed to ‘holes in the net’ (i.e. predators) (31%), handling (20%), technical malfunction (14%), weather (14%), damage by propellers (11%), and

Environmental Impacts and Environmental Capacity

181

other causes (10%). Recent records from Scotland suggest that around 10–20 incidents occur per annum, the most common causes being damage by storms or predators, resulting in the release of 50 000–100 000 ﬁsh per year (FRS 2002b). Such ﬁgures, of course, take no account of the regular losses of small numbers of ﬁsh that occur during routine farm operations. Estimates from Chile suggest that as many as 900 000 caged Atlantic salmon escape each year. By contrast, with fresh waters where many exotic species have become established (Welcomme 1988; Holcik 1991; Bartley & Casal 1998), few species introductions in the marine environment appear to have become established (Baltz 1991). While there is no evidence of exotic species completely supplanting indigenous species, feral animals can cause serious negative impacts. Habitat damage is rare; disruption of communities through competition or predation is much more widespread. Many of the characteristics that make species ideally suited to aquaculture – high fecundity and rapid early development, ﬂexible phenotypes, wide environmental tolerances, catholic habitat preferences and feeding habits – are found in invasive species (Welcomme 1988). However, analysis of available data by Bartley & Casal (1998) suggests little correlation between the two. It has also been assumed that the introduction of top carnivores would cause more serious negative ecological impacts than that of omnivores or herbivores (Welcomme 1988; Moyle & Light 1996). However, Bartley & Casal (1998) show that 81% of introductions of omnivores caused negative ecological impacts compared with around 60% associated with carnivores and herbivores. Guidelines to assess risks of introductions have been available for some time (Turner 1988; Barg 1992; Pullin et al. 1993) and have been successfully applied (CostaPierce & Soemarwoto 1990). While the farming of native species reduces risks of habitat modiﬁcation and inter-speciﬁc interactions it may aggravate risks to indigenous strains through the introduction of non-adaptive genes with consequent reductions in ﬁtness. There are particular fears in this regard in relation to Atlantic salmon in view of the relative size of farmed and wild populations, the numbers of farmed ﬁsh that escape, and the fact that in the wild, genetically distinct, non-interbreeding populations occur within catchments. In 1999, for example, the biomass of farmed Atlantic salmon in the North Atlantic area was 620 000 t compared to around 3000 t caught by commercial ﬁsheries (Youngson et al. 2001). Research in Norway has established that escaped farmed Atlantic salmon tend to follow the prevailing currents along the coast, entering rivers or being intercepted by coastal net ﬁsheries. Over the past decade, on average 49% of salmon caught in Norwegian coastal net ﬁsheries have been of farmed origin (Youngson et al. 2001). Between 1989 and 1996, 21–38% of spawning salmon in Norwegian rivers were escaped farmed ﬁsh; in some rivers the ﬁgure is as high as 90% (Youngson et al. 1998, 2001; Fleming et al. 2000). However, much lower numbers of escaped farmed Atlantic salmon appear in Scottish or Irish net and rod ﬁsheries. While escaped farmed salmon have relatively poor reproductive success, male Atlantic salmon that mature precociously in fresh water have higher breeding and fertilization success than wild and hybrid ﬁsh (Garant et al. 2003). Early maturing males may thus promote the introgression of domesticated traits into wild populations with long-term impacts on genetic integrity.

182

Chapter 5

Further research is needed to fully understand the impacts of escaped farmed salmon on wild stocks, but preliminary results have shown signiﬁcant gene ﬂow from farmed to wild salmon can occur. While impacts are likely to prove situation speciﬁc (i.e. number and age of escapes; proximity to river; size of river; time of incident; genetic composition of wild and farmed stock, etc.), there is nonetheless a consensus that more must be done to contain farmed salmon. In 1994, contracting parties (i.e. all countries with Atlantic salmon stocks, including Norway, the UK, Canada and the United States) to the North Atlantic Salmon Conservation Organization (NASCO; http://www.nasco.org.uk) adopted a resolution to minimize impacts from salmon aquaculture. Key components included the establishment of cage design standards, guidance on surveillance and the development of contingency plans. These were rapidly translated into industry Codes of Practice in countries such as Scotland (http://www.scottishsalmon.co.uk/pdfs/contain.pdf), although it is a moot point how effective this has been, given the continuing high level of escapes in many countries. The suggestion that salmon farms only rear local strains is not practical: moreover, having been brought into the hatchery stocks inevitably pass through a series of selection bottlenecks, either accidental (e.g. disease) or deliberate (selection for growth, shape, colour), reducing heterozygosity within a few generations. Farming of sterile animals may be an answer in some instances (Benfey 2001; Beveridge 2001; Youngson et al. 2001). The impacts observed with feral salmon are most likely to apply to species that share similar characteristics: small populations with little migration (e.g. sea bass, sea bream). Domesticated triploid steelhead trout (Oncorhynchus mykiss), for example, show a high degree of site ﬁdelity, even when released some distance from the cage site (Bridger et al. 2001).

5.3.5

Conclusions

While quantities are usually small compared to inputs from other anthropogenic sources and while they have fallen on a per unit production basis (Fig. 5.12), waste nutrients from intensive cage aquaculture can be substantial in both local and regional terms (see Folke & Kautsky 1989; Mäkinen 1991; Enell & Ackefors 1992; Folke et al. 1994; Wu et al. 1994; Kelly & Elberizon 2001). Although there is little ﬁeld evidence that waste food, faecal and urinary products from ﬁsh farms cause problems, it has been demonstrated in the laboratory that ﬁsh farm wastes stimulate dinoﬂagellate growth (Nishimura 1982). It has also been proposed that biotin, found in ﬁsh farm wastes, triggers toxin production in marine dinoﬂagellates (Roberts et al. 1983; Graneli et al. 1993). However, others argue that the impacts of uneaten food and faecal material on sediments are non-toxic and readily reversible. In view of their effects on the environment and lack of knowledge, escapes of ﬁsh and release of chemotherapeutants are judged particularly worthy of concern (Beveridge et al. 1994b, 1997b). In Norway, industry and government are working together to address the issues, focusing on the development of new cage technology, development of technical standards for nets, regular equipment testing and training.

Environmental Impacts and Environmental Capacity

183

Fig. 5.12 Changes in quantities of N and P wastes discharge into Scottish coastal waters by salmon farms in relation to production trends (modiﬁed from Beveridge et al. 1997b).

Through uneaten food and faecal wastes, cage ﬁsh farming can result in the discharge of chemicals to the environment. Where there is concern about a particular chemical, environmental quality objectives (EQO) and environmental quality standards (EQS) can be established to ensure that unacceptable environmental impacts do not occur. However, putting the appropriate regulations in place can be difﬁcult, time consuming and expensive. Moreover, such regulation does not yet exist in many countries. The cage ﬁsh farming industry has at last become aware of the problems resulting from reliance on chemicals to control parasites and as a result they are increasingly used as part of integrated sea lice management strategies (see section 7.2). This section serves to show the range of environmental problems posed by cage aquaculture. The application of environmental risk assessment (ERA) methodologies – a four-step process involving identiﬁcation and characterization of hazards, assessment of exposure and characterization of risk – to the cage aquaculture industry is beginning to be explored.

5.4

MODELLING ENVIRONMENTAL CAPACITY

The term ‘environmental capacity’ is used here to describe the aquaculture production that can be sustained by an environment within certain deﬁned criteria. Irrespective of environment or type of cage culture, all environmental capacity models must consider:

184

• • • •

Chapter 5

what determines the productivity of the environment; what the farmed ﬁsh consume/produce in terms of food/wastes; how the environment responds to waste loadings; how much change is permissible.

Predictive models that generate estimates of environmental capacity for cage aquaculture can be used both to help match cage aquaculture development with societal objectives and to reduce risks to ﬁsh farmers. Note that models are generally based on considerations of trophic status rather than the release of toxic substances into the environment or consideration of pathogens or feral ﬁsh. At marine sites, the question of scale of impacts is particularly pertinent (Silvert 1992). In this section the focus is on local impacts; regional and global impacts will be considered in section 5.4.3.

5.4.1

Cage aquaculture in inland waters

Extensive cage aquaculture is a net consumer of primary production while intensive aquaculture stimulates productivity. Semi-intensive cage culture can either stimulate or reduce productivity. Different models are thus required to estimate environmental capacity.

Intensive cage aquaculture The rationale behind the following model is detailed in Beveridge (1984a). It is based on the assumptions that algal population densities are negatively correlated with water quality in general, and growth and survival of ﬁsh stocks in particular, and that phosphorus (P) is the limiting nutrient that controls phytoplankton abundance in most lakes and reservoirs. The well-established link between water quality – particularly DO levels – and algal densities is a cornerstone of modern water management philosophy (e.g. OECD 1982) while the relationship between water quality and ﬁsh health is discussed elsewhere in this book. The concept of a limiting nutrient is based on the fact that while a range of nutrients is required by planktonic algae, if the supply of any one nutrient is less than the demand, then it will limit growth. In most freshwater ecosystems P is limiting since it is the rarest element with respect to algal and higher plant demand (Vallentyne 1974). Phosphorus is an essential element required by all ﬁsh for normal growth and bone development, maintenance of acid–base regulation, and lipid and carbohydrate metabolism (Lall 1979; Takeuchi & Nakazoe 1981). It is derived principally from dietary sources (Nose & Arai 1979). While P requirements are species-speciﬁc (Table 5.3) most diets developed for intensive culture contain P surplus to requirements or in a form that is partially unavailable to the ﬁsh. Surplus P is excreted while unavailable P is voided in the faeces. The other major source of P is uneaten food (see also above). A mass balance approach is used to estimate total wastes, as explained in section 5.3.1. Using data on the P-content of feeds, FCR values and carcass

Environmental Impacts and Environmental Capacity

185

Table 5.3 Dietary phosphorus requirements of ﬁsh, expressed as percentage weight of diet (from Beveridge 1984a). Species Anguilla japonica Salmo trutta S. salar Onchorynchus mykiss O. keta Cyprinus carpio Ictalurus punctatus Chrysophrys major Oreochromis niloticus

Requirement % 0.29 0.71 0.30 0.70–0.80 0.50–0.60 0.60–0.80 0.45–0.80 0.68 0.90

Source Arai et al. 1975 McCartney 1969 Ketola 1975 Ogino & Takeda 1978 Watanabe et al. 1980a Ogino & Takeda 1976 Andrews et al. 1973; Lovell 1978 Sakomoto & Yone 1980 Watanabe et al. 1980b

Box 5.1 Calculations of total-P losses to the environment during intensive cage culture (modiﬁed from Beveridge 1984a). (a) Rainbow trout P content of food 0.9% Therefore, 1 t feed contains 9.0 kg FCR = 1.0 : 1 Pfood = 9.0 kg FCR = 1.5 : 1 Pfood = 13.5 kg FCR = 2.0 : 1 Pfood = 18.0 kg P content of trout = 0.48% wet weight of ﬁsh = 4.8 kg t-1 P losses to the environment Penv = Pfood - Pﬁsh Therefore, for FCR 1.0 : 1 Penv = 9.0 - 4.8 = 4.2 kg FCR 1.5 : 1 Penv = 13.5 - 4.8 = 8.7 kg FCR 2.0 : 1 Penv = 18.0 - 4.8 = 13.2 kg (b) Tilapia P content of food 1.3% Therefore, 1 t feed contains 13.0 kg FCR = 1.5 : 1 Pfood = 19.5 kg FCR = 2.0 : 1 Pfood = 26.0 kg FCR = 2.5 : 1 Pfood = 32.5 kg P content of tilapia = 0.34% wet weight of ﬁsh = 3.4 kg t-1 P losses to the environment Penv = Pfood - Pﬁsh Therefore, for FCR 1.5 : 1 Penv = 19.5 - 3.4 = 16.1 kg FCR 2.0 : 1 Penv = 26.0 - 3.4 = 22.6 kg

P-content, it is possible to estimate total-P loadings per tonne caged ﬁsh production (Box 5.1). Several models have been developed to predict the response of aquatic ecosystems to increases in P loadings. Most are empirical and have been calibrated and tested, veriﬁed and modiﬁed using various databases. The two most widely used and tested models are those of Dillon & Rigler (1974) and the OECD (1982). The former is a modiﬁcation of Vollenweider’s original model (Vollenweider

186

Chapter 5

1968, 1975) and states that the concentration of total-P in a water body, [P], is determined by the P loading, the size of the lake (area, mean depth), the ﬂushing rate (i.e. the fraction of the water body lost annually through the outﬂow) and the fraction of P lost permanently to the sediments. At steady state:

[P] =

L(1 - R) zr

where [P] is in g m-3 total-P; L = the total-P loading in g m-2 per year; z = is the mean depth in m; R = the fraction of total-P retained by the sediments; and r = the ﬂushing rate in volumes per year. The OECD model states that the total-P concentration in the lake is a function of the concentration of total-P in the inﬂows [P]j and the residence time, T(w):

[P] =

[P]j (1 + T (w))

where [P]j is in mg m-3 and T(w) in years. This model has been tested and veriﬁed in marine and freshwater environments in Sweden (Johansson & Nordvarg 2002; Nordvarg & Johansson 2002). Assessment of environmental capacity using the Dillon & Rigler model is best described in a series of steps. Step 1: Determine the steady-state total-P concentration in the water body to be farmed. In temperate waters this should be determined at the time of spring overturn when the waters are well mixed. For tropical lakes and reservoirs [P] should be taken as the annual mean total-P concentration of surface waters, and thus should be based on a number of samples taken during the year. Step 2: The development capacity of a lake or reservoir for intensive cage culture is the difference between the productivity of the water body prior to exploitation and the desired/acceptable level of productivity. Although [P] is used to assess productivity it is, of course, the corresponding levels of algal biomass that are of interest, and since ﬁsh are usually cultured throughout the year it is the peak level of algal biomass, as measured by chlorophyll a content, [chl], that is of concern. The desired/acceptable [chl] is dependent upon the species being cultured and whether or not the water body is multipurpose. Phosphorus concentrations corresponding to permissible [chl] for different species and water bodies with more than one use are given in Table 5.4 and Fig. 5.13. Values of [P] can be related to algal biomass using the regression equations in Table 5.5. For tropical water bodies [P] must be related to annual mean chlorophyll a levels [chl], using an equation such as that of Walmsley & Thornton (1984): 0.675

[chl ] = 0.416[P]

;

r = 0.84;

n = 16

Environmental Impacts and Environmental Capacity

187

Table 5.4 Tentative values (see text) for maximum acceptable [P] in lentic water bodies used for cage aquaculture (from Beveridge 1984a). Water body category

Species

Temperate

Salmonid Carp Carp and tilapia

Tropical

Tentative maximum acceptable [P] (mg l-1) 60 150 250

Table 5.5 Regression equationsa relating annual mean chlorophyll levels [chl] and peak chlorophyll levels [chl] to each other and to mean in-lake total phosphorus concentrations [P] for temperate water bodies (data from OECD 1982). [chl] = 0.61 [P]0.69; n = 99; r = 0.75; SE = 0.335 [chl] = 1.77 [P]0.67; n = 65; r = 0.70; SE = 0.375 [chl] = 2.86 [P]1.03; n = 73; r = 0.93; SE = 0.199 a

Regression based on unscreened data, i.e. no corrections for light or nitrogen limitation included.

Fig. 5.13 Suggested acceptable (dashed line) and ideal (solid line) P concentrations associated with freshwater bodies used for different purposes (from Beveridge 1984a).

188

Chapter 5

Step 3: The capacity of the water body for intensive cage ﬁsh culture is the difference, D[P], between [P] prior to exploitation, [P]i, and the desired/acceptable [P] once ﬁsh culture is established, [P]f: D[P] = [P]f - [P]i D[P] is related to P loadings from the ﬁsh cages, Lﬁsh, the size of the lake, A, its ﬂushing rate and the ability of the water body to handle the loadings (i.e. the fraction of Lﬁsh retained by the sediments): D[P] =

Lfish (1 - Rfish ) zr

Lfish =

D[P]zr (1 - Rfish )

The acceptable/desirable change in [P], D[P] (mg m-3) is determined as described in Step 2 above, and z can be calculated from hydrographic data obtained either from the literature or from survey work: z=

V A

where V = volume of water body (m3) and A = surface area (m2). The ﬂushing rate, r (per year), is equal to QoV, where Qo is the average total volume (m3) ﬂowing out of the lake/reservoir each year. Qo can be calculated by direct measurement of outﬂows, or in some circumstances can be determined from published data on total long-term average inﬂows from catchment area surface runoff (Ad.r), precipitation (Pr) and evaporation (Ev), such that: Qo = Ad ◊ r + A(Pr - EV ) (see Dillon & Rigler 1975 for details). Rﬁsh is the most difﬁcult parameter to estimate. Phillips et al. (1985b) estimated that at least 45–55% of the total-P wastes from cage rainbow trout are likely to be permanently lost to the sediments as a result of solids (faeces and food) deposition, and thus only 45–55% of the totalP loadings are in the form of dissolved P. In the absence of any other data, these values are also used for cage tilapia and carp. A fraction of the dissolved totalP component is also lost to the sediments and it is suggested that the most appropriate formula in Table 5.6 is used to calculate this. Values of Rﬁsh are, therefore, much greater than R for conventional P loadings, and can be summarized as: Rfish = x + [1 - x]R where x = the net proportion of total-P lost permanently to the sediments as a result of solids deposition (i.e. 0.45–0.55) and R = proportion of dissolved totalP lost to the sediments calculated from Table 5.6.

c

b

a

0.83 0.80a

d = 0.129 (L/z)0.549 R = 1/(1 + 0.747r0.507)

—

0.83 0.80a

d = 0.114 (L/z)0.589 R = 1/(1 + 0.515r0.551)

R = 0.201 exp(-0.0425qs) + 0.574 exp(-0.00949qs)

0.81 0.79a 0.79a 0.73a 0.71a 0.79 0.79 0.71 0.68 0.68 0.66

Correlation coefﬁcient

d = 0.129 (L/z)0.549 R = 1/(1 + 0.614r0.491) d = 0.94 V = 2.99 + 1.7qsb V = 5.3 R = 1/(1 + r0.5) d = 0.65 R = 0.426 exp(-0.271qs) + 0.574 exp(-0.00949qs) V = 11.6 + 1.2qs d = 10/z V = 12.4

Model

Coefﬁcients recalculated by Canﬁeld & Bachmann (1981) using their database. qs = areal water loading (mg-1). r = ﬂushing rate (volumes per year).

53

151

Natural lakes

Lakes with slow ﬂushing rates

210

73

704

Size of database

Reservoirs, North American

General; USEPA database and several European lakes and reservoirs

Model type

Ostrofsky 1978

Canﬁeld & Bachmann 1981 Larsen & Mercier 1976

Canﬁeld & Bachmann 1981 Larsen & Mercier 1976

Canﬁeld & Bachmann 1981 Larsen & Mercier 1976 Jones & Bachmann 1976 Reckow 1983 Chapra 1975 Larsen & Mercier 1976 Jones & Bachmann 1976 Kirchner & Dillon 1975 Reckow 1983 Vollenweider 1975 Chapra 1975

Source

Table 5.6 Empirical models for calculating the sedimentation rate, d, the retention coefﬁcient, R(1/d), and the sedimentation coefﬁcient, V of phosphorus for both general and speciﬁc categories of temperate water bodies (from Beveridge 1984a).

Environmental Impacts and Environmental Capacity 189

190

Chapter 5

Fig. 5.14 The relationship between response time and water residence time, T(w), for water bodies with different mean depths z (from OECD 1982).

The response time of a water body to increases in P loading is a non-linear function of the water residence time (T(w)) and mean depth ( z ). The expected 95% response time, t(M)95, which is used as an approximation to the full response time, can be calculated from Fig. 5.14. Step 4: Once the permissible/acceptable total-P loading, Lﬁsh, has been calculated, then the intensive cage ﬁsh production (tonnes per year) can be estimated by dividing Lﬁsh by the average total-P wastes per tonne ﬁsh production (see Box 5.1). A worked example is given in Appendix 5.1. Extensive cage aquaculture The principles of extensive cage culture in inland waters are discussed in Chapter 2. Fish production and thus environmental capacity are almost entirely dependent upon plankton production. On the basis of studies carried out with tilapias in inorganically fertilized ponds it has been suggested that yields from extensive cage culture are between 1 and 3% of primary production, depending on the level of primary production (see Beveridge 1984a). Environmental capacity can be estimated as follows: Step 1: Determine the annual gross primary production, SPP(g C m-2 per year) of the site. Since many tropical inland water bodies exhibit seasonality in the

Environmental Impacts and Environmental Capacity

191

Table 5.7 Conversion efﬁciencies of SPP to areal ﬁsh yields for water bodies of different productivities (from Beveridge 1984a). SPP (g C m-2 per year) <1000 1000–1500 1500–2000 2000–2500 2500–3000 3000–3500 3500–4000 4000–4500 >5000

% conversion to areal ﬁsh yields (g ﬁsh C m-2 per year) 1.0–1.2 1.2–1.5 1.5–2.1 2.1–3.2 3.2–2.1 2.1–1.5 1.5–1.2 1.2–1.0 ~1.0

pattern of primary production (Melack 1976), regular measurements may have to be made. Step 2: Convert SPP to potential annual ﬁsh yield, Fy, using Table 5.7 to convert plankton carbon content to ﬁsh carbon content and assuming fresh ﬁsh carbon content = 10% net weight of ﬁsh (Gulland 1970). Step 3: Production planning depends on a number of variables. The number of crops per year and the size of the ﬁsh at harvest should be decided. If, for example, tilapia are being farmed, then two crops of 160-g ﬁsh (6 ﬁsh per kg) may be desirable. However, seasonality of primary production may mean that one crop takes longer to grow. In order to reach target harvest size, the sum of primary production during the crop 1 growth period, SPPc l, should approximate that of crop 2, SPPc2, although this ignores changes in ﬁsh feeding efﬁciencies with changes in algal density. A worked example is given in Appendix 5.2. Semi-intensive cage aquaculture The principle of semi-intensive cage culture is that low-quality feeds are given to supplement the intake of natural food (see Chapter 2). The environmental capacity of inland waters for semi-intensive culture depends on the productivity of the water body and the amount of natural food available, and the quantity and quality of supplementary food used. Computation can be carried out as follows: Step 1: Determine the primary production, SPP, of the site being considered, as described in Appendix 5.2. Step 2: Calculate the annual ﬁsh yield, Fy, from the site using the conversion ﬁgures in Table 5.7, and a fresh ﬁsh carbon content = 10% wet weight.

192

Chapter 5

Step 3: Calculate the average annual amount of the various feedstuffs available, SFood, and estimate the FCR from the literature (e.g. see Table 30, Beveridge 1984a) in order to determine the ﬁsh yield attributable to the supplementary food. Step 4: Calculate the total-P loadings associated with the use of supplementary feedstuffs, Lﬁsh and using the model described above calculate the resultant increase in dissolved total-P. The increase in total-P can then be used to calculate increases in primary production, SPPﬁsh, attributable to ﬁsh culture. Step 5: Estimate the ﬁsh yields due to SPPﬁsh using the conversion efﬁciencies given in Table 5.7 (see Step 2 above), and calculate the total ﬁsh yields from semi-intensive culture, SFy, as: SFy = (aSPP) + (SFood ¥ FCR) + (bSPPfish ) where a and b are the expected conversion efﬁciencies of primary production to ﬁsh biomass obtained from Table 5.7. A worked example is given in Appendix 5.3.

5.4.2

Cage culture in coastal waters

A similar type of modelling process to that described above could be used to calculate the environmental capacity of marine sites, except that nitrogen, which is generally regarded as being the growth-limiting nutrient in the sea (Dugdale, 1967), would be substituted for phosphorus. In Scotland, the estimated enhancement of dissolved nutrient concentrations above background levels in sea lochs, attributable to cage ﬁsh farming, has been modelled in very simple terms by: ECE =

SM Q

where ECE = equilibrium concentration enhancement (kg m-3); S = the rate at which nutrient nitrogen is discharged (kg t-1); M = the total consented biomass of all farms in a sea loch; q = ﬂushing rate of the loch (m3 per year) (FRS 2002a). However, because of the much higher ﬂushing rates associated with marine coastal sites – days or weeks rather than months or years – the focus has been on modelling the impacts of intensive marine cage ﬁsh farming on the benthos rather than on the water column. Monitoring and regulation of impacts has also tended to follow this approach. Two types of model have been developed: one that predicts wastes dispersion and impacts (see Gowen et al. 1989, 1994; Panchang et al. 1993; Gillibrand et al. 2002; Pérez et al. 2002) and one that deals with assimilative or environmental capacity (Silvert 1992; Panchang et al. 1993; Hargrave 1994; Silvert & Sowles 1996; Cromey et al. 2000).

Environmental Impacts and Environmental Capacity

193

Solid wastes dispersion models Most solid wastes dispersion models incorporate assumptions regarding the quantities of solids (uneaten food, faeces) produced, settling rates of particles, integrity of solids falling through the water column, variations in depth at the site, current speeds, and direction and post-depositional changes. The model developed by Pérez et al. (2002) comprises three stages, the ﬁrst two in a spreadsheet, the third in a geographical information system (GIS) environment: • • •

quantiﬁcation of solid wastes; calculation of the distribution of the wastes; generation of waste contour plots.

Solid waste outputs, expressed in terms of carbon (C), are derived from a mass balance model that uses data on production per unit time, FCR, water content of feed, C-content of feed, food loss, C respired and C-content of ﬁsh tissue. Faecal-C-content is estimated as the difference between C consumed and C used for growth and respiration. Horizontal distribution of a particle (X, Y), falling through the bottom of a cage, is calculated from the equations of Gowen et al. (1989): X=

d V sin q +x u

X=

d V cos q +y u

(a)

(b) X0Y0

X0Y0

Spreading

X1Y1

t0(V0q0) t1(V1q1)

Cage Fig. 5.15 (a) Representation of horizontal displacement of particulate carbon from cages and (b) the effect of the smoothing function in the GIS-based model in transforming the point value to a cell equivalent to the cage area (see text) (modiﬁed from Pérez et al. 2002).

194

Chapter 5

where d = depth under each cage (m); V = mean current speed (m s-1); q = current direction; u = settling velocity (m s-1); and x, y = the position of each cage (Fig. 5.15a). For modelling purposes, the cages are located within a 500 ¥ 500 cell array, each cell representing 1 m2. From the literature, Pérez et al. (2002) use settling velocities (u) of 0.128 m s-1 and 0.04 m s-1 for uneaten food and faeces, respectively. The C deposition co-ordinates are then exported to a GIS software package, IDRISI 32®, and interpolation between values is carried out to generate a complete surface. The interpolation process, a particular strength of modelling within a GIS environment, redresses the problems of determining current speeds and directions over deﬁned time periods (e.g. 1 h), which results in wastes produced in 1 h being deposited at one set of co-ordinates (X0, Y0), while outputs over the subsequent hour are deposited at another set of coordinates (X1, Y1). Unlike most models, the GIS-based model also allows use of a ﬁlter to spread C inputs to the sea bed over an area equal to the cage area rather than at a single point (Fig. 5.15b). A second ﬁltering technique is then used to take account of laboratory-derived estimates of post-settlement redistribution of wastes caused by currents. An example of the resultant C deposition contour map is shown in Fig. 5.16. Pérez et al. (2002) compared modelled data with data from the ﬁeld and found a signiﬁcant correlation between predicted and actual sediment loadings. Pérez et al. (2002) suggested that differences in bathymetry over the waste dispersion area and in sediment type accounted for the differences between modelled and real values. Gowen et al. (1994) modiﬁed their original model to take account of changing bathymetry and varying currents with depth. However, there are likely to be other reasons. Silvert (1994) used a sub-model that allows for temperature-dependent changes in food consumption, while Panchang et al. (1993) incorporated an exponential decay function to account for nutrient loss. Assimilative capacity models It is one thing to predict sediment waste loadings from a cage farm but quite another to determine acceptable loadings. There is, as yet, no universally accepted approach. Most models developed to date that deal with environmental or assimilative capacity incorporate assumptions regarding the grazing rate of the benthos and the rate at which bottom currents disperse settled wastes (Silvert 1992; Panchang et al. 1993; Hargrave 1994). A further problem is that at sites with limited water exchange it may take several years for the benthos to equilibrate and achieve long-term stability. Hargrave (1994) developed a benthic enrichment index (the product of organic C-content and redox potentials in surface sediments) as a measure of ecosystem health for temperate marine areas. He related this to organic C sedimentation rates and proposed that where organic C sedimentation rates exceed 1 g C m-2 per day, C loss through respiration is insufﬁcient to prevent C accumulation and the development of anoxia. This has not been further substantiated in the literature. In Japan, guidelines for sediment assimilative capacity are based on modelling impacts of feed and faecal inputs on acid volatile sulphide sulphur (AVS-S) concentrations (Yokoyama 2000). According to Pawar et al. (2002), AVS-S

Environmental Impacts and Environmental Capacity

Grid

metres 100.00

195

North

0.00 0.89 1.77 2.66 3.54 4.43 5.31 6.20 7.08 7.97 8.85 9.74 10.62 11.51 12.39 13.28 14.16 15.05 15.93 16.82 17.70 Kg C/ m2/Yr

Cage Fig. 5.16 Waste loading contours (kg C m-2 per year) from cages. These are predictions from a model (courtesy Stirling Environmental Services).

correlates well with seasonal C inputs, although feed pellet size also has an important effect on the feed input–sediment quality relationship.

5.4.3

Conclusions

The models proposed here for predicting the environmental capacity of different types of site for cage ﬁsh culture are still undergoing testing and veriﬁcation and should be used as guides, rather than hard-and-fast rules, to development (Hargrave 1994; Kelly 1995). The main problems associated with each model are summarized below. Inland waters For intensive cage culture in inland waters, the setting of desirable/acceptable water quality criteria presents a major difﬁculty. Although various agencies (e.g. OECD, USEPA) have set tentative water quality criteria, these relate to mini-

196

Chapter 5

mizing nuisance blooms in multi-use water bodies. However, the establishment of objectives for lakes and reservoirs where ﬁsh culture is the primary or sole objective is made difﬁcult by our poor understanding of the relationships between algae and water quality and water quality and stress, growth, disease and mortality. In view of this the values given in Table 5.4 must be used with caution and amended through experience and in the light of information collected from environmental monitoring. The model, too, suffers from a number of shortcomings. The proportion of P wastes from ﬁsh farms that is consumed by wild ﬁsh and that is lost to the sediments requires further research. Håkanson et al. (1998) attributed differences between modelled and actual total-P concentrations in Swedish lakes with cage ﬁsh farms to poor estimates of these two processes. Adopting an empirical approach based on a set of eight ﬁsh farm lakes, Johansson & Nordvarg (2002) determined that the estimate used by Beveridge (1984a) was at least a factor of two too high. A further issue is that the model was developed for use in P-limited water bodies. However, some shallow lakes and reservoirs are light-limited while many others, especially in the tropics, are nitrogen-limited (Horne & Goldman 1994). Kelly (1995) has also shown that the model is most applicable to small, well-mixed water bodies where cage ﬁsh farming is the major anthropogenic source of nutrients. The model is less successful at predicting changes at large lentic sites with many bays, at sites where cages are situated near an outﬂow or where cages represent only one of a number of perturbations occurring in the catchment (e.g. in a recently afforested and fertilized catchment) (Fig. 5.17). Nevertheless, if the model is used with care, and in the appropriate conditions, predictions should still be sufﬁcient to act as a management guide to permissible levels of production and can be adjusted in the light of water quality data collected when the farm is in operation. The importance of instigating a water quality monitoring scheme, particularly on large farms, cannot be stressed too highly. The model for extensive culture is based on limited data for phytoplankton grazing by tilapias in ponds. Although this type of model may also be applied to other phytoplankton feeders, such as silver carp, it cannot be used to predict appropriate production levels for extensively reared ﬁshes that primarily feed on zooplankton (e.g. bighead carp). Bayne et al. (1992) tried to apply the model to cage culture of carp hybrids and criticized it as being over-simplistic, particularly with regard to the exploitation of hypereutrophic conditions. Further work is needed. Because it is a hybrid of both the extensive and intensive models, the model for semi-intensive cage culture involves the errors associated with both. Moreover, it is difﬁcult to ﬁnd data on the P-content and conversion efﬁciencies of materials used for supplementary feedstuffs, and it is even more difﬁcult to assess how ingestion in combination with natural food (algae, detritus, zooplankton) affects these values. Coastal waters Although increasingly widely used in assessing impacts of proposed new farms or increases in production, solid wastes loading models (see Panchang et al. 1993; Gowen et al. 1994; Cromey et al. 2000; Pérez et al. 2002) have yet to be

Environmental Impacts and Environmental Capacity

197

Fig. 5.17 Modelled (hatched) and observed (blank) changes in water column P concentrations in two Scottish lochs (a) where the model ﬁts well and (b) where it does not (from Kelly 1995).

extensively tested and veriﬁed. Assimilative capacity models are still being developed. Measurements of organic matter decomposition in sediments under cages in the Gulf of Aqaba suggest that assimilative capacity may be higher in warm than temperate waters (Angel et al. 1992). Debate continues on whether benthic capacity models are entirely appropriate for highly enclosed bays (Cranston 1994); models that predict ammonia concentration in the water column, for example, may be better. Others argue that other biological terms, such as infection pressure from sea lice on farmed ﬁsh, or even aesthetic and amenity factors, should be included in any concept of environmental capacity.

198

Chapter 5

While models are proving useful in predicting impacts and assimilative capacity at a particular cage farm site, regulatory authorities have yet to decide the approach required to control cage aquaculture on regional or national scales. The approach taken in Norway to consider assimilative capacity on a wholecountry basis (LENKA) is unique to that country (see Kryvi et al. 1991).

APPENDIX 5.1 Example of intensive cage rainbow trout production assessment for a temperate natural lake (see section 5.4.1) (modiﬁed from Beveridge 1984a) Site Surface area of lake, A, = 100 ha (calculated from map). Mean depth, z , = 10 m (from hydrographical survey). Flushing coefﬁcient, r, = 1 per year (determined from sampling outﬂows). Method Step 1: Determine steady state [P], [P]i, prior to development. 15 mg m-3 as determined from monitoring programme. Step 2: Set maximum acceptable [P], [P]f, following the introduction of ﬁsh cages. Assuming no other developments or criteria take precedence, then 60 mg m-3 is chosen as target [P]. Step 3: Determine DP DP = [P]f - [P]i = 45 mg m -3 since

P=

L fish =

Lfish (1 - Rfish ) zr DPzr

(1 - Rfish )

Rﬁsh = x + [(1 - x)R]; where R is calculated from Table 5.6, and x is assumed to be 0.5. Thus, Lﬁsh = 45 ¥ 10 ¥ 1/0.23 = 1957 mg m-2 per year = 1.957 g m-2 per year Step 4: Since the lake has a surface area of 106 m2, the total acceptable loading = 1.957 ¥ 106 g per year Thus, the tonnage of ﬁsh that can be produced, assuming a P loading of 17.7 kg t-1 (see Table 5.3.), = 1.957 ¥ 106 g/17 700 g = 111 t per annum

Environmental Impacts and Environmental Capacity

199

This value should be used as a pre-development guide to the environmental capacity of the lake. A monitoring programme must be implemented and production levels adjusted in the light of information collected on water quality – principally algal biomass and DO levels.

APPENDIX 5.2 Example of extensive cage tilapia production for a tropical reservoir (see section 5.4.1) (from Beveridge 1984a) Site Surface area = 100 ha Method Step 1: Calculate the annual gross primary production, SPP. SPP = 1200 g C m-2 per year, as determined by regular measurement. Step 2: Convert to annual ﬁsh yields, using Table 5.7. i.e. = 1.3% S PP Æ ﬁsh = 15.6 g ﬁshC m -2 per year = 15.6 g ﬁsh m -2 per year = 156 t annual ﬁsh production for whole lake. Step 3: Assuming two crops per year, determine culture periods. SPPc1 ∫ SPPc2, in order for ﬁsh to reach target market size. SPP( Nov – May ) = 570 g C m -2 SPP( Jun – Oct) = 630 g C m -2 One seven-month, and one ﬁve-month cycle are chosen. Assume 25 g ﬁsh stocked. Assume 8 ﬁsh per kg target market size (i.e. 125-g each) Therefore, each ﬁsh grows 100 g during culture period. Therefore, stocking requirements = 156 t/100 g = 1.56 ¥ 106 ﬁngerlings.

APPENDIX 5.3 Example of semi-intensive cage tilapia production assessment for a tropical lake (see section 5.4.1) (modiﬁed from Beveridge 1984a) Site Surface area = 100 ha Mean depth, z, = 10 m Flushing coefﬁcient, r, = l per year

200

Chapter 5

Method Step 1: Calculate the annual gross primary production, SPP. SPP = 1600 g C m-2 per year, as determined by regular measurement. Step 2: Convert to annual ﬁsh yields, using Table 5.7. i.e. ~ 1.3% SPP Æ ﬁsh = 156 t annual production for whole lake. Step 3: Assume 100 t of cotton seed meal and 20 t of soya meal are available for feed each year. Using FCR values of 2.7 : 1 and 3.0 : 1, respectively, (from Beveridge, 1984a, Table 30), 37.2 t can be grown from cotton seed meal, and 6.6 t can be grown from soya meal. Step 4: Total-P loadings from ﬁsh grown on supplementary food are: (37.2 ¥ 23.77) + (6.6 ¥ 16.97) = 996.24 kg (P-content of feeds from Beveridge, 1984a) The resultant increase in [P] can be calculated from:

[P] =

L(1 - Rfish ) zr

where L is the areal loading from the ﬁsh cages (996.24 kg 106 m-2 = 996.24 mg m-2); Rﬁsh = x + [(l - x)R], where x = 0.50 and R = 0.54 calculated from the equation of Larsen & Mercier (1976) for natural laker (Table 5.6) and therefore Rﬁsh = 0.77; [P] = 996.24[(1 - 0.77)/(10 ¥ 1)] = 22.9 mg m-3 Using the formula: SPPﬁsh = 31.1 [P]0.54 (OECD, 1982) to relate increase in [P] to primary production, SPPﬁsh = 31.1 ¥ 22.90.54 = 34.7 g C m-2 per year increase. Step 5: Fish yield due to SPPﬁsh can be calculated using the conversion efﬁciencies in Table 5.7: SPPfish Æ fish = 0.3 g fish C m -2 per year = 3 g fish m -2 per year = 3 t fish production for whole lake. SFy, the total ﬁsh yield can now be calculated from: Fy = (0.013 ¥ 1200 ¥ 10) + [(100 2.69) + (20 3.04)] + (0.01 ¥ 34.7 ¥ 10) = 203 t fish per annum.

Cage Aquaculture, Third Edition Malcolm C. M. Beveridge Copyright © 1996, 2004 by Blackwell Publishing Ltd

Chapter 6

Management

Fish farm managers are charged with maximizing proﬁt. While farm managers in general have little control over economic factors such as ﬁsh and feed prices, those managing cage farms are further disadvantaged in having little control over environmental variables (see also Bjørndal & Uhler 1993). Cage farm management strategies are directed towards minimizing stock losses and promoting good growth while controlling costs, which in practice means: • stocking at densities appropriate to the site, species and method of rearing; • feeding in the most cost-effective manner; • ensuring the best possible water quality within cages; • maintaining cages, moorings, anchors and ancillary gear; • regular monitoring of stocks for signs of disease, removal of mortalities and treatment of infected stock. Increasingly, however, it is realized that various routine farm operations, such as transport, stocking, grading, feeding, sample weighing and harvesting, have implications for ﬁsh welfare (FSBI 2002). Fish welfare has both ethical and economic dimensions, the two of course being inextricably linked, impinging not only on production through stress, food conversion and disease susceptibility, but also on product quality and marketability. Finally, managers are also responsible for ensuring that working conditions for staff are as safe and comfortable as possible. The following sections summarize the general routine of cage ﬁsh farming.

6.1 TRANSPORT AND STOCKING Although cages can be used for production of seed for ﬁsh (e.g. milkﬁsh – Marte 1988 – and tilapia – Beveridge 1984b; Little & Hulata 2000; Ariyaratne 2001) and for penaeid shrimp (Tomiyama 1973; Peña & Prospero 1984; Walford & Lam 1986), supplies of juveniles are usually either from land-based hatcheries or from the wild.

6.1.1

Transport

Reviews of ﬁsh transport have been made by Solomon & Hawkins (1981), Rengaswamy et al. (1999) and Erikson (2001) among others. Fish are often deprived of food for several hours prior to transport in order to clear the diges201

202

Chapter 6

Fig. 6.1 Sealed plastic bags of oxygenated water containing tilapia fry, ready for distribution. Note outer bags of woven palm leaves that help protect and keep ﬁsh cool.

tive tracts of food, thus minimizing fouling of the transport system with excreta and excretory products, and reducing oxygen consumption. Injured or weak ﬁsh are discarded and, if necessary, treatment for ectoparasites is carried out (see Martyshev 1983, for example, for details). The processes of capture, handling, loading and transport are highly stressful to ﬁsh, resulting not only in physical damage (e.g. scale removal) but also in changes in blood chemistry, increased oxygen consumption, osmoregulatory problems and heightened susceptibility to disease (Solomon & Hawkins 1981; Bandeen & Leatherland 1997; Cubero & Molinero 1997; Pickering 1998; Svobodová et al. 1999; Barton 2000; Congleton et al. 2000). There is a great deal of interspeciﬁc variability in response with some species, such as silver carp, being particularly difﬁcult to transport (Horvath et al. 1984). Nevertheless, handling during transport should be kept to a minimum. Small quantities (a few thousand) of ﬁsh may be transported over short distances to cage sites with relatively few problems. Plastic bags, which should be one-third ﬁlled with water and the remaining space ﬁlled with oxygen prior to sealing, can be used (Fig. 6.1). The alternative is to use an insulated transport box that holds several thousand litres and is mounted on a trailer or truck (Fig. 6.2). Tanks should have rounded corners to minimize damage to ﬁsh and are usually connected to aeration or oxygenation equipment. For larger consignments, tankers with a capacity of 20 000 l or more, equipped with refrigeration and oxygenation/aeration equipment, are available from specialist transport companies (see Fig. 6.5). Transport vehicles without water puriﬁcation equipment are limited in the length of time they can carry ﬁsh (see below). Recommended ﬁsh densities for transport are given in Table 6.1.

Management

203

Fig. 6.2 Small tank, equipped with a ﬁsh pump, for short distance road transport of smolts (from Edwards 1978). Table 6.1 Recommended transport conditions for various ﬁsh (data from Piper et al. 1983; Horvath et al. 1984; Karlsen 1993; D. Grifﬁth, pers. comm.). Species

Size

Stock. density (g l-1)

Duration (h)

Temperature (°C)

40 mm 60 mm

60–120 120–240

8–10 8–10

5–10

100–130 mm

240–360

8–10

5–10

<50 g

40 100

— —

20 5

200–280 mm

300–420

8–10

5–10

100 g 10 g 4g 2g

350–600 250–400 200–350 150–200

8–16 8–16 8–16 8–16

18 18 18 18

20 g (100 mm) 5 g (75 mm) 1 g (50 mm)

120 80 60

12 12 12

18–30 18–30 18–30

Common and bighead carp

<100 g

280 50

— —

5 30

Silver carp

<100 g

90 25

— —

5 25

0.5–200 g

100–200

24

8–28

Chinook salmon Coho salmon Atlantic salmon Trout (rainbow, brook, brown, etc.) Channel catﬁsh

Largemouth bass

Tilapia

204

Chapter 6

Fig. 6.3 Well-boat transporting smolts to farms in the remote north-west of Norway (courtesy D. A. Robertson).

Transportation of ﬁsh in the holds of boats is common practice today. In Southeast Asia, grouper fry are carried in the water-ﬁlled live-bait compartments of traditional pole-and-line ﬁshing vessels from around the Chinese coast to stock the cages of Hong Kong ﬁsh farmers (Tseng 1983). In Norway, ‘well-boats’ developed for carrying live ﬁsh such as cod, have been used for many years to transport smolts to remote salmon farms, and the practice is now widely employed through the marine cage ﬁsh farming sector (Edwards 1978; Karlsen 1993; Beaz Paleo et al. 2000) (Fig. 6.3). Because of the value of the cargo and disease regulations, purpose-built designs (20–40 m in length), capable of transporting up to 50 t of live ﬁsh, have been developed. Most permit either seawater circulation or closed circulation and carry aeration and/or oxygenation equipment. Closed circulation is desirable when passing through areas of known disease risk or where poor water quality conditions prevail. Purpose-built boats have facilities to refrigerate sea water and have a propulsion system designed to minimize noise, vibration and thus stress, and even have ozonation systems to reduce bacterial

Management

205

Fig. 6.4 Towable cage for transporting salmon smolts (redrawn from Karlsen 1993).

loads. Directional propellers and bow-thrusters enhance manoeuvrability (Beaz Paleo et al. 2000). Smolts can be successfully transported over many hundreds of kilometres in this way although low speeds must be maintained during long journeys. Rough weather can also severely affect ﬁsh transported by sea and result in high mortalities (Tseng 1983). Towable cages are widely used in the transfer of ﬁsh, especially of young ﬁsh captured in ﬁshing nets or traps, to cage production sites (Kreiberg & Solmie 1987; Rottiers 1991; Karlsen 1993; Brown et al. 1995; Cabello 2000; Gooley et al. 2000) (Fig. 6.4). The principal advantages are low costs and reductions in stress associated with loading and unloading. In southern Australia, small southern blueﬁn tuna are captured by purse-seiners in the Great Australian Bight, and towed some 400 km to the cage production facilities near Port Lincoln (Gooley et al. 2000). Bridgestone-type or plastic Norwegian cages that can hold up to 165 t each are used. At a towing speed of <2 km h-1, it typically takes 5 weeks to transport the ﬁsh. However, such marathon journeys are not without incident. In early 2003, for example, 5000 tuna, weighing some 130 t, died or were lost when a large tow cage failed near Port Lincoln. Elsewhere, farmers have recommended towing speeds of 0.5–1 knot. Air transportation, especially using helicopters, is widely practised among salmon farmers. Tanks with capacities of up to several thousand litres, and holding 20 000 Atlantic salmon smolts, are suspended from the helicopter and

206

Chapter 6

Fig. 6.5 Helicopter delivering Atlantic salmon smolts at a cage farm.

can be used to transport ﬁsh for several hours (Black et al. 1992b; Karlsen 1993) (Fig. 6.5). Transport problems may be aggravated by high temperatures and salinities. In tropical countries transportation at night or in insulated containers is recommended (Tattanon & Maneewongsa 1982). If ﬁsh have to be moved considerable distances toxic metabolic wastes (e.g. carbon dioxide, ammonia) can accumulate and there may be dramatic increases in bacterial numbers. In order to minimize mortalities, various modiﬁcations to transport practices have been suggested, including a lowering of metabolic rate and thus oxygen consumption and waste production through a combination of light sedation and hypothermia (Solomon & Hawkins 1981; Rothbard 1988; Cubero & Molinero 1997; Ross & Ross 1999; Sandodden et al. 2001), and absorption of ammonia and carbon dioxide through the addition of natural zeolite and a buffer to the transport media (Amend et al. 1982). However, these methods have not been adopted in commercial-scale ﬁsh transport operations.

Management

207

Fig. 6.6 Large transport truck transferring Atlantic salmon smolts to a ﬁsh cage, Scotland. The cage will be towed to the farm site and the smolts redistributed to other cages (courtesy D. A. Robertson).

6.1.2

Stocking

Prior to transfer, care should be taken to ensure that the temperature of the ﬁsh approximates that of their new environment. Such problems do not arise with ﬁsh delivered in well-boats or towable cages. Bags of ﬁsh should be placed in the cages and temperatures allowed to equilibrate prior to release. In warm climates, transfer to cages should be carried out in late evening or early morning, when temperatures are lower. Again, handling should be kept to a minimum, particularly when transferring ﬁsh from fresh to sea water. Fish transported by well-boat or in plastic bags can be taken directly to where the cages are moored. If large transport tanks are used, then cages may be towed to the nearest suitable landing site and the ﬁsh piped directly into the cage (Fig. 6.6); alternatively, transport boxes may be taken to the cages. Bagged ﬁsh should be transferred by gently decanting into the cages, whereas for transport tanks and well-boats, the volume of water is reduced prior to ﬁsh being transferred by hand or by ﬁsh pump (Fig. 6.7). If nets are used, they should be of knotless mesh to minimize damage. A number of studies, however, point to the stress associated with transferring ﬁsh by net or pump (Flagg & Harrell 1990; Wagner & Driscoll 1994). Although ﬁsh can be counted by eye or by counting board, infrared and digital video-based systems have become the standard in intensive cage aquaculture. Commercial systems can count up to 200 000 smolts per hour. Cabello (2000) describes the transfer of ﬁsh from a transport cage to a growout cage. One side of the production cage can be detached from the collar, the

208

Chapter 6

Fig. 6.7 Smolt delivery at a Norwegian salmon farm. Fish are netted out of the hold, counted and transferred to the cages (courtesy D. A. Robertson).

transport cage swum inside, the side of the production cage re-attached to the collar and the ﬁsh gently decanted from the transport cage to the production cage. The empty transport cage is then removed by crane. In Australia, the transfer of tuna from transport cage to production cage is generally done by exploiting swim-through doors (Gooley et al. 2000). Feeding ﬁsh immediately after transfer to cages is generally not recommended, although farmers may wish to gauge the health of stock from the feeding response. Some ﬁsh, such as tilapias, recover relatively quickly from handling and regular feeding can commence 3–4 h after transfer. Monitoring of blood chemistry following transport suggests that ﬁsh take several days to recover (Aldrin et al. 1979; Cubero & Molinero 1997; Pickering 1998). More stresssensitive ﬁshes, such as salmonids, are best left for 12–24 h before feeding. Two stocking strategies prevail: farmers may either stock the number of ﬁsh required to give the desired production per cage, taking account of likely mortalities, or stock high numbers of young ﬁsh that are redistributed among other cages as they grow. The former strategy is often adopted by small-scale producers, especially in the tropics where the time from stocking to harvesting may be only a few months (Schmittou 1993; McAndrew et al. 2002). The latter strategy tends to result in better growth and less size variation among stock (Duarte et al. 1994), is a much more cost-effective use of space and facilitates disease control strategies such as separation of year classes. It is thus the preferred method at intensive operations despite increased labour and risks associated with handling (for further discussion see Bjørndal 1990; Boghen 1995). Fish are moved when stocking densities are judged to be adversely affecting growth and

Management

209

Table 6.2 A simple production cycle for Atlantic salmon farm (modiﬁed from Heen 1993). Year 1

Year 2

Year 3

(a) Smolts released in one-third of the cage capacity

(a) First-year smolts half through the production process (b) Smolts released in one-third of the cage capacity

(a) First-year smolts slaughtered (b) Second-year smolts half through the production process (c) Smolts released in one-third of the cage capacity

production (Table 6.2). However, while it is known that stocking densities affect growth, incidence of injury and mortality (e.g. Konikoff & Lewis 1974; Ewing & Ewing 1995), there have been few systematic attempts to determine optimum stocking densities for farmed species. Most trials have limited consideration of the impact of stocking density on growth and production, ignoring issues such as welfare or economics or its inﬂuence on ﬂesh quality. While stocking densities of intensively farmed Atlantic salmon and yellowtail have fallen over the past 30 years to around one-ﬁfth of what had been accepted practice, the change has been driven by experience and production management considerations rather than any scientiﬁc analysis. It is now known that high densities lead to welfare problems in some species, such as Atlantic salmon and trout, but improve it in others (e.g. Arctic charr Salvelinus alpinus, halibut) (Jørgensen et al. 1993; FSBI 2002). Information on ﬁsh biomass is necessary not only for management of cage densities but also for feeding and medication (see section 6.2 below).

6.2

FEEDS AND FEEDING

Extensive methods of cage ﬁsh farming rely solely on natural food, although illumination may be used to attract food organisms. Semi-intensive and intensive operations, however, involve feeding and since feed costs represent the single largest component of operating costs, accounting for 25–50% at intensive ﬁsh farms, it follows that the quality of feed and the manner in which it is used are signiﬁcant determinants of proﬁtability. In addition to fulﬁlling nutritional requirements, feeds must also meet an increasing number of other criteria, including that of reducing pollution. Fish foods can be divided into two types: semi-intensive and intensive.

6.2.1

Semi-intensive feeds

Semi-intensive feeds are relatively low in protein and made from low-cost, readily available, local materials (Coche 1982). Fish grown in semi-intensive culture situations remain heavily reliant on natural feed, which is comparatively high in

210

Chapter 6

Fig. 6.8 Feeding a moist ball of home-made feed to caged catﬁsh, Bangladesh (courtesy K. McAndrew and L. G. Ross).

protein content, supplemented by foods high in carbohydrate or fat that ‘spare’ protein from being used as an energy source, allowing it instead to be utilized for growth. This type of aquaculture is conducted with herbivores, planktivores, detrivores and omnivores grown in conditions where there is a plentiful supply of natural food (i.e. productive water bodies). Semi-intensive cage culture is largely conﬁned to freshwater species and prevails in tropical and sub-tropical countries. The range of feedstuffs used in semi-intensive culture is enormous. Reviews of materials and their nutritional value are given in Coche (1982), ADCP (1983), Lovell (1989), McVey (1991), Wilson (1991), New et al. (1993), NRC (1993) De Silva & Anderson (1995) and Jauncey (2000). Feedstuffs are fed either singly or in combination but because of seasonal variations in costs and availability, use may be sporadic. Moist feeds are sometimes compounded using dry ingredients, water, and a suitable binder (e.g. dried green banana powder, cooked plantain, potato, etc.) to form a ball that can be fed fresh or partially dried (Fig. 6.8). Care must be taken with materials containing anti-nutrients that adversely affect ﬁsh (Wilson 1991; Tacon 1992; Jauncey 2000).

6.2.2

Intensive feeds

Intensive feeds are used principally in the culture of carnivorous species, although omnivores/herbivores such as the tilapias may be reared intensively where water

Management

211

resources are scarce or where the growing season is restricted. Unlike semiintensive feeds, intensive feeds must fulﬁl all nutritional requirements and include the appropriate quantities and types of proteins, fats, carbohydrates, minerals and vitamins (for reviews see McVey 1991; Wilson 1991; Jobling 1993; New et al. 1993; Jobling et al. 2001). From the beginnings of intensive aquaculture in the latter half of the 19th century until 40 years ago the only option available for growing trout or other carnivores was to use slaughterhouse wastes, ﬁsh or shellﬁsh. In the mid-1950s the ﬁrst formulated moist diets were compounded for salmonids and by the 1960s dry diets had been developed. Intensive diets have continued to develop and improve as our understanding of ﬁsh nutritional requirements has grown and ﬁsh feed technology progressed. Trash ﬁsh Economic factors and problems with diet formulation, feed storage and distribution have resulted in fresh or frozen minced and chopped trash ﬁsh remaining the mainstay of a number of cage ﬁsh farming industries, including yellowtail and sea bream in Japan (Ikenoue & Kafuku 1988; Fukusho 1991; Nakada & Murai 1991), snakehead in Thailand (Wee 1982; New et al. 1993), tuna in Australia (Gooley et al. 2000) and grouper in Asia (Tseng 1983; New et al. 1993) (see Fig. 6.12). Cod, too, are reportedly much less likely to reject trash ﬁsh than dried pellets, especially at low temperatures (Clark et al. 1995). While trash ﬁsh may seem ideal, it suffers from a number of disadvantages. Many readily available species of trash ﬁsh, especially sardines and mackerel, have excessive fat contents making them unsuitable for some cultured species while others contain high levels of thiaminase that can lead to thiamin deﬁciency if not heat treated (Ikenoue & Kafuku 1988; Nakada & Murai 1991; Tacon 1992). Quality may vary considerably during the year: capelin, for example, contains 14% fat during January–February, falling to 8% after spawning (Jobling 1993). Availability may vary with season and ﬁsh may be reluctant to switch diet, adversely affecting growth for several weeks (Edwards 1978). Trash ﬁsh also has a high moisture content and is expensive to transport, and is therefore best suited to farming operations sited close to ﬁsh landing or processing centres. A further consideration is the high degree of waste associated with this type of diet and its effects on water quality (Beveridge 1984a; Wu et al. 1994). A pigmented diet may also have to be used to improve ﬂesh colour. Raw feeds can also act as a source of bacterial infection (Ross & Johnson 1962; Freidman & Shibko 1972; Shepherd 2001). Moist diets Moist feeds are usually prepared on the farm from minced fresh or frozen ﬁsh or hydrolysed ﬁsh wastes (silage) (see section 6.3.2) mixed with a commercially produced dry meal containing protein (ﬁsh meal), carbohydrate (cooked starch), vitamins, a binding agent (alginate, carboxymethylcellulose) and perhaps a colorant (e.g. shrimp meal). If white ﬁsh is used, ﬁsh oil may be added in order to spare protein from being utilized as an energy source. The paste is then either

212

Chapter 6

fed as a moist ball or pelleted to the required size. Details of moist feed manufacture are given in Edwards (1978), Lovell (1989), McVey (1991) and New et al. (1993). Although moist feeds are superior to trash ﬁsh in terms of nutritional qualities and stability in water, they suffer from many of the disadvantages of trash ﬁsh diets – seasonal ﬂuctuations in quality, transport and storage problems and associated high pollution loadings. Use is largely limited to a few species, such as yellowtail, red sea bream (Fukusho 1991; Nakada & Murai 1991) or early life stages that ﬁnd dry diets unpalatable or to parts of the world where commercial dry diets are unavailable. For the latter reason on-farm feeds, many of which are fed in moist form, are still used by many ﬁsh farmers in Asia (New et al. 1993; Santiago et al. 1996). Dry diets Dry diets have a number of advantages over moist or trash ﬁsh diets: they are more stable in water and thus less polluting and are readily eaten by most species of cultured ﬁsh. Moreover, they are often more digestible and commercially prepared diets also have fewer anti-nutrients owing to the methods used in processing. Dry diets, while being relatively cheap to transport and easy to store, are more expensive, the cost of processing adding up to 20% to costs of feed preparation (ADCP 1983; New et al. 1993; Schmittou 1993; Goddard 1996). Feed formulae for commonly farmed species are widely available (ADCP 1983; Lovell 1989; McVey 1991; Wilson 1991; Jobling 1993; Schmittou 1993). Lovell (1989), Goddard (1996) and Jobling et al. (2001) review feed manufacture. The simplest way of producing a pelleted dry diet is to use a commercial food mixer to produce a wet dough that is then passed through a pellet mill with an appropriately sized die and subsequently dried. This is often adequate for small operations where commercial feeds are unavailable (for review see New et al. 1993). Large-scale operations, however, demand bulk supplies of high-quality feed ingredients of uniform quality. A typical feed production scheme is illustrated in Fig. 6.9. Following mixing and milling, dietary ingredients are pretreated by steam or hot water to improve handling and pelleting, to increase nutritional value and to help destroy anti-nutrients. Most sinking diets are extruded using low-pressure steam, high-pressure steam conditioning prior to pelleting resulting in an expanded or extruded diet that is comparatively low in density and that consequently ﬂoats or slowly sinks. Floating pellets were ﬁrst developed for catﬁsh in 1960 (Avault 1981) and became popular because ﬁsh could be observed feeding. Floating, expanded pellets also have superior water stability properties, are more easily digested and can incorporate higher levels of oil than sinking pellets (Jackson 1988; Jobling 1993; Jobling et al. 2001). Disadvantages include costs and high losses of some vitamins during processing. While it is obvious that sinking pellets should be used for species that use tactile means to locate their food (e.g. sturgeon; Romanycheva & Barybina 1979) or that remain on the cage bottom to feed (e.g. turbot, halibut) the question of whether to use sinking or ﬂoating feeds in cage ﬁsh farming remains a matter of debate. Two factors are important: minimizing losses and ensuring as many

Management

213

Fig. 6.9 Flow diagram to illustrate industrial manufacture of dry, pelleted feeds (courtesy K. Jauncey).

of the caged ﬁsh population as possible have access to food. This is particularly important when feeding ﬁsh in cages. Cages have smaller surface to volume ratios than other systems, food is rapidly carried out of reach by currents and, because of the depth of the cage bag, light levels may be low throughout much of the system. Environmental conditions, especially light levels, change with season and time of day, affecting vertical ﬁsh distribution. In a study of feeding in caged Atlantic salmon involving the use of light and normal (dark) coloured pellets, Petrell & Ang (2001) found that ﬁsh prefer to eat at low light levels, provided the pellets are readily accessible and detectable. Stocking density, body size and age affect social structure and access to food (Anras et al. 2001). Finally, food type is important, as is method of feeding (see section 6.2.4 below). Coche (1979, 1982) argued that losses of ﬂoating pellets are more easily controlled than losses of sinking pellets, and concluded that ﬂoating feeds are superior for tilapias reared in small cages. Konikoff & Lewis (1974) claim that as far as caged channel catﬁsh are concerned, ﬂoating pellets give a more uniform size/weight frequency distribution at harvest than sinking pellets, although the densities of ﬁsh used in the experiment may have inﬂuenced results. Conversely, Newton (1980) states that catﬁsh grown in cages in clear water do not come to the surface to feed while Schmittou (1993) claims that catﬁsh, common carp and tilapia all readily learn to take either type of food. With regard to trout, sinking feeds are said to be superior and result in less size variation among stock, it having been observed that dominant ﬁsh monopolize food areas. Other arguments relate to food quality. However, while the availability of the carbohydrate

214

Chapter 6

Table 6.3 A summary of the advantages and disadvantages of different types of diets. Type of feed

Composition

Advantages

Disadvantages

Trash ﬁsh

Trash ﬁsh species, scraps from ﬁsh processing. Fresh/frozen; whole/chopped/minced

Cheap ingredients; highly palatable

High moisture content, expensive to transport and store, variable quality, high pollution loadings. Anti-nutrients may be present. Coloration of farmed ﬁsh may be a problem

Moist pellet/dough

Trash ﬁsh and binder meal which includes vitamins, minerals and added protein

Improved nutritional and water stability properties over trash diets

Requires regular fresh/ frozen ﬁsh supply, must be used immediately when pelleted. High pollution loadings. Limited choice of feeding system

Stable in water, convenient, consistent quality, relatively low transport and storage costs, long shelf-life

Expensive. Unpalatable to a few species

Dry diets

fraction of steam-pelleted ﬂoating feeds can be increased to such an extent that liver function in salmonids is impaired (Hilton et al. 1981), problems can be avoided by maintaining dietary carbohydrate inclusion levels of less than 20% (Jobling 1993). Studies suggest that the suitability of pellet type depends on site, species, cage size and stocking density. Advantages and disadvantages are summarized in Table 6.3.

6.2.3

Storage of feeds

Since feeds are usually delivered in bulk, most cage ﬁsh farming operations require storage facilities. Storage facilities must maintain feed quality: humidity, heat, insects, rodents, fungi, dirt and other contaminants can destroy or greatly damage materials, rendering them unpalatable, less nutritious and even toxic to ﬁsh (Jobling 1993; Schmittou 1993). Larger cage farms in the Philippines sometimes have guard houses on the water where feedstuffs are stored (Fig. 6.10). Boats, moored alongside cage rafts, may also serve as feed stores. Storing unprotected feed bags on walkways is not recommended as it attracts gulls and other opportunistic birds that take feed from bags, spilling a great deal in the process. The last decade has seen the introduction and widespread use of purpose-built feed barges by the salmon farming industry in Western Europe (for review see

Management

215

Fig. 6.10 Guard house at a lake-based tilapia hatchery and ﬁsh pen/ﬁsh cage operation, Laguna de Bay, Philippines.

Beaz Paleo et al. 2000) (Fig. 6.11). Feed barges, made by a number of companies, typically hold 50–400 t of food and deliver up to 80 t of feed each day. Many are equipped with cranes, silos, generators and fuel tanks. They have centralized feed distribution systems that deliver food directly to the cages via blowers, thereby minimizing the need to handle food, representing substantial savings in production costs. They are highly stable in poor sea conditions. Beaz Paleo et al. (2000) also argue that such barges, which hold almost all of the food below the water surface, help store feed in relatively cool conditions in Mediterranean climates. Trash ﬁsh may arrive at the farm in a frozen or unfrozen state and since ﬁsh spoils rapidly it should be checked for freshness before being stored, smell and appearance being reasonable indicators of quality. If ﬁsh are put into cold storage they will keep for many months and remain free from pest damage. Temperatures must be low enough to prevent oxidation of fats and the longer the ﬁsh are kept, the colder temperatures must be. Facilities for extended storage are expensive and costly to run. An alternative is to ensile ﬁsh, which can prove cheaper than freezing and can be used to produce a highly palatable moist feed (Raa & Gildberg 1982; Raa et al. 1983). Silage is mixed 60 : 40 or 50 : 50 with a commercial binder meal containing extra protein, vitamins and a binding agent to form a water-stable moist pellet with a shelf-life of several days.

216

Chapter 6

(a)

(b) Fig. 6.11 (a) Feed barge, Shetland. Note feed hopper and control room on deck and feed distribution lines extending from the barge to the feeders (courtesy J. Deverill). (b) Feed being transferred to feed barge (courtesy J. Deverill).

Dry feeds are delivered either in bulk by road tanker or, more commonly, packed in plastic-lined 25-kg or 50-kg sacks stacked on pallets. Bagged dry compound feeds and other feedstuffs must be kept in clean, dry, cool facilities used solely for that purpose in order to avoid contamination with materials such as insecticides and pesticides, pharmaceuticals or petroleum products. Both temperature and humidity have a great inﬂuence on the rates at which chemical changes take place and on the growth of fungi and insects. High humidity is conducive to the destruction of vitamin C, while the combination of high humidity and high temperatures increases peroxidation with the consequent destruction

Management

217

of vitamin E and other fat-soluble vitamins (Wilson 1991; Hawkins et al. 2002). Insects and rodents not only eat considerable quantities of food, they also spoil much more than they consume through spillage and contamination. Faecal contamination of feeds is a source of Salmonella infection in farmed ﬁsh (Freidman & Shibko 1972). Several fungae that invade feedstuffs produce mycotoxins and some feedstuffs, such as cottonseed meal, contain compounds that exacerbate mycotoxicosis in ﬁsh (Wilson 1991; Tacon 1992). Aﬂatoxins secreted by Aspergillus spp. produce hepatomas in salmonids and cause a haemorrhagic syndrome in tilapias (Roberts & Sommerville 1982; Wilson 1991). Small quantities of propionic or formic acid can be added to diets to inhibit mould growth. Deterioration of stored dry feeds and feedstuffs can be minimized as follows (see also New et al. 1993; Jauncey 1998): • • •

• •

•

Feeds should be examined before being put into the store. Any infested materials should be disposed of or fumigated ﬁrst to avoid contamination. Feeds should be stored in bags and raised off the ground – pallets often sufﬁce. Ventilation in storage facilities should be maximized and temperatures kept as low as possible. Galvanized iron sheeting is not recommended as a building material in hot climates. Spilled material should be swept up. Rodents and insects must be kept under control. Traps and not poisons should be used. Double-skinned buildings in the tropics may become infested and should be regularly inspected. Feeds should be checked frequently, and any change in colour and texture (lumpiness indicates fungal attack), friability, or smell (staleness, rancidity) taken as evidence of spoilage. Moisture content should also be assessed. The moisture content of dry feeds is around 10%. If it increases to 13–16%, then it becomes much more susceptible to spoilage. The presence of moulds, insects and rodents must be carefully watched for, and infested materials should be isolated as soon as possible and fumigated.

The shelf-lives of feeds are summarized in Table 6.4. Fish fed contaminated feeds may exhibit abnormal behaviour, loss of appetite, poor growth and loss of condition.

6.2.4

Feeding

There have been relatively few studies of food intake by caged ﬁsh (Juell 1995; Anras et al. 2001; Petrell & Ang 2001; Andrew et al. 2002), with the result that there is still a poor understanding of how feeding methods and feeding rates interact with environmental factors and stocking densities to determine food intake, growth and production. This affects production efﬁciency, proﬁtability and environmental impacts. Fish will not feed, or should not be fed, during extremes of cold and heat or during rough weather. In the former USSR, cages of common carp are overwintered below the ice surface (see Fig. 7.33) and are not fed until the spring

218

Chapter 6

Table 6.4 Summary of storage and shelf-life of various feeds. Specialist publications (e.g. Wilson 1991, Tacon 1992) should also be consulted. Moreover, many commercially manufactured feeds have expiry dates printed on the bags. Feed type

Storage and duration

Dry, supplementary feedstuffs (e.g. rice bran, wheat middlings)

Providing mositure content is <10% and materials are stored in a cool, dry and pest-free environment, they can remain in storage for several (tropical) to many (temperate) months

Trash ﬁsh (frozen)

High fat content – up to 3 months at -20°C Low fat content – 1 year or more at -20°C

Silage

6–8 months, providing it has a low fat content and contains sufﬁcient antioxidants

Pelleted, commercial intensive feeds

In temperate countries where there are usually surpluses of antioxidants and vitamins present, feeds can be kept for 9 months or more. In the tropics, where feeds often do not have the same levels of antioxidants or vitamins and where higher temperatures and moisture conditions prevail, feed should be stored for as little as possible – 1–3 months

when the ice melts and the cages are lifted (Martyshev 1983), while in Western Europe, feeding of Atlantic salmon may be temporarily halted if temperatures rise above 18°C. In Japan, cages of yellowtail are sometimes lowered below the sea surface during typhoons, and while some cages are adapted so that feeding may continue during periods of submersion (see Fig. 7.29), others are not. Feeding can be done by hand or by one of three types of feeder: • • •

automatic, sometimes referred to as ﬁxed feed ration systems; demand, or self-feeders; interactive feedback systems.

Hand feeding Most cage farms that use supplementary feeds, trash ﬁsh or moist feeds, feed by hand as it is cheaper and easier, at least on a small scale. Frozen trash ﬁsh is sometimes thawed ﬁrst, either at ambient room temperature or by pumping sea water over the blocks, chopped and minced if necessary, and simply broadcast over the surface with a shovel or scoop. However, research in Japan suggests that feeding frozen trash ﬁsh to caged stock reduces bacterial infections and water pollution (Taniguchi 1983). This practice prevails at tuna farms in South Australia, where blocks of frozen pilchard are fed (Fig. 6.12a). Dry supplementary feeds and moist pelleted feeds are also usually fed by hand. Moist balls of feed, which are sometimes used in semi-intensive cage culture, can either be dropped into the cage or placed on the top net of the cage and gently lowering the net into the water (see Fig. 6.8).

Management

219

(a)

(b) Fig. 6.12 (a) Feeding caged tunas with frozen pilchards, Australia. (b) Feeding caged snappers, East Flores, Indonesia (courtesy M. J. Phillips).

220

Chapter 6

The principal advantage claimed for hand feeding is that it is possible to assess ﬁsh appetite and adjust feeding rates, at least for species that come to the surface to feed (Fig. 6.12b). The health status of stock can be readily monitored, since sick or stressed ﬁsh tend to lose appetite. Thorpe et al. (1990) observed that hand feeding results in a more even distribution of food among salmonids (see also below). For cage farms rearing ﬁsh semi-intensively, hand feeding may indeed be best since operations are usually small and the quantities and qualities of the materials used may be highly variable. However, hand feeding is labour intensive and is not always the best way to deliver dry food at large intensive farms or if farmers are rearing ﬁsh on a part-time basis. Only a proportion of ﬁsh appear at the surface during feeding. Ang & Petrell (1997) used mechanical feeders to feed ﬁsh to satiation according to comparisons of observations of surface feeding behaviour (control) and sub-surface feeding behaviour made by video camera. They found that using camera monitoring of sub-surface feeding behaviour resulted in improved food conversion, growth and production compared to the control. Automatic feeders With automatic feeders a measured feed ration is delivered at pre-set intervals. There are a number of different types of automatic feeder, designed to handle either dry or moist pellets. Dry pellet automatic feeders typically consist of a plastic feed hopper surmounting a triggering device that releases or ﬁres a quantity of pellets each time it is activated, and are usually operated either by batteries or by compressed air. The former type is shown in Fig. 6.13.

Fig. 6.13 Automatic feeder suspended over the centre of a small cage.

Management

221

Electrically operated systems either consist of a centralized power supply with control unit linked to a number of feeders or have feeders that have their own power source and control device and operate independently. The former system typically uses a 12 V battery that operates a dozen or so units. The control unit determines the frequency and length of the impulses sent to the feeders so that the number and duration of feeds can be programmed. A photocell ensures that the feeders only operate during daylight. In addition, information on temperature, DO, salinity and currents can be incorporated to modify the feeding regime. The system is commonly installed on a group of cages, cabling being run from the battery to the feeders along the walkways and cage superstructure. The feeders are suspended above the middle of the cages on poles. Electrical impulses from the battery drive a vibrating plate ﬁtted to the base of the hopper, releasing feed into the water. When the hoppers are empty the feeders can be pulled to the sides of the cage and reﬁlled. Because all feeders are linked to the same power source and timing device, all ﬁre simultaneously and for the same duration. Feeders that operate independently cost considerably more but are ideal for small numbers of cages or at farms where cages are not grouped together. Compressed air systems can be used where cages are moored close to the shore and can thus utilize a three-phase, high-voltage power supply needed to drive the compressor. Feed barge-based feeding systems use a diesel generator to drive the compressor. In some systems, pressurized bursts of air, controlled by a timing device, are fed via high-pressure pipelines run along the walkways to the feeders on the cages. The feeders are attached to the sides of the cages. They usually consist of a hopper from which pellets fall into a horizontal delivery pipe connected to the compressed air line, sometimes via a subsidiary air chamber attached to the feeder. When air is released from the air chamber, pellets in the delivery pipe are blown out into the cages. The amount of feed delivered is determined not only by the frequency with which air is released from the chamber, but also by the size of the aperture through which the feed falls into the delivery pipe. More common these days, especially at large-scale cage farming operations, are systems that operate from feed barges, often incorporating some sort of feedback on food consumption to regulate feed (see below). Compressed air feeders, operated from a boat, have been used to distribute feed at large, offshore cages. While this is less labour intensive than hand feeding, monitoring of sediments suggests that much feed is wasted, resulting in higher FCR values and feed costs, and a deterioration in environmental conditions. Pressurized water systems have also been used to deliver food from a centralized unit to individual cages along pipelines. However, there have been a number of problems, including poor growth and mortalities among stock, some of which have been attributed to loss in food quality. For example, two-thirds of vitamin C may be lost from pellets after only 10 seconds of immersion in water (Slinger et al. 1979). The simplest type of automated moist pellet feeding system used at cage farms involves little more than transporting pelleting equipment on a boat and pelleting the food directly into the cages. Such systems, although costly in terms of man-hours and fuel, were once fairly common at Norwegian and Japanese cage farms.

222

Chapter 6

Fig. 6.14 Cross-sectional view of home-made demand feeder for cage culture of tilapia with side view of support structure superimposed. Key: a = demand feeder; b = PVC support structure; c = cover; d = 19-l plastic bucket; e = feed chamber; f = brass nut and bolt; g = plastic funnel; h = horizontal brass rod; i = key ring swivel; j = brass nuts; k = plexiglass plate; l = vertical brass rod (redrawn from Hargreaves et al. 1988).

Demand feeders Demand feeders, or self-feeders as they are sometimes called, work on the principle that the ﬁsh are able to regulate the level of feeding. Demand feeders are also relatively inexpensive. The simplest types consist of a feed hopper ﬁtted with a plate that is connected to a pendulum rod that projects down into the water (Fig. 6.14, 6.15). The feeders are usually mounted over the middle of the cage and when the pendulum is touched, the plate moves, releasing small quantities of food. In electronically controlled systems, the pendulum motion generates an electric pulse that is transmitted to a feed dispenser. The merits of using demand feeders for different species of ﬁsh have been reviewed by a number of authors (Avault 1981; Boydston & Patterson 1982; Hargreaves et al. 1988; Lovell 1989; Alanärä 1992a, b; Juell et al. 1993; Thomassen & Lekang 1993). While a number of ﬁsh are known to be able to learn how to operate demand feeders (Andrew et al. 2002), it is not known whether all species are able to do this. Moreover, it can take several weeks to learn. Proponents claim less size variability among stock, better FCRs, faster growth, higher production, improved water quality and fewer disease problems. Many claims, however, have not been

Management

223

Fig. 6.15 Simple demand feeder in operation at Lake Kossou, Côte d’Ivoire, in a 6-m3 cage (from Coche 1979).

rigorously tested or apply to particular species, circumstances and methods of culture. Demand feeders seem to work well with channel catﬁsh and rainbow trout, although time restrictions on food availability may improve FCR values (Alanärä 1992a, b). Hargreaves et al. (1988) found no signiﬁcant differences in growth rates between hand-fed and demand-fed groups of tilapias. For Atlantic salmon, evidence is contradictory. Thomassen & Lekang (1993) claim growth is suppressed in comparison to groups fed by automatic feeders. Juell et al. (1993), using a different system, claim superior growth although poorer FCR values among stock. Andrew et al. (2002), again using a different system, found improved growth and production efﬁciency. Like hand feeding, demand feeders operate on the principle that when a ﬁsh is no longer hungry it will stop feeding and that the ﬁsh knows best when to stop (see Blyth 1992 for discussion). This assumes that (a) triggering of the feeder is entirely a function of appetite and that (b) appetite is largely controlled by the hypothalamus, which responds to stretch receptors in the stomach wall or fore-

224

Chapter 6

end of the gut and possibly to blood sugar levels (Fange & Grove 1978). Thus, appetite and feeding are suppressed as the stomach or fore-end of the gut ﬁlls with food, and returns as food passes into the intestine or hind-end of the gut. However, these mechanisms have evolved to deal with food encountered in the wild and if ﬁsh are fed to satiation with artiﬁcial food that has much higher protein and lipid content then incomplete digestion may occur, resulting in poor FCRs and increased waste production (Goddard & Scott 1980; Jobling 1986). The problem, however, may be addressed by adjusting the amounts delivered when the feeder is triggered, by ﬁlling the feeder with a ration sufﬁcient to supply the caged stock for one day or by restricting the time during which food is available from the feeder (Alanärä 1992a, b). At exposed cage sites, currents or waves may trigger the pendulum, releasing more food than ﬁsh can consume (Pitt et al. 1977). However, Coche (1979) and Balarin & Haller (1982) report that this can be turned to advantage in tilapia cage culture by loading hoppers with small quantities of food that, when triggered by waves, release the ration continually over a period of several hours, a feeding pattern suited to their feeding habits. Demand feeders also tend to have a small hopper capacity, requiring frequent ﬁlling. Interactive feedback feeders Interactive feedback feeding systems employ a range of devices to monitor uneaten food or ﬁsh feeding activity, thereby allowing feeding rate to be matched with appetite (Juell 1991; Blyth 1992; Bjordal et al. 1993; Blyth et al. 1993; Juell et al. 1993; Foster et al. 1995; Ang & Petrell 1997; Myrseth 2000) (Fig. 6.16). All systems rely on some method or other to detect uneaten food, including camera, Doppler or infrared detection systems or systems that pump water up from the ﬂoor of the cage. In simpler systems the information is used by farm operators to stop feeding the ﬁsh. In more sophisticated integrated interactive feedback feeding systems, observations of uneaten food are interpreted by software that controls when the next meal will occur and the level of ration that will be fed. The feeding system thus track ﬁsh appetite. While such systems might appear to suffer from the same deﬁciencies as hand feeding or demand feeders, i.e. the ﬁsh are the best judge of when they have had enough to eat, the systems measure the response of the whole cage population rather than the few ﬁsh observed feeding at the surface or the few dominant ﬁsh within a population that may be responsible for the majority of feeding events. While there are still arguments to be resolved about appetite and feeding, interactive feedback feeding systems have been shown to be highly effective in reducing FCR values with a range of ﬁsh species. Myrseth (2000) analysed the performance of interactive feedback feeding systems in commercial situations in Norway. Using only a camera, food conversion was improved by 8.3% compared to controls, while one of the more sophisticated integrated interactive feeding systems reduced FCR values from an average of 1.21 to 0.97, an improvement of 20%. Although expensive to install, such systems are sufﬁciently cost-effective to have become widely adopted at salmon farms throughout the world.

Management

225

Fig. 6.16 Schematic diagram of the Aquasmart adaptive feeding system and PC communications (from Blyth et al. 1993).

Conclusions, problems and new initiatives Until recently there had been little study of feeding behaviour of caged ﬁsh. Research has focused on the salmonids and suggests that marked improvements in feed conversion, in waste losses to the environment and in production economics can be achieved through a better understanding of feeding behaviour (Phillips 1985; Kadri et al. 1991; Blyth 1992; Thorpe & Huntingford 1992; Blyth et al. 1993; Smith et al. 1993; Fernö et al. 1995; Jobling 1995; Beveridge & Kadri 2000; Petrell & Ang 2001; Andrew et al. 2002). Although there remains some controversy about the interpretation of results, marked seasonal changes in appetite, related more to changes in day length than temperature, exist. Diurnal patterns in feeding behaviour, which change with season, are also apparent. The use of artiﬁcial lights to modify diurnal feeding patterns, extend feeding periods, reduce early sexual maturation, increase growth and improve FCRs and ﬂesh quality is currently being explored in salmon (Oppedal et al. 2001). Few feeders are yet designed with caged ﬁsh in mind. Goddard & Scott (1980) speculated that owing to the relatively small surface area to volume ratio of cages, compared with ponds or raceways, caged surface-feeding ﬁsh should be fed over longer periods of time. Although it is known that the more restricted or defensible the food supply, the greater the competition and the more pronounced the disparity in food acquisition among individuals (Jobling 1995), little attention has been paid to the importance of distribution of feed, most feeders dispersing feed over a comparatively small area of the cage. Thomassen &

226

Chapter 6

Lekang (1993) compared the growth of groups of Atlantic salmon fed using three types of automatic feeder that distributed food over 1%, 5–8% and 30–40% of the cage surface area. Results show that growth rate was strongly correlated with feed distribution area. However, there is also circumstantial evidence from FCR values and other data that feed losses from cages are higher than from other systems (see Chapters 2 and 5) and this has been attributed to feed being rapidly carried out of the cages by water currents and to poorer visibility at depth. Thus, feeding methods may also have to take into account the environmental conditions that prevail at a site. Thomassen & Lekang (1993) also investigated novel feed delivery systems for large salmon cages (90 m circumference, 20 m deep) in which feed was delivered by tube at depths of 1.5 m and 4 m below the water surface. Despite the limited nature of the trials, there was strong evidence that feed delivered at a depth of 1.5 m resulted in signiﬁcantly higher growth rates than feed delivered either at the surface or at 4 m. Several studies have investigated the use of feeding rings, trays or net curtains to reduce feed losses. Plastic feeding rings, 25–150 cm in diameter, are ﬂoated on the surface in the centre of the cage so that feed is prevented from ﬂoating out through the walls of the cage. Cage walls near the surface may also be fabricated from ﬁne mesh materials. Both are used in conjunction with ﬂoating or slow sinking pellets and, while there is evidence that they may be effective under certain circumstances (e.g. in fast-ﬂowing waters), they have not been thoroughly evaluated. The solid-bottomed cages used in marine ﬂatﬁsh culture not only facilitate feeding but also prevent loss of feed, and feeding trays, typically suspended 20–100 cm below the cage surface, operate in a similar manner. However, like feeding rings, they have limitations. Coche (1979), for example, has expressed concern that small trays that only partially cover the cage area will beneﬁt the more aggressive ﬁshes while Ibrahim et al. (1976) observed that trays provide an unwanted spawning substrate for tilapias. Other designs for use in small cages are discussed by Schmittou (1993).

6.3 6.3.1

ROUTINE MANAGEMENT Monitoring water quality

Water quality parameters relevant to cultured ﬁsh are discussed in section 4.1. Monitoring is particularly important at intensively managed cage farms that have a high production in relation to the size and nature of the site. Monitoring can help: • • • • •

avoid losses caused by changes in water quality; evaluate siting and conﬁguration of cages; maintain optimum stocking and feeding rates; help evaluate stress levels among caged stock so that operations that might exacerbate stress, such as grading, can be avoided; gain information of long-term changes in water quality at a site so that proposed changes in production may be properly evaluated (see Chapter 5).

Management

227

The most important data are those on DO and temperature. Ideally, measurements should be made daily and at times when they are likely to be highest and lowest (i.e. at dawn and mid-day, and at slack tide), and readings both inside and outside of the cages and at the cage surface and bottom should be made. However, this is too time consuming for most farms; instead it is suggested that periodic measurements be taken throughout the year but that the frequency of observations is increased during warm periods. A single temperature and oxygen determination from the middle of a cage within a group is sufﬁcient to indicate levels experienced by stocks. Data on nitrogen (ammonia, nitrate, nitrite) and dissolved phosphorus levels, pH, Secchi disc and chlorophyll give a more complete picture of what is happening in the cage farm environment, alerting farmers to dangerous levels of toxins (ammonia, nitrite) and to the effects of the farming operation on algal populations (chlorophyll levels, Secchi disc). Again, tests should be carried out at regular intervals, but are of particular importance during periods of calm, hot weather. While pH and Secchi disc readings can be readily made using simple equipment, other tests rely on chemical methods but are nevertheless straightforward, requiring minimum skills and facilities. Details of sampling methods, sample storage and treatment are given in Mackereth et al. (1978), Boyd (1979), Parsons et al. (1984), Stirling (1985) and APHA (1989). Portable test kits, suitable for use by ﬁsh farmers, are widely available (see Boyd 1980).

6.3.2

Fish husbandry and management

Weight/size determination Samples of ﬁsh should be taken at regular intervals and weighed so that the growth of stock can be monitored. The information is needed to determine stocking and feeding policies, and when harvesting should be carried out. Information is usually gleaned from sample weighings, a stressful and inaccurate method in which ﬁsh are crowded by lifting the cage bag and samples are netted, anaesthetized and weighed and measured, and the data used in conjunction with records on mortalities to estimate cage biomass. While estimates of 5% are acceptable for most purposes, bias and inaccuracies confound efforts so that estimates are often in the order of ±15–25% (Naiberg et al. 1993; Petrell et al. 1993; Treasurer 2000b). Freeze branding or tagging is also used at some salmon farms to identify siblings or facilitate assessment of growth of individuals. Optical technology-based systems for measuring the sizes of individual ﬁsh are available, although they require ﬁsh to swim through a sensing enclosure, a system not ideally suited to all ﬁsh species or cage systems. Non-invasive video and hydroacoustic systems that monitor ﬁsh size, distribution and biomass within cages are increasingly widely used in intensive ﬁsh farms (Dunn & Dalland 1993; Naiberg et al. 1993; Petrell et al. 1993; Ruff et al. 1994; Beddow et al. 1996; Treasurer 2002b). Manufacturers recommend that something like 500 ﬁsh have to be sampled to achieve an accuracy of 95% (Treasurer 2002b).

228

Chapter 6

Fig. 6.17 Grading ﬁsh with a hand grader (from Edwards 1978).

Grading Total biomass and size variations among caged populations increase with time: if ignored, growth, food conversion and water quality may suffer. Separation of stock into different size classes also facilitates production planning and reduces post-harvest grading. Redistribution is usually based on size differences within stocks. The simplest method of assessing ﬁsh size is by eye; however, at most intensive operations much of the routine grading is mechanized. Hand graders may be used to grade small ﬁsh (Fig. 6.17) while automatic graders are used at large on-growing sites (Fig. 6.18). Machines typically consist of a hopper, into which the ﬁsh are loaded, mounted above a sloping set of bars that is designed so that the gaps between adjacent bars increase from top to bottom. Fish fall from the hopper, slide along the bars and fall through into tanks whenever the gaps are wide enough. Most machines can handle ﬁsh of 50–500 g and can be adjusted to sort ﬁsh into four or ﬁve different size groups. The operation can be greatly speeded up by using

Management

229

Fig. 6.18 Fish grading machine sitting on a cage raft.

live ﬁsh pumps to transfer ﬁsh from the cage to the grader and the graded ﬁsh to new cages. Grading and moving of caged ﬁsh, however, is stressful and time consuming (Flagg & Harrell 1990; Wagner & Driscoll 1994; Lazur 1996; Huguenin 1997). In-cage grading systems are much less stressful to stock and are beginning to be used in Norway and Scotland (Fig. 6.19) (Karlsen 1993). Disease monitoring Monitoring of ﬁsh stock health is essential, not least because it is an indication of poor state of welfare, often an indication of some poor underlying problem with the environment or farm conditions. However, as Roberts & Shepherd (1997) state, ‘Before a ﬁsh farmer can hope to diagnose and control disease outbreaks, (s)he must be able to appreciate when the stock is healthy and thriving’. This can only be learned through experience. Often the ﬁrst signs that something is wrong can be surmised from changes in behaviour. Farmers must become accustomed to observing their ﬁsh without undue disturbance and form a general picture of how stock behave under the usual cycle of environmental conditions at the site, i.e. at dawn/mid-day/dusk; slack/ﬂood tide; feeding/non-feeding. Feeding behaviour in particular is a good indicator of health. If a problem is suspected, samples of ﬁsh should be taken for further examination; changes in appearance (deformed spine), skin (colour, presence of lesions, rashes, spots or lumps, excessive mucus), eyes (bulging eyes, cloudy lens), ﬁn and tail (erosion) all being signs that something is amiss. In Scotland and Norway industry Codes

230

Chapter 6

Fig. 6.19 In-cage grading system. The screen is pushed in the direction of the arrow, ﬁsh larger than the mesh size of the screen becoming separated from the smaller ﬁsh (redrawn from Karlsen 1993).

of Practice necessitate regular inspection of ﬁsh to determine sea lice infestation levels, treatment being necessary when the number of lice increases beyond a threshold value. While some indication of disease status can be gained from examination of moribund ﬁsh netted from the surface, studies at Canadian salmon farms have shown this to be an unreliable indicator of disease status, in terms of both incidence and type of disease (Stephen & Ribble 1995). More information on disease diagnosis and appropriate treatment methods is given in Chapter 7. Removal and disposal of dead ﬁsh Disease outbreaks apart, unexplained deaths always occur at ﬁsh farms. Removal of dead ﬁsh is not only a precaution against the spread of disease, it also helps reduce phosphorus and nitrogen waste loadings (Beveridge 1984a). In Norway in 1991, for example, some 20 000 t of dead ﬁsh were recovered from farms (Kaasa 1995). Removal involves not only netting out ﬁsh ﬂoating on the surface but also daily lifting of nets and removal of dead ﬁsh lying on the cage bottom. Because lifting nets is both time consuming and stressful – even injurious – to stock, divers are sometimes used. A number of dead ﬁsh collector designs have also been developed (Fig. 6.20). Recording numbers of dead ﬁsh is important as changes in the incidence of mortalities can warn of the onset of a disease outbreak and gives farmers valuable information on the performance of different stocks or management strategies (stocking densities, feeding rates, etc.). Records are also essential for insurance claims. Dead ﬁsh should be removed from the site and carefully disposed of ashore, preferably in a pit and covered with lime – particularly if a disease outbreak is suspected. Since 1989 the Norwegian State Pollution Control Authority has

Management

231

Fig. 6.20 Dead ﬁsh collector. System also recovers uneaten food (redrawn from Ervik et al. 1994a).

taken measures to encourage the recycling of dead ﬁsh and more than 75% of mortalities are now ensiled and used as food mostly in the fur industry. Ensiled ﬁsh contaminated with antibiotics are burnt at a high temperature waste incineration plant (Kaasa 1995). Equipment used to transfer dead or diseased ﬁsh must be disinfected. Harvesting Harvesting of ﬁsh is either done continually or in batches, depending on the production cycle. In some industries, decision-making has become very sophisticated, involving the use of computer-based production planning models (see Bjørndal 1990; Duarte et al. 1994; Boghen 1995). Harvesting of cage farmed ﬁsh evolved from the traditional methods used in the ﬁshing sector. However, tremendous efforts have been put into developing methods appropriate to processing of high volumes of high market value products. Today, many producer organizations, especially in salmon farming, have introduced quality assurance schemes detailing how ﬁsh must be harvested as well as product speciﬁcations. Prior to harvesting, ﬁsh should be starved, partly in order to ﬁrm the ﬂesh and partly to give the gut time to evacuate since the presence of food, enzymes and

232

Chapter 6

Fig. 6.21 Harvesting ﬁsh (courtesy J. A. Stewart).

bacteria can accelerate deterioration and increase the risk of carcass contamination during processing (Erikson 2001). Starvation periods are species and temperature dependent; Atlantic salmon, for example, are starved for 10–14 days at water temperatures <10°C and for 7 days at >10°C. Fish can be harvested in situ, or the cages towed to a quay where the netting operation may be carried out more efﬁciently with the aid of mechanized lifting gear, trucks, etc. In Indo-China cages are towed to market (Pantulu 1979). Although few cages have been designed with harvesting in mind, the process is simple: in most cage designs the net is hoisted until the ﬁsh are concentrated in a small volume of water and netted out with dip nets (Fig. 6.21). Power-assisted hoists, which can be transported to the cages and operated from a ﬂoating platform or boat, are sometimes used to operate brailers (Fig. 6.22). Fish may have to be netted out by hand from a boat if cages do not have walkways. Harvesting of the rotating cage shown in Fig. 2.2b is carried out by stretching a net across a cage section and then rotating the cage about its axis (Grave 1975). For most rigid designs, ﬁsh must be crowded at one end of the cage and then scooped out (Huguenin et al. 1981). However, in the design shown in Fig. 6.23, the cage is simply rotated and lifted out of the canal, the ﬁsh being concentrated in one corner. In a number of Southeast Asian countries, species of grouper, snapper, sea bream and even tilapia, are cultured for the live ﬁsh market where they fetch considerably higher prices. At harvest, ﬁsh are transferred directly from the cages to boats equipped with holding tanks. In Hong Kong, ﬁsh are either transported directly to a designated landing point by the grower and sold to the wholesaler

Management

233

Fig. 6.22 Floating platform with power hoist, ﬁsh harvesting bins and grading machine at an Atlantic salmon farm, Ireland.

Fig. 6.23 Rotating rigid mesh cages sitting in a heated water efﬂuent canal at a power station (courtesy I. H. MacRae).

234

Chapter 6

or, alternatively, the ﬁsh may be collected from the farm by restaurant owners and transported live to the restaurants (Wong 1982). Fish are usually either killed prior to dispatch from farms or transported live to a processing plant. Oxygenation may be necessary in well-boats. Fish should be handled gently as undue stress causes a build-up of adenosine triphosphate (ATP) in the muscle that decreases shelf-life. Fish are also easily damaged, causing not only ugly skin blemishes but also bruising of the ﬂesh that hastens deterioration in quality (Connell 1995). All slaughter methods are stressful, some more so than others (FSBI 2002). Netting ﬁsh then dumping them into containers and leaving them to asphyxiate may seem the most practical way to deal with small ﬁsh. However, it is an anathema to many and it increases lactic acid content of the muscles, accelerating the degenerative processes. Killing by electrical stunning is being explored for smaller ﬁsh, such as rainbow trout (Robb & Roth 2003). Larger, more valuable ﬁsh, such as yellowtail and salmon, are usually killed individually. Anaesthetizing with carbon dioxide prior to killing, once widely recommended, is increasingly questioned (Ottera et al. 2001). Alternative techniques, including electrical stunning and lowering of body temperature prior to slaughter, are believed to be more effective in both reducing the physiological stress response and in improving ﬂesh quality (Ottera et al. 2001; Skjervold et al. 2001; Robb & Roth 2003). Killing by bleeding is also increasingly believed to be both stressful and to accelerate the onset of rigor (Ottera et al. 2001). Moreover, there are problems with disposal of blood. Clubbing is time consuming and can cause bruising. Alternative methods of destroying the brain, such as spiking, show greatest promise, resulting in lower plasma glucose and lactate levels than in ﬁsh killed by alternative methods (Ottera et al. 2001; Poli et al. 2002, in FSBI 2002). Purpose-built processing facilities are recommended. Fish should be chilled and dispatched as rapidly as possible after harvesting to ensure the freshness of the product to the consumer. Large ﬁsh are often packed whole between layers of ice in insulated boxes (Fig. 6.24) and dispatched. Sometimes ﬁsh are gutted and cleaned before leaving the farm, while at other farms ﬁsh may be smoked or even dried or frozen prior to dispatch. Although fresh ﬁsh always fetches a higher price than frozen, freezing is sometimes necessary if farmers have been forced to harvest ﬁsh at a bad time or if the market has taken an unexpected downturn.

6.3.3

Maintenance of cages and gear

Irrespective of the damage that can be caused by storms, predators, drifting objects, shipping, boat propellers, poachers and vandals, cage materials have a ﬁnite life-span and will eventually wear out (see Chapter 3). Cages, nets and moorings, therefore, must be checked regularly for signs of damage and wear and tear, and repaired or replaced if necessary, as cages, stock and human lives could be at risk. Increasingly, industry Codes of Conduct, recommending tagging of nets and regular checking of gear, are being developed. Cage nets are checked for damage during cleaning. However, since cleaning may be infrequent (see section 7.4), inspection of net bags in situ should also be

Management

235

Fig. 6.24 Farmed yellowtail packed in insulated boxes with ice, ready for dispatch (courtesy T. Sano).

Table 6.5 Recommended mesh sizes for cage culture of tilapia. Size of ﬁsh

Purpose

Fry (<12 g) Fingerling (12–30 g) 30–200 g 200+ g

Nursery Grow-out Grow-out Grow-out

Mesh size (mm) 1–3 4–8 10–20 20–25

carried out. In clear water this is often possible from a boat or from the cage walkway, while in turbid conditions divers may have to go down and check nets every week or so. Rotating cages, of course, may be readily checked without recourse to using divers, while predator nets may simply be lifted for inspection. Although small tears may be able to be repaired in situ by lifting the damaged side of the net and hooking it on to the cage superstructure until repairs are completed, more extensive damage may mean the bag having to be changed (see later) and repaired on shore. Several manuals give details of net mending methods (see Libert et al. 1987). At most sites, particularly in the marine environment, bags become fouled and, unless a rotating design is used, have to be changed and cleaned. Nets are also often exchanged for those of a larger mesh size as the ﬁsh grow (Table 6.5). Care must be taken over mesh size: if too small water exchange is restricted and if too large some ﬁsh species (e.g. sea bream) will chew the bars of the net. The

236

Chapter 6

frequency of net changing varies from once per week to once per year, depending upon site location, materials used, season, and management and design of cage. In northern Norway one change per year may sufﬁce if farmers wait until after the period of maximum larval tunicate and mussel settlement in July (Sutterlin & Merrill 1978). The frequency of net changing may also be reduced if the top 2 m – the part of the net where fouling is heaviest – of any one side is pulled out of the water and left to dry. At least two people are involved in the net changing process, and although there are variations, the principle is the same for most cages. The rigging is released in order to free one side of the bag from the cage collar. The freed side is then drawn towards the opposite side and gathered up so that the ﬁsh are concentrated in a small section of the bag near the surface, the rigging on the other two sides being released from the collar as necessary. One side of the new bag is then attached to the free corners and drawn under the old bag. The ﬁsh are then gently tipped out of the old bag into the new one prior to the fouled net being removed for cleaning. There are a number of variations on this method. On some farms one or two cages within a group are left empty throughout the summer and each week ﬁsh are gently decanted from one cage into an adjacent empty cage so that the fouled net may be readily exchanged with the minimum of stress to stock (see also Yu et al. 1979). The entire net changing process usually takes from 30 min to 2 h, depending on the degree of fouling (i.e. how heavy the net is), the size and design of the cage, and the weather. As a large farm can have as many as 50 or 60 cages, routine net changing can fully occupy two members of staff for much of the year. Several of the more innovative designs of industrial cages have tried to address this by the partial mechanization of net changing. There are two basic methods used to clean fouled nets: chemical and mechanical. For rotating designs the cage is turned so that fouled panels are lifted out of the water and exposed to the air. The organisms are then left to dehydrate and die, a process that typically takes a week or so in temperate climates (Blair et al. 1982), and then removed by brush. Blair et al. (1982) compared the maintenance time required to reduce biofouling to a minimum on rotating and conventionally designed cages. They estimated that the work involved with the rotating design was reduced to around 5% of that required to maintain the conventional cage. In the past, ﬁsh farmers have stacked fouled nets in a heap, covered them with black plastic sheeting and left the organisms to decay prior to cleaning, while others have recommended that marine nets be submerged to allow starﬁsh and other predators to remove mussel encrustations (Sutterlin & Merrill 1978). At most farms today, however, the practice is to hang or lay the nets out to dry for a few days (Fig. 6.25), making them easier to clean. Hard bristle brushes, sticks, or high-pressure hoses are used to dislodge adhering material, the latter method being more effective at removing some of the more stubborn fouling organisms such as ascidians. If high-pressure hoses are used, then cleaning should be carried out in an area with a sloping concrete ﬂoor (Fig. 6.26) so that the debris and water can drain away. In the past, chemical cleaning methods, such as soaking nets for 2–3 days in a 3% formic acid plus 9% copper sulphate solution, or for

Management

237

Fig. 6.25 Cage bag laid out to dry.

Fig. 6.26 Cleaning cage bag with a power hose.

3 h or so in a sodium hypochlorite solution, have been used (Sutterlin & Merrill 1978; Møller 1979). While this apparently dissolved mussel byssal threads, such methods are no longer recommended. Net washing machines are also increasingly commonly used. Although expensive to buy they can clean a net in less than 1 h, saving considerably on manpower. In situ automated cage cleaning devices have been developed (see Fig. 6.27), but have proven ineffective, leaving behind a surface that is only partly cleaned and that is rapidly re-colonized (Hodson et al. 1997).

238

Chapter 6

Fig. 6.27 In situ automated cage net cleaning system, Australia (developed from Hodson et al. 1997).

(a) Fig. 6.28 (a) Cu–Ni cage after four years’ immersion in temperate marine conditions (courtesy J. E. Huguenin). (b) Pressure cleaning a Cu–Ni cage (courtesy J. E. Huguenin).

Cage nets are costly and easily damaged, and are particularly vulnerable during cleaning. Secretan (1979) estimated that 4–5% of losses reported by cage ﬁsh farmers concerned escapes of ﬁsh through holes in nets (see section 5.3.4 for more recent estimates of losses of ﬁsh through holed nets). Nets not in use should be carefully stored in clean, dry conditions. New cage nets that have been treated with antifouling compounds may require to be soaked before for at least 24 h in clean ﬂowing water prior stocking with ﬁsh (see also section 7.4). If nets are

Management

239

(b) Fig. 6.28 Continued.

re-treated with antifoulants, farmers must follow the manufacturer’s recommendations before using them again. The rate of biofouling on PP (polypropylene) or metal alloy rigid mesh cages is much slower than on net cages and cleaning needs to be carried out far less frequently (Ansuini & Huguenin 1978; Milne 1979; Huguenin et al. 1981). Cages can be cleaned in situ using a hard-bristle brush to dislodge weed and accumulated debris, although care must be taken with regard to accumulation of wastes in the sediments below cages. Ansuini & Huguenin (1978) estimated that cleaning need only be carried out four times per year and take a total of eight man-hours, a saving of more than 90% on the cleaning of conventional cages. Periodically, however, rigid mesh cages must be lifted, checked for damage and repaired if necessary. While undergoing maintenance they should be thoroughly cleaned (Fig. 6.28). Moorings must be checked regularly by divers, particularly after storms. Mooring lines should be kept relatively free from fouling and worn shackles should be replaced. Again, there are increasing industry-led moves to include mooring maintenance details in Codes of Practice.

Cage Aquaculture, Third Edition Malcolm C. M. Beveridge Copyright © 1996, 2004 by Blackwell Publishing Ltd

Chapter 7

Problems

As discussed in Chapter 2, cages suffer by comparison with other aquaculture systems in being particularly vulnerable to both environmental variables and anthropogenic hazards. The site selection process is concerned with choosing the best environment so that mortalities are minimized, growth and production maximized and the venture is as proﬁtable as possible (see Chapters 4 and 5). Unfortunately it is often not possible to choose the ideal site. Farmers may be faced with trying to rear a species in a sub-optimal climate (e.g. trout in the subtropics; tilapia in Europe). Often, too, the best sites have gone, forcing farmers to choose sites with increased risks. Neither is it always possible to foresee the problems that might occur or to examine the risks as thoroughly as one might like. Damage caused by drifting objects and pollution owes much to chance and it is usually impractical to investigate factors such as fouling. All sites are to some extent a compromise; the price for good water exchange may be currents that deform nets and make moorings expensive and difﬁcult to install, while the trade-off for unpolluted water may be the prevalence of predators. Insurance can cover all risks associated with farming, from the import of fry to dispatch of the ﬁnished product (Secretan 1980; Simmonds 1996; Nash et al. 2000), and can be categorized into: • • •

what is needed; what is desirable; what is affordable.

In many countries some cover is mandatory; insurance of employees, for example, falls into this category. Third party insurance, which covers damage caused by a ﬁsh farmer’s operation to someone else’s property (e.g. drifting cages colliding with yachts), products liability, loss of buildings, cages and stock, and key-man insurance (i.e. medical insurance for manager/owner) are regarded by most as desirable. Insurance costs are typically between 1 and 10% of the value of the insured stock, depending on the extent of cover and the risks involved. Charges are based on the detailed proposal form completed by the client. Companies that specialize in insurance of ﬁsh farms advertise regularly in the aquaculture press. In the following section, some of the problems associated with growing ﬁsh in particular climates or types of site are discussed and advice given to mitigate effects where possible. 240

Problems

7.1

241

CURRENTS

It is generally recommended that currents at cage sites do not exceed 60 cm s-1; much greater deformation of nets, excessive strain on moorings and cage collars, and unacceptable losses of feed and wastes can occur (see section 4.2.3). High ﬂow rates also increase metabolic expenditure on swimming. Flow regimes for caged species remain a matter of debate. According to Russian work, summarized by Privol’nev (1975), ﬂow rates inside cages should not exceed 20 cm s-1; above this, energy expenditure on swimming increases markedly, and growth and survival are adversely affected (see also section 4.2.3). A number of approaches have been taken to overcome problems associated with excessive current velocities. Stocking densities can be reduced to take account of loss of volume through cage bag deformation. An alternative is to use as large a mesh size as possible and to suspend additional weights from the cage walls (see Fig. 3.16). However, increasing vertical loading on the netting greatly increases the horizontal forces acting on the net (Tomi et al. 1979) and may cause tearing, particularly under poor weather conditions. This also does nothing to reduce feed losses or other associated problems. In Indonesia and Vietnam traditional submerged carp and catﬁsh cages used in running water are made from wood (see Fig. 2.2i); that not only does not deform but also greatly reduces the current regime within the cage. In Indonesia, the weight of the cages is sufﬁcient to keep them anchored on the bottom of rivers and canals, even during ﬂoods (Vass & Sachlan 1957). The ﬂoating designs used in Indo-China (see Fig. 1.4) are also well-suited to the problems associated with rearing ﬁsh in running water. In Russia, 6–l0-m3 ﬂoating cages, designed for use in rivers, are constructed from 3–4-cm wooden slats spaced 0.75–2 cm apart, nailed to a stout wooden frame and ﬁtted with a galvanized wire mesh lid (Fig. 7.1). The

Fig. 7.1 Wooden cage for trout culture in ﬂowing waters, Russia (from Martyshev 1983).

242

Chapter 7

Fig. 7.2 Small, plastic-coated steel mesh cage used for experimental culture of rainbow trout in the outﬂow of Lake Hichukhota, Bolivia.

cages are secured by wooden pilings (Privol’nev 1975; Martyshev 1983). When stocked with 20–25-g ﬁsh, netting is ﬁtted to prevent ﬁsh escaping and removed a month or so later. Much of the bottom of the cage is solid and acts as a feeding platform. Cages are cheap to construct, long lasting (10 years) and currents are reduced to acceptable levels within the cage. Rigid mesh cages are used in lotic sites such as irrigation canals (Budhabhatti & Maughan 1993, 1994). In Bolivia, cage culture of rainbow trout was established in the outﬂow of Lake Hichukhota in 2-m3 cages constructed from 1-cm plastic-coated wire mesh attached to a wooden frame (Fig. 7.2). However, the ﬂow of water through the cages made feeding difﬁcult. In Italy and the Netherlands, where ﬂoating plastic-coated wire mesh rigid cages have been used to culture trout and cyprinids in heated water power station discharges, systems of ‘Venetian blinds’ that can be adjusted to deﬂect water past cages during periods of high ﬂow, have been installed upstream (Bronzi & Ghittino 1981; Hogendoorn 1984). Metal mesh cages have also been used at marine sites in France and the Netherlands where currents are high (3–6 knots; 1.5–3.0 m s-1) (Boeuf & Harache 1980; Anon 1985).

Problems

7.2

243

DISEASE

There are many excellent textbooks on ﬁsh diseases (see Roberts & Shepherd 1997; Roberts 2001; Woo et al. 2002). The prevention and treatment of diseases among caged ﬁsh is considered only brieﬂy in this book.

7.2.1

Disease agents

Diseases encountered in ﬁsh may be classiﬁed as follows (from Kinne 1980): • • • • •

genetic; nutritional; environmental, including abiotic factors such as light, temperature, salinity and oxygen, as well as natural and man-made pollutants; physical injuries; parasites and microorganisms.

Many are non-communicable and, if diagnosed correctly and the causes determined, then changes in husbandry and management practices will often eradicate the problem. For example, genetic disorders, such as spinal defects, are comparatively rare and avoided by prudent stock selection. Physical injuries can be minimized by handling ﬁsh with care, use of the appropriate choice of netting materials and mesh sizes, the exclusion of predators and by avoiding overexposed sites where excessive cage motion can damage ﬁsh. Risks of environmental hazards can be greatly reduced by prudent site selection, rational site development and good husbandry. Nevertheless, farmers can and do have environmental problems and have to resort to some form of remedial action. Mitigative measures that can be used against adverse temperature, light, pH and oxygen conditions are discussed in sections 7.6 and 7.9. Nutritional disorders are often difﬁcult to diagnose and treat without specialist knowledge (Tacon 1992; Roberts 2001), and even when available there may still be problems since, for some species at least, research into nutritional requirements is still at an early stage. Using commercial feeds with a proven reputation minimizes the risk of problems although care must be taken over storage and shelf-life of the feed as well as in the method of feeding (see section 6.2). For semi-intensive operations a varied diet of fresh, uncontaminated feeds is least likely to give rise to problems. Communicable diseases are caused by parasites, bacteria, fungi and viruses. Parasites spend part or all of their life cycle extracting nourishment from – and at the expense of – other living organisms. Fish parasites include members of both the protozoa and metazoa, the latter including such diverse groups as acanthocephalans, nematodes, cestodes, trematodes, leeches, crustaceans, bivalve glochidia and even other ﬁshes (lampreys). Most pathogenic microorganisms occur naturally and are widely distributed, living on the dead and decaying organic matter universally present in aquatic systems. The change from saprophyte to pathogen is usually precipitated by a stress factor(s) or wound that increases the susceptibility of the ﬁsh to infection.

244

Chapter 7

Fungi are also saprophytic organisms that are very common in the aquatic environment and under certain circumstances can invade living ﬁsh, often through lesions caused by some other agent. Commercially important fungal pathogens of ﬁsh include Saprolegnia spp., Ichthyophonus hoferi and Branchiomyces spp. (Roberts & Shepherd 1997; Roberts 2001). Viruses are obligate intracellular parasites, incapable of independent replication, that invade cells and direct them to manufacture viral-speciﬁc proteins that are assembled to form new viruses that go on to invade other cells. Infected cells often die. There are around a dozen commercially important ﬁsh viruses, including IHN (infectious haematopoietic necrosis), IPN (infectious pancreatic necrosis), VHS (viral haemorrhagic septicaemia) and SVC (spring viraemia of carp).

7.2.2

Cage farms and disease

Wild ﬁsh can carry a wide range of parasites and other organisms that rarely prove fatal unless the ﬁsh are stressed by some other factor, such as pollution, or the balance between host and pathogen shifts in favour of the latter through the introduction of a novel host or strain of pathogen. Mass mortalities are thus rare in nature. However, the host/pathogen relationship in ﬁsh farms is very different. Fish may be moved from one site to another, carrying the risk of introduction of a new or exotic disease. Moreover, farmed ﬁsh may be subject to chronic environmental stressors such as low oxygen, high temperatures or light levels, or acute stressors such as handling or physical damage, lowering resistance to disease (Pickering 1993, 1998; Schreck et al. 1993), while high stocking densities and the close proximity of farms facilitate the rapid spread of disease. Cage farms appear to suffer by comparison with other systems. The establishment of a cage ﬁsh farm in a water body can disrupt parasite life cycles by increasing the number of hosts, promoting parasite transmission from wild to caged stocks (McGuigan & Sommerville 1985). The cage structure itself can harbour pathogens. Nets trap parasite eggs, infective larvae and waste food, attracting carrier/reservoir feral ﬁsh (Lio-Po & Lim 2002). Free-living stages of sea lice (Lepeophtheirus salmonis), for example, can live in heavily fouled cage nets, acting as a reservoir of infection. The risks of physical damage in cages may also be greater than in land-based systems. Nets treated with certain antifoulants are hard and can abrade skin, particularly during storms or netchanging operations. Fin erosion has been attributed to damage caused by cage nets (Boydstun & Hopelain 1977; Moring 1982; Furevik et al. 1993), although there is some debate with regard to the role of stocking density and the effects of selective breeding on behavioural traits that lead to aggressive interactions (Soderberg & Krise 1987; Ruzzante 1994). Disease remains an important issue in aquaculture. Few systematic studies have been published comparing disease incidence among wild and caged ﬁsh. Leong & Wong (1988) found a greater variety of parasites and higher intensities of infestation among caged than wild groupers.

Problems

7.2.3

245

Minimizing disease risks

There are a number of measures that can be taken to minimize disease risks. •

• • • •

• •

• • • • •

Some species or strains may be more resistant to disease than others. During cage trials in the Côte d’Ivoire, for example, the local species Tilapia heudelotii proved less susceptible to myxosporidian infections than the more widely farmed Oreochromis niloticus (see Pullin & Lowe-McConnell 1982). Careful site selection is important (see section 4.1.3). Healthy ﬁsh of the appropriate size should be selected for stocking. Size is particularly important when moving ﬁsh from fresh to sea water. Separation of year classes, to different sites if possible, is recommended. Regular observation of ﬁsh should be carried out and unusual behaviour or changes in colour or appearance taken as a sign that something may be wrong. Regular sampling and scrutiny of ﬁsh is also often recommended as part of a disease control strategy (see des Clers 1994). Fish should not be overcrowded. Water quality must be monitored and maintained: site selection and appraisal of environmental capacity are important in this respect. Fallowing of sites is not only useful as part of an environmental management policy (see section 7.8.2) but can also play a part in the control of sea lice L. salmonis on both farmed and wild ﬁsh (Bron et al. 1993; Wheatley et al. 1995). Sites are typically fallowed for several months between production cycles. Food should be as fresh as possible and free from fungal or other contaminants. Fish should not be overfed. Mortalities should be removed regularly and destroyed. Equipment should be regularly disinfected. Predators should be discouraged. Handling of ﬁsh should be done as gently as possible and kept to a minimum, especially during adverse environmental conditions.

Vaccines have been successfully developed for a number of important bacterial diseases affecting salmonids, such as cold-water vibriosis, and have made signiﬁcant contributions both to reducing mortalities and to reducing the use of infeed antimicrobials and their release into the environment (see Fig. 5.8).

7.2.4

Diagnosis and treatment

Despite precautions disease outbreaks occur, but before any action can be taken a diagnosis must be made. A picture of the pattern of mortality can help since diseases develop in different ways; viral epidemics tend to develop rapidly whereas bacterial, fungal and parasitic outbreaks may take many days to become apparent. Although external examination of affected ﬁsh may be sufﬁcient for diagnosis, in most instances post-mortem examination is necessary. Tissue samples may also be required for bacteriology and virology. While experienced farm staff may be able to make a preliminary diagnosis, conﬁrmation often

246

Chapter 7

requires specialist skills and techniques. Laboratories with trained staff are now widely available throughout Europe, North America and Southeast Asia although they are still scarce in Africa and Latin America. Diseased stock may eventually have to be treated with chemicals and there are four basic methods (Roberts & Shepherd 1997): • • • •

bath treatment, in which chemicals are added to the water; direct application of chemicals to the ﬁsh; administration of chemicals in the food; biological control.

Bath treatment Soluble chemicals are sometimes used in bath-form to treat ﬁsh with skin and gill diseases. While this is relatively simple in ponds and tanks, chemicals added to cages are rapidly dispersed and, once used, are usually released into the environment (see Chapter 5) (Dobson & Tack 1991). For this reason bath treatments are banned in some countries and require consents to discharge in others. During treatment the cage bag is lifted to concentrate ﬁsh in a smaller volume of water and enclosed or partially enclosed by a plastic or canvas skirt. Purpose-built enclosures for small cages have also been designed (Kleinholz & Luker 1994). In marine environments treatment is carried out at slack tide. If large numbers of cages of ﬁsh have to be treated the process can be expensive and time consuming. An alternative is to use a small purpose-built treatment cage. The simplest consists of a weighted canvas bag with net liner, suspended from a ﬂoating collar, which is moored alongside the cage of affected ﬁsh. Fish are netted out and transferred to the treatment cage and left for the recommended time before being returned to the production cage. Although this saves on chemicals, the transfer of ﬁsh is stressful and can result in mortalities. Brandal & Egidius (1979) designed a treatment cage in which the bag is attached to the collar on three sides only. One side of the production net is then unfastened, passed under the collar to the outside and sewn on to the freed side of the net inside the treatment cage. Fish are transferred by raising the production net bag and depressing the free side of the tarpaulin under the water by 40 cm or so (Fig. 7.3) so that the ﬁsh are encouraged to swim in. Once ﬁsh have been transferred the tarpaulin is again raised. Following treatment, the net bag inside the treatment cage is lifted and the ﬁsh gently decanted back inside the production cage. Brandal & Egidius (1979) claim that 8–10 small cages of salmon (25–40 t) can be treated for sea lice infestations in a single day (treatment time 20 min per cage) by four or ﬁve staff. While treatment cages are impractical for large cages, and offer few advantages in terms of costs or labour over conventional tarpaulin skirt methods (Roth et al. 1993), they minimize the discharge of chemicals to the environment. Rotating cages can be readily adapted for bathing ﬁsh. The cage shown in Fig. 7.4, for example, can be set in a half-submerged position, a canvas cover attached to the exposed half, and the cage rotated so that the ﬁsh are isolated within the

Problems

247

Fig. 7.3 Diagram showing how a treatment cage is attached to a production cage. Fish are transferred by submerging the net walls at the point indicated (modiﬁed from Brandal & Egidius 1979).

Fig. 7.4 Rotating cage, Scotland (courtesy J. F. Muir).

now submerged canvas covered section. Chemicals are then added and once the treatment has been completed the cage is turned over and the cover removed. If the cage is isolated with a tarpaulin, water exchange is halted and an oxygen diffuser or aeration device should be used, especially since oxygen consumption

248

Chapter 7

increases during treatment. Aerators also help in the rapid dispersal of the chemical following treatment. A number of other precautions should be taken (Roberts & Shepherd 1997): • • •

•

• •

•

Fish should not be fed for 24 h prior to treatment. Plastic, and never galvanized, containers should be used to mix chemicals. The volume of water that the ﬁsh will be held in during treatment must be calculated carefully so that the dose concentration is accurate. The computed dosage should be checked as it is easy to miscalculate by a factor of ten. While the use of the dye rhodamine-B may be useful in assessing how well the treatment chemical is dispersed in the cage, it may not be licensed for use in this way. Treatment during high temperatures should be avoided if possible. During summer months, or in tropical climes, treatment should be carried out in the morning provided that dissolved oxygen (DO) levels are satisfactory. If it is the ﬁrst time a treatment has been used at a farm, a trial should be carried out ﬁrst. If successful, treatment can be carried out 12–24 h later. Stock should be monitored closely during treatment and the farmer should be prepared to rapidly lower the cage net and remove the tarpaulin skirt or increase the oxygen supply if ﬁsh become distressed. It is important that courses of treatment are completed, otherwise the disease may return with greater severity and the cost of the initial treatment will have been wasted (see later).

Details of bath treatments used for various diseases and ﬁsh species are given in Brandal & Egidius (1979), IDRC (1983), Roth et al. (1993), Kleinholz & Luker (1994), Roberts & Shepherd (1997) and Speare (2002). Cleaning symbiosis offers an alternative to chemotherapy for treatment of ectoparasites (see below). Direct application Where valuable ﬁsh or broodstock are concerned, individual treatment is sometimes feasible. Injections of individual ﬁsh are sometimes necessary (e.g. to induce spawning of egg-bound ﬁsh), although in some countries this can only be done under veterinary supervision. In-feed administration Some diseases can only be treated by using medicines incorporated in the feed. Examples include internal parasites, such as acanthocephalans and some of the cestodes, internal protozoan infections (e.g. Octomitus) and septicaemias caused by vibriosis and Aeromonas infections. There are, however, a number of disadvantages to this approach. It can be difﬁcult to persuade manufacturers to incorporate medicines in small quantities of commercially prepared feed (~1 t), leaving the farmer to mix the medicines with the feed. When using trash ﬁsh or a wet pellet diet, medicines are simply mixed with the feed prior to feeding, whereas

Problems

249

dry diets are coated with either a gelatine mixture or corn oil slurry containing the medicine and left to dry. These methods should only be carried out immediately prior to feeding as some antibiotics degrade within 24 h. The most critical calculation in the oral administration of medicines is the dose rate, and it is often recommended that the feeding rate for medicated diets is reduced by 0.5% below usual feeding rates; thus, if ﬁsh are feeding at 3% body weight per day, the feeding rate with a medicated diet is 2.5% body weight per day. Details of medicine incorporation rates are given in IDRC (1983) and Roberts & Shepherd (1997). However, sick ﬁsh may well have poor appetites, while medicines can affect palatability (e.g. Poe & Wilson 1989), further suppressing ingestion. Not only can this render treatment expensive or ineffective, but repeated treatments with sub-therapeutic levels of antibiotics may also help promote drug resistance. In many countries the use of antibiotics requires a veterinary prescription. There are strict regulations concerning withdrawal periods for ﬁsh sold for human consumption (Michel & Alderman 1992; Boghen 1995; Weston 1996; Jahncke & Schwarz 2002). In Norway, uneaten medicated feed must be recovered using systems such as the LiftUp feed collector (see Ervik et al. 1994a; Fig. 6.20) and incinerated (Kaasa 1995). Biological control Sea lice (L. salmonis, Caligus elongatus) infestations are a problem at many Atlantic salmon cage farms. While baths of the chemotherapeutic compound Cypermethrin® (dichlorvos) may still be used under some form of temporary registration in Scotland, Canada, Norway, Ireland and the United States, the technique has many disadvantages: it is ineffective against chalimus lice stages, stressful to salmon and toxic to marine organisms (for reviews see Roth et al. 1993; Sayer et al. 1996; Brooks et al. 2002). Wrasse (e.g. Crenilabrus melops, Ctenolabrus rupestris), which readily feed on ectoparasites of caged ﬁsh, especially where clean nets are maintained, have proved effective lice control agents (Costello 1993; Deady et al. 1995; Sayer et al. 1996; Treasurer 2002a). Wrasse are caught from the wild and stocked with salmon at ratios of around 1 : 25–250. The efﬁcacy of wrasse in controlling lice is well documented and they can be a cheaper, more effective lice control method than chemotherapy, albeit that the supply from the wild has continued to be a constraint. In 1998, something like 30% of Atlantic salmon smolts in Scotland were stocked with wrasse. The outbreak of infectious salmon anaemia (ISA) that year has since deterred many Scottish farmers from stocking with wild ﬁsh, although wrasse are still widely used in Norway, especially as part of integrated pest management programmes (Treasurer 2002a). Tropical cleaner ﬁsh have also been used to control ectoparasite infestations on tilapias (Cowell et al. 1993). Discussion Disease outbreaks often stem from poor management or husbandry and treatment should be very much seen as a last resort. Many diseases cannot be tackled

250

Chapter 7

by therapeutic means; control of most viral diseases, for example, is best achieved by removal of stressors such as low oxygen, crowding, etc. Provided that the farmer can estimate likely mortalities, the economics of treatment versus permitting the disease to run its course should also be considered. Treatment of caged ﬁsh is a long-established practice. In Cambodia, for example, cow dung, tree bark and salt, mud and plant mixtures were once used to treat parasitic and fungal diseases of caged catﬁsh murrels and gobies (Pantulu 1979). The range of drugs and chemicals that ﬁsh farmers have at their disposal to treat diseases is small and unlikely to increase. Resistance among ﬁsh pathogens to some medicines is growing in some areas of the world and there are fears that antimicrobial resistance might be transmitted to microorganisms of public health concern (Michel & Alderman 1992; Jahncke & Schwarz 2002). In recent years there have been tremendous advances in our understanding of the environmental requirements of ﬁsh, the links between environment, stress and disease, and the ﬁsh immune system and how it functions. This has translated into better husbandry (e.g. separation of year classes, fallowing, lower stocking densities) and the development of effective vaccines. In Norway, the result has been a dramatic decrease in antimicrobial use, from 48 t in 1987 to less than 1 t a decade later, despite a doubling of production (see Fig. 5.8).

7.3

DRIFTING OBJECTS

Most cages are vulnerable to damage by drifting objects that can obstruct water ﬂow, tear holes in the netting and even destroy cages. Logs and other large ﬂoating objects are hazards faced by farmers everywhere and cage ﬁsh farmers are advised to keep shorelines adjacent to sites clear of objects which can be lifted off the beach and swept out to the cages during high tides and storms. Jellyﬁsh are common in coastal waters during certain periods of the year and can deform nets and restrict water ﬂow and so should be removed regularly. Jellyﬁsh have been implicated in mortalities of caged ﬁsh in the Baltic and Scotland (Koops 1976 in Coche 1983), although mortalities have not always been caused by suffocation alone. A problem faced by cage ﬁsh farmers in inland tropical lakes is drifting masses of ﬂoating plants, such as water hyacinth (Eichhornia crassipes) (Fig. 7.5). Plants can cover substantial portions of a lake’s surface and in strong wind conditions, ﬂoating masses weighing many tonnes, can destroy cages and moorings. Fixed cages are particularly vulnerable. In the Philippines, some farmers have abandoned bamboo pole constructions and instead use anahaw palms (Livingstonia rotundifolia) (Fig. 7.6). Although they cost 4–8 times as much as bamboos, only a fraction of the number are used, poles being driven into the lake substrate at 2-m intervals. They are strong, long-lasting and offer less resistance to the ﬂoating hyacinth mass. Top nets are ﬁtted to many of the cages so that during stormy periods, the whole cage may be submerged. An alternative solution is to build a barricade or deﬂective fence that can divert weeds and keep the cage area clear. Again, this strategy has been adopted by a number of farmers in the Philippines.

Problems

251

Fig. 7.5 Wind-driven accumulation of water hyacinths, Laguna de Bay, Philippines. Water buffalo can be seen in the foreground.

Fig. 7.6 Lake-based tilapia hatchery, Laguna de Bay, Philippines. Fixed, submersible cages constructed from ‘anahaw’ palms and ﬁtted with top nets.

7.4

FOULING

Fouling occurs to some degree at all cage sites. As discussed in section 4.1, marine sites are worse than fresh water, especially if the environment is warm and eutrophic and if prevailing currents, which greatly inﬂuence settlement, are low.

252

Chapter 7

The degree of fouling is also dependent upon the material used (see section 3.1.3), the mesh size and colour, and whether the mesh is knotted or knotless. It may even be inﬂuenced by the species being cultured. Biofouling is not usually a problem in the farming of herbivores such as the tilapias (Pantastico & Baldia 1981). In Japan, yellowtail cages accumulate fewer fouling organisms than those holding blueﬁn tuna or red sea bream (Pagrus major) owing to the action of the ﬁsh, which continually brush past the netting (Kuwa 1984). Studies of the rates of fouling and fouling communities of net cages have been detailed for many different parts of the world (Milne 1970; Moring & Moring 1975; Cheah & Chua 1979; Wee 1979; Huguenin & Ansuini 1981; Kuwa 1983, 1984; Santhanam et al. 1984; Greenland et al. 1988; Lai et al. 1993; Hodson & Burke 1994; Dubost et al. 1996; Hodson et al. 2000). Fouling increases vertical forces on cage structures and netting, and reduces mesh size, thereby limiting water exchange and increasing drag (Milne 1970, 1976; Inoue 1972; Wee 1979; Kuwa 1983, 1984; Santhanam et al. 1984; Greenland et al. 1988; Aarsnes et al. 1990; Lai et al. 1993; Løland 1993a, b). Milne (1970), for example, used immersion trials to establish changes in weight and in twine diameter of panels of materials (see Tables 3.5, 3.6). Increases in current forces can be as much as 12-fold (see Fig. 3.14) while increases in weight can be up to 200-fold, affecting both routine operations and ﬂotation. A number of net failures at commercial marine cage ﬁsh farms have been attributed to fouling (Milne 1976; Huguenin & Ansuini 1978; Huguenin 1997). If fouling at a site is severe, there are number of options: • • •

management procedures can be modiﬁed to cope with the problem; the fouling can be tackled using chemical or biological control agents; a fouling-resistant or rotating design can be used.

7.4.1

Management

The ﬁrst option is discussed in section 6.3, which describes net changing and cleaning. If fouling is severe, then this procedure may have to be carried out more frequently.

7.4.2

Chemical control

Most antifoulants contain a slow-leaching biocide so that the surface is permanently surrounded by a thin layer of toxic material that prevents the susceptible juvenile planktonic stages of fouling organisms from gaining a foothold (Lovegrove 1979). Ideally, the compound should release the toxin at a controlled rate over an extended period. There are two principal types in use: (1) soluble matrix or soft ﬁlm antifouling and (2) contact leaching, tough-ﬁlm antifouling. Soluble matrix antifoulants contain both a water-soluble biocide and a part water-soluble resin. When immersed in sea water the paint gradually becomes porous, allowing access to the biocide in the deeper layers. This type of antifoulant has the advantage that fairly low levels of biocide are included in the paint. The latter type uses a water-resistant media that must contain much higher

Problems

253

levels of water-soluble biocide in order that adequate leaching occurs. Pigments (e.g. iron oxide) are often added to help retard UV damage (see below). One of the oldest types of antifoulant, tannin, extracted from mangrove wood (Rhizophora spp.), is estimated to reduce fouling by about 20%. However, it is difﬁcult to apply to netting and is not in widespread use in aquaculture (Lai et al. 1993). The most widely used antifoulants in the marine environment are copper or copper oxide-based, sometimes with other compounds present, such as zinc oxide or 2,4,5,6-tetrachloro isophthalonitrile, and are sold in a variety of forms under different brand names (Henderson & Davies 2000; Costello et al. 2001). Copper is toxic, even at moderately low concentrations, particularly to the larval stages of invertebrates. Antifoulants are regarded as biocides in Europe as they are not used directly on food-producing ﬁsh. As such, they do not fall under the maximum residue limit (MRL) system, despite the fact that ﬁsh may be exposed to the compounds for months (Costello et al. 2001). However, based on a commonly used commercial copper-based antifouling product, models run under North American conditions suggest that EPA (Environmental Protection Agency) copper water-quality criteria are unlikely to be exceeded at most cage farms (Brooks et al. 2002). A number of alternatives are being developed, including both toxic plant-based compounds and non-toxic silicone coatings, which have no known toxicants present (Costello et al. 2001). The latter appear to work by proving difﬁcult for fouling organisms to settle on and are also easy to clean (Hodson et al. 2000). Antifoulants can either be applied at the factory (Fig. 7.7) or at the farm. Commercial formulations usually require nets to be dipped in the antifoulant

Fig. 7.7 Copper-based antifouling material being applied to cage nets (courtesy W. & J. Knox Ltd).

254

Chapter 7

solution for some minutes and hung up in a fully open position so that the meshes do not become clogged. Manufacturers recommend that copper-treated nets are immersed as soon as possible after drying and that farmers postpone stocking for at least 24 h. Treated nets provide a measure of protection for many months although the degree of protection declines rapidly after 6 months or so. In part, this is due to paint being dislodged from the net. Copper has been found in sediments underlying cages that have been treated with copper-based antifoulants (Lewis & Metaxas 1991; Uotila 1991; Brooks et al. 2002). In temperate regions nets may have to be treated each year, newly treated nets being installed in April– May to give protection over the summer months when larval settlement is greatest. The cost of treatment can add 20–25% to the cost of a knotless nylon cage net. Treatment with copper-based antifoulants, however, has been found to promote amoebic gill disease in salmon cage farms in Tasmania (DouglasHelders et al. 2003). Although many questions remain to be answered the working hypothesis is that nets treated with copper-based antifoulants harbour many times more bacteria than untreated nets. The bacteria act as a source of food for the amphizoic protozoan Neoparamoeba pemaquidensis, thereby encouraging increased numbers of the pathogen on treated nets.

7.4.3

Biological control

While the organisms that foul net cages may augment the diets of cultured herbivores or detrivores (see above), they are only occasionally ingested by carnivorous species such as salmonids and play a negligible role in their diet (Moring & Moring 1975; Black et al. 1992). However, herbivores are sometimes stocked alongside commercially more important species in an attempt to reduce biofouling. In Palau, Hasse (1974) successfully used siganids (Siganus canaliculatus, S. lineatus) at densities of approximately 2 ﬁsh m-3 to reduce fouling in cages holding strings of oysters (Ostrea nomades), while Chua & Teng (1977) found siganids effective in controlling fouling in cages of grouper and carangids. The striped and spotted knifejaws Oplegnathus spp. have been used in cage mariculture in Japan (Kuwa 1983, 1984) (Fig. 7.8). Low densities (1 ﬁsh 5 m-3) of rohu (Labeo rohita) have proved effective in maintaining the cleanliness of carp cages in Nepal (Sharma 1979), while in China the stocking of tilapia or common carp at 3–5% of total cage biomass is recommended (Li 1994). Marine prawns have also been stocked in salmon cages and mullets with pompano and other species (Swingle 1971; Tatum 1974; Rensel & Prentice 1979). However, there can be problems: siganids have been found to damage cage netting as a result of grazing (Ben Yami 1974) and knifejaws have been observed attacking the tails and ﬁns of sick yellowtail (Kuwa 1984).

7.4.4

Fouling-resistant designs

Farmers may consider the use of fouling-resistant materials in the construction of the cages. Some companies manufacture PE netting inlaid with copper wire,

Problems

255

Fig. 7.8 Spotted knifejaws browsing on fouling in a yellowtail cage (courtesy M. Kuwa).

and while this may be readily fabricated into conventional net bag designs, it is much more expensive, costing at least twice as much as a nylon net bag of similar dimensions. Moreover, the copper gives limited protection and nets must either be replaced every year or two or treated with conventional antifouling compounds after the copper wire has corroded. Rigid cages have also been designed using fouling-resistant materials such as galvanized steel, PVC-coated wire and copper or copper–nickel compounds (Milne 1970; Swingle 1971; Powell 1976; Ansuini & Huguenin 1978; Ojeda & Strawn 1980; Huguenin et al. 1981; Kuwa 1983, 1984; Ravindran 1983; Huguenin 1997; see Figs 3.12, 3.19). As might be expected, copper and copper alloy mesh cages are considerably more expensive (3–5 times) than conventional nylon bag systems (Ojeda & Strawn 1980; Huguenin et al. 1981). Although copper leaching rates have been estimated (Ravindran 1983), there have been no investigations of copper accumulation in caged ﬁsh. Rotating designs can be effective in controlling fouling and in minimizing cleaning (Grave 1975; Geffen 1979; Porter 1981; Blair et al. 1982; Schneider et al. 1990; Willinsky et al. 1991; Huguenin 1997). However, in countries where UV levels are high, prolonged exposure of netting to intense sunlight can accelerate the degradation of the nets (Porter 1981). Semi-submersible cages, such as the Farm Ocean design (see Fig. 7.28), are also readily cleaned without recourse to net changing. Commercially available designs of semi-submersible or rotating cage tend to be expensive, costing several times as much per unit volume as conventional cage designs.

256

Chapter 7

7.4.5

Discussion

The majority of commercial ﬁsh farmers use nylon or PE net cages and either accept the increased labour involved in cleaning (some managers believe that frequent checking of nets for fouling results in a greater awareness of the status of ﬁsh health on the farm), or resort to the use of antifouling compounds. While the savings in manpower offered by rotating designs or galvanized steel cages may not justify higher initial capital expenditure, they have a number of other advantages over conventional net bag systems that should be carefully assessed by prospective farmers (see below).

7.5

OXYGEN

The dissolved oxygen (DO) regime in a ﬁsh cage is the balance between various sources and sinks. Oxygen inputs by diffusion across the air/water interface are negligible – Kils (1979) estimated that it provides somewhere in the region of 0.5% of the oxygen respired by caged ﬁsh – and the major source of DO is the water ﬂux through the cage caused by external currents and the movements of the caged ﬁsh. Mass balance models for cage DO levels, based on DO concentration in external water, current velocity, transmission rate and oxygen consumption of caged ﬁsh, can be readily assembled. At most sites the DO content of surface waters approximates saturation levels and as long as cages are maintained relatively free from fouling there should be few problems (see section 4.1). High temperatures reduce the absolute amount of oxygen that water can hold in solution and increase the oxygen demands of the ﬁsh. Provided that the farmer monitors DO levels and reduces feeding rates, and avoids handling (i.e. keeps oxygen consumption to a minimum), the event usually passes with only a small interruption in production. DO levels can fall far below saturation in natural water bodies under the following circumstances: •

• •

Where the phytoplankton community is large (i.e. under bloom conditions), night-time respiration can cause a marked decline in DO, reaching a minimum around dawn. At the end of a phytoplankton bloom, bacterial decomposition of algae can lead to anoxia. Upwelling of anoxic water from the hypolimnion to the surface can occur at certain sites.

Problems can be minimized during site appraisal by rejecting eutrophic or strongly stratiﬁed water bodies that mix infrequently. Limiting ﬁsh production to levels appropriate to the site further reduces risks at intensive farms. Nevertheless, this is not always possible or even desirable (e.g. extensive ﬁsh farms are best sited in eutrophic waters) and problems have been reported at both marine and freshwater sites in temperate and tropical environments (Inoue 1972; Koops 1972; Kils 1979; Landless 1985; Zoran et al. 1994; Santiago 1994; Bagarinao & Lantin-Olaguer 1998; Beveridge & Stewart 1998; Yambot 2000; Huy et al. 2001; Nelson et al. 2001a; Costa Pierce 2002).

Problems

7.5.1

257

Aeration and oxygenation systems

Remedial action involves either increasing the DO content of the water and/or the rate of water ﬂow through cages. The factors that govern the solubility of oxygen in water are discussed in section 4.1. The transfer of oxygen to water is dependent upon the difference between the maximum level of oxygen that can be dissolved in the water under speciﬁc conditions of temperature, salinity and atmospheric pressure, and prevailing DO levels. Transfer is also inﬂuenced by the area of contact between the water and air or oxygen supply, such that the rate of transfer (R) is determined by: R = KL a(CS - C ) where KL = the mass transfer coefﬁcient (determined experimentally); a = the interfacial area between the gas and liquid; CS = the equilibrium concentration of DO; and C = the DO concentration of the oxygenated water. Oxygenation is achieved by maximizing the area of contact between water and air or oxygen either by introducing ﬁne bubbles of gas into the water or by breaking the water into ﬁne droplets in air. Air can be introduced to the water from an air blower, via coarse bubble (large volumes at low pressure) or ﬁne bubble (small volumes at high pressure) diffusers and air-lifts, while droplets of water can be generated using various mechanical devices including surface aerators, surface agitators, venturis and submerged pumps (Fig. 7.9). Details of how systems work are given in Colt & Tchobanoglous (1981), Sowerbutts & Forster (1981), Colt & Watten (1988), Boyd & Watten (1989), Colt & Orwicz (1991), Lawson (1995) and Boyd & Tucker (1998). The most common source of oxygen used at ﬁsh farms is bulk liquid oxygen, which is more than 90% pure and is stored under pressure in tanks or cylinders. However, pressure swing adsorption (PSA) devices that produce 90–95% pure oxygen by removing nitrogen from air as it is forced through a series of zeolite ‘molecular sieves’ are gaining in popularity since, according to manufacturers, they can be more economical than alternative oxygenation systems for small– medium size production units (<15 m3 O2 h-1). Most studies of aeration and the measures used to rate aerators are from the water treatment industry. The standard oxygen transfer rates (SOTR) describe the rate of oxygen transfer per hour at 20°C when the DO concentration is zero: SOTR = (KL a) 20 (CS ) 20 (V )(10 -3 ) where (KLa)20 = overall oxygen transfer coefﬁcient at 20°C; (CS)20 = DO concentration at saturation and 20°C; and V = volume of water in test basin. Another key comparative measure is the standard aeration efﬁciency (SAE), deﬁned as the quantity of DO transferred per unit energy expended (kg O2 kWh-1): SAE =

SOTR Power input

and is usually expressed in kg O2 kWh-1. The SAE values for various devices are summarized in Table 7.1.

258

Chapter 7

Fig. 7.9 Various designs of aerator (redrawn from Sowerbutts & Foster 1981).

Useful as such comparisons are, aeration devices must operate in ﬁsh farm conditions where the driving force (CS - C) is much lower than that used in standard tests, where surfactants affect surface tension, where salinity, dissolved solids and suspended solids inﬂuence DO saturation and where ﬁsh, phytoplankton and other animals are respiring (Colt & Orwicz 1991; Lawson 1995; Boyd & Tucker 1998). These factors can be accounted for by alpha (a; the ratio of KLa for ﬁeld conditions to KLa for clean water at the same temperature), beta (b; the ratio of CS for ﬁeld conditions to CS for clean water) and theta (q; the

Problems

259

Table 7.1 Typical oxygen transfer rates of various devices used in ﬁsh culture systems (from Colt & Tchobanoglous 1981; Colt & Orwicz 1991). Transfer rate kg O2 kWh-1 Aeration system

Standarda

6 mg l-1 O2b

Surface aerators Low speed High speed Vertical pump

1.2–2.4 1.2–2.4 0.7–1.8

0.25–0.80 0.25-0.50 —

1.2–2.0 1.0–1.6 0.6–1.2 2.0–3.3

0.25–0.42 0.21–0.34 0.13–0.25 0.25–0.50

—

1.20–1.80

Subsurface aerators Diffused air system Fine bubble Medium bubble Coarse bubble Venturi aerator Pure oxygen system Fine bubble a b

20°C, tap water, 0 mg l-1, a = b = 1.0. 20°C, a = 0.85, b = 0.9, DO = 6 mg l-1.

factor used to correct KLa for changes in viscosity, surface tension and diffusion constants as affected by temperature) factors: a=

(KL a) 20 cage site water (KL a) 20 clean water

b=

CS cage site water CS clean water

A value of 1.24 for q is recommended in fresh water (Colt & Orwicz 1991); however, since salinity determines surface tension forces, which in turn determine bubble size, transfer efﬁciencies in sea water are higher (Kils 1977). Transfer efﬁciencies for ﬁeld conditions can be computed using the following equation:

(OTR)f =

a(SOTR) 20 qT - 20 (b CS - C m ) 9.092

where (OTR)f = ﬁeld oxygen transfer rate (kg h-1); CS = clean water DO saturation at test temperature and pressure; Cm = measured DO concentration at cage site (mg l-1) and 9.092 is the value for CS at 20°C and 760 mmHg (Lawson 1995). In Table 7.1 mass transfer rates have also been calculated for freshwater pond/tank systems assuming T = 20°C, a = 0.85, b = 0.9 and Cm = 6.0 mg 1-l (i.e. Cm = 70% saturation). Values are around 20% of those derived under

260

Chapter 7

standard conditions. However, to achieve 90% saturation the mass transfer rate falls to around 10% of the standard values, i.e. it takes twice as much energy to achieve 90% saturation as it does to achieve 70% saturation under given conditions. The depth of water in a cage is greater than that in most tanks and ponds, and there is little restriction in exchange between the cage environment and that outside. Cage ﬁsh farmers have a problem in trying to oxygenate the cage volume without having to aerate the external environment as well. Surface aerators and agitators are unlikely to be very efﬁcient when used in cages since they aerate only the top few metres; submerged devices (pumps, venturis, air-lifts, air stones) are probably best. The formula given above for calculating oxygen transfer efﬁciencies and the values shown in column 2 of Table 7.1 apply to freshwater pond and tank conditions, although in the absence of any data for cages they can be used as a general guide. Surprisingly, studies of aeration in deep tanks have shown that the efﬁciency of oxygen transfer is unaffected by the depth (1–8 m) at which ﬁne bubble aerators are installed (Lister & Boon 1973). The depth at which aeration devices are installed in cages (5–10 m compared with 1–2 m in ponds and tanks) increases contact time between air and water that may offset the increased energy costs of delivering air to those depths. In choosing and designing a system the cage ﬁsh farmer must establish when problems are likely to occur and their likely severity. A balance must be struck between a design that can cope with the worst possible conditions and costly over-engineering. Using temperature, species and ﬁsh size data, the oxygen consumption of the farmed ﬁsh during the crisis period can be computed (Table 7.2). Oxygen requirements per cage per unit time can be derived from stocking density and cage size data. Using the oxygen mass transfer efﬁciencies given in Table 7.1, the power requirements for providing all or part of the oxygen requirements can then be calculated. The lower value should be used where the DO gradient (C* - C; where C* = DO concentration at saturation and C = actual DO concentration) is small and the more optimistic ﬁgure used where (C* - C) is large. A worked example is given in Appendix 7.1. The power needed to supply all of the oxygen requirements of the ﬁsh is around 1.2 kW t-1, approximating the ruleof-thumb value of 1 kW t-1 for rainbow trout held at 15°C given by Sowerbutts & Forster (1981). Cage ﬁsh farmers tend to install systems with power outputs much less (5–50%) than this and rely on aeration merely to top up available supplies when necessary. At some farms, aeration equipment is switched on and left running for the entire risk period, while at others the system includes an oxygen transducer, the signal from which switches on or off the equipment at pre-determined levels. A typical aeration system consists of a diesel or electrically driven blower or compressor connected to a PVC hose that distributes air to an aerator or system of aerators suspended several metres below the water surface. Compressors are cheaper and are often easier to obtain in tropical countries although they are less efﬁcient and give a lower oxygen transfer rate. An oil-free air supply is important and so an air ﬁlter or oil trap may also be required. Blowers utilize either a single (low pressure) or multiple (high pressure) fan system, choice

Problems

261

Table 7.2 Some oxygen consumption values for farmed ﬁsh. Species

Size (g)

Temperature (°C)

Oncorhynchus mykiss

100 100

15 15

? production levels

0.3 0.3

Oncorhynchus nerka

29

15

unfed

0.23

29

15

0.28

Ictalurus punctatus

100

30

3% body weight per day unfed

0.56

Cyprinus carpio

100 100

30 10

satiation fed

0.81 0.17

100 100 15

20 25 20

fed fed —

0.48 0.70 0.20

240

23

fed

0.25

50

25

unfed

0.16

50 50

30 35

unfed unfed

0.24 0.40

Hypophthalmichthys molitrix

Oreochromis niloticus

a

Feed rate

Oxygen consumption (g O2 kg ﬁsh-1 h-1)

Source

Liao 1971 Muller-Fuega et al. 1971 Brett & Zala 1975

Andrews & Matsuda 1975 Beamish 1964

Muhamedova 1977a Vetskanov 1975a Ross & Ross 1999

From ADCP (1983).

depending on the extent and nature of the pipe work involved. If the cages are attached to the shore via a walkway, then the blower/compressor and motor can be installed on land and the airline run out across the walkway. If the cages are moored offshore then it must be decided whether the blower/compressor and motor remain on shore and a ﬂoating airline used to deliver air to the cages or whether the blower compressor and motor are installed at the cages, either on the cage raft or on a boat moored alongside. The longer and narrower the pipeline, the greater the drop in pressure and the lower the overall efﬁciency of oxygen transfer. The capital cost of an aeration system is inﬂuenced by the size of the engine or blower and whether the system is purchased ‘off-the-shelf’ or assembled by the farmer. Second-hand diesel engines and ex-milking parlour blowers can be used with considerable savings in costs. As indicated above, there has been little study of aeration in cages and it is thus difﬁcult to assess the effectiveness, let alone which aeration devices (airstones, venturis, air-lift pumps, submerged pumps) are best. Landless (1985) cites a number of examples where air-lift devices have proved effective in maintaining good DO levels at cage farms. Hargreaves et al. (1991) found that aeration with submerged airstones had no signiﬁcant effect on either production or food

262

Chapter 7

conversion of tilapias held in intensively stocked, small (1 m3) cages. Kils (1979) stated that the use of submerged aeration devices could even make oxygen conditions worse by creating a rising stream of oxygenated water that rapidly ﬂows across the water surface and out of the cage, creating an upwelling of deoxygenated hypolimnetic water which, if the cages are located near the lake or sea bottom, may also carry cage wastes. Nelson et al. (2001a) examined the costs and beneﬁts associated with two air-lift aeration regimes – continuous and partial (9 h per night, plus emergencies) – versus none (control) in catﬁsh cages. They modelled the results, but concluded that in order to demonstrate which aeration regime was superior in terms of production and proﬁtability, an impractical number of replicates would be needed.

Fig. 7.10 (a) Diagram of cage aerator, showing casing, motor with fan and hollow rotor. Water is sucked in through intakes at the bottom of the casing and aerated water is discharged via outlets in the side. The transverse plate reduces agitation of the water while the air space at the top acts as a foam trap (redrawn from Kils 1979). (b) Hollow rotor of aerator (redrawn from Kils 1979).

Problems

263

Fig. 7.11 Submerged venturator (redrawn from Landless 1985).

Kils (1979) designed and tested a modiﬁed submerged pump (Fig. 7.10a, b) under ﬁeld conditions and reported that it gave excellent results with a very high oxygen transfer efﬁciency (2.3–3.2 kg O2 kWh-1). However, the design has yet to be tested in large or more conventional cage designs. No details of venturis (which require pumped water) having been used in cages are available and most practical experience has been with air-lift devices (Fig. 7.11). Either ﬁne or coarse discharge holes are used, the former being more effective. However, capital costs

264

Chapter 7

Fig. 7.12 Mixer installed at Atlantic salmon farm to improve water ﬂow through cages (courtesy J. F. Muir).

for ﬁne-bubble systems are higher and the aerators are also more difﬁcult to clean and maintain. While air-lifts appear to be successful, once again there is no good evidence concerning effectiveness. Submerged aeration devices in cages can be used to help maintain ice-free conditions (section 7.9) and to reduce water temperatures, and hence improve oxygen content through de-stratiﬁcation of temperate lakes used for salmonid culture (e.g. Landless 1985). Submerged aeration devices are also used to combat harmful algal blooms. Aeration affects caged ﬁsh behaviour (Kils 1979). The alternative to aeration is to use oxygen. However, there are no published accounts of whether such systems are practical. As costs of oxygenation on ﬁsh farms are highly site-speciﬁc (Colt & Orwicz 1991), no advice on costeffectiveness can be given.

7.5.2

Alternative systems

The alternative to aeration or oxygenation is to increase the rate of water exchange through the cage using a low-speed propeller device. Low-speed propellers (Fig. 7.12), initially used at marine salmon farms in Norway and Scotland to aid wastes dispersion, can improve DO conditions in cages. They should be used with caution in fresh water, however, as devices installed on or

Problems

265

close to the lake bottom may re-suspend sedimented wastes. There are no published accounts of the effectiveness of mixing devices.

7.6

SECURITY

Cages are more vulnerable to poaching and vandalism, and to storm damage, than other aquaculture systems. Farms sited some distance from shore-based facilities can easily be visited without observation, particularly at night. Cages are also difﬁcult to protect since they cannot be fenced and conventional electronic devices such as infra-red (IR) beams are easily triggered by the motion of the cages. Fish may be quickly removed by hand net while vandals can cause many thousands of pounds worth of damage and loss of valuable stock in minutes using nothing more sophisticated than a sharp knife. Security is often a major consideration in site selection and proximity to the owner’s house is undoubtedly a deterrent to poachers. In many parts of Southeast Asia guard houses are built on the cage rafts and nightwatchmen with dogs employed to guard stock (IDRC/SEAFDEC 1979; Tseng 1983) (see Fig. 6.10). Rigid mesh cages cannot be vandalized so readily and it is also more difﬁcult to remove ﬁsh without the appropriate equipment. A number of devices have been developed to help protect cages. Lockable lids can be incorporated into smaller cage designs. Sonar-based systems are also widely marketed. These have programmed decision-making capabilities and are able to determine the intent of boats in the vicinity of the cages, contact farm staff by telephone link and activate deterrents such as lights or alarms if necessary (Jackman & Ace-Hopkins 1993). The reasons why poaching and vandalism at cage and pen ﬁsh farms are commonplace in many parts of the world are discussed in Beveridge (1984a), Zerner (1992), McAndrew et al. (2000), Brugere et al. (2001), Hambrey et al. (2001a, b), Costa-Pierce 2002). Conﬂicts over the utilization of a resource often lie at the root of persistent problems. In Laguna de Bay in the Philippines, for example, the unchecked expansion of the ﬁsh pen and ﬁsh cage industry during the late 1970s and early 1980s disrupted traditional livelihoods of ﬁshing and duck rearing, and the free movement of lakeside dwellers around the lake. Many ﬁsh pen owners were from outside the area and imported labour to protect their investments, leading to friction and even violence between the local communities and the ﬁsh pen operators. At other lakes in the Philippines, where development was better regulated and local people were encouraged to establish their own farm units, there were far fewer problems. The key to reducing poaching and vandalism is effective planning.

7.7

PREDATORS AND SCAVENGERS

The range of predatory and scavenging species reported at cage ﬁsh farms includes squid, ﬁsh, turtles, lizards, sea snakes, birds and mammals (Table 7.3). They pose problems for the ﬁsh farmer by:

Trout

Maine

Up to 40% mortalities caused at one site

Pemberton & Shaughnessy 1993

Chua & Teng 1980

Jee 1978

Pitt et al. 1977

Möller 1979

Beveridge 1988

Mills 1979; Beveridge 1988; Ross 1988; NCC 1989, 1990; Carss 1993a, b

Rueggeberg & Booth 1989

Henderson 1980

Lindbergh 1979

Mahnken 1975

References

Squalus acanthius; b Larus argentatus, L. marinus; c Ardea herodias; d Phalacrocorax auritus; e Phoca vitulina; f Eumetopias jubatus, Zalophus californianus; Lutra candensis; h Mustela vison; i Procyon lotor; j Megaceryle alcyon; k Larus glaucescens, L. philadelphia; l Phalacrocorax auritus, P. penicillatus; m Mergus serrator;n Mergus mergarner; o Sula bassana; p Larus spp.; q Podiceps cristatus; r Ophiocephalus sp.; s Lutra sumatrana; t Sphaeroides sp.

g

a

Seals

Otters, gulls, puffer ﬁsht

Grouper

Salmon

Snakeheadsr, water snakes

Marble goby

Australia Tasmania

Lots of small ﬁsh also found in cages Damage to nets and mortalities in fresh water Nets damaged by puffer ﬁsh Dogs used to deter otters

Moray eels, squid

Sparus sp.

Malaysia

More than 50% losses at some farms attributed to predators

Damage to nets by diving birds

List derived from many surveys

Israel

Herons, gulls, cormorants, seals

Seals, otters, cormorants, shags, mergansersm, goosandersn, guillemots, gannetso, gullsp Gulls, grebesq, cormorants

Sealse, sealionsf, ottersg, minkh, racoonsi, heronsc, kingﬁshersj, gullsb,k, cormorantsl, etc.

Large numbers of unrecorded losses attributed to birds 22 species of birds and 7 species of mammal involved

Particular problems with larger mesh sizes Metal mesh cages used

Comments

Salmon

Trout

Salmon, trout

Salmon

Gullsb, heronsc, cormorantsd

Mink, otters

Dogﬁsha

Predator

Norway

England

UK Scotland

British Columbia

Salmon

Salmon

United States Wahington

Canada New Brunswick

Farmed species

Country

Table 7.3 Predators at cage ﬁsh farms. Latin names are given where possible (see also text).

266 Chapter 7

Problems

• • • •

267

killing or wounding ﬁsh; damaging nets and fouling equipment; stressing ﬁsh so that feeding is disrupted and resistance to disease is reduced; spreading disease.

Predatory and scavenging species are present at most farms. In Scotland in the 1970s and 1980s it was estimated that between 60 and 90% of Scottish ﬁsh farms were visited by piscivorous birds, while around 80% of cage ﬁsh farms suffered attacks by seals (Mills 1979; Ross 1988). In a postal-based survey carried out in Scotland in 2002, 55% of cage ﬁsh farm managers, equivalent to 70% of those that responded, reported predation problems (Quick et al. 2004). Twelve species of predator were involved, the most important being seals: 81% of sites reported seals as a problem. No signiﬁcant increases in predator problems were recorded in the 20 or so years between studies, although Quick et al. (2004) deduced that gull-related problems may have increased signiﬁcantly. In British Columbia, Canada, Rueggeberg & Booth (1989) estimated that up to 60% of farms have problems with birds or mammals. Although there are great differences between one farm and another, and between different parts of the world, marine farms tend to have more problems and attract a greater range of predators. Despite the fact that there have been a number of studies of predation, there are few reliable ﬁgures on economic impacts because, as Meyer’s study (1980) of heron predation at trout ponds has demonstrated, it is difﬁcult – even by careful observation – to reliably estimate the numbers of ﬁsh taken let alone establish secondary effects on growth or disease. Rueggeberg & Booth (1989) estimated losses through predation at salmon farms in British Columbia to be around 1.5% of ﬁsh stocked, plus an additional Can$22 000 worth of damage to equipment. In a study of the Scottish salmon farming industry, Ross (1988) estimated that predator-related losses in 1987 were £1.4–4.8 million. Kennedy (1994), however, has stated that predation was not a major cause of insurance claims. From the ﬁgures published in Norway for the period 1994–99, some 31% of the 2.5% of all ﬁsh stocked (i.e. 0.8%) are lost due to holes in the net, which may be attributed to predators (see section 5.3.4). Nash et al. (2000) estimate losses for the United States and Canada run to many millions of dollars per year. The US National Marine Fisheries Service estimated losses in excess of US $50 million in one year in the Gulf of Maine alone (Würsig & Gailey 2002). Impacts of cage ﬁsh farming on wildlife are discussed in Chapter 5.

7.7.1

Why predators and scavengers occur at cage farms

The reason may seem obvious: because of the caged ﬁsh (Würsig & Gailey 2002). However, while this is undoubtedly the principal cause, scavengers are attracted by the presence of food or dead ﬁsh and predators may be drawn by wild ﬁsh congregating outside cages to feed on waste food (NCC 1989, 1990). The cage structure attracts ﬁsh in the same way that ﬁsh are attracted to FADs (ﬁsh aggregation devices; see Beveridge 1984a; Costa-Pierce & Bridger 2002), while the cage superstructure serves as a roost or observation site for opportunistic scav-

268

Chapter 7

engers. Predators are also sometimes inadvertently introduced when cages are stocked or when nets are changed (Reksalegora 1979). Why should predators plague certain farms but not others? Farm size appears to be unimportant; large farms are no more attractive to predators than small farms. While many piscivores are highly mobile, Rueggeberg & Booth (1989) found that frequency of attack by mammals and birds was correlated to proximity to colonies or haul-out sites, a ﬁnding conﬁrmed by the study of seal (Arctocephalus pusillus doriferus) predation at salmon farms in Tasmania by Pemberton & Shaughnessy (1993). Rueggeberg & Booth (1989) also report that certain cage designs are less susceptible to damage (see below) while particular species of ﬁsh are more vulnerable to predator attack, possibly because of feeding behaviour. Management, too, plays a role as will be discussed below.

7.7.2

Predator visit patterns

Predation is usually highly seasonal (Beveridge 1988). For example, cormorants (Phalacrocorax carbo) in Scotland usually begin to arrive at freshwater cage farms in August, increase in numbers over the autumn and winter months, and disappear when the breeding season begins in April (Mills 1965; Ranson & Beveridge 1983; Carss 1993a). Ranson & Beveridge (1983) observed that the daily number of cormorants visiting one particular cage farm increased to an average of 15 per day at the end of December, although more than 70 birds were observed on one occasion. A study of cormorant distribution by Mills (1965) has shown that the number of birds visiting fresh waters depends on the availability of food at sea and thus, to some extent, the weather. Marked seasonal patterns in the number of herons (Ardea cinerea) at ﬁsh farms have also been observed (Meyer 1980; Carss 1993b). Most predators, particularly birds and mammals, exhibit diurnal patterns of activity, preferring to attack at dawn and dusk when staff are absent from the farm. Clear, calm nights are preferred.

7.7.3

Predator-related problems

If mesh size is sufﬁciently large, squid, sea snakes and predatory ﬁsh will enter and consume as much as they can, although they may be unable to leave. At the end of a milkﬁsh trial in the Philippines more than 12 species of ﬁsh, including croakers and barracuda, were discovered in the cages, the ﬁsh having entered while being small enough to pass through the meshes. Herons take ﬁsh from unprotected cages by ﬁshing from the walkways, while mink (Mustela lutreola) and otters (Lutra spp.) clamber over cage collars and into the net bags. The attempts of some species to catch caged ﬁsh, however, are more destructive. Seals and sea lions typically attack by charging the side or bottom of the cage, causing ﬁsh to panic and crowd into a corner where the predator then tries to grasp mouthfuls of ﬁsh through the net. Seal attacks usually result in a large number of ﬁsh being killed or wounded, often by a ventral bite behind the gills and/or claw marks along the ﬂanks (Ross 1988; Rueggeberg & Booth 1989; Pemberton & Shaughnessy 1993) (Fig. 7.13). Serious damage to

Problems

269

Fig. 7.13 Dead Atlantic salmon, recovered from ﬁsh cage, showing seal bite (courtesy R. Collins).

nets occurs, often with the loss of ﬁsh through the tears. Predatory ﬁsh may also damage nets. Chua & Teng (1980) reported that puffer ﬁsh (Sphaeroides spp.) damaged the net bags of cages of grouper resulting in a 20% loss of stock. In Lake Kariba, Zimbabwe, tiger ﬁsh (Hydrocyon goliath) frequently attack caged tilapias, tearing nets in the process. Predators have distinct size preferences for prey – herons, for example, tend to select ﬁsh of <300 g (Carss 1993a) – smaller ﬁsh possibly being more vulnerable to certain types of predator because of feeding behaviour. A high proportion of ﬁsh in poor condition tend to be taken and, not surprisingly, adult birds are more successful predators than juveniles (Beveridge 1988). The success of predators is also to some extent dependent upon mesh size, larger mesh nets – and hence bigger and more valuable ﬁsh – sometimes being more vulnerable. Ranson & Beveridge (1983) concluded that diving birds such as cormorants rarely manage to extract ﬁsh from a cage although they may wound a great many. During a survey at a freshwater rainbow trout cage farm as many as 6% of caged ﬁsh showed signs of wounding by birds (Ranson & Beveridge 1983) (Fig. 7.14).

7.7.4

Predators, stress and disease

The presence of birds and mammals at a cage farm increases disease risks. The heron is an important intermediary host in the life cycle of the cestode Diphyllobothrium sp., a parasite prevalent at a number of UK cage trout farms, while otters are ﬁnal hosts for the digenean Haplorchis sp., a common parasite of tilapias (Roberts & Sommerville 1982). Ranson & Beveridge (1983) suggested

270

Chapter 7

Fig. 7.14 Bird damage to ﬁsh (from Ranson & Beveridge 1983).

that wounds (Fig. 7.14) may provide a site for invasion by pathogenic fungi and bacteria. Birds are also known to be vectors of pathogenic bacteria and viruses (Beveridge 1988; Willumsen 1989; McAllister & Owens 1992), although their signiﬁcance as vectors remains unknown. Even unsuccessful attacks on caged ﬁsh may be damaging, stressing ﬁsh so that feeding is disturbed and the immune system is suppressed, increasing susceptibility to disease. In a study of feeding rhythms in caged salmon, Blyth (1992) found that the presence of dolphins altered feeding patterns and considerably reduced feed intake (see also Pemberton & Shaughnessy 1993).

7.7.5

Antipredator and scavenger measures

Farmers can and do retaliate against predators and scavengers, some methods being more successful than others. Measures can be summarized as aversive conditioning, exclusion, deterrence/harassment or removal.

Problems

271

Aversive conditioning Lithium chloride, used for many years to induce taste aversion in coyotes predating on sheep, has also been used to deter rough-toothed dolphins (Steno bredanensis), Stellas sea lions (Eumetopias jubatus) and Australian fur seals (Arctocephalus pusillus) from predating on ﬁsh (Würsig & Gailey 2002). However, as Würsig & Gailey (2002) point out, there are doubts about the effectiveness and safety of using LiCl and its effectiveness when dealing with large numbers of predators.

Exclusion methods Top nets and curtain nets are widely used and have proved highly successful in deterring predators. A number of suggested schemes are shown in Fig. 7.15. Top nets are regarded by cage farm managers in Scotland as moderate to very effective (Quick et al. 2004), particularly at deterring gulls and herons. They are made from strong, 5–10-cm square mesh netting, treated to retard the effects of weathering. Nylon monoﬁlament and PE are superior to woven nylon; these materials are cheaper and as they do not absorb water are less likely to sag in the middle. Coloured netting may help reduce entanglement problems. It is perhaps stating the obvious, but top nets must not only be secured ﬁrmly but also should not be permitted to touch the water surface in the cage as they may damage stock. Groups of birds have also been known to stand on slack top nets, weighing them down to gain access to the ﬁsh. Securing the top net to the freeboard netting is recommended where otters or mink are a problem. For large cages, a supporting pole mounted on a ﬂoat in the middle of the cage may be necessary (Fig. 7.16). In Canada, electric fences have been used to exclude otters and mink with some success (Rueggeberg & Booth 1989). Tensioning of cage bag walls can help exclude predators (Rueggeberg & Booth 1989; Huguenin & Willinsky 1996) while antipredator nets provide a considerable degree of added protection (Quick et al. 2004). The material, design, size and mesh size of antipredator nets used depends on the species of predator involved. In Scotland, cage ﬁsh farmers generally use curtains of 10-cm mesh PE nets, 5–10 m deep that are ﬁxed approximately 1–2 m from the cages to protect caged stock from seals and diving birds. In Tasmania, 35-mm PP netting of 4-mm bar diameter or steel mesh netting, hung 1.5 m from the cage bag wall, is recommended for excluding seals (Pemberton & Shaughnessy 1993). Small mesh antipredator nets may be used in Shetland where guillemots (Uria aalgae) are common predators. Many birds (e.g. cormorants) can dive to depths of 20–30 m and come up under the cages. Dogﬁsh (Squalus spp.), too, scavenge for dead ﬁsh by attacking the bottom of cages. However, to use nets that extend to these depths or that protect the cage bottom is expensive. Alternatives include the use of an outer ‘false bottom’ made of reinforced material that separates stock from predators. An alternative is to suspend a cone net from the cage bottom panel (for more details see Quick et al. 2004). Like other nets in use on a cage ﬁsh farm, antipredator nets need to be serviced regularly.

272

Chapter 7

Fig. 7.15 Antipredator measures for cages. (a) Plan view of cages; (b) system with curtain-type antipredator nets extending well below the cage bottom; (c) completely enclosed antipredator net system; (d) tent-type antipredator nets.

Problems

273

Fig. 7.15 Continued.

A number of alternative solutions to predation problems at cage farms are available. Several companies now market galvanized steel mesh cages, especially in Australia, as effective in excluding predators (see Fig. 3.12). Würsig & Gailey (2002) have also proposed that bubble curtains, which are effective in acoustically masking noise in the environment, may also deter predators from exploring cage sites and thus attacking caged stock.

274

Chapter 7

Fig. 7.16 Anti-bird lines stretched across the top of a large, ﬂexible frame cage. Note pole in centre to help support system.

Deterrence/harassment Effective harassment measures include the use of dogs to deter mink and rats (IDRC/SEAFDEC 1979; Rueggeberg & Booth 1989). The presence of staff, too, deters predators. Scarecrows, ﬂashing lights and pursuit of predators may reduce the incidence of attacks but will not stop them. There are two types of acoustic devices: acoustic deterrence devices are designed to alert predators to the presence of netting, while acoustic harassment devices generally emit louder noises and are designed to scare potential predators away. In practice, devices that produce irritating sounds or the sounds of natural predators of seals are used and a number of commercial devices are available (see Jackman & Ace-Hopkins 1993; Quick et al. 2004). While some systems in some situations have proved effective, elsewhere they have proved to be of only limited success as predators appear to habituate to the signal (Mate & Harvey 1986; Beveridge 1988; Pemberton & Shaughnessy 1993; Würsig & Gailey 2002; Quick et al. 2004). In their survey of Scottish salmon farms, Quick et al. (2004) found that only 23% of site managers considered seal scarers to be highly effective, although the majority of managers believed them to be moderately effective. In six cases (7%), seal scarers were deemed ineffective.

Problems

275

Removal Live traps are sometimes employed against mink and otters, but although successful in the short term, trapping has rarely proved effective for long. Trapping also may require a permit or licence. Shooting is sometimes carried out by farmers. However, there are two points to bear in mind. First, shooting of wild birds and mammals is illegal in many countries. In the United States, for example, birds are protected by federal and state laws and cannot be killed without both a permit issued by the Law Enforcement Division of the Fisheries and Wildlife Service and a State permit. Federal permits are only issued after all non-lethal control methods have been exhausted. In their survey of antipredator devices used on Scottish salmon farms Quick et al. (2004) reported that 62% of site managers believed shooting to be very effective. However, a review carried out by Beveridge (1988) found little scientiﬁc evidence to support this claim.

7.7.6

Discussion

Prevention of a problem is always better than having to ﬁnd a solution. Avoidance of sites where predators occur can help reduce predation, especially with regard to seals and sea lions (Pemberton & Shaughnessy 1993). However, increases in seal populations in many parts of the world have resulted in an expansion in the numbers of haul-out sites. Many piscivores are also highly mobile. Choice of system can help reduce problems. Rigid mesh cages, for example, are much superior to conventional cage designs in terms of resistance to damage and inﬁltration by predators (Huguenin et al. 1981; see also section 3.3.1). Management can also signiﬁcantly minimize problems. Cage nets should be rigged and weighted so that the netting remains taut, thereby reducing the success rate of attacks. Because scavenging ﬁsh such as dogﬁsh are attracted by dead ﬁsh on cage bottoms, mortalities should be removed. Bags of feed should not be left on cage walkways. Properly secured top nets and antipredator net curtains should be used where problems are signiﬁcant. Many industry Codes of Practice now incorporate advice of this sort.

7.8

WASTES

Results from a number of studies have demonstrated that the quantities of material sedimenting under ﬁsh cages at intensively managed farms may be as much as an order of magnitude higher than those recorded at control sites or in undisturbed water bodies. As most material is composed of uneaten food, higher sedimentation rates occur at the more intensively farmed units. However, there is enormous variability in the rates of waste accumulation due to local site conditions, species, feed type and management. The closer the cage bottom is to the lake or sea bed the greater the proportion of wastes that is likely to sediment directly under the cages, and where prevailing currents are low, few of the wastes – especially the larger, denser waste feed particles – will be carried away (see

276

Chapter 7

section 5.3.1). Anchoring rafts of cages on single-point moorings (see Figs 3.42, 3.43) results in the distribution of the wastes over a much greater area and hence greatly reduces the rate of areal waste accumulation.

7.8.1

Effects on caged ﬁsh

Accumulated sedimented wastes may adversely affect caged ﬁsh production in a number of ways. Increases in the rates of supply of organic material to sediments stimulate microbial and macroinvertebrate activity which in turn increases sediment oxygen demand. Where the rate of waste accumulation is high, the oxygen supply may be insufﬁcient to meet the respiratory demands of the macrobenthic and microbial community. Sediments become anoxic, the benthic community structure becomes dominated by species tolerant of low oxygen conditions and anaerobic species, and the end result is the production of reduced inorganic and organic compounds such as lactate, ammonia, methane, hydrogen sulphide and reduced metal complexes (see section 5.3). The presence of anoxic sediments under and around ﬁsh cages has been reported in many studies (see Arizono & Suizo 1977; Braaten et al. 1983; Nishimura 1983; Gowen 1990; Kupka-Hansen et al. 1991; Holby & Hall 1992, 1994; Kelly 1992; Sowles et al. 1994; Berg et al. 1996; Findlay & Watling 1997; McCaig et al. 1999; Pearson & Black 2001; Brooks et al. 2002). The implications for ﬁsh health of a build-up of wastes around ﬁsh cages are not fully understood. High rates of sediment oxygen consumption can cause deoxygenation of overlying water, as has been observed during the summer months at Japanese yellowtail farms (Nishimura 1983; Tsutsumi & Kikuchi 1983) and at Scottish salmon farms (Lumb 1989), posing a threat to ﬁsh health. Upwelling of deoxygenated waters can also occur, causing ﬁsh kills. In studies of Israeli dual-purpose reservoirs used for cage culture, Zoran et al. (1994) reported the occurrence of daily ﬁsh kills at certain times of the year because of wind-induced tilting of the thermocline which allowed the inﬂux of deoxygenated hypolimnetic water into cages. Mass mortalities also occurred at the end of prolonged periods of stratiﬁcation as a result of upwelling of anoxic bottom waters (see also section 4.1.1). Fish kills in cages caused by mixing in stratiﬁed lakes have also been reported from the Philippines by Santiago (1994) and Bagarinao & Lantin-Olaguer (1998). Hydrogen sulphide (H2S) produced during anoxia suffers a number of fates. Since it is highly soluble in water it is present in solution in sediment pore water. However, it is readily precipitated as ferrous sulphide, giving anoxic sediments their characteristic black coloration. Hydrogen sulphide may also be utilized by sulphur bacteria present at the sediment/water interface, while under certain conditions H2S gas may be released from the sediments. Although H2S is known to be highly toxic to ﬁsh (Reynolds & Haines 1980; Black et al. 1994; Santiago 1994; Bagarinao & Lantin-Olaguer 1998), much of the gas may be oxidized before it reaches the caged stock. Nevertheless, methane/H2S bubbles have been implicated in outbreaks of gill disease at Norwegian salmon farms (Braaten et al. 1983; Black et al. 1994) and in ﬁsh kills in cages in tropical fresh waters (Santiago 1995; Bagarinao & Lantin-Olaguer 1998). There may also be a build-

Problems

277

up of facultative pathogens in the sediments under cages; hence, disease outbreaks are reportedly more common at long-established farms (Ikenoue & Kafuku 1988). In summary, the relationship between sediment conditions, farmed ﬁsh health and mortalities has been little studied and remains open to speculation. Nevertheless, there is evidence of a link. In a study of yellowtail farms, Arizono & Suizo (1977) demonstrated that disease outbreaks, where more than 1% of stock was lost, were strongly correlated with conditions in the sediments, as classiﬁed by EI (environmental index): EI = TS ◊ DO ¥ 100 where TS = concentration of sulphides in the mud (mg g-1 dry mud) and DO = dissolved oxygen concentration of the water immediately above the sediments (ml l-1) (Fig. 7.17). Waste-rich sediments also act as an important reserve of nutrients. Under appropriate conditions nitrogen and phosphorus are released into the overlying

Fig. 7.17 Environmental index versus number of disease outbreaks at Japanese cage ﬁsh farms (redrawn from Arizono & Suizo 1977).

278

Chapter 7

water, stimulating phytoplankton growth (Gowen & Bradbury 1987; Kelly 1992; Santiago 1994).

7.8.2

Prevention and remediation

The accumulation of wastes under and around ﬁsh cages should, therefore, be of concern to ﬁsh farmers and regulators. The following may be taken as signs of growing problems with waste accumulation (Braaten et al. 1983): • • • •

the periodic smell of sulphur and methane, especially during summer and autumn; the appearance of gas bubbles at the water surface; the rapid accumulation of anoxic (black) sediments under cages, with pockets of trapped gas (divers should periodically examine the muds under the cages); extensive gill damage – gills may appear pale, sometimes with dark patches; ﬁlaments are swollen, fused, or badly damaged.

Again, there are two approaches to addressing the problem: prevention and cure (collection, dispersal). Prevention In recent years ﬁsh feed manufacturers have produced a range of ‘low-pollution’ diets for the salmon and trout industries, diets that in essence contain lower levels of phosphorus and that are more digestible. Better husbandry – lower stocking rates, improved feeding regimes – and better feed technology are also helping to reduce wastes, and in some countries the adoption of such measures has been encouraged through regulation of feed use (see Fig. 5.12). In Hong Kong scavenging red (Pagrus major) and black (Mylio macrocephalus) sea breams have been stocked inside grouper cages in an effort to reduce waste feed losses (Tseng 1983), while in Saguling Reservoir, Indonesia, cage culture of tilapias that utilize detrital uneaten food and faeces from intensive cage carp production has been successfully developed (Costa-Pierce 2002). Judicious site selection can help ensure waste accumulation in the vicinity of cages is minimized. Sediment carbon loadings of <1 g C m-2 per day in temperate marine conditions should ensure that sediments remain oxygenated (Hargrave 1994) (note that sediments in warmer climates may be capable of higher loadings; see Angel et al. 1992). The choice of deep, well-ﬂushed sites minimizes waste accumulation while the use of single-point moorings can greatly increase the dispersion of wastes. However, it is increasingly argued that the ‘solution-to-pollution-is-dilution’ approach is undesirable. Collection There have been numerous attempts to collect solid wastes as they fall through the cage bottom towards the sediments. The ﬁrst trials were conducted in Poland, where large ﬁlter-funnel shaped collectors completely enclosing the cage bottom

Problems

279

Fig. 7.18 PC-controlled waste collection and uneaten food retrieval system, Sweden (from Enell et al. 1984).

were slung under rainbow trout cages and the sedimented wastes periodically (weekly–monthly) recovered by pump (Tucholski et al. 1980a, b; Tucholski & Wojno 1980). Forty-ﬁve percent of solid wastes were recovered, although the reductions in nitrogen and phosphorus loadings were only 15–20%. In Sweden, Enell et al. (1984) developed a PC-controlled system at a 20-t commercial rainbow trout cage farm. Large PVC funnels, similar in design to those used in Poland, were attached to the bottoms of ﬁsh cages. Immediately following feeding, submerged pumps were switched on for several minutes and the waste feed and faeces pumped to the surface where the food was separated and returned to the cages (Fig. 7.18). Results showed that 70% of waste phosphorus was recovered while savings in lost feed resulted in a 25–30% reduction in feed costs. Although cost-beneﬁt analysis demonstrates that the system pays for itself within 2–3 years, the assumption that the recovered feed would be fully utilized by ﬁsh proved wrong. Waste collection systems have also been developed for use in lakes in the United States (Behmer et al. 1993; Costa-Pierce 1996). Behmer et al. (1993) reported that food recovery varied with currents, ranging from 16% under calm conditions to less than 5% when high currents prevail. Kelly & Elberizon (2001) also report that high levels of dissolved wastes are lost, even when such systems are used. Several commercial waste feed recovery systems are now available, including the Lift-Up system developed by Refa A/S (Behmer et al. 1993; Juell et al. 1993;

280

Chapter 7

Fig. 7.19 Diagram of low-speed mixer installed on the bottom of a cage site to disperse wastes. The mixer can also be raised or the angle changed to improve water ﬂow through cages.

Ervik et al. 1994a; Kelly & Elberizon 2001) (see Fig. 6.20). However, they are not widely used. Dispersal and sediment management strategies Submerged, electrically driven mixers (Fig. 7.19) have been employed to disperse wastes. Braaten et al. (1983) found that dispersal was most effective if the mixer was operated from a boat so that it could be moved around the farm and if the device could be positioned above the sea bed during operation. In a three-month trial 40 cm of accumulated wastes were reduced by 60–75%. Annual operational costs of preventing further accumulation at the 35-t production site were estimated to be around £1800 (1983 prices). Again, however, waste dispersal strategies only move the problem elsewhere and are not a solution in lakes or reservoirs. Fallowing of marine sites is widely advocated as an effective means of controlling the accumulation of wastes (O’Connor et al. 1994; McGhie et al. 2000; Brooks et al. 2003). On removal of ﬁsh cages from a site, sediments will recover, although recovery rates can vary taking from less than a year to several years, depending on site, extent of impact and how recovery is assessed (Johannessen et al. 1994; Karakassis et al. 1999; McGhie et al. 2000; Mazzola & Sarà 2001; Pearson & Black 2001). O’Connor et al. (1994) advocated the use of harrowing to accelerate sediment recovery but recommended that attention was given to timing as re-suspension of sediments might cause clogging of gills and encourage the formation of algal blooms. However, implicit in any fallowing strategy is the ability of farmers to move production from one site to another, and planning regulations or economics do not readily facilitate this. No recovery of Posi-

Problems

281

donia beds has been observed following the removal of cages (Delgado et al. 1999), at least in the short term. Sedimented wastes can also be removed using submersible pumps (Braaten et al. 1983). However, while aquaculture wastes, being rich in organic minerals and nutrients, can be used as fertilizers for crops (Willett & Jacobsen 1986; Bergheim et al. 1993), they are highly dilute, expensive to move and are of little value in many parts of the world where disposal of cattle and pig manures already poses a problem. Marine sediments will also be contaminated with salt. Discussion The best and most cost-effective solution to problems associated with waste accumulation under cages is to pursue a strategy of waste reduction. Recent improvements in feed and feeding methods show that this is possible; FCR values in European salmon farming have fallen by more than 20% in 5 years, for example (see Fig. 5.12). Collection may be effective in removing solids and may be particularly useful during chemotherapy or to recover mortalities, but collection deals only with reducing part of the loadings. For this reasons, among others, Martin & Heard (1987) and Fast (1991a, b) have developed solid cages and ﬂoating raceways, structures through which there is no free exchange of water, thereby facilitating waste collection (see also Kelly & Elberizon 2001). Such designs still beg the question, however, of how to dispose of the collected material. Dispersal and fallowing do not address excessive feed use although they may help prevent the assimilative capacity of sediments being exceeded.

7.9 7.9.1

WEATHER AND CLIMATE Rough weather conditions

Certain sites, especially those offshore, are prone to violent storms or adverse sea conditions by virtue of location or topography (see section 4.2). During storms, high winds blowing over exposed stretches of water cause wave heights to increase and wavelengths to decrease. The resulting wave action can not only damage cage structures and moorings but can also injure ﬁsh through the pitching motion of nets, and can make routine jobs such as net changing and harvesting hazardous. In some parts of the world the best marine sites have already been developed and if more exposed locations are to be utilized, then either the environment must be modiﬁed so that the effects of waves are reduced, or cages able to withstand poor conditions or that can take evasive action by submerging during bad weather, must be used. Breakwaters Breakwaters are often used to reduce the effects of waves on coastal marine installations such as harbours and marinas, and a number have also been installed to protect coastal aquaculture projects in various parts of the world

282

Chapter 7

(Tazaki et al. 1975; Kato et al. 1979; Kogan & Romanycheva 1979). There are two basic types: ﬁxed bottom-resting and ﬂoating. The former are usually substantial concrete or stone-built structures that reﬂect wave energy. They have an extended life-span and low maintenance costs but are expensive, costing at least US$3000 m-1 (extrapolated from De Young 1978 to 2003 prices) and suffer from a number of disadvantages from a cage ﬁsh farming point of view: • • • • •

they reduce or interfere with prevailing currents; they are largely restricted to shallow waters, otherwise costs become prohibitive; they are difﬁcult to install where bottom sediments are soft; once constructed they are difﬁcult to modify to meet changing circumstances; they are ‘over-engineered’ if protection is only required for short periods during the year.

Floating breakwaters, on the other hand, have been suggested as being more appropriate for use in cage ﬁsh farming; they are relatively inexpensive, can be moored in deep or shallow water, do not interfere with currents and can be readily modiﬁed as farms expand or cages are moved (Tazaki et al. 1975; McGregor 1978; Kato et al. 1979; Kerr et al. 1980; Twu et al. 1986). Floating breakwaters redistribute wave energy in a number of ways (McGregor 1978): • • • • •

reﬂection; dissipation within the structure; transmission of unsuppressed energy; diffraction (bending) of the wave train passing the end of the breakwater into the sheltered lee of the structure; generation of waves by the movement of the breakwater itself that interfere with the incident waves.

The relative importance of each effect is design dependent. At one extreme are ﬂoating pontoon designs that reﬂect most of the energy, while at the opposite extreme are ﬂoating sheets that absorb and dissipate energy by friction (Kato et al. 1979). Floating pontoons, while very effective, are difﬁcult to anchor securely and ﬂoating sheets are impractical since in order to obtain a sufﬁcient damping effect, it is necessary for the sheet to be more than ten times the wavelength of the incident waves. Floating structures that incorporate the properties of the two designs (i.e. redistribute incident wave energy both by reﬂection and absorption) without suffering from the disadvantages of either are ideal. Much research and development has gone into ﬂoating breakwaters and a number of commercial designs are available. In Japan, for example, Bridgestone breakwaters, developed over the past 35 years, have been used in several areas to protect aquaculture projects. The design consists of ﬂoating synthetic rubberﬁlled ﬁbreglass modules (12 ¥ 7 m or 10 ¥ 7 m) linked together as shown in Fig. 7.20 on the windward side of a ﬁsh farm. Effectiveness depends on incident

Problems

283

Fig. 7.20 Bridgestone breakwater system, Japan (courtesy J. Kato).

wavelength and the deployment of modules. Details of commercial ﬂoating breakwaters and installation can be obtained from manufacturers. Although a number of radical designs of ﬂoating breakwater have been designed for use in aquaculture (see Twu et al. 1986), the most widely used remains the ﬂoating tyre breakwater or FTB. FTBs were ﬁrst developed in the United States in the early 1970s. Most use ﬂexible mats based on the Kowalski or Goodyear module (see Kowalski 1974, 1976) in which 18 tyres are tied together to form a module that ﬂoats with all tyres in the vertical plane (Fig. 7.21a, b). Air trapped in the crowns of the tyres gives each module sufﬁcient buoyancy to support the weight of a man. Wave-induced movement of the modules is sufﬁcient to replenish air pockets and keep the structure aﬂoat, unless prolonged periods of calm occur during which trapped air may slowly be dissipated in the sea water. Lacerations and heavy fouling also adversely affect buoyancy. Both truck/tractor and car tyres can be used to construct the modules. Because truck and tractor tyres are considerably larger, fewer modules are required to effect comparable wave attenuation, and since wave attenuation is to some extent dependent upon draft, they are more effective than car tyres in deeper (>10 m) water. However, they are less buoyant since most truck and tractor tyres are tubed and unlike car tyres do not have airtight butyl or chlorobutyl linings. The heavy-duty steel beading used around the rims of tractor and truck tyres makes them much more abrasive and thus much more likely to cause problems with tying materials (see later). Moreover, they are heavy and difﬁcult to handle.

284

Chapter 7

Fig. 7.21 Floating tyre breakwater (from several sources).

The ﬁnal choice can be derived from computer simulations of wave characteristics at the site (see later), but more often than not is dictated by availability of materials and economics. Nylon, PP (polypropylene) and dacron ropes, stainless steel wire closed and open link galvanized chain and conveyer belting have been used as tying materials. However, trials have demonstrated that ropes tend to abrade and wear out very quickly and that either 35-mm wide strips of 3-ply synthetic rubber conveyer belting or 8-mm galvanized chain are best. Chain is expensive, heavy and prone to wear inside the links, while conveyer belting is considerably less durable. De Young (1978) describes a simple-to-manufacture tyre rack to aid construction of the modules. However, he estimates that two workers can put together a module in 5–10 min by stacking the tyres on the ground via a 3-2-32-3-2-3 conﬁguration, standing on top, and securing the bindings in such a way that the module is rigid without the tyres being distorted. Greased high-tension shackles or closing links work best with chain, while black nylon or stainless steel bolts or shackles may be used to secure conveyor belting (Fig. 7.22). Assembly of the FTB from modules requires that the four outside tyres be swung out as shown in Fig. 7.21c, d. Four additional tyres are then used to link a group of four modules together. Small sections are put together on land, launched and linked up with other units in shallow water before being towed to the site and connected to form completed sections (Fig. 7.23). The size of the sections completed on land is dictated by the available equipment. A 4 ¥ 10 module unit can easily be towed by a boat ﬁtted with a 40 hp engine during calm weather.

Problems

285

One method of mooring an FTB is to secure a windward corner of a completed section to an anchor and anchorline, tow it to the site using the anchorline as the towrope, and to drop the anchor when the section ﬂoats into the correct position. The section can then be secured with as many anchors as necessary. The second section is then towed into position and attached to the ﬁrst

Fig. 7.22 Method of joining scrap conveyer belt ties (modiﬁed from White Fish Authority design).

(a) Fig. 7.23 (a), (b) FTB modules being assembled prior to being towed out to the installation site on the west coast of Scotland.

286

Chapter 7

(b) Fig. 7.23 Continued.

prior to the anchors being dropped. The procedure is repeated until all of the sections are in place. FTB design is site-speciﬁc, depending on wave heights, wavelengths and wave periods, the size and orientation of the farm, and the size and type of cages used. It must also take account of navigational regulations. Consultant engineers can carry out design trials for a particular site using wave tanks, model tyres and a wave generator that can simulate the wave spectra of the site based on appropriate meterological and topographical data. FTBs developed in this way have performed well (McGregor 1978; McGregor & Gilbert 1983). However, farmers may assemble their own FTBs using some simple rules of thumb. Because the breadth of an FTB largely determines its wave damping characteristics, and since the most destructive wavelengths are between 0.5 and 1.25 times the length of the structure (see also section 3.3), Kerr et al. (1980) suggest that the breadth of the breakwater is around 1.5 times the width of the structure to be protected. Thus, a group of 4-t production cages, exposed to gale force winds over an effective fetch of 3–4 km, should be protected by an FTB four to ﬁve modules in breadth. The length of the FTB should be at least as long as the cage raft(s) it is protecting since there will be some refraction around the edges. The anchoring system is designed to withstand peak mooring loads and depends upon the breadth and draft of the structure, the depth of water, the wave steepness (the ratio of wave height : wavelength) and the substrate. Numerous methods of mooring marine structures can be found in the literature (see also section 3.3.4) and details with regard to FTBs are given in De Young (1978). McGregor (1978) has estimated that average mooring loads are approximately

Problems

287

Fig. 7.24 Suggested mooring scheme for an FTB.

250 N m-1, with extreme values around ﬁve times greater than this. McGregor & Gilbert (1983) used 250-kg block anchors every 9 m on the windward side of a three-module broad FTB anchored in soft sediments in shallow water, while De Young (1978) estimated around twice that holding power was necessary for an eight-module broad structure installed in a similar depth of water. Anchorage on the leeward side should be around 20–25% that on the windward side (see section 3.3 for details of anchor types). A buoy is placed at the top of each riser chain for support and to facilitate removal and maintenance of the structure. Each buoy should be connected to at least two modules in order to distribute the mooring load as evenly as possible (see Fig. 7.24). FTBs should be well marked and navigation lights installed if necessary. Best protection is achieved by orienting the breakwater at 90° with respect to the most destructive waves. However, if more than one fetch is involved then there must be a compromise. As a general guide, De Young (1978) recommends that cages are moored on the leeward side within four wavelengths of the breakwater, although the ﬁnal position of the FTB relative to the farm is best decided by trial and error. One advantage of FTBs is that they can be moved relatively easily. Assuming that the tyres are scrap, capital costs are around 10–20% that of a ﬁxed breakwater, depending on the breadth, the type of anchors and the tying material used, and excluding the costs of transport. However, maintenance costs can be up to 30% of capital outlay per annum as ties inevitably wear out or fail, and it is recommended that moorings and ties be checked every month or so and that the whole structure be beached regularly – each year – and serviced. Fouling should be removed and damaged tyres or ties replaced. Modules that lose buoyancy and sink can be temporarily re-ﬂoated using compressed air, and the crowns of leaking tyres ﬁlled with high-density polyurethane or polyethylene foam (McGregor & Gilbert 1983). There may also be problems during winter with the formation of ice on the leeward side of the structure and in severe

288

Chapter 7

climates the movement of ice ﬂoes has caused a great deal of damage. Maintenance of the FTB can be assigned to a contractor. A number of manuals on the design and construction of FTBs are available (De Young 1978; Bishop 1980). Several cage manufacturing companies also offer breakwater construction and installation services. Pneumatic barriers Another method for modifying wave regimes is discussed brieﬂy by Milne (1970, 1979) and involves using a pneumatic barrier, produced by pumping compressed air through a submerged perforated pipeline (Fig. 7.25). Limited trials conducted by Milne demonstrated that both wave height and turbulence could be ameliorated to some extent although temperature/salinity proﬁles were also disrupted. There have been no further developments of this system for the protection of cages. Floating cage designs The alternatives to using FTBs or bubble curtains are to use cages that can withstand bad weather/open sea conditions or cages that can be lowered below the surface. There are two approaches to building cages that can withstand rough seas – either rigid collar cages built to high speciﬁcations using strong materials or ﬂexible collar designs. The former owe much to offshore oil industry technology. They are large and often incorporate accommodation, storage for feed, a slaughterhouse and cold store facilities. However, few have been built or have remained in operation for long (Oltedal et al. 1988; Svealv 1988; Bjerke 1990; Polk 1996; Huguenin 1997; Lien 2000; Muir & Basurco 2000). Flexible collar cages of the type shown in Fig. 7.26 were ﬁrst developed in Japan in the early 1980s and are now widely used in Western Europe, North America, New Zealand, Australia and in the former USSR. Commercial designs are typically constructed from 10–16-m lengths of hollow synthetic rubber tubing, 380–420 mm in diameter, which are bolted together to form 4-, 6-, 8- or 10-sided or circular collars. The units are ﬁlled with gas at high pressure to aid ﬂotation and increase rigidity. In some designs, two concentric rings of tubing are used in order to increase stiffness and reduce movement in the horizontal plane and to provide a base for a walkway. In most designs the cage bag is supported by its own ﬂotation system so that collar and bag move independently of one another. Advantages and disadvantages are summarized in Table 7.4. Several ﬂoating rigid collar designs have also been built (Polk 1996; Muir & Basurco 2000). They are generally large structures, constructed from steel, and incorporate a variety of management-related features, such as feed stores, harvest cranes and fuel stores (Scott & Muir 2000). They are costly and have not been as widely used as ﬂexible collar designs (see Table 7.5). Mooring requires careful consideration and is installed only after detailed study of site conditions. The simplest mooring systems use 10–30-t block anchors attached to each node point or, in the case of circular designs, at a number of equidistant points around the collar. Mooring systems reliant on plough anchors

Problems

289

Fig. 7.25 Pneumatic barriers used (a) to induce water currents, (b) to reduce wave heights, (c) to prevent incursion of salt water and (d) to contain oil and debris (from Milne 1972).

290

Chapter 7

Fig. 7.26 Bridgestone ﬂexible frame cage (courtesy B. Whelan).

Table 7.4 Advantages and disadvantages of ﬂexible frame cages (modiﬁed from Scott & Muir 2000). Advantages

Disadvantages

Highly resilient to wave forces, with long service life (>10 years); relatively good impact resistance Effective and proven net hanging system Variety of conﬁgurations possible Relatively cheap at higher volumes Most widely used commercial offshore system

Stanchions may cause problems (e.g. twisting, turning) Relatively expensive at lower volumes Limited walkway access Top net and feed systems difﬁcult to place Large service vessels necessary

Table 7.5 Advantages and disadvantages of ﬂoating rigid collar cages (modiﬁed from Scott & Muir 2000). Advantages

Disadvantages

Stable working platform for all husbandry and management operations

Large and heavy structures requiring good port facilities and/or expensive towing to install May be susceptible to structural failure in extreme conditions Large mass may require heavier mooring systems Relatively high capital cost; steel structures require protection/maintenance Limited commercial track record

Potential for integral feeding and harvesting systems; may be used to service other cages Ship mortgages may be available Potentially improved operator safety and efﬁciency Construction and repair facilities may be carried out at conventional shipyards

Problems

291

(~750 kg) are easier to deploy. High-speciﬁcation ground chains (5-cm stud link) and braided rope (50 mm) are typically used (Christensen 2000). In rough weather, the rubber collars ride out heavy seas by absorbing and dissipating wave energy through internal friction in much the same way as an FTB (see earlier), while the cage bag takes advantage of the shelter afforded by the collar and moves comparatively little. The cages appear to be relatively easy to operate, even in parts of the world where nets must be frequently changed, it taking three men 3–5 h to change nets using power-assisted lifting gear. Flexible frame cages in use on the east coast of Japan reportedly have survived 7-m waves and winds of 29 m s-1 that occurred during a typhoon. Cages of over 1000 m2 have been built and, according to the manufacturers, performance in adverse weather conditions is largely independent of size. Details of construction and operation can be found in Gunnarsson (1988, 1993), Fearn (1990), Slaattelid (1990), Polk (1996), Lien (2000) and Muir & Basurco (2000). Semi-submersible and submersible cage designs Semi-submersible and submersible cage designs, which differ only in the degree to which they can be submerged, are used in commercial aquaculture in a number of countries, usually as a solution to exploitation of exposed sites but also to reduce problems with ice (see later). The principal reason for using such designs is that the water movements produced by waves are largely a surface phenomenon. As a deepwater wave moves over the sea, the water particles do not move forward with the wave, but instead rotate in circular orbits (Fig. 7.27). At the surface, the diameter of the orbit is equal to H, the wave height. However, the orbit of the particles decreases exponentially with depth, such that:

Fig. 7.27 Rotational properties of water particles in a wave (redrawn from Pond & Pickard 1983).

292

Chapter 7

Dz = H ( 2 z L) where Dz = diameter of the orbit; z = depth; and L = wavelength (Pond & Pickard 1983). At a depth of around L/9, Dz is approximately halved, and at L/2, Dz has decreased to 0.04 H. Thus submerging cages, particularly during storms when the waves increase in steepness (wave height : wave length ratio), can dramatically reduce wave forces and the pitching motion of the cages. Farm Ocean cages are among the most widely used of the commercially available semi-submersible rigid collar designs (http://www.farmocean.se/; Fig. 7.28). They were designed in Sweden and ﬁrst deployed in 1986 (Scott & Muir 2000); today, there are more than 40 deployed in Northern Europe and the Mediterranean alone. Cage volumes range from 2500 m3 to 6000 m3. The most widely deployed has a volume of 3500 m3 and consists of a 25-m diameter tubular

(a) Fig. 7.28 A Farm Ocean semi-submersible cage (a) located in a semi-exposed site on the west of Scotland. In operational mode (b) the ﬂotation system is situated 3 m or so below the water surface while for servicing (c) the ballast tanks are emptied of water and ﬁlled with air so that the cage rises out of the water.

Problems

(b)

(c) Fig. 7.28 Continued.

293

294

Chapter 7

hexagonal steel pontoon to which the six-panelled net bag is attached. Six ballast tanks in the pontoon control the position of the cage in the water column. During operation the ﬂotation system lies 3 m below the surface, while for servicing air is pumped into the ballast tanks so that the pontoon is lifted slightly out of the water. The bottom of the cage bag is weighted to maintain shape. A 3-t capacity feed hopper sits on top of the pontoon. Moorings are attached at three points to the collar. According to the Norwegian Veritas speciﬁcation, the cage is designed for semi-exposed rather than exposed conditions, tolerating wave heights up to 5.5 m. The Refa cage is a tension-leg semi-submersible design, in which a positive buoyancy plastic-supporting frame is held in place by vertical mooring ropes attached to concrete blocks on the sea bed and to sub-surface buoys (Lisac & Muir 2000; Scott & Muir 2000) (www.versaware.net/Acuicultura/Documentos/ TLC.Instalaciones.pdf). Cage volume is up to 10 000 m3. They are simple, relatively inexpensive and are claimed to deform little in strong currents or during storms. However, sub-surface feeding systems are required and the heavy block anchors are difﬁcult to install. There are both ﬁxed and ﬂoating types of submersible design. Fixed submersible cages have been used since the late 1970s in several lakes in the Philippines to culture tilapias (Beveridge 1984b) (see Fig. 7.6). Cages are ﬁtted with top nets and in the event of a storm the rigging is untied and the cage bags secured a metre or so under the water. The posts that anchor the cage to the substrate and project above the surface offer little resistance to wind or waves. In Japan, ﬂoating submersible cages have been used to culture yellowtail, and red and black sea bream at exposed sites since the early 1970s. Early designs were very simple. In one example, the head rope of the cage bag is attached to surface mooring buoys and during normal operating conditions ﬁxed so that the cage remains 2–5 m below the surface. Fish are fed via a tube. During storm conditions the feeding tube is tied shut and the ropes lowered so that the cage is submerged to a depth of 10 m or so (Fig. 7.29). According to Fujiya (1979), submerged units of up to 100 m2 have been used and are secured together in groups of 10–12. A more sophisticated design, which uses variable-buoyancy synthetic rubber ﬂoats (Fig. 7.30a) that can be ﬁlled or emptied with compressed air or sea water, was developed by Tomi et al. (1979). Floats are linked by highcompression rubber hose-line and the cages conﬁgured as shown in Fig. 7.30b. Cages are only submerged during rough weather. However, problems of net deformation due to deep-water currents have been reported. Submersible designs have been on trial in other parts of the world for 20 years now. In Martinique, sea bass, sea bream and sar (Diplodus sargus) have been cultured in rotating cages, which could be swiftly sunk to the sea bottom during typhoons (Rene 1984). In Europe, submersible designs have been used to culture salmonids, sea breams and sea bass (Boeuf & Harache 1980; Woods Hole Engineering Associates 1984; Cook et al. 1984). In France, a submersible design was used for several years at exposed coastal rainbow trout farms in Brittany. The cages (64 m3) were raised each day to feed the ﬁsh and then lowered again to a depth of 15–20 m (G. Boeuf, pers. comm.). The Sadco submersible rigid cage design has been in development since the early 1980s (Bugrov 1996). Today,

Problems

295

Fig. 7.29 Submersible cage for yellowtail (redrawn from Fujiya 1979).

Fig. 7.30 (a) Diagram of a pneumatic submersible ﬂoat (from Tomi et al. 1979). (b) Layout of cages and deployment of submersible ﬂoats (from Tomi et al. 1979).

296

Chapter 7

Fig. 7.30 Continued.

models of 1200, 2000 or 4000 m3 are produced, consisting of a net bag whose shape is maintained by a ballasted upper steel superstructure and a weighted steel lower frame (Fig. 7.31). Cages can be lowered to a depth of 30 m and the manufacturers claim they can withstand 15-m wave heights. Cages are supplied with different types of feeding system, including manual feeding, self-contained underwater feeders and an automated remote-controlled feeding system. Cages are currently installed in the Caspian and Black Seas and in the Mediterranean (http://www.sadco-shelf.sp.ru/english/ep02_pr.htm). Spherical, geodesic dome design Trident cages have been used to farm salmon in the Bay of Fundy and Bras D’Or Lakes of eastern Canada since the early 1990s (Willinsky et al. 1991, 1995a; Huguenin & Willinsky 1996). Sizes range

Problems

297

Fig. 7.31 Sadco submersible rigid cage, 1200–4000 m3 in volume, complete with self-contained feed store (courtesy Sadco Shelf).

from 1000 to 5500 m3. In the normal operating position, 81% of the cage volume is below the water surface although during storms or ice conditions the entire cage can be sunk beneath the surface. Compressed air is used to control ballast water in the aluminium foam-ﬁlled tubular ﬂotation system. There is little deformation. One cage, moored so that two-thirds of its volume was below the water surface, reportedly withstood waves in excess of 3.5 m wave height (Scott & Muir 2000). The Ocean Spar Sea Station is another rigid collar (frame) design (Loverich & Croker 1993; Loverich & Swanson 1993; Willinsky 1995a) (http://www.oceanspar.com). It takes advantage of the reduced wave motion afforded by locating the ﬂotation system below the surface. The basic design is built around a central, vertically positioned steel spar buoy that provides buoyancy and distributes the loads to the net and the circular tubular rim via tensioned lines that serve as spokes. The tubular rim can be moved to sit either above or below the mid-point of the spar buoy, thereby adjusting the net in the water column to accommodate a range of species. Other versions are under development. The design is simple, distorts little in strong currents or storm conditions and is easily towed and moored. For information on other submersible designs see Dahle et al. (1989), Dahle & Oltedal (1990), Lien (1993), Polk (1996), Huguenin (1997) and Muir & Basurco (2000). Submerged cage designs A number of submerged cage designs have been proposed, largely in response to the problems posed by farming in exposed, offshore environments or to over-

298

Chapter 7

wintering of ﬁsh (see also sections 2.1, 7.9.2). However, few designs have developed beyond the drawing board. In Belle Isle, Brittany, in the early 1980s rainbow trout were maintained in cages submerged at 20 m. Shortly after ﬁsh were introduced to the cages it was observed that many appeared ‘distorted’. The problem was solved using an inverted bell containing air that the ﬁsh could access (G. Boeuf, pers. comm.). As is also discussed in section 7.9.2, there remains some controversy regarding whether ﬁsh require access to the air to regulate swim bladders (Mikheyev et al. 1970; Fosseidengen et al. 1982; Ablett et al. 1989; Rubach & Svendsen 1993; Saunders 1995; Huguenin & Willinsky 1996). According to Juell & Westerberg (1993) salmon maintained in large cages swam at depths of 2–7 m for 25 days without contact with the water surface. Rubach & Svendsen (1993) kept salmon in submerged cages for 42 days. They proposed that in this instance the ﬁsh were given a sufﬁciently large cage to allow normal swimming behaviour and that problems arise when small, submerged cages, which exacerbate stress and induce hypoxia, are used.

7.9.2

Ice

Ice can be a problem in both marine and freshwater environments where ﬁsh are grown in cages throughout the year. Ice occurs in many maritime temperate lakes and reservoirs for short periods during the winter months but can easily be broken up and dispersed. The movement of the caged ﬁsh, too, helps keep the environment ice-free (Paetsch 1977; Harvey 1988). An airlift or submerged propeller device may be used to transport warm, deep water to the surface, keeping ice from forming (see Fig. 7.11). Care must be taken to regulate the airﬂow in order to avoid splashing, which can cause icing of the freeboard and top nets (Kiekhafer 1983). Disturbance of sedimented wastes may also be a problem with this type of system. In the continental lakes of Europe and North America, however, thick ice cover may be present for several months during winter. Although the area immediately around the cages may be kept ice-free, problems can occur in spring when large ice sheets begin to break up and move. A 10-cm sheet of ice covering several hundred square metres will weigh tens of tonnes and can destroy cages caught in its path. Drifting sheets of ice frozen around mooring lines can also drag anchors and disrupt moorings. Siting cages on the windward side of a lake so that ice is carried away from cages by the prevailing winds reduces the risk of damage. In central Europe iron bars or anchors are trapped in the ice and secured to trees on the shore until it thaws sufﬁciently and no longer poses a threat (Kiekhafer 1983; Behrendt 1984). Temperatures inside cages in lakes with winter ice cover can be extremely low, resulting in poor or even negative growth. In the former Soviet Union two types of submersible cage have been used to overcome such problems; one for species such as rainbow trout that require feeding during the winter months, and one for those that do not (e.g. catﬁsh, carp) (Martyshev 1983). In the autumn, rainbow trout were transferred to cages of the type shown in Fig. 7.32. The cages measured 4 ¥ 4 ¥ 1 m, and were ﬁtted with a 1 ¥ 1 ¥ 1.2 m plexiglass funnel that

Problems

299

Fig. 7.32 Submerged trout cage, former Soviet Union (from Martyshev 1983).

served not only to support the cage when the ice cover formed, but also as a window through which light and air could penetrate and food could be delivered. The surface of the plexiglass funnel was maintained free from ice and snow. A cage of the above dimensions held 50 000 10–15 g trout at stocking densities of 31–47 kg m-3 throughout the winter. Fish were fed with a vitaminenriched diet at a rate of 1% body weight twice per week. The cages were freed from the ice prior to the spring thaw to prevent them being damaged by drifting ice. The design shown in Fig. 7.33 was used to over-winter carp and catﬁsh at depths of 1–1.2 m below the surface of the ice. Stock were not fed. Freezing is rarely a problem in the marine environment unless sites are located in a coastal area with a substantial freshwater input where, owing to differences in density, a fresh or brackish water layer sits above the sea water. Such areas exist on the Atlantic seaboard of Canada, for example (Page & Robinson 1992; Saunders 1995). Despite the controversy surrounding submersible cages and salmonids (see above), submersible cages have been successfully used to overwinter salmon in Canada (Willinsky et al. 1991, 1995a) and in Murmansk, Russia (Dushkina 1994). Drifting ice ﬂoes also pose a threat (see section 4.2.1) and have been responsible for a number of insurance claims (Secretan 1980).

300

Chapter 7

Fig. 7.33 Submerged carp/catﬁsh cage, former Soviet Union (from Martyshev 1983).

Farmers may choose to move cages before winter sets in. A ﬂoating boom can also help deﬂect ice.

7.9.3

Light

Light levels at farms are something that ﬁsh farmers should be concerned about, both because light is an important component in the ﬁshes’ environment and because of its effects on many of the materials used. As the light from the sun passes through the Earth’s atmosphere there is a selective scattering and absorption of wavelengths, particularly at either end of the spectrum, so that the light reaching the ground is greatly modiﬁed (Fig. 7.34). The intensity and quality of light arriving at any particular point is determined by the angle of incident light, the distance through the atmosphere it must travel and the prevailing atmospheric conditions, resulting in geographic and seasonal variations. In general, however, light intensity is much higher in the tropics than in the mid or low latitudes and increases with altitude (Fig. 7.35; see Kirk 1994 for discussion). The most damaging wavelengths in terms of the materials used in aquaculture and on ﬁsh and other living organisms are at the lower (i.e. ultraviolet) end of the electromagnetic spectrum. Ultraviolet radiation (UVR) is harmful because it is composed of comparatively high-energy photons. UVR can be divided into three types in order of decreasing wavelengths: UV-A (400–320 nm), UV-B (320–280 nm) and UV-C (280–200 nm). Only UV-A and UV-B reach the Earth’s surface, the former in much greater quantities. However it is the group that encompasses the shorter wavelengths, UV-B, that causes greatest damage (Bullock 1988; Häder et al. 1995).

Problems

301

Fig. 7.34 Solar energy as it reaches the Earth’s atmosphere at ground level. The hatched area shows the sensitivity of the human eye (400–700 nm). Thermal radiation (5000–14 000 nm, not shown) together with infrared radiation accounts for about half of the solar energy at the lake surface (from Horne & Goldman 1994).

Weathering of materials The cage collar and netting components that appear above the waterline – top nets, exposed areas of cage bags, etc. – are prone to weathering, a term used to describe the combined effects of UVR, rain, wind, industrial smokes and gases, and microbial attack, although UVR probably has the single greatest deleterious effect (Klust 1982). The mechanisms of photodegradation are complex: there is a loss in molecular weight of materials with time and an increase in density resulting from photochemical-induced ﬁssion of chemical bonds, reorientation of polymer chains and/or the formation of cross-linking (Kanehiro & Kasu 1988). Synthetic materials differ in susceptibility, and tests designed to measure deterioration in breaking strength of undyed netting materials have been carried out in different parts of the world and under different conditions. PVC ﬁbre has been shown to be the most resilient of netting materials while PA (nylon) monoﬁlaments and PVA, PVC and PES ﬁbres show moderate resistance. Normal nylon ﬁbres (continuous ﬁlament and staple) have a resistance similar to cotton and other natural ﬁbres, while PP ﬁlaments, even with the inclusion of antioxidants and radiation absorbers, are least resistant (Fig. 7.36) (Klust 1982). Largediameter netting twine is more resistant than smaller diameter materials while knotted netting shows less loss of strength with exposure than knotless netting (Dahm et al. 1990).

302

Chapter 7

Fig. 7.35 Annual variations of daily integrals of global radiation reaching the ground on clear days. Northern hemisphere – upper set of curves; southern hemisphere – lower set of curves (from Strasˇkraba 1980).

The life-span of netting materials can be extended by the addition of dyestuffs that absorb the light or antioxidants (Klust 1982; Kasu 1988; Dahm et al. 1990). Dyeing is carried out by the ﬁbre producers, netmaker or by the ﬁsh farmer. PE and PP are usually dyed by the ﬁbre manufacturer while nylon netting can be treated either by the farmer or by the net manufacturer. One of the commonest groups of dyestuffs used to treat nylon netting is coal derivatives, such as bitumen or coal tar, which in addition to retarding fouling improves resistance to degradation by light by up to 20% (Dahm et al. 1990). Unfortunately bituminous products can increase net stiffness, increasing the risks of abrasive damage to stock. In New Zealand nylon netting used for cages is often dyed black because of high prevailing UVR levels.

Problems

303

Fig. 7.36 Remaining breaking strength of netting yarns of 2–3 mm diameter, expressed in terms of initial breaking strength, after 1500 h exposure to sunlight (from Klust 1982).

Fibre manufacturers add dyestuffs to PE and PP polymers before extrusion as monoﬁlaments. The pigments used are very ﬁne particles (1 mm) of synthetic compounds, which, once incorporated, cannot be washed out. Not all dyes retard weathering. In Fig. 7.36, the breaking strains of dyed and un-dyed netting yarns after exposure to 1500 h of sun are shown. PE samples dyed orange and green are relatively well protected while the orange and blue pigments used to treat PP samples had a photo-catalytic effect, accelerating deterioration (Klust 1982). The colour of synthetic ﬁbre ropes is of less concern; they do not readily succumb to photo-degradation, harmful UVR being absorbed in the outer few millimetres (Klust 1983). Polystyrene is another material widely used in cage construction that suffers badly from weathering (Fig. 7.37). Although it is highly resistant to microbial attack it is prone to photo-degradation owing to impurities, such as acetophone groups, which are formed by thermal oxidation during manufacture (Geuskens 1975). Under intense sunlight, the outer surface becomes yellow and brittle and is easily eroded, exposing the underlying material to UVR. However, weathering is also due to physical damage by wind, waves, etc., and resistance to attack is largely dependent upon how the material was manufactured. Where weather-

304

Chapter 7

Fig. 7.37 Polystyrene block following several years weathering at a temperate, marine site.

ing is a severe problem, it is recommended that polystyrene be encased in glassreinforced concrete or epoxy-resin, plastic drums, oil drums, or tyres. Effects on ﬁsh Only around 5% of light striking the water surface is reﬂected, the rest being transmitted through the water column (Kirk 1994). As UVR travels through water, it is gradually attenuated, the rate of attenuation being primarily dependent upon turbidity and colour. The humic acid compounds that commonly stain dystrophic upland waters yellow are especially efﬁcient at absorbing UVR (Kirk 1994). However, it is a fallacy that UVR is absorbed within the ﬁrst few centimetres; the depth at which surface levels of UV-B are reduced to 10% varies from 15.4 m in clear oceanic water to 0.66 m in more turbid coastal waters (Bullock 1988). Although synthetic ﬁbre netting gains some protection from UVR when submerged, levels as low as 0.01% of surface UV-B levels can elicit a response in ﬁsh skin (Bullock 1988). Fish skin is very different in organization from that of higher vertebrates, the protective pigment layer being located in the upper dermis as opposed to the epidermis, leaving the outermost layer of the skin highly vulnerable to UVR damage (Roberts 2001). This is perhaps of little consequence in the wild, where ﬁsh can avoid harmful light levels by seeking shaded areas or deep water. However, as a number of studies have demonstrated, ﬁsh constrained in shallow aquaculture facilities such as tanks or cages can experience sunburn especially if prevailing light levels are high (Bullock 1988). Typical gross evidence of UVR damage includes greyish focal thickening in dorsal areas of the ﬁsh, particularly along the outermost region of the dorsal ﬁn or on the head (Fig.

Problems

305

(a)

(b) Fig. 7.38 (a) Characteristic UV damage to the skin on the head of a trout (courtesy A. M. Bullock). (b) UV damage to the dorsal spine region of ﬁsh (courtesy A. M. Bullock).

306

Chapter 7

Fig. 7.39 Floating trout cage, Lake Titicaca, Bolivia, ﬁtted with canvas sheet to protect ﬁsh from UVR.

7.38a, b). In many cases the areas of damage spread and are followed by necrosis and sloughing, sometimes of the entire dorsal ﬁn, exposing the underlying dorsal musculature. Histopathological examination of the centre of the lesions reveals the characteristic presence of sunburn cells in the outer layer of the epidermis (for reviews see Bullock 1988; Speare 2002). Summer syndrome in Atlantic salmon smolts is believed to occur because the rapidly growing ﬁsh have many epidermal cells undergoing DNA synthesis or mitosis, making them especially vulnerable (Speare 2002). Although there is evidence that ﬁsh have repair mechanisms which, given favourable conditions, can deal with moderate levels of radiation damage in the outer epidermis (Bullock & Coutts 1985), severe lesions are often invaded by opportunistic pathogens. Prolonged exposure to high levels of sunlight are increasingly regarded as stressful to ﬁsh, affecting swimming and feeding behaviour and perhaps suppressing resistance to disease (Huse et al. 1990; Furevik et al. 1993; Huse & Holm 1993; Fernö et al. 1995; Oppedal et al. 2001). Light levels should be considered in a similar manner to temperature and water quality, as an important component of the farm environment. Where light levels are high, some form of shade should be provided for caged ﬁsh. Huse et al. (1990) demonstrated enhanced growth and reduced sea lice infestation of caged salmon in summer in Norway by reducing light levels by 76 and 44%, although they also report suppressed growth where covers were used in winter. There is an approximate 4% increase in UVR levels for every 300-m increase in elevation and trout grown at high altitudes in Bolivia, Kenya and Mexico have been found to suffer from UVR-induced damage (Bullock 1988).

Problems

307

There has been no work carried out on the materials or cover required to give adequate protection against UVR damage. The alternatives are to use a small mesh top net to cover the entire cage surface or to restrict shading to certain areas of the cage (see Fig. 7.39), the former being preferable.

Appendix 7.1 Example of calculation for aeration system design for a freshwater rainbow trout cage, assuming airlift pumps are employed Expected problem period: Predicted worst conditions: Size of ﬁsh: Estimated respiration rate: Size of cage: Stocking density: Cage biomass: Oxygen requirements:

late summer 20°C, 50% DO saturation 100 g 300 mg O2 kg-1 h-1 5 ¥ 6 ¥ 5m 10 kg m-3 1500 kg 0.45 kg O2 h-1

Assuming aeration must supply all of the oxygen requirements of the ﬁsh, and a poor mass transfer rate (0.25 kg O2 kWh-1), power requirements can be computed as: 0.45 = 1.8 kWh ∫ 2.5hp 0.25 Assuming the farm has ten cages, total power requirements = 18 kWh or 25 hp.

Cage Aquaculture, Third Edition Malcolm C. M. Beveridge Copyright © 1996, 2004 by Blackwell Publishing Ltd

References Aarsnes, J. V., Rudi, H. & Loland, G. (1990) Current forces on cage, net deﬂection. In: Engineering for Offshore Fish Farming, pp. 137–52. Thomas Telford, London. Ablett, R. F., Marr, C. R. & Roberts, J. D. (1989) Inﬂuence of chronic subsurface retention on swimming activity of Atlantic salmon (Salmo salar) in cold temperate conditions. Aquacultural Engineering, 8, 1–13. ADCP (1983) Fish Feeds and Feeding in Developing Countries. United Nations Development Programme and Food and Agriculture Organization of the United Nations. ADCP/REP/83/18. FAO, Rome. Ahlgren, M. O. (1998) Consumption and assimilation of salmon net pen fouling debris by the red sea cucumber Parastichopus californicus: implications for polyculture. Journal of the World Aquaculture Society, 29, 133–9. Alabaster, J. S. (Ed.) (1982) Report of the EIFAC Workshop on Fish Farm Efﬂuents, Silkeborg, Denmark, 26–28 May, 1981. EIFAC Technical Paper 41. FAO, Rome. Alabaster, J. S. & Lloyd, R. (1980) Water Quality Criteria for Freshwater Fish, 2nd edition. Butterworths, London. Alanärä, A. (1992a) Demand feeding as a self-regulating feeding system for rainbow trout (Oncorhynchus mykiss) in net-pens. Aquaculture, 108, 347–56. Alanärä, A. (1992b) The effect of time-restricted feeding on feeding activity, growth and feed conversion in rainbow trout (Oncorhynchus mykiss). Aquaculture, 108, 357–68. Albright, L. J., Yang, C. Z. & Johnson, S. (1993) Sub-lethal concentrations of the harmful diatoms Chaetoceros concavicornis and C. convolutus increase mortality rates of penned Paciﬁc salmon. Aquaculture, 117, 215–25. Alderman, D. J., Rosenthal, H., Smith, P., Stewart, J. & Weston, D. P. (1994) Chemicals Used in Mariculture. ICES Cooperative Research Report, No. 202. ICES, Copenhagen. Aldrin, J. F., Messager, J. L. & Mevel, M. (1979) Essai sur le stress de transport chez le saumon coho juvenile (Oncorhynchus kisutch). Aquaculture, 17, 279–89. Alferez, V. N. (1982) Fishpen design and construction. In: Report of the Training Course on Small-Scale Pen and Cage Culture for Finﬁsh, pp. 23–51. SCS/GEN/ 82/34. South China Sea Fisheries Development and Co-ordinating Programme, Manila. Amend, D. F., Croy, T. R., Goven, B. A., Johnson, K. A. & McCarthy, D. H. (1982) Transportation of ﬁsh in closed systems: methods to control ammonia, carbon dioxide, pH and bacterial growth. Transactions of the American Fisheries Society, 111, 603–11. Andersen, R. J., Luu, H. A., Chen, D. Z. X., Holmes, C. F. B., Kent, M. L., Le Blanc, F., Taylor, F. J. R. & Williams, D. E. (1993) Chemical and biological evidence links microcystins to salmon netpen liver disease. Toxicon, 31, 1315–23. Anderson, D. P. (1990) Immunological indicators: effects of environmental stress on immune protection and disease outbreaks. American Fisheries Society Symposium, 8, 38–50. Andrew, J. E., Noble, C., Kadri, S. A., Jewell, H. & Huntingford, F. A. (2002) The effect of demand feeding on swimming speed and feeding responses in Atlantic salmon Salmo 308

References

309

salar L., gilthead sea bream Sparus aurata L. and European sea bass Dicentrarchus labrax L. in sea cages. Aquaculture Research, 33, 501–507. Andrew, J. W., Murai, T. & Campbell, C. (1973) Effects of dietary calcium and phosphorus on growth, food conversion and hematocrit levels in catﬁsh. Journal of Nutrition, 103, 766–71. Andrews, J. Y. & Matsuda, Y. (1975) Inﬂuence of various culture conditions on the oxygen consumption of channel catﬁsh. Transactions of the American Fisheries Society, 104, 322–7. Ang, K. P. & Petrell, R. J. (1997) Control of feed dispensation in seacages using underwater video monitoring: effects on growth and food conversion. Aquacultural Engineering, 16, 45–62. Angel, D., Krost, P., Zuber, D., Mozes, N. & Neori, A. (1992) The turnover of organic matter in hypertrophic sediments below a ﬂoating ﬁsh farm in the oligotrophic Gulf of Eilat (Aqaba). Bamidgeh, 44, 143–4. Anh, B. T. & Son, N. H. (2001) Cage rearing of fry to ﬁngerlings of carp species in large reservoirs in northern Vietnam. In: Reservoir and Culture-Based Fisheries: Biology and Management. Proceedings on an International Workshop, Bangkok, Thailand, 15–18 February 2000. ACIAR Proceedings No. 98 (Ed. by S. S. De Silva), pp. 347–58. ACIAR, Canberra. Anon (1982) Big freeze strikes chill in salmon producers. Fish Farming International, 9 (1), 15. Anon (1985) Dutch company develops high energy pellets. Fish Farming International, 12 (3), 21. Anon (1988) Net cages increase farm yields. Fish Farming International, 15 (9), 2. Anon (1992) A ﬁrst for yellowﬁn spawning. INFOFISH International, 5/92, 37. Anon (1999) Snapper behind bars. Fisheries NSW, Autumn 1999, 13. Anras, M.-L. B., Beauchaud, M., Juell, J.-E., Covès, D. & Lagardère, J.-P. (2001) Environmental factors and feed intake: rearing systems. In: Food Intake in Fish (Ed. by D. Houlihan, T. Boujard & M. Jobling), pp. 157–88. Blackwell, Oxford. Ansuini, F. J. & Huguenin, J. E. (1978) The design and development of a foulingresistant marine cage system. Proceedings of the Annual Meeting of the World Mariculture Society, 9, 737–45. Aoki, T. (1992) Chemotherapy and drug resistance in ﬁsh farms in Japan. In: Diseases in Asian Aquaculture (Ed. by M. Shariff, R. P. Subasinghe & J. R. Arthur), pp. 519–29. Asian Fisheries Society, Manila. APHA (1989) Standard Methods for the Examination of Water and Wastewater, 17th edition. American Public Health Association, Washington, DC. Aquino, L. V. & Nielsen, B. H. (1983) Primary production and growth rate of Tilapia nilotica in cages in Sampaloc lake, Philippines. In: Proceedings of the International Symposium on Tilapia in Aquaculture (Ed. by L. Fishelson & Z. Yaron), pp. 455–63. Hebrew University of Jerusalem, Jerusalem. Aragon, C. T., de Lim, M. M. & Tioseco, R. L. (1985) Economics of cage culture in Laguna Province. In: Proceedings of the PCARRD/ICLARM Tilapia Economics Workshop, U.P. Los Banos, Laguna, Philippines, 10–13 August 1983 (Ed. by I. R. Smith, E. B. Torres & E. O. Tan), pp. 66–82. ICLARM, Manila. Arai, S., Nose, T. & Kawatsu, H. (1975) Effects of minerals supplemented to the ﬁshmeal diet on growth of eel Anguilla japonica. Bulletin of the Freshwater Fisheries Research Laboratory, Tokyo, 24, 95–9.

310

References

Ariyaratne, M. H. S. (2001) Performance of cage reared ﬁngerlings of commonly cultured ﬁsh species in response to different feeds. In: Reservoir and Culture-Based Fisheries: Biology and Management. Proceedings on an International Workshop, Bangkok, Thailand, 15–18 February 2000. ACIAR Proceedings No. 98 (Ed. by S. S. De Silva), pp. 359–66. ACIAR, Canberra. Arizono, M. & Suizo, H. (1977) Environmental deterioration in yellowtail farms. Bulletin of the Yamaguchi Prefecture Gakai Fisheries Experimental Station, 15, 43–56. Arzul, G., Segues, M. & Clément, A. (2001) Effects of marine animal excretions on differential growth of phytoplankton species. ICES Journal of Marine Science, 58, 386–90. Aune, T., Skulberg O. M. & Underdal, B. (1992) A toxic phytoﬂagellate bloom of Chrysochromulina cf. leadbeateri in coastal waters in the North of Norway, May–June 1991. Ambio, 21, 471–4. Aure, J. & Stigebrandt, A. (1990) Quantitative estimates of the eutrophication effects of ﬁsh farming on fjords. Aquaculture, 90, 135–56. Austin, B. (1993) Environmental issues in the control of bacterial diseases. In: Environment and Aquaculture in Developing Countries. ICLARM Conference Proceedings 31 (Ed. by R. S. V. Pullin, H. Rosenthal & J. L. MacLean), pp. 237–51. ICLARM, Manila. Avault, J., Jr (1981) Feeding methods affect production and proﬁt. Aquaculture Magazine, 7 (4), 38–59. Backiel, T., Kokurewicz, B. & Ogorzalek, A. (1984) High incidence of skeletal anomalies in carp, Cyprinus carpio, reared in cages in ﬂowing water. Aquaculture, 43, 369–80. Bagarinao, T. & Lantin-Olaguer, I. (1998) The sulﬁde tolerance of milkﬁsh and tilapia in relation to ﬁsh kills in farms and natural waters in the Philippines. Hydrobiologia, 382, 137–50. Baird, D. J., Beveridge, M. C. M., Kelly, L. A. & Muir, J. F. (Eds) (1996) Aquaculture and Water Resources Management. Blackwell, Oxford. Balarin, J. D. & Haller, R. D. (1982) The intensive culture of tilapia in tanks, raceways and cages. In: Recent Advances in Aquaculture, Vol. 1 (Ed. by J. F. Muir & R. J. Roberts), pp. 265–356. Croom Helm, London. Baltz, D. M. (1991) Introduced ﬁshes in marine systems and inland seas. Biological Conservation, 56, 151–77. Bandeen, J. & Leatherland, J. F. (1997) Transportation and handling stress of white suckers raised in cages. Aquaculture International, 5, 385–96. Bangen, M., Grave, K., Nordmo, R. & Søli, N. E. (1994) Description and evaluation of a new surveillance programme for drug use in ﬁsh farming in Norway. Aquaculture, 119, 109–18. Bardach, J. E., Ryther, J. H. & McLarney, W. O. (1972) Aquaculture. The Farming and Husbandry of Freshwater and Marine Organisms. John Wiley, New York. Barg, U. (1992) Guidelines for the Promotion of Environmental Management of Coastal Aquaculture Development. FAO Fisheries Technical Paper 328. FAO, Rome. Barker, C. J. (1990) Classiﬁcation society rules for ﬁsh farms. In: Engineering for Offshore Fish Farming, pp. 31–8. Thomas Telford, London. Barnaby, R. & Adams, S. (2002) Aquaculture: opportunity or threat to traditional capture ﬁshermen? In: Responsible Marine Aquaculture (Ed. by R. R. Stickney & J. P. McVey), pp. 71–9. CAB International, Cambridge. Barret, E. C. (1974) Climatology from Satellites. Methuen, London. Bartley, D. & Casal, C. V. (1998) Impacts of introductions on the conservation and sustainable use of aquatic biodiversity. FAO Aquaculture Newsletter, 20, 15–19.

References

311

Barton, B. A. (2000) Salmonid ﬁshes differ in their cortisol and glucose reponses to handling and transportation stress. North American Journal of Aquaculture, 62, 12–18. Bascom, W. (1964) Waves and Beaches. Anchor Books, New York. Bautista, A. M. (1984) Tilapia Cage Farming in Lakes. Aquaculture Technology Module No. 1. SEAFDEC, Tigbauan, Philippines. Bayne, D. R., Rai, A. K., Joshi, P. L. & Williams, J. C. (1992). Limnological factors inﬂuencing growth of cage-cultured bighead carp (female) ¥ silver carp (male) hybrids. Journal of Applied Aquaculture, 1 (4), 29–50. Beamish, F. W. H. (1964) Respiration of ﬁshes with special emphasis on standard oxygen consumption. II. Inﬂuence of weight and temperature on respiration of several species. Canadian Journal of Zoology, 42, 177–94. Beaz Paleo, J. D., Muir, J. & Turner, R. (2000) Offshore mariculture: workboats. In: Mediterranean Offshore Mariculture (Ed. by J. F. Muir & B. Basurco), pp. 179–90. Options Méditerranéennes, Series B, No. 30. CIHEAM, Zaragoza. Beddow, T. A., Ross, L. G. & Marchant, J. A. (1996) Predicting salmon biomass using a digital stereo-imaging technique. Aquaculture, 146, 189–203. Beem, M. D., Gebhart, G. E. & Maughan, O. E. (1988) Winter polyculture of channel catﬁsh and rainbow trout in cages. The Progressive Fish-Culturist, 50, 49–51. Behmer, D. J., Greil, D. W., Greil, D. C. & Fessell, B. P. (1993) Evaluation of cone-bottom cages for removal of solid wastes and phosphorus from pen-cultured rainbow trout. The Progressive Fish-Culturist, 55, 255–60. Behrendt, A. (1984) Eastern Europe solves some problems. Fish Farmer, 7 (3), 28–9. Bell, A. K. & Barr, A. G. (1990) Computerised mathematical modelling in the assessment of offshore ﬁsh farms. In: Engineering for Offshore Fish Farming, pp. 49–61. Thomas Telford, London. Bell, J. D., Hart, A. M., Foyle, T. P., Gervis, M. & Lane, I. (1997) Can aquaculture help restore and sustain production of giant clams? In: Developing and Sustaining World Fisheries Resources. The State of Science and Management (Ed. by D. A. Hancock, D. C. Smith, A. Grant & J.-P. Beumer), pp. 509–13. CSIRO, Collingwood, Australia. Ben Yami, M. (1974) Gnawing at ﬁsh netting – a problem in cage raising of herbivorous ﬁsh. Aquaculture, 3, 199–202. Ben Yami, M. (1978) Tuna Fishing with Pole and Line. Fishing News Books, Oxford. Benfey, T. J. (2001) Use of sterile triploid Atlantic salmon (Salmo salar L.) for aquaculture in New Brunswick, Canada. ICES Journal of Marine Science, 58, 263–8. Berg, H., Michelsen, P., Folke, C., Kautsky, N. & Troell, M. (1996) Managing aquaculture for sustainability in tropical Lake Kariba. Ecological Economics, 18, 141–59. Bergheim, A., Kristiansen, R. & Kelly, L. (1993) Treatment and utilization of sludge from landbased farms for salmon. In: Techniques for Modern Aquaculture (Ed. by J.-K. Wang), pp. 486–95. American Society of Agricultural Engineers, St Joseph, MI. Berry, T. (1981) Mariculture Equipment Data File, with Special Reference to Raft and Cage Design. National Board for Science and Technology, Dublin. Bevan, D. J. & Kramer, D. L. (1986) The effect of swimming depth on respiratory behaviour of the honey gourami Colisa chuna (Pisces, Belentiidae). Canadian Journal of Zoology, 64, 1893–6. Beveridge, M. C. M. (1984a) Cage and Pen Fish Farming. Carrying Capacity Models and Environmental Impact. FAO Fisheries Technical Paper 255. FAO, Rome. Beveridge, M. C. M. (1984b) Lake-based or land-based tilapia hatcheries? ICLARM Newsletter, 7 (1), 10–11.

312

References

Beveridge, M. C. M. (1988) Problems caused by birds at inland and freshwater ﬁsh farms. Literature review. In: Prevention and Control of Bird Predation in Aquaculture and Fisheries (Ed. by R. Welcomme), pp. 34–73. EIFAC Technical Paper 52. FAO, Rome. Beveridge, M. C. M. (2001) Aquaculture and wildlife interactions. In: Environmental Impact Assessment of Mediterranean Aquaculture Farms. Proceedings of the Seminar of the CIHEAM Network TECAM, Zaragoza, Spain, 17–21 January 2000 (Ed. by A. Uriarte & B. Basurco), pp. 57–66. Options Méditerranéennes, Series Cahiers. Zaragoza: Vol. 55. CIHEAM/FAO, Rome. Beveridge, M. C. M. (2002) Overview of cage aquaculture. In: Diseases and Disorders of Finﬁsh in Cage Culture (Ed. by P. T. K. Woo, D. W. Bruno & L. H. S. Lim), pp. 41–60. CABI Publishing, Wallingford. Beveridge, M. C. M. & Baird, D. J. (2000) Diet, feeding and digestive physiology. In: Tilapias: Biology and Exploitation (Ed. by M. C. M. Beveridge & B. J. McAndrew), pp. 59–87. Kluwer, Dordrecht. Beveridge, M. C. M. & Haylor, G. (1998) Warm-water farmed species. In: Biology of Farmed Fish (Ed. by K. D. Black & A. D. Pickering), pp. 389–412. Shefﬁeld Academic Press, Shefﬁeld. Beveridge, M. C. M. & Kadri, S. (2000) Reducing feed losses in aquaculture. International Aquafeed, 1, 27–9. Beveridge, M. C. M. & Little, D. C. (2002) Aquaculture in traditional societies. In: Ecological Aquaculture (Ed. by B. A. Costa-Pierce), pp. 3–29. Blackwell, Oxford. Beveridge, M. C. M. & Muir, J. F. (1982) An evaluation of proposed cage culture in Loch Lomond, an important reservoir in central Scotland. Canadian Water Resources Journal, 7, 107–13. Beveridge, M. C. M. & Muir, J. F. (1999) Environmental impacts and sustainability of cage culture in Southeast Asian lakes and reservoirs. In: Fish and Fisheries of Lakes and Reservoirs in Southeast Asia and Africa (Ed. by W. L. T. van Densen & M. J. Morris), pp. 383–96. Westbury Publishing, Otley, UK. Beveridge, M. C. M. & Phillips, M. J. (1988) Aquaculture in reservoirs. In: Reservoir Fishery Management and Development in Asia (Ed. by S. S. De Silva), pp. 234–43. IDRC, Ottawa. Beveridge, M. C. M. & Phillips, M. J. (1993) Environmental impact of tropical inland aquaculture. In: Environment and Aquaculture in Developing Countries. ICLARM Conference Proceedings 31 (Ed. by R. S. V. Pullin, H. Rosenthal & J. L. Maclean), pp. 213–36. ICLARM, Manila. Beveridge, M. C. M. & Stewart, J. A. (1998) Cage culture: limitations in lakes and reservoirs. In: Inland Fishery Enhancements (Ed. by T. Petr), pp. 263–79. FAO Fisheries Technical Paper 374. FAO, Rome. Beveridge, M. C. M., Phillips, M. J. & Clarke, R. M. (1991) A quantitative and qualitative assessment of wastes from aquatic animal production. In: Aquaculture and Water Quality. Advances in World Aquaculture, Vol. 3 (Ed. by D. E. Brune & J. R. Tomasso), pp. 506–27. World Aquaculture Society, Baton Rouge, LA. Beveridge, M. C. M., Baird, D. J., Rahmatullah, S. M., Lawton, L. A., Beattie, K. A. & Codd, G. A. (1993) Grazing rates on toxic and non-toxic cyanobacteria by Hypophthalmichthys molitrix and Oreochromis niloticus. Journal of Fish Biology, 43, 901–907. Beveridge, M. C. M., Martinez-Palacios, C. A. & Flores Nava, A. (1994a) Sistemas de crecimiento para Cichlasoma urophthalmus. In: Biologia y Cultivo de la Mojarra Lati-

References

313

noamericana (Ed. by C. A. Martinez-Palacios & L. G. Ross), pp. 135–63. Consejo Nacional de la Ciencia y Tecnologia, Mexico City. Beveridge, M. C. M., Ross, L. G. & Kelly, L. A. (1994b) Aquaculture and biodiversity. Ambio, 23, 497–502. Beveridge, M. C. M., Ross, L. G. & Mendoza, A. Q. M. (1994c) Geographic Information Systems (GIS) for coastal aquaculture site selection and planning. In: Ecology of Marine Aquaculture (Ed. by K. Koop), pp. 26–47. International Foundation for Science, Stockholm. Beveridge, M. C. M., Phillips, M. J. & Macintosh, D. C. (1997a) Aquaculture and the environment: the supply and demand for environmental goods and services by Asian aquaculture and the implications for sustainability. Aquaculture Research, 28, 101–11. Beveridge, M. C. M., Ross, L. G. & Stewart, J. A. (1997b) The development of mariculture and its implications for biodiversity. In: Marine Biodiversity: Patterns and Processes (Ed. by R. F. G. Ormond, J. Gage & M. Angel), pp. 372–93. Cambridge University Press, Cambridge. Beverton, R. J. H. & Holt, S. J. (1957) On the dynamics of exploited ﬁsh populations. Fisheries Investigations, London, Series II, 19, 1–533. Billington, N. & Herbert, P. D. N. (Eds) (1991) The ecological and genetic implications of ﬁsh introductions. Canadian Journal of Fisheries and Aquatic Sciences, 48 (Supplement 1). Bishop, C. T. (1980) Design and Construction Manual for Floating Tyre Breakwaters. Hydraulic Division, National Water Research Institute, Burlington, Ontario. Bjerke, K. S. (1990) Offshore ﬁsh-farming platforms – development, design, construction and operation: the SEACON and SULAN concepts. In: Engineering for Offshore Fish Farming, pp. 153–70. Thomas Telford, London. Bjordal, Å., Juell J. E., Lindem, T. & Fernö, A. (1993) Hydroacoustic monitoring and feeding control in cage rearing of Atlantic salmon (Salmo salar L.). In: Fish Farming Technology (Ed. by H. Reinertsen, L. A. Dahle, L. Jørgensen & K. Tvinnereim), pp. 203–208. A. A. Balkema, Rotterdam. Bjørndal, T. (1990) The Economics of Salmon Aquaculture. Blackwell, Oxford. Bjørndal, T. & Uhler, R. S. (1993) Salmon sea farm management: basic economic concepts and applications. In: Salmon Aquaculture (Ed. by K. Heen, R. L. Monaghan & F. Utter), pp. 239–54. Fishing News Books, Oxford. Black, E. A. & Truscott, J. (1994) Strategies for regulation of aquaculture site selection in coastal areas. Journal of Applied Ichthyology, 10, 295–306. Black, E. A., Gillis, D. J., Hay, D. E., Haegele, C. W. & Levings, C. D. (1992a) Predation by caged salmon in British Columbia. Bulletin of the Aquaculture Association of Canada, 92–3, 58–60. Black, E. A., Redfern, B. & Rosenthal, H. (1992b) New developments in air transport of salmon smolts. Journal of Applied Ichthyology, 7, 249–52. Black, K. D. (1998) The environmental interactions associated with ﬁsh culture. In: Biology of Farmed Fish (Ed. by K. D. Black & A. D. Pickering), pp. 284–326. Shefﬁeld Academic Press, Shefﬁeld. Black, K. D., Ezzi, I. A., Kiemer, M. C. B. & Wallace, A. J. (1994) Preliminary evaluation of the effects of long-term periodic sublethal exposure to hydrogen sulphide on the health of Atlantic salmon (Salmo salar L.). Journal of Applied Ichthyology, 10, 362–7.

314

References

Blair, A., Campbell, R. & Grant, P. T. (1982) A submersible ﬁsh cage that can be rotated on the surface to remove biofouling and for other purposes. Aquaculture, 29, 177–84. Blakstad, F. (1993) The economic aspects of land based versus ocean based Atlantic salmon farming. In: From Discovery to Commercialization (Ed. by M. Carrillo, L. A. Dahle, J. Morales, P. Sorgeloos, N. Svennevig & J. Wyban). Published abstracts from World Aquaculture ’93. European Aquaculture Society Special Publication, 19, 305. Blyth, P. J. (1992) Boosting proﬁts with adaptive feeding: letting the ﬁsh decide when they’re hungry. AustAsia Aquaculture, 6 (5), 33–8. Blyth, P. J., Purser, G. J. & Russell, J. F. (1993) Detection of feeding rhythms in seacaged Atlantic salmon using new technology. In: Fish Farming Technology (Ed. by H. Reinertsen, L. A. Dahle, L. Jørgensen & K. Tvinnereim), pp. 209–16. A. A. Balkema, Rotterdam. Boeuf, G. & Harache, Y. (1980) Present status of salmonid cage rearing in western France and prospects for development. In: Proceedings of the Institute of Fisheries Management Cage Rearing Symposium, University of Reading, 26–27 March 1980, pp. 50–57. Janssen Services, London. Boghen, A. D. (Ed.) (1995) Cold-Water Aquaculture in Atlantic Canada, 2nd edition. Tribune Press, Sackville, New Brunswick. van Boven, C. J. P. (1968) Design considerations for permanent-type offshore structures. In: Offshore Europe 1968, pp. 51–5. Scientiﬁc Surveys (Offshore), London. Boxaspen, K. & Holm, J. C. (2001) The development of pyrethrum-based treatments against the ectoparasitic salmon lice Lepeophtheirus salmonis in sea cage rearing of Atlantic salmon Salmo salar L. Aquaculture Research, 32, 701–7. Boyd, C. E. (1979) Water Quality in Warm Water Fishponds. Auburn University Agricultural Experimental Station, Auburn, AL. Boyd, C. E. (1980) Reliability of water analysis kits. Transactions of the American Fisheries Society, 109, 239–43. Boyd, C. E. & Tucker, C. S. (1998) Pond Aquaculture Water Quality Management. Kluwer, Dordrecht. Boyd, C. E. & Watten, B. J. (1989) Aeration systems in aquaculture. Reviews in Aquatic Science, 1, 425–72. Boydston, H. & Patterson, T. (1982) The pros and cons on the use of self-feeders. Salmonid, Jan–Feb, 14–16. Boydstun, L. B. & Hopelain, J. S. (1977) Cage rearing of steelhead rainbow trout in a freshwater impoundment. The Progressive Fish-Culturist, 39, 70–75. Braaten, B. R. & Saetre, R. (1973) Rearing of Salmonids in Norwegian Coastal Waters – Environment and Types of Installations. Translation Series of the Fisheries Research Board of Canada 2759. Fisheries Research Board of Canada, Ottawa. Braaten, B., Ervik, A. & Boje, E. (1983) Pollution problems on Norwegian ﬁsh farms. Aquaculture Ireland, 14, 6–10. Brandal, P. O. & Egidius, E. (1979) Treatment of salmon lice (Lepeophtheirus salmonis Kroyer) with Nevugon – description of method and equipment. Aquaculture, 18, 183–8. Brett, J. R. & Zala, C. A. (1975) Daily patterns of nitrogen excretion and oxygen consumption of sockeye salmon (Oncorhynchus nerka) under controlled conditions. Journal of the Fisheries Research Board of Canada, 32, 2479–86. Bridger, C. J., Booth, R. K., McKinley, R. S. & Scruton, D. A. (2001) Site ﬁdelity and dispersion patterns of domestic triploid steelhead trout (Oncorhynchus mykiss) released in the wild. ICES Journal of Marine Science, 58, 510–16.

References

315

Brittain, D. (1996) Dunlop ‘Tempest’ ﬁsh cages. In: Opean Ocean Aquaculture. Proceedings of an International Conference, Portland, Maine, 8–10 May 1996 (Ed. by M. Polk), pp. 255–68. Report No. UNHMP-CP-SG-96-9. New Hampshire/Maine Sea Grant College Program, Portland, ME. Bron, J. E., Sommerville, C., Wootten, R. & Rae, G. H. (1993) Fallowing of marine Atlantic salmon, Salmo salar L., farms as a method for the control of sea lice, Lepeophtheirus salmonis (Kroyer, 1837). Journal of Fish Diseases, 16, 487–93. Bronisz, D. (1979) Selective exploitation of lake zooplankton by coregonid fry in cage culture. Special Publication of the European Mariculture Society, 4, 301–7. Bronzi, P. & Ghittino, P. (1981) Floating cages and/or raceways: two different systems of utilising warm waters discharged by thermal power stations in inland ﬁsh culture. Schriften der Bundesforschungsanstalt für Fischerei, 16/17, 289–300. Brooks, K. M., Mahnken, C. & Nash, C. (2002) Environmental effects associated with marine netpen waste with emphasis on salmon farming in the Paciﬁc Northwest. In: Responsible Marine Aquaculture (Ed. by R. R. Stickney & J. P. McVey), pp. 149–203. CAB International, Cambridge. Brooks, K. M., Stierns, A. R., Mahnken, C. V. W. & Blackburn, D. B. (2003) Chemical and biological remediation of the benthos near Atlantic salmon farms. Aquaculture, 219, 355–77. Brown, E. E. (1983) World Fish Farming: Cultivation and Economics, 2nd edition. AVI Publishing, Westport, CT. Brown, J., Helm, M. & Moir, J. (1995) New-candidate species for aquaculture. In: ColdWater Aquaculture in Atlantic Canada, 2nd edition (Ed. by A. D. Boghen), pp. 341–60. Tribune Press, Sackville, New Brunswick. Brown, J. R., Gowen, R. J. & McLusky, D. S. (1987) The effect of salmon farming on the benthos of a Scottish sea loch. Journal of Experimental Marine Biology and Ecology, 109, 39–51. Brugere, C., McAndrew, K. & Bulcock, P. (2001) Does cage aquaculture address gender goals in development? Results of a case study in Bangladesh. Aquaculture Economics and Management, 5, 179–90. Bruno, D. W. & Stone, J. (1990) The role of saithe Pollachius virens L. as a host for the sea lice Lepeophtheirus salmonis Kroyer and Caligus elongatus Nordmann. Aquaculture, 89, 201–8. Bruno, D. W., Dear, G. & Seaton, D. D. (1989) Mortality associated with phytoplankton blooms among farmed Atlantic salmon, Salmo salar L., in Scotland. Aquaculture, 78, 217–22. Brylinsky, M. (1980) Estimating the productivity of lakes and reservoirs. In: The Functioning of Freshwater Ecosystems (Ed. by E. D. Le Cren & R. H. Lowe-McConnell), pp. 411–54. Cambridge University Press, Cambridge. Brzeski, V. & Newkirk, G. (1997) Integrated coastal food production systems – a review of current literature. Ocean and Coastal Management, 34, 55–71. Budhabhatti, J. & Maughan, O. E. (1993) Production of channel catﬁsh Ictalurus punctatus in ﬂowing agricultural waters in Arizona. Journal of the World Aquaculture Society, 24, 493–8. Budhabhatti, J. & Maughan, O. E. (1994) Cage designs for ﬁsh culture in irrigation waters. The Progressive Fish-Culturist, 56, 147–8. Bugrov, L. Y. (1996) Underwater ﬁsh-farming technology for open sea areas: review of a 10-year experience. In: Open Ocean Aquaculture. Proceedings of an International Conference, Portland, Maine, 8–10 May 1996 (Ed. by M. Polk), pp. 269–96. Report No.

316

References

UNHMP-CP-SG-96-9. New Hampshire/Maine Sea Grant College Program, Portland, ME. Bullock, A. M. (1988) Solar ultraviolet radiation: a potential environmental hazard in the cultivation of farmed ﬁnﬁsh. In: Recent Advances in Aquaculture, Vol. 3 (Ed. by J. F. Muir & R. J. Roberts), pp. 139–224. Croom Helm, London. Buras, N. (1993) Microbial safety of produce from wastewater-fed aquaculture. In: Environment and Aquaculture in Developing Countries. ICLARM Conference Proceedings 31 (Ed. by R. S. V. Pullin, H. Rosenthal & J. L. Maclean), pp. 285–95. ICLARM, Manila. Burroughs, P. A. (1986) Principles of Geographic Information Systems for Land Resources Assessment. Clarendon Press, Oxford. Butler, M. J. A. (1982) Plight of the blue ﬁn tuna. National Geographic Magazine, 162, 220–29. Cabello, L. (2000) Production methods for offshore ﬁsh management. In: Mediterranean Offshore Mariculture (Ed. by J. F. Muir & B. Basurco), pp. 191–202. Options Méditerranéennes, Series B, No. 30. CIHEAM, Zaragoza. Cairns, J. & Linfoot, B. T. (1990) Some considerations in the structural engineering of sea-cages for aquaculture. In: Engineering for Offshore Fish Farming, pp. 63–77. Thomas Telford, London. Canﬁeld, D. E. & Bachmann, R. W. (1981) Prediction of total phosphorus concentrations, chlorophyll a and Secchi depths in natural and artiﬁcial lakes. Canadian Journal of Fisheries and Aquatic Sciences, 38, 414–23. Carmichael, W. W. (1994) The toxins of cyanobacteria. Scientiﬁc American, 270, 78–86. Carss, D. N. (1990) Concentrations of wild and escaped ﬁshes immediately adjacent to ﬁsh farm cages. Aquaculture, 90, 29–40. Carss, D. N. (1993a) Cormorants Phalacrocorax carbo predation at cage ﬁsh farms in Argyll, western Scotland. Seabird, 15, 19–25. Carss, D. N. (1993b) Grey heron, Ardea cinerea L., predation at cage ﬁsh farms in Argyll, western Scotland. Aquaculture and Fisheries Management, 24, 29–45. Carss, D. N. (1994) Killing of piscivorous birds at Scottish ﬁn ﬁsh farms, 1984–87. Biological Conservation, 68, 181–8. Chacon Torres, A., Ross, L. G. & Beveridge, M. C. M. (1988a) The effects of ﬁsh behaviour on dye dispersion and water exchange in small net cages. Aquaculture, 73, 283–93. Chacon Torres, A., Ross, L. G. & Beveridge, M. C. M. (1988b) The use of remote sensing in water quality investigations for aquaculture. In: Aquaculture Engineering Technologies for the Future. Institute of Chemical Engineers Symposium Series, Vol. 111, pp. 21–44. Institute of Chemical Engineers, Rugby. Chamberlain, G. W. (1978) The Use of Caged Fish for Mariculture and Environmental Monitoring in a Power Plant Cooling Water System. MSc thesis, Texas A & M University, College Station, TX. Chamberlain, G. W. (1993) Aquaculture trends and feed projections. World Aquaculture, 24 (1), 19–29. Chamberlain, G. W. & Strawn, K. (1977) Submerged cage culture of ﬁsh in supersaturated thermal efﬂuent. Proceedings of the Annual Meeting of the World Mariculture Society, 8, 625–45. Chang, W. B. (1989) Integrated lake farming for ﬁsh and environmental management in large, shallow Chinese lakes: a review. Aquaculture and Fisheries Management, 20, 441–52.

References

317

Chapra, S. C. (1975) Comment on ‘An empirical method of estimating the retention of phosphorus in lakes’ by W. B. Kirchner and P. J. Dillon. Water Resources Research, 11, 1033–4. Cheah, S. H. & Chua, T. E. (1979) A preliminary study of the tropical marine fouling organisms on ﬂoating net cages. Malaysian Nature Journal, 33, 39–48. Chen, F.-Y. (1979) Progress and problems of netcage culture of grouper (Epinephelus tauvina) in Singapore. Proceedings of the Annual Meeting of the World Mariculture Society, 10, 260–71. Chen, Y.-S., Beveridge, M. C. M. & Telfer, T. C. (1999a) Physical characteristics of commercial pelleted Atlantic salmon feeds and consideration of implications for modeling of waste dispersion through sedimentation. Aquaculture International, 7, 89–100. Chen, Y.-S., Beveridge, M. C. M. & Telfer, T. C. (1999b) Settling rate characteristics and nutrient content of the faeces of Atlantic salmon Salmo salar L., feeds and the implications for modeling of solid waste dispersion. Aquaculture Research, 33, 395–8. Chen, Y.-S., Beveridge, M. C. M. & Telfer, T. C. (2000) Derived data for more effective modeling of solid wastes dispersion from sea cage farms. In: Cage Aquaculture in Asia. Proceedings of the First International Symposium on Cage Aquaculture in Asia, Turgkang, Taiwan, 2–6 November 1999 (Ed. by I. C. Liao & C. K. Lin), pp. 157–66. Asian Fisheries Society, Manila and World Aquaculture Society – Asia Branch, Bangkok. Chen, Y.-S., Beveridge, M. C. M., Telfer, T. C. & Roy, W. J. (2003) Nutrient leaching and settling rate characteristics of the faeces of Atlantic salmon (Salmo salar L.) and the implications for modelling of solid waste dispersion. Journal of Applied Ichthyology, 19, 1–4. Chettleburgh, P. (1991) Batten down the farm for stormy weather. Canadian Aquaculture, 8 (2), 64–5. Chong, C. V. Y. (1977) Properties of Materials. McDonald & Evans, Plymouth. Chong, K.-C. (1993) Economics of on-farm aquafeed preparation and use. In: Farm-Made Aquafeeds. Proceedings of the FAO/AADCP Regional Expert Consultation on Farm-Made Aquafeeds, 14–18 December 1992, Bangkok, Thailand (Ed. by M. B. New, A. G. J. Tacon & I. Csavas), pp. 24–60. FAO-RAPA/AADCP, Bangkok. Chopin, T., Yarish, C., Wilkes, R., Belyea, E., Lu, S. & Mathieson, A. (1999) Developing Porphyra/salmon integrated aquaculture for bioremediation and diversiﬁcation of the aquaculture industry. Journal of Applied Phycology, 11, 463–72. Chopin, T., Buschmann, A. H., Halling, C., Troell, M., Kautsky, N., Neori, A., Kraemer, G. P., Zertuche-González, Yarish, C. & Neefus, C. (2001) Integrating seaweeds into marine aquaculture systems: a key toward sustainability. Journal of Phycology, 37, 975–86. Chow, V. T. (Ed.) (1959) Handbook of Applied Hydrology. McGraw Hill, New York. Christensen, I. (2000) Nets for offshore aquaculture. In: Mediterranean Offshore Mariculture (Ed. by J. F. Muir & B. Basurco), pp. 173–8. Options Méditerranéennes, Series B, No. 30. CIHEAM, Zaragoza. Christensen, M. S. (1989) Techniques and Economics of Intensive Cultivation of Jelawat and Lempam Carp in Floating Cages. A Handbook for Extension Workers and Farmers. GTZ, Eschborn, Germany. Chua, T.-E. (1979) Site selection, structural design, construction, management and production of ﬂoating cage culture systems in Malaysia. In: Proceedings of the IDRC/SEAFDEC International Workshop on Pen and Cage Culture of Fish, Tigbauan, Iloilo, Philippines, 11–22 February 1979, pp. 65–80. SEAFDEC, Iloilo, Philippines.

318

References

Chua, T.-E. & Tech, E. (2002) Introduction and history of cage culture. In: Diseases and Disorders of Finﬁsh in Cage Culture (Ed. by P. T. K. Woo, D. W. Bruno & L. H. S. Lim), pp. 1–39. CABI Publishing, Wallingford. Chua, T.-E. & Teng, S. K. (1977) Floating ﬁsh pens for rearing ﬁshes in coastal waters, reservoirs and mining pools in Malaysia. Fisheries Bulletin of the Ministry of Agriculture, Malayasia, No. 20. Ministry of Agriculture, Melaka, Malaysia. Chua, T.-E. & Teng, S. K. (1978) Effects of feeding frequency on the growth of young estuary grouper Epinephelus salmoides Maxwell, cultured in ﬂoating net cages. Aquaculture, 14, 31–47. Chua, T.-E. & Teng, S. K. (1980) Economic production of estuary grouper Epinephelus salmoides Maxwell, reared in ﬂoating cages. Aquaculture, 20, 187–228. Ciesluk, A. F. (1974) Dissolved Gas Supersaturation Associated with the Thermal Efﬂuent of an Electric Generating Station and Some Effects on Fish. MSc thesis, Texas A & M University, College Station, TX. Clark, D. S., Brown, J. A., Goddard, S. J. & Moir, J. (1995) Activity and feeding behaviour of Atlantic cod (Gadus morhua) in sea pens. Aquaculture, 131, 49–57. des Clers, S. (1994) Sampling to Detect Infections and Estimate Prevalence in Aquaculture. Pisces Press, Stirling. Cobham Resource Consultants (1987) An Environmental Assessment of Fish Farms. Final Report to the Countryside Commission for Scotland, Crown Estate Commissioners, Highlands and Islands Development Board and the Scottish Salmon Growers Association. Coche, A. G. (1979) A review of cage culture and its application in Africa. In: Advances in Aquaculture (Ed. by T. V. R. Pillay & W. A. Dill), pp. 428–41. Fishing News Books, Oxford. Coche, A. G. (1982) Cage culture of tilapias. In: Biology and Culture of Tilapias (Ed. by R. S. V. Pullin & R. H. Lowe-McConnell), pp. 205–46. ICLARM, Manila. Coche, A. G. (1983) The Cultivation of Fish in Cages. A General Bibliography. FAO Fish Circular, 714 (Revision 1). FAO, Rome. Collins, R. A. & Delmendo, M. N. (1979) Comparative economics of aquaculture in cages, raceways and enclosures. In: Advances in Aquaculture (Ed. by T. V. R. Pillay & W. A. Dill), pp. 472–7. Fishing News Books, Oxford. Colt, J. E. & Orwicz, C. (1991) Aeration in intensive culture. In: Aquaculture and Water Quality Advances in World Aquaculture, Vol. 3 (Ed. by D. E. Brune & J. R. Tomasso), pp. 198–271. World Aquaculture Society, Baton Rouge, LA. Colt, J. E. & Tchobanoglous, G. (1981) Design of aeration systems for aquaculture. In: Proceedings of the Bioengineering Symposium on Fish Culture, Traverse City, Michigan, 16–18 October 1979 (Ed. by L. J. Allen & E. C. Kinney), pp. 138–48. Fish Culture Section of the American Fisheries Society, Bethesda, MA. Colt, J. E. & Watten, B. J. (1988) Applications of pure oxygen in ﬁsh culture. Aquacultural Engineering, 7, 397–441. Congleton, J. L., LaVoie, W. J., Schreck, C. B. & Davis, L. E. (2000) Stress indices in migrating juvenile chinook salmon and steelhead of wild and hatchery origin before and after barge transportation. Transactions of the American Fisheries Society, 129, 946–61. Connell, J. J. (1995) Control of Fish Quality, 4th edition. Fishing News Books, Oxford. Cook, J.-M., Cristal, G. & Lamare, J.-P. (1984) Technologie des structures d’élevage en mer (cages ﬂottantes et immergées). In: L’Aquaculture du Bar et des Sparides (Ed. by G. Barnabe & R. Billard), pp. 395–402. INRA, Paris.

References

319

Cornel, G. E. & Whoriskey, F. G. (1993) The effects of rainbow trout (Oncorhynchus mykiss) cage culture on the water quality, zooplankton, benthos and sediments of Lac du Passage, Quebec. Aquaculture, 109, 101–17. Costa-Pierce, B. A. (1996) Environmental impacts of nutrients discharged from aquaculture: towards the evolution of sustainable, ecological aquaculture systems. In: Aquaculture and Water Resources Management (Ed. by D. J. Baird, M. C. M. Beveridge, L. A. Kelly & J. F. Muir), pp. 81–113. Blackwell, Oxford. Costa-Pierce, B. A. (2002) Sustainability of cage aquaculture ecosystems for largescale resettlement from hydropower dams: an Indonesian case study. In: Ecological Aquaculture (Ed. by B. A. Costa-Pierce), pp. 286–313. Blackwell, Oxford. Costa-Pierce, B. A. & Bridger, C. J. (2002) The role of aquaculture facilities as habitats and ecosystems. In: Responsible Marine Aquaculture (Ed. by R. R. Stickney & J. P. McVey), pp. 105–44. CAB International, Cambridge. Costa-Pierce, B. A. & Effendi, P. (1988) Sewage ﬁsh cages of Kota Cianjur, Indonesia. NAGA, 11 (2), 7–9. Costa-Pierce, B. A. & Soemarwoto, O. (Eds) (1990) Reservoir Fisheries and Aquaculture Development for Resettlement in Indonesia. ICLARM Technical Report 23. ICLARM, Manila. Costa-Pierce, B. A., Rusydi Safari, A. & Atmadja, G. W. (1989) Culture of Common Carp in Floating Net Cages. ICLARM Education Series 7. ICLARM, Manila. Costello, M. J. (1993) Controlling sea-lice infestations on farmed salmon in Northern Europe: options and the use of cleaner ﬁsh. World Aquaculture, 24, 49–55. Costello, M. J., Grant, A., Davies, I. M., Cecchini, S., Papoutsoglou, S., Quigley, D. & Saroglia, M. (2001) The control of chemicals used in aquaculture in Europe. Journal of Applied Ichthyology, 17, 173–80. Couper, A. D. (1983) The Times Atlas of the Oceans. Times Books, London. Cowell, L. E., Watanabe, W. O., Head, W. D., Grover, J. J. & Shenker, J. M. (1993) Use of tropical cleaner ﬁsh to control the ectoparasite Neobenedenia melleni (Monogenea: Capsalidae) on seawater-cultured Florida red tilapia. Aquaculture, 113, 189–200. Coyne, R., Hiney, M., O’Connor, B., Kerry, J., Cazabon, D. & Smith, P. (1994) Concentration and persistence of oxytetracycline in sediments under a marine salmon farm. Aquaculture, 123, 31–42. Cranston, R. (1994) Dissolved ammonium and sulfate gradients in surﬁcial sediment pore water as a measure of organic carbon burial rate In: Modelling Benthic Impacts of Organic Enrichment from Marine Aquaculture (Ed. by B. T. Hargrave), pp. 93–120. Canadian Technical Report on Fisheries and Aquatic Sciences No. 1949. Department of Fisheries and Oceans, Ottawa. Cromey, C. J., Nickell, T. D. & Black, K. D. (2000) DEPOMOD – modelling the deposition and biological effects of waste solids from marine cage farms. Aquaculture, 214, 211–39. Csanady, G. T. (1975) Hydrodynamics of large lakes. Annual Review of Fluid Mechanics, 7, 357–85. Cubero, L. & Molinero, A. (1997) Handling, conﬁnement and anaesthetic exposure induces changes in the blood and tissue immune characteristics of gilthead sea bream. Diseases of Aquatic Organisms, 31, 89–94. Dahle, L. A. & Oltedal, G. (1990) Norwegian research and industrial development of ﬂoating structures for salmon ﬁsh farming. In: Engineering for Offshore Fish Farming, pp. 185–96. Thomas Telford, London.

320

References

Dahl, E. & Tangen, K. (1990) Gyrodinium aureolum bloom along the Norwegian coast in 1988. In: Proceedings of the International Conference on Toxic Marine Phytoplankton, Lund (Sweden), 26–30 June 1989 (Ed. by E. Graneli, B. Sundstroem, L. Edler & D. M. Anderson), pp. 123–37. Elsevier, New York. Dahle, L. A., Jorgensen, L., Vangen, K. & Aarsnes, J. V. (1989) An underwater offshore cage. In: Aquaculture – A Biotechnology in Progress (Ed. by N. de Pauw, E. Jaspers, H. Ackefors & N. Wilkins), pp. 1001–7. European Aquaculture Society, Bredene, Belgium. Dahm, E., Suuronen, P. & Lehtonen, E. (1990) Weathering experiments with netting material for stationary ﬁshing gear. Finnish Fisheries Research, 11, 17–24. Davies, I. M., Gillibrand, P. A., McHenery, J. G. & Rae, G. H. (1998) Environmental risk of ivermectin to sediment-dwelling organisms. Aquaculture, 163, 263–75. Davies, I. M., Rodger, G. K., Redshaw, J. & Stagg, R. M. (2001) Targeted environmental monitoring for the effects of medicines used to treat sea lice infestation on farmed ﬁsh. ICES Journal of Marine Science, 58, 477–85. Deady, S., Varian, S. J. A. & Fives, J. M. (1995) The use of cleaner-ﬁsh to control sea lice on two Irish salmon (Salmo salar) farms with particular reference to wrasse behaviour in salmon cages. Aquaculture, 131, 73–90. Dela Cruz, C. R. (1980) Capture and culture ﬁsheries in Chinese lakes. ICLARM Newsletter, 3 (4), 8–9. Delgado, O., Ruiz, J., Perez, M., Romero, J. & Ballesteros, E. (1999) Effects of ﬁsh farming on seagrass (Posidonia oceanica) in a Mediterranean bay: seagrass decline after loading cessation. Oceanologica Acta, 22, 109–17. De Magalhaes, V. F., Soares, R. M. & Azevedo, S. M. F. O. (2001) Microcystin contamination in ﬁsh from the Jacarepagua Lagoon (Rio de Janeiro, Brazil): ecological implications and human health risk. Toxicon, 39, 1077–85. De Silva, S. S. (1993) Supplementary feeding in semi-intensive aquaculture systems. In: Farm-Made Aquafeeds. Proceedings of the FAO/AADCP Regional Expert Consultation on Farm-Made Aquafeeds, 14–18 December 1992, Bangkok, Thailand (Ed. by M. B. New, A. G. J. Tacon & I. Csavas), pp. 87–99. FAO–RAPA/AADCP, Bangkok. De Silva, S. S. & Anderson, T. A. (1995) Fish Nutrition in Aquaculture. Chapman & Hall, London. De Silva, S. S. & Davy, F. B. (1992) Strategies for ﬁnﬁsh nutrition research for semiintensive aquaculture in Asia. Asian Fisheries Science, 5, 237–41. De Sylva, D. P. (1999) Global warming and potential range extensions of poisonous and dangerous marine organisms. World Resource Review, 2, 174–89. De Young, B. (1978) Enhancing Wave Protection With Floating Tyre Breakwaters. New York College of Agriculture and Life Sciences, Cornell University, Ithaca, NY. Diaz, M. M., Temporetti, P. F. & Pedrozo, F. L. (2001) Response of phytoplankton to enrichment from cage ﬁsh farm waste in Alicura Reservoir (Patagonia, Argentina). Lakes and Reservoirs: Research and Management, 6, 151–8. Dickman, M. D. (2000) Hong Kong’s worst red tide. In: Proceedings of the International Symposium on Environmental Hydraulics, Hong Kong, 16–18 December 1998 (Ed. by J. H. W. Lee, A. W. Jayawardena & Z. Y. Wang), pp. 641–6. A. A. Balkema, Rotterdam. Dillon, P. J. & Rigler, F. H. (1974) A test of a simple nutrient budget model predicting the phosphorus concentration in lake water. Journal of the Fisheries Research Board of Canada, 31, 1771–8. Dillon, P. J. & Rigler, F. H. (1975) A simple method for predicting the carrying capacity

References

321

of a lake for development, based on lake trophic status. Journal of the Fisheries Research Board of Canada, 32, 1519–31. Dobrowolski, Z. (1987) The application of benthic indices for evaluation of caged trout cultures inﬂuence on the littoral of Letoskie Lake. Polskie Archiv für Hydrobiologie, 34, 579–91. Dobson, D. P. & Tack, T. J. (1991) Evaluation of the dispersion of treatment solution of dichlorvos from marine salmon pens. Aquaculture, 95, 15–32. Doi, M. (1991) Yellowtail culture in Japan – an industry at cross-roads. INFOFISH International, 9/91, 42–6. Douglas-Helders, G. M., Tan, C., Carson, J. & Nowak, B. F. (2003) Effects of copperbased antifouling treatments on the presence of Neoparamoeba pemaquidensis Page, 1987 on nets and gills of reared Atlantic salmon (Salmo salar). Aquaculture, 221, 13–22. Doumenge, F. (1996) L’aquaculture des thons rouges. Biologia Marine Méditerranée, 3, 258–88. Downing, J. A., Watson, S. B. & McCauley, E. (2001) Predicting cyanobacteria dominance in lakes. Canadian Journal of Fisheries and Aquatic Sciences, 58, 1905–8. Duarte, S. A., Nelson, R. G. & Masser, M. P. (1994) Proﬁt-maximising stocking rates for channel catﬁsh cage culture. Journal of the World Aquaculture Society, 25, 442–7. Dubost, N., Masson, G. & Moreteau, J. C. (1996) Temperate freshwater fouling on ﬂoating net cages: method of evaluation, model and composition. Aquaculture, 143, 303–18. Dugdale, R. C. (1967) Nutrient limitation in the sea: dynamics, identiﬁcation and signiﬁcance. Limnology and Oceanography, 12, 685–95. Dunn, M. & Dalland, K. (1993) Observing behaviour and growth using the Simrad FCM 160 ﬁsh cage monitoring system. In: Fish Farming Technology (Ed. by H. Reinertsen, L. A. Dahle, L. Jørgensen & K. Tvinnereim), pp. 269–74. A. A. Balkema, Rotterdam. Dushkina, L. A. (1994) Farming of salmonids in Russia. Aquaculture and Fisheries Management, 25, 121–6. Dyer, K. R. (1997) Estuaries – A Physical Introduction, 2nd edition Wiley, New York. Easton, M. D. L., Luszniak, D. & Von der Geest, E. (2002) Preliminary examination of contaminant loadings in farmed salmon, wild salmon and commercial salmon feed. Chemosphere, 46, 1053–74. EDAW & CH2M/Hill (1986) Aquaculture Siting Study. A Report Prepared for the Department of Ecology, State of Washington. Seattle, WA. Edwards, A. & Edelsten, D. J. (1976) Marine ﬁsh cages – the physical environment. Proceedings of the Royal Society of Edinburgh (B), 75, 207–21. Edwards, D. J. (1978) Salmon and Trout Farming in Norway. Fishing News Books, Oxford. Edwards, P. (1992) Reuse of Human Wastes in Aquaculture. A Technical Review. UNDPWorld Bank Water and Sanitation Program, The World Bank, Washington, DC. Efﬁrd, K. D. (1975) Interrelation of corrosion and fouling for metals in seawater. In: Proceedings of Corrosion 75, pp. 232–41. National Association of Corrosion Engineers, Houston, TX. Egidius, E. & Moester, B. (1987) Effect of nevugon and nuvan treatment on crabs (Cancer pagrus, C. moenas), lobster (Homarus gammarus) and the blue mussel (Mytilus edulis). Aquaculture, 60, 165–8. Eikrem, W. & Throndsen, J. (1998) Morphology of Chrysochromulina leadbeateri (Prymnesiophyceae) from northern Norway. Phycologia, 37, 292–9.

322

References

Eley, R. L., Carroll, J. H. & De Woody, D. (1972) Effect of cage catﬁsh culture on water quality and community metabolism of a lake. Proceedings of the Oklahoma Academy of Sciences, 52, 10–15. Eller, L. L. (1975) Gill lesions in freshwater teleosts. In: The Pathology of Fishes (Ed. by W. E. Ribelin & G. Migaki), pp. 305–30. University of Wisconsin Press, Madison, WI. Ellis, M. M. (1944) Water purity standards for freshwater ﬁshes. Special Scientiﬁc Report of the US Fisheries and Wildlife Service, Vol. 2. US Fisheries and Wildlife Service, Washington, DC. Enell, M. (1982) Changes in sediment dynamics caused by cage culture activities. In: Proceedings of the 10th Nordic Symposium on Sediment, Otaniemi, Finland, 1982 (Ed. by I. Bergstrom, J. Kettunen & M. Stenmark), pp. 72–88. Division of Water Engineering, Helsinki University of Technology, Helsinki. Enell, M. & Ackefors, H. (1992) Development of Nordic salmonid production in aquaculture and nutrient discharges into adjacent sea areas. Aquaculture Europe, 16 (4), 6–11. Enell, M. & Lof, J. (1985) Changes in sediment phosphorus, iron and manganese dynamics caused by ﬁsh farming impact. In: Proceedings of the 11th Nordic Symposium on Sediments, Stockholm, Sweden, 1985 (Ed. by T. R. Gulderbrandsen & S. Samin), pp. 80–89. Division of Water Engineering, Helsinki University of Technology, Helsinki. Enell, M., Lof, J. & Bjorkland, T.-L. (1984) Fiskkasseodling med rening teknisk beskrivning och reningseffekt. Institute of Limnology, University of Lund, Sweden. Engle, C. R. (1982) Growth of fed and unfed bighead carp in cages at two stocking densities. The Progressive Fish-Culturist, 44, 216–17. Erikson, U. (2001) Potential effects of preslaughter fasting, handling and transport. In: Farmed Fish Quality (Ed. by S. C. Kestin & P. D. Warriss), pp. 202–19. Blackwell, Oxford. Ervik, A., Samuelsen, O. B., Juell, J. E. & Sveier, H. (1994a) Reduced environmental impact of antibacterial agents applied in ﬁsh farms using the LiftUp feed collector system or a hydroacoustic feed detector. Diseases of Aquatic Organisms, 19, 101–4. Ervik, A., Tehorsen, V., Eriksen, B. T., Lunestad, B. T. & Samuelsen, O. B. (1994b) Impact of adminstering antibacterial agents on wild ﬁsh and blue mussels (Mytilus edulis) in the vicinity of ﬁsh farms. Diseases of Aquatic Organisms, 18, 45–51. Escover, E. M. & Claveria, R. L. (1985) Economics of cage culture in Bicol freshwater lakes. In: Proceedings of the PCARRD/ICLARM Tilapia Economics Workshop, U.P. Los Banos, Laguna, Philippines. 10–13 August 1983 (Ed. by I. R. Smith, E. B. Torres & E. O. Tan), pp. 50–64. ICLARM, Manila. Ewing, R. D. & Ewing, S. K. (1995) Review of the effects of rearing density on the survival to adulthood for Paciﬁc salmon. The Progressive Fish-Culturist, 57, 1–25. Falconer, I. R. (Ed.) (1993) Algal Toxins in Seafood and Drinking Water. Academic Press, London. Fange, R. & Grove, D. J. (1978) Fish digestion. In: Fish Physiology, Vol. VIII (Ed. by D. J. Randall, W. Hoar & J. Brett), pp. 162–260. Academic Press, London. FAO (1983) Freshwater Aquaculture Development in China. FAO Fisheries Technical Paper 215. FAO, Rome. FAO (2003) The State of World Fisheries and Aquaculture 2002. FAO, Rome. (http://www.fao.org/DOCREP/003). Fast, A. W. (1991a) A ﬂoating ﬁsh cage with solid plastic membrane and pumped water exchange. Journal of Applied Aquaculture, 1, 99–110.

References

323

Fast, A. W. (1991b) Some ﬁsh cage designs with forced water exchange. In: Proceedings of the National Science Foundation Workshop on Offshore Engineering and Mariculture, 25–30 September 1991 (Ed. by G. N. Hirata), pp. 523–36. National Science Foundation, Honolulu, HA. Fearn, P. T. (1990) The development of an offshore ﬁsh cage. In: Engineering for Offshore Fish Farming, pp. 107–17. Thomas Telford, London. Fedoruk, A. & Srisuwantach, V. (1984) Sand Goby Culture in Thailand. FAP/WP-17. Department of Fisheries, Bangkok. Fermin, A. C. & Seronay, G. A. (1997) Effects of different illumination levels on zooplankton abundance, feeding periodicity, growth and survival of the Asian sea bass, Lates calcarifer (Bloch), fry in illuminated ﬂoating nursery cages. Aquaculture, 157, 225–35. Fernö, A., Huse, I., Juell, J.-C. & Bjordal, Å. (1995) Vertical distribution of Atlantic salmon (Salmo salar L.) in net pens: trade-off between surface light avoidance and food attraction. Aquaculture, 132, 285–96. Fernandes, T. F., Miller, K. L. & Read, P. A. (2000) Monitoring and regulation of marine aquaculture in Europe. Journal of Applied Ichthyology, 16, 138–43. Findlay, R. H. & Watling, L. (1994) Towards a process level model to predict the effects of salmon netpen aquaculture on the benthos. In: Modelling Benthic Impacts of Organic Enrichment from Marine Aquaculture (Ed. by B. T. Hargrave), pp. 47–77. Canadian Technical Report on Fisheries and Aquatic Sciences No. 1949. Department of Fisheries and Oceans, Ottawa. Findlay, R. H. & Watling, L. (1997) Prediction of benthic impact for salmon net-pens based on the balance of benthic oxygen supply and demand. Marine Ecology Progress Series, 155, 147–57. Flagg, T. A. & Harrell, L. W. (1990) Use of water-to-water transfer to maximise survival of salmonids stocked directly to seawater. The Progressive Fish-Culturist, 52, 127–9. Fleming, I. A., Hindar, K., Mjolnerod, I. B., Johnsson, B., Balstad, T. & Lamberg, A. (2000) Lifetime success and interactions of farm salmon invading a native population. Proceedings of the Royal Society of London (Series B), 267, 1717–24. Folke, C. (1988) Energy economy of salmon aquaculture in the Baltic Sea. Environmental Management, 12, 525–37. Folke, C. & Kautsky, N. (1989) The role of ecosystems for a sustainable development of aquaculture. Ambio, 18, 234–43. Folke, C. & Kautsky, N. (1992) Aquaculture with its environment: prospects for sustainability. Ocean and Coastal Management, 17, 5–24. Folke, C., Kautsky, N. & Troell, M. (1994) The costs of eutrophication from salmon farming: implications for policy. Journal of Environmental Management, 40, 173–82. Fosseidengen, J. K., Boge, E. & Huse, I. (1982) A survey with rainbow trout and salmon in submergible cages (in Norwegian). Norsk Fiskeoppdrettingen, 10, 24–5. Foster, M., Petrell, R. J., Ito, M. R. & Ward, R. (1995) Detection and counting of uneaten food pellets in a sea cage using image analysis. Aquacultural Engineering, 14, 251–69. Frediriksson, D. W., Swift, M. R., Irish, J. D., Tsukrov, I. & Celikkol, B. (2003) Fish cage and mooring system dynamics using physical and numerical models with ﬁeld measurements. Aquacultural Engineering, 27, 117–46. Freidman, L. & Shibko, S. I. (1972) Non-nutrient components of the diet. In: Fish Nutrition (Ed. by J. E. Halver), pp. 182–255. Academic Press, London. FRS (2002a) Locational Guidelines for the Authorization of Marine Fish Farms in Scottish Waters. Scottish Executive, Edinburgh.

324

References

FRS (2002b) Scottish Fish Farms. Annual Production Survey, 2001. Scottish Executive, Edinburgh. FSBI (2002) Fish Welfare. Brieﬁng Paper 2. Fisheries Society of the British Isles, Cambridge. Fujiya, M. (1979) Coastal culture of yellowtail (Seriola quinqueradiata) and red sea bream (Pagrus major) in Japan. In: Advances in Aquaculture (Ed. by T. V. R. Pillay & W. A. Dill), pp. 428–41. Fishing News Books, Oxford. Fukusho, T. (1991) Red sea bream culture in Japan. In: Finﬁsh Aquaculture. Handbook of Mariculture, Vol. II (Ed. by J. P. McVey), pp. 73–87. CRC Press, Boca Raton, FL. Furevik, D. M., Bjordal, Å., Huse, I. & Fernö, A. (1993) Surface activity of Atlantic salmon (Salmo salar L.) in net pens. Aquaculture, 110, 119–28. Furness, R. W. (1996) Interactions between seabirds and aquaculture in sea lochs. In: Aquaculture and Sea Lochs (Ed. by K. D. Black), pp. 50–55. Scottish Association for Marine Science, Oban. Gace, L., Celikkol, B. & Savage, G. (1996) Engineering analysis and design of cage systems for exposed locations. In: Open Ocean Aquaculture. Proceedings of an International Conference, Portland, Maine, 8–10 May 1996 (Ed. by M. Polk), pp. 467–94. Report No. UNHMP-CP-SG-96-9. New Hampshire/Maine Sea Grant College Program, Portland, ME. Gaigher, I. G. & Krause, J. B. (1983) Growth rate of Mozambique tilapia (Oreochromis mossambicus) and silver carp (Hypophthalmichthys molitrix) without artiﬁcial feeding in ﬂoating cages in plankton-rich waste water. Aquaculture, 31, 361–7. Gaigher, I. G. & Toerien, D. F. (1985) Cage culture of Mozambique tilapia, Oreochromis niloticus, without artiﬁcial feeding in maturation ponds of the Phuthaditjhaba sewage system. Water SA, 11, 19–24. Garant, D., Fleming, I. A., Einum, S. & Bernatchez, L. (2003) Alternative male life history tactics as potential vehicles for speeding introgression of farm salmon traits into wild populations. Ecology Letters, 6, 541–9. Garner, J. (1989) How to Make and Set Nets. Fishing News Books, Oxford. Garvey, S. & Bennett, J. (1991). Aquaculture and rural development in the west coast of Ireland. In: Aquaculture and the Environment (Ed. by N. de Pauw & J. Joyce), pp. 421–59. EAS Special Publication No. 16. European Aquaculture Society, Ghent. Gasca-Leyva, E., León, C. J., Hernández, J. M. & Vergara, J. M. (2002) Bioeconomic analysis of production location of sea bream (Sparus aurata) cultivation. Aquaculture, 213, 219–32. Gausen, D. & Moen, V. (1991) Large-scale escapes of Atlantic salmon (Salmo salar) into Norwegian rivers threaten natural populations. Canadian Journal of Fisheries and Aquatic Sciences, 48, 426–8. Gavine, F. M., Phillips, M. J. & Murray, A. (1995) Inﬂuence of improved feed quality and food conversion ratio on phosphorus loadings from cage culture of rainbow trout Oncorhynchus mykiss (Walbaum) in freshwater lakes. Aquaculture Research, 26, 483–95. Geffen, A. (1979) Rotating cage to prevent fouling. Aquaculture, 16, 83–5. GESAMP (1997) Towards Safe and Effective Use of Chemicals in Coastal Aquaculture. Report Study of the IMO/FAO/UNESCO-IOC/WMO/WHO/IAEA/UN/UNEP Joint Group of Experts on the Scientiﬁc Aspects of Marine Environmental Protection. Reports and Studies 65. FAO, Rome. GESAMP (2001) Planning and Management for Sustainable Coastal Aquaculture Development. Reports and Studies 68. FAO, Rome. (http://gesamp.imo.org/publicat.htm)

References

325

Getabu, A. (1987) The Culture and Food of Oreochromis niloticus in Waste-Water Lagoons. MSc thesis, University of Nairobi, Kenya. Geuskens, G. (Ed.) (1975) Degradation and Stabilization of Polymers. Elsevier Applied Science Publishers, New York. Gillibrand, P. A., Gubbins, M. J., Greathead, C. & Davies, I. M. (2002) Scottish Executive Locational Guidelines for Fish Farming: Predicted Levels of Nutrient Enhancement and Benthic Impact. Scottish Fisheries Research Report 63. Fisheries Research Services, Aberdeen. Goddard, S. (1996) Feed Management in Intensive Aquaculture. Chapman & Hall, London. Goddard, S. & Scott, P. (1980) Understanding appetite. Fish Farmer, 3 (6), 40–41. Gonzales, E. (1984) Small-scale tilapia cage technology adopted in ﬁshing villages in Laguna Lake, Philippines. Aquaculture, 41, 161–9. Gooley, G. J., De Silva, S. S., Hone, P. W., McKinnon, L. J. & Ingram, B. A. (2000) Cage aquaculture in Australia: a developed country perspective with reference to integrated aquaculture development within inland waters. In: Cage Aquaculture in Asia. Proceedings of the First International Symposium on Cage Aquaculture in Asia, Tungkang, Taiwan, 2–6 November 1999 (Ed. by I. C. Liao & C. K. Lin), pp. 21–37. Asian Fisheries Society, Manila and World Aquaculture Society – Asia Branch, Bangkok. Goudey, C. A., Loverich, G., Kite-Powell, H. & Costa-Pierce, B. A. (2001) Mitigating the environmental effects of mariculture through single point moorings (SPMs) and drifting cages. ICES Journal of Marine Science, 58, 497–503. Gowen, R. J. (1984) The Ecology of Phytoplankton in Scottish Coastal Waters With Particular Reference to Toxic Species and Their Importance to Mariculture. Scottish Marine Biological Association, Dunstaffnage, Oban. Gowen, R. J. (1990) An Assessment of the Impact of Fish Farming on the Water Column and Sediment Ecosystems of Irish Coastal Waters. A report prepared for the Department of the Marine, Ireland. Department of the Marine, Dublin. Gowen, R. J. & Bradbury, N. B. (1987) The ecological impact of salmonid farming in coastal waters: a review. In: Oceanography and Marine Biology: An Annual Review (Ed. by M. Barnes), Oceanography and Marine Biology Reviews, 25, 563–75. Gowen, R. J., Bradbury, N. B. & Brown, J. R. (1985) The ecological impact of salmon farming in Scottish coastal waters: a preliminary appraisal. Proceedings of ICES CM 1985/F:35. ICES, Copenhagen. Gowen, R. J., Bradbury, N. B. & Brown, J. R. (1989) The use of simple models in assessing two of the interactions between ﬁsh farming and the marine environment. In: Aquaculture – A Biotechnology in Progress (Ed. by N. de Pauw, E. Jaspers, H. Ackefors & N. Wilkins), pp. 1071–80. European Aquaculture Society, Ghent. Gowen, R. J., Weston, D. P. & Ervik, A. (1991) Aquaculture and the benthic environment: a review. In: Nutritional Strategies and Aquaculture Waste. Proceedings of the First International Symposium on Nutritional Strategies in the Management of Aquaculture Waste (Ed. by C. B. Cowey & C. Y. Cho), pp. 187–205. University of Guelph, Guelph, Ontario. Gowen, R. J., Smyth, D. & Silvert, W. (1994) Modelling the spatial distribution and loading of organic ﬁsh farm waste to the seabed. In: Modelling Benthic Impacts of Organic Enrichment from Marine Aquaculture (Ed. by B. T. Hargrave), pp. 19–30. Canadian Technical Report on Fisheries and Aquatic Sciences No. 1949. Department of Fisheries and Oceans, Ottawa.

326

References

Graneli, E., Sundstroem, B., Edler, L. & Anderson, D. M. (Eds) (1991) Proceedings of the International Conference on Toxic Marine Phytoplankton, Lund (Sweden), 26–30 June 1989. Elsevier, New York. Graneli, E., Paasche, E. & Maestrini, S. Y. (1993) Three years after the Chrysochromulina polylepis bloom in Scandanavian waters in 1988: some conclusions of recent research and monitoring. In: Toxic Phytoplankton Blooms in the Sea (Ed. by T. J. Smayda & Y. Shimizu), pp. 23–32. Elsevier, Amsterdam. Grave, H. (1975) A new type of net cage for ﬁsh culture used in Kiel Fjord. Meeresforschung, 24, 209–11. Grave, H. (1981) Net cage culture under extreme hydrological and meteorological conditions. Special Publication of the European Mariculture Society, 6, 25–36. Grave, K., Engelstad, M., Søli, N. E. & Hastein, T. (1990) Utilization of antibacterial drugs in salmonid farming in Norway during 1980–88. Aquaculture, 86, 347–58. Greenland, D. C., Newton, S. H. & Faucette, R. F., Jr. (1988) Effects of cage encrustation by the bryozoan Plumatella casmiana production of channel catﬁsh. The Progressive Fish-Culturist, 50, 42–5. Grimaldi, E. (1973) Diffusa infezione branchiale da funghi attribiti al genera Branchimyces Plehn (Phycomycetes saprolegniales) a carico dell’ittiofauna di laghi situati a nord e a sud delle Alpi. Memoria del Istituto di Italien Idrobiologia, 30, 61–96. Grizzle, J. M., Horowitz, S. A. & Strength, D. R. (1988) Caged ﬁsh as monitors of pollution: effects of chlorinated efﬂuent from a wastewater treatment plant. Water Resources Bulletin, 25, 951–9. Guerrero, R. D., III (1981) An Introduction to Fish Culture in the Philippines. Philippine Education Co., Inc., Manila. Guldberg, B., Kittelsen, A., Rye, M. & Åsgard, T. (1993) Improved salmon production in large cage systems (Abstract). In: Fish Farming Technology (Ed. by H. Reinertsen, L. A. Dahle, L. Jørgensen & K. Tvinnereim), p. 241. A. A. Balkema, Rotterdam. Gulland, J. A. (1970) Food chain studies and some problems in world ﬁsheries. In: Marine Food Chains (Ed. by J. H. Steele), pp. 296–315. University of California Press, Berkeley, CA. Gunnarsson, J. (1988) Bridgestone Hi-Seas ﬁsh cage design and use. In: Aquaculture Engineering Technologies for the Future. Institute of Chemical Engineers Symposium Series Vol. 111, pp. 133–41. Institute of Chemical Engineers, Rugby. Gunnarsson, J. (1993) Bridgestone Hi-Seas ﬁsh cage: design and documentation. In: Fish Farming Technology (Ed. by H. Reinertsen, L. A. Dahle, L. Jørgensen & K. Tvinnereim), pp. 199–201. A. A. Balkema, Rotterdam. Gunnarsson, J. (1996) Open ocean aquaculture: Bridgestone Hi-Seas ﬁsh cage. In: Open Ocean Aquaculture. Proceedings of an International Conference, Portland, Maine, 8–10 May 1996 (Ed. by M. Polk), pp. 389–98. Report No. UNHMP-CP-SG-96-9. New Hampshire/Maine Sea Grant College Program, Portland, ME. Häder, D.-P., Worrest, R. C., Kumar, H. D. & Smith, R. C. (1995) Effects of increased solar ultraviolet radiation on aquatic ecosystems. Ambio, 24, 174–80. Hair, C., Bell, J. & Doherty, P. (2002) The use of wild-caught juveniles in coastal aquaculture and its application to coral reef ﬁshes. In: Responsible Marine Aquaculture (Ed. by R. R. Stickney & J. P. McVey), pp. 327–53. CAB International, Cambridge. Håkanson, L., Carlson, L. & Johansson, T. (1998) A new approach to calculate the phosphorus load to lakes from ﬁsh farm emissions. Aquacultural Engineering, 17, 149–66.

References

327

Hall, P. O. J., Anderson, L. G., Holby, O., Kollberg, S. & Samuelsson, M.-O. (1990) Chemical ﬂuxes and mass balances in a marine ﬁsh cage farm. I. Carbon. Marine Ecology Progress Series, 61, 61–73. Hall, P. O. J., Holby, O., Kollberg, S. & Samuelsson, M.-O. (1992) Chemical ﬂuxes and mass balances in a marine ﬁsh cage farm. IV. Nitrogen. Marine Ecology Progress Series, 89, 81–91. Halvasen, O. & Hartvigsen, R. (1989) A review of the biogeography and epidemiology of Gyrodactylus salaris. Norsk Institut für Naturforsken, Utredning, 2, 1–41. Hambrey, J., Beveridge, M. C. M. & McAndrew, K. (2001a) Aquaculture and poverty alleviation. I. Cage culture in freshwater in Bangladesh. World Aquaculture Magazine, 32 (1), 50–55, 67. Hambrey, J., Le, A. T. & Ta, K. T. (2001b) Aquaculture and poverty alleviation. II. Cage culture in the coastal waters of Vietnam. World Aquaculture Magazine, 32 (2), 26–9. Hambrey, J. & Roy, M. (2002) Final Project Review of the One-Year Extension of the CAGES Project. Report for CARE-Bangladesh. Nautilus Consultants, Edinburgh. Hansen, P. K., Lunestad, B. T. & Samuelsen, O. B. (1992) Effects of oxytetracycline, oxolinic acid and ﬂumequine on bacteria in an artiﬁcial marine ﬁsh farm sediment. Canadian Journal of Microbiology, 38, 1307–12. Hardy, R. W. & Tacon, A. G. J. (2002) Fish meal: historical uses, production trends and future outlook for sustainable supplies. In: Responsible Marine Aquaculture (Ed. by R. R. Stickney & J. P. McVey), pp. 311–25. CAB International, Cambridge. Hargrave, B. T. (1994) A benthic enrichment index. In: Modelling Benthic Impacts of Organic Enrichment from Marine Aquaculture (Ed. by B. T. Hargrave), pp. 31–46. Canadian Technical Report on Fisheries and Aquatic Sciences No. 1949. Department of Fisheries and Oceans, Ottawa. Hargreaves, J. A., Rakocy, J. E. & Nair, A. (1988) An evaluation of ﬁxed and demand feeding regimes for cage culture of Oreochromis aureus. In: The Second International Symposium on Tilapia in Aquaculture (Ed. by R. S. V. Pullin, T. Bhukuswan, K. Tonguthai & J. L. Maclean), pp. 335–9. Department of Fisheries, Bangkok and ICLARM, Manila. (ICLARM is now known as the WorldFish Center.) Hargreaves, J. A., Rakocy, J. E. & Bailey, D. S. (1991) Effects of diffused aeration and stocking density on growth, feed conversion, and production of Florida red tilapia in cages. Journal of the World Aquaculture Society, 22, 24–9. Harvey, P. (1988) In: Salmon and Trout Farming (Ed. by L. M. Laird & T. Needham), pp. 190–201. Ellis Horwood, Chichester. Hasse, J. (1974) Utilisation of rabbit ﬁshes (Siganidae) in tropical oyster culture. The Progressive Fish-Culturist, 36, 160–62. Hastings, T. S. & McKay, A. (1987) Resistance of Aeromonas salmonicida to oxolinic acid. Aquaculture, 60, 133–41. Hawkins, W. E., Fournie, J. W. & Chansue, N. (2002) Non-infectious disorders of warmwater ﬁsh. In: Diseases and Disorders of Finﬁsh in Cage Culture (Ed. by P. T. K. Woo, D. W. Bruno & L. H. S. Lim), pp. 283–304. CABI Publishing, Wallingford. Haya, K., Burridge, L. E. & Chang, B. D. (2001) Environmental impact of chemical wastes produced by the salmon aquaculture industry. ICES Journal of Marine Science, 58, 492–6. Haylor, G. S. (1993) Fish production from engineered water systems in developing countries. In: Recent Advances in Aquaculture, Vol. V (Ed. by J. F. Muir & R. J. Roberts), pp. 1–103. Blackwell, Oxford.

328

References

Heen, K. (1993) Comparative analysis of the cost structure and proﬁtability in the salmon aquaculture industry. In: Salmon Aquaculture (Ed. by K. Heen, R. L. Monaghan & F. Utter), pp. 220–38. Fishing News Books, Oxford. Heilskov, A. C. & Holmer, M. (2001) Effects of benthic fauna on organic matter mineralization in ﬁsh-farm sediments: importance of size and abundance. ICES Journal of Marine Science, 58, 427–34. Henderson, A. R. & Davies, I. M. (2000) Review of aquaculture, its regulation and monitoring in Scotland. Journal of Applied Ichthyology, 16, 200–208. Henderson, E. B. (1980) Salmonid cage rearing trials in the Bay of Fundy. In: Proceedings of the Institute of Fisheries Management Cage Fish Rearing Symposium, University of Reading, 26–27 March 1980, pp. 76–97. Janssen Services, London. Henriksson, S.-H. (1991) Effects of ﬁsh farming on natural Baltic ﬁsh communities. In: Marine Aquaculture and Environment (Ed. by T. Mäkinen), pp. 85–104. Nordic Council of Ministers, Copenhagen. Herbert, D. W. M. & Merkens, J. C. (1961) The effect of suspended mineral solids on the survival of trout. International Journal of Air & Water Pollution, 5, 46–55. Herbert, D. W. M. & Richards, J. M. (1963) The growth and survival of ﬁsh in some suspensions of solids of industrial origin. International Journal of Air & Water Pollution, 7, 297–302. Herwig, R. P., Gray, J. P. & Weston, D. P. (1997) Antibacterial resistant bacteria in surﬁcial sediments near salmon net-cage farms in Puget Sound, Washington. Aquaculture, 49, 263–83. Hey, R. J. & Pagan, A. V. (1983) Metals. In: Civil Engineering Materials, 3rd edition (Ed. by N. Jackson), pp. 1–65. Macmillan Publishers, London. Hickling, C. F. (1962) Fish Culture. Faber & Faber, London. Higgs, D. A. (1997) Nutritional strategies for cost effective salmon production In: Proceedings of the First Korea–Canada Joint Symposium in Aquatic Biosciences, Pukyong National University, 16 October 1997, pp. 67–92. Institute of Fisheries Science, Pukyong National University, Korea. Hilton, J. W., Cho, C. Y. & Slinger, S. J. (1981) Effect of extrusion, processing and steam pelleting diets on pellet durability, pellet water absorption and the physiological response of rainbow trout. Aquaculture, 25, 185–94. Hisoaka, M., Nogami, K., Takeuchi, O., Suzuki, M. & Sugimoto, H. (1966) Studies on seawater exchange in ﬁsh farm. II. Exchange of sea water in ﬂoating net. Bulletin Naikai Regional Fisheries Research Laboratory, 115, 21–43. Ho, J.-S. (2000) The major problem of cage aquaculture in Asia relating to sea lice. In: Cage Aquaculture in Asia. Proceedings of the First International Symposium on Cage Aquaculture in Asia, Tungkang, Taiwan, 2–6 November 1999 (Ed. by I. C. Liao & C. K. Lin), pp. 13–19. Asian Fisheries Society, Manila and World Aquaculture Society – Asia Branch, Bangkok. Hodson, S. L. & Burke, C. (1994) Microfouling of salmon cage netting: a preliminary investigation. Biofouling, 8, 93–105. Hodson, S. L., Lewis, T. E. & Burke, C. M. (1997) Biofouling of ﬁsh-cage netting: efﬁcacy and problems of in situ cleaning. Aquaculture, 152, 77–90. Hodson, S. L., Burke, C. M. & Bissett, A. P. (2000) Biofouling of ﬁsh-cage netting: the efﬁcacy of silicone coating and the effect of netting colour. Aquaculture, 184, 277–90. Hogendoorn, I. H. (1984) Dutch multiple ﬁsh culture. Fish Farmer, 7 (1), 18–20. Holby, O. & Hall, P. O. J. (1992) Chemical ﬂuxes and mass balances in a marine ﬁsh cage farm. II. Phosphorus. Marine Ecology Progress Series, 70, 263–72.

References

329

Holby, O. & Hall, P. O. J. (1994) Chemical ﬂuxes and mass balances in a marine ﬁsh cage farm. III. Silicon. Aquaculture, 120, 305–18. Holcik, J. (1991) Fish introductions in Europe with particular reference to its central and eastern part. Canadian Journal of Fisheries and Aquatic Sciences, 48 (Suppl. 1), 13–23. Holdgate, M. W. (1979) A Perspective of Environmental Pollution. Cambridge University Press, Cambridge. Holm, J.-C. (1992) Salmonids in duo-culture: growth rate gain and behavioural changes. In: The Importance of Feeding Behaviour for the Efﬁcient Culture of Salmonid Fishes (Ed. by J. E. Thorpe & F. A. Huntingford), pp. 49–52. World Aquaculture Society, Baton Rouge, LA. Holm, J.-C. & Møller, D. (1984) Growth and prey selection by Atlantic salmon yearlings reared on live freshwater zooplankton. Aquaculture, 43, 401–12. Horbund, H. M. & Freiberger, A. (1970) Slime ﬁlms and their role in marine fouling. A review. Ocean Engineering, 1, 631–4. Horne, A. J. & Goldman, C. R. (1994) Limnology, 2nd edition. McGraw-Hill, New York. Horvath, L., Tamas, G. & Tolg, I. (1984) Special Methods in Pond Fish Husbandry. Halver Corporation, Seattle, WA. Howard, K. T. & Kingwell, S. J. (1975) Marine ﬁsh farming: development of holding facilities, with particular reference to work by the WFA on the Scottish west coast. In: Proceedings of Oceanology International 1975, 16–21 March 1975, Brighton, UK, pp. 183–90. BPS Exhibitions, London. Howarth, W. (1991) The Law of Aquaculture. The Law Relating to the Farming of Fish and Shellﬁsh in Britain. Fishing News Books, Oxford. Howells, G. (Ed.) (1994) Water Quality for Freshwater Fish. Gordon & Breach, Reading. Howgate, P., Bunting, S., Beveridge, M. C. M. & Reilly, A. (2002) Aquaculture associated public, animal and environmental health issues in non-industrialized countries. In: Public, Animal and Environmental Aquaculture Health Issues (Ed. by M. L. Jahncke, S. Garrett, A. Reilly, R. E. Martin & E. Cole), pp. 21–65. Wiley, New York. Hu, B.-T. (1994) Cage culture development and its role in aquaculture in China. Aquaculture and Fisheries Management, 25, 305–10. Huchette, S. M. H. & Beveridge, M. C. M. (2003) Technical and economical evaluation of periphyton-based cage culture of tilapia (Oreochromis niloticus) in tropical freshwater cages. Aquaculture, 218, 219–34. Huchette, S. M. H., Beveridge, M. C. M., Baird, D. J. & Ireland, M. (2000) The impacts of grazing by tilapias (Oreochromis niloticus L.) on periphyton communities growing on artiﬁcial substrate in cages. Aquaculture, 186, 45–60. Huet, M. (1956) Apercu de la pisiculture en Indonesie. Bulletin Agricole Congo Belge, 47, 901–57. Huguenin, J. E. (1997) The design, operation and economics of age culture systems. Aquacultural Engineering, 16, 167–203. Huguenin, J. E. & Ansuini, F. J. (1978) A review of the technology and economics of marine ﬁsh cage systems. Aquaculture, 15, 151–70. Huguenin, J. E. & Ansuini, F. J. (1981) Marine biofouling of synthetic and metallic screens. In: Proceedings of Ocean 81, pp. 545–9. Marine Technology Society and Institute of Electrical and Electronics Engineers, Boston, MA. Huguenin, J. E. & Rothwell, G. N. (1979) The problems, economic potentials and system design of large future tropical marine cage systems. Proceedings of the Annual Meeting of the World Mariculture Society, 10, 162–81.

330

References

Huguenin, J. E. & Willinsky, M. (1996) Conceptual, engineering and operational frameworks for submersible cage systems. In: Open Ocean Aquaculture. Proceedings of an International Conference, Portland, Maine, 8–10 May 1996 (Ed. by M. Polk), pp. 41–92. Report No. UNHMP-CP-SG-96-9. New Hampshire/Maine Sea Grant College Program, Portland, ME. Huguenin, J. E., Fuller, S. C., Ansuini, F. J. & Dodge, W. T. (1981) Experience with a fouling-resistant modular marine ﬁsh cage system. In: Proceedings of the Bioengineering Symposium for Fish Culture, Traverse City, Michigan, 16–18 October 1979 (Ed. by L. J. Allen & E. C. Kinney), pp. 201–9. Fish Culture Section of the American Fisheries Society, Bethesda, MA. Hull, S. T. & Edwards, R. D. (1979) Progress in farming turbot, Scophthalmus maximus, in ﬂoating sea cages. In: Advances in Aquaculture (Ed. by T. V. R. Pillay & W. A. Dill), pp. 466–72. Fishing News Books, Oxford. Hurlbert, S. (1984) Pseudoreplication and the design of ecological ﬁeld experiments. Ecological Monographs, 54, 187–211. Hurtado-Ponce, A. Q. (1992) Cage culture of Kappaphycus alvarezzi var. tambalang (Gigartinales, Rhodophyceae). Journal of Applied Phycology, 4, 311–13. Huse, I. & Holm, J. C. (1993) Vertical distribution of Atlantic salmon Salmo salar as a function of illumination. Journal of Fish Biology, 43 (Suppl. A), 147–56. Huse, I., Bjordal, Å., Fernö, A. & Furevik, D. M. (1990) The effect of shading in pen rearing of Atlantic salmon (Salmo salar). Aquacultural Engineering, 9, 235–44. Hussain, S. M. & Khatoon, Z. (2000) Preliminary studies on cage culture of Lutjanus johni (snapper) and Pomadasys kaakan (grunt) marine ﬁshes. Pakistan Journal of Zoology, 32, 85–91. Huy, P. T., Phue, P. D., Vinhj, N. N. & Sollows, J. D. (2001) Cage ﬁsh trials in Ea Soiup Reservoir, Vietnam. In: Reservoir and Culture-Based Fisheries: Biology and Management. Proceedings on an International Workshop, Bangkok, Thailand, 15–18 February 2000. ACIAR Proceedings No. 98 (Ed. by S. S. De Silva), pp. 367–74. ACIAR, Canberra. Ibrahim, K. H., Nozawa, T. & Lema, R. (1976) Preliminary observations on cage culture of Tilapia esculenta (Graham) and Tilapia zillii (Gervais) in Lake Victoria waters at the Freshwater Institute, Nyegezi, Nigeria. African Journal of Tropical Hydrobiology and Fisheries, 4, 121–7. Ibrekk, H. O., Kryvi, H. & Elvestad, S. (1993) Nationwide assessment of the suitability of the Norwegian coastal zone and rivers for aquaculture. Coastal Management, 21, 53–73. ICE (1979) Corrosion in Civil Engineering. Institute of Civil Engineering, London. IDRC (1983) Fish Quarantine and Fish Diseases in Southeast Asia. Report of Workshop held in Jakarta, Indonesia, 7–10 December 1982. IDRC, Ottawa. IDRC/SEAFDEC (1979) International Workshop on Pen and Cage Culture of Fish. SEAFDEC Aquaculture Department, Tigbauan, Philippines. Ikenoue, H. & Kafuku, T. (1988) Modern Methods of Aquaculture in Japan, 2nd edition. Elsevier, Amsterdam. Inoue, H. (1972) On water exchange in a net cage stocked with the ﬁsh Hamachi. Bulletin of the Japanese Society of Scientiﬁc Fisheries, 38, 167–76. Iwama, G. K. (1991) Interactions between aquaculture and the environment. Critical Reviews in Environmental Control, 21, 177–216. Jackman, S. & Ace-Hopkins, J. (1993) CAGEGUARD: a sonar-based system for the security of offshore ﬁsh farm installations and the detection and repelling of seals. In: Fish

References

331

Farming Technology (Ed. by H. Reinertsen, L. A. Dahle, L. Jørgensen & K. Tvinnereim), pp. 265–8. A. A. Balkema, Rotterdam. Jackson, A. (1988) Growth, nutrition and feeding. In: Salmon and Trout Farming (Ed. by L. M. Laird & T. Needham), pp. 202–16. Ellis Horwood, Chichester. Jäger, T. (1986) Economic aspects of coregonid culture in illuminated net cages. Archiv für Hydrobiologie und Limnologie, 22, 231–9. Jäger, T. & Kiwus, A. (1980) Aufzucht von tiechtzechtlingen in erleucheten netzgehegen. Fischerei und Teichwirt, 11, 323–6. Jahncke, M. L. & Schwarz, M. H. (2002) Public, animal and environmental aquaculture health issues in industrialized countries. In: Public, Animal and Environmental Aquaculture Health Issues (Ed. by M. L. Jahncke, S. Garrett, A. Reilly, R. E. Martin & E. Cole), pp. 67–102. Wiley, New York. Jangkuru, A. & Djajadiredja, R. (1979) Common carp in ﬂoating net cage culture. In: Proceedings of the IDRC/SEAFDEC International Workshop on Pen and Cage Culture of Fish, Tigbauan, Iloilo, Philippines, 11–22 February 1979, pp. 55–60. SEAFDEC, Iloilo, Philippines. Janssen, J. J. A. (1981) Bamboo in building structures. PhD thesis, University of Eindhoven, Germany. Jarrams, P., Starkie, A., Easton, K. W. & Templeton, R. G. (1980) Salmonid rearing in ﬂoating net cages in freshwater reservoirs owned by the Severn–Trent Water Authorities. Fisheries Management, 11, 63–79. Jauncey, K. (1998) Tilapia Feeds and Feeding. Pisces Press, Stirling. Jauncey, K. (2000) Nutritional requirements. In: Tilapias: Biology and Exploitation (Ed. by M. C. M. Beveridge & B. J. McAndrew), pp. 327–75. Kluwer, Dordrecht. Jayaraj, E. G., Devaraj, K. V. & Chowdary, G. P. (1998) Rearing of common carp (Cyprinus carpio) fry to ﬁngerlings in pens. Journal of Aquaculture in the Tropics, 13, 245–53. Jee, A. K. (1980) Some problems in the cage culture of marble goby (Oxyeleotris marmorata Bleeker). Aquaculture, 20, 229. Jobling, M. (1986) Gastrointestinal overload – a problem with formulated feeds. Aquaculture, 51, 257–63. Jobling, M. (1993) Nutrition, diet formulation and feeding practices. In: Salmon Aquaculture (Ed. by K. Heen, R. L. Monaghan & F. Utter), pp. 83–126. Fishing News Books, Oxford. Jobling, M. (1995) Simple indices for the assessment of the inﬂuences of social environment on growth performance, exempliﬁed by studies on Arctic charr. Aquaculture International, 3, 60–5. Jobling, M., Baardvik, B. M., Christiansen, J. S. & Jørgensen, E. H. (1993) The effects of prolonged exercise training on growth performance and production parameters in ﬁsh. Aquaculture International, 1, 95–111. Jobling, M., Gomes, E. & Dias, J. (2001) Feed types, manufacture and ingredients. In: Food Intake in Fish (Ed. by D. Houlihan, T. Boujard & M. Jobling), pp. 25–48. Blackwell, Oxford. Johannessen, P. J., Botnen, H. B. & Tvedten, Ø. F. (1994) Macrobenthos: before, during and after a ﬁsh farm. Aquaculture and Fisheries Management, 25, 55–66. Johansson, T. & Nordvarg, L. (2002) Empirical mass balance models calibrated for freshwater ﬁsh farm emissions. Aquaculture, 212, 191–211. Johansson, T., Håkanson, L., Borum, K. & Persson, J. (1998) Direct ﬂows of phosphorus and suspended matter from a ﬁsh farm to wild ﬁsh in Lake Southern Bullaren, Sweden. Aquacultural Engineering, 17, 111–37.

332

References

Johnsen, R. I., Grahl-Nielsen, O. & Lunestad, B. T. (1993) Environmental distribution of organic waste from a marine ﬁsh farm. Aquaculture, 118, 229–44. Johnssen, B. O. & Jensen, A. J. (1994) The spread of furunculosis in salmonids in Norwegian rivers. Journal of Fish Biology, 29, 233–41. Jones, J. R. & Bachmann, R. W. (1976) Prediction of phosphorus and chlorophyll levels in lakes. Journal of the Water Pollution Control Federation, 48, 2176–82. Jones, K. J., Ayres, P., Bullock, A. M., Roberts, R. J. & Tett, P. (1982) A red tide of Gyrodinium aureolum in sea lochs of the Firth of Clyde and associated mortality of pondreared salmon. Journal of the Marine Biological Association of the UK, 62, 771–82. Jones, K. J., Gowen, R. J. & Tett, P. (1984) Tidal mixing and summer phytoplankton distribution in the Sound of Jura. Journal of Experimental Marine Biology and Ecology, 78, 269–90. Jones, T. O. & Iwama, G. K. (1991) Polyculture of the Paciﬁc oyster Crassostrea gigas (Thunberg) with Chinook salmon Oncorhynchus tshawtyscha. Aquaculture, 92, 313–22. Jørgensen, E. H., Christiansen, J. S. & Jobling, M. (1993) Effects of stocking density on food intake, growth performance and oxygen consumption in Arctic charr (Salvilinus alpinus). Aquaculture, 110, 191–204. Juell, J.-E. (1991) Hydroacoustic detection of food waste – a method to estimate maximum food intake of ﬁsh populations in sea cages. Aquacultural Engineering, 10, 207–17. Juell, J.-E. (1995) The behaviour of Atlantic salmon in relation to efﬁcient cage rearing. Reviews in Fish Biology and Fisheries, 5, 320–35. Juell, J.-E. & Westerberg, H. (1993) An ultrasonic telemetric system for automatic positioning of individual ﬁsh used to track Atlantic salmon (Salmo salar L.) in a sea cage. Aquacultural Engineering, 12, 1–18. Juell, J.-E., Furevik, D. M. & Bjordal, Å. (1993) Demand feeding in salmon farming by hydroacoustic food detection. Aquacultural Engineering, 12, 155–67. Kaartvedt, S., Johnsen, T. M., Aksnes, D. L., Lie, U. & Svendsen, H. (1991) Occurrence of the toxic phytoﬂagellate Prymnesium parvum and associated ﬁsh mortality in a Norwegian fjord system. Canadian Journal of Fisheries and Aquatic Sciences, 48, 2316–23. Kaasa, T. (1995) Handling of dead ﬁsh from ﬁsh farms in Norway. In: Abstracts, Aquaculture and Water Resources Management. International Conference, University of Stirling, 22–26 June 1994 (Ed. by D. J. Baird, M. C. M. Beveridge, L. A. Kelly & J. F. Muir), p. 31. Institute of Aquaculture, University of Stirling, Stirling. Kadowaki, S., Kasedo, T., Nakazono, T. & Hirata, H. (1978) Continuous records of DO contents by cruising in the coastal culture farms. II. Diffusion of suspended particles by feeding. Memoirs of the Faculty of Fisheries of Kagoshima University, 27, 281–8. Kadri, S., Metcalfe, N. B., Huntingford, F. A. & Thorpe, J. E. (1991) Daily feeding rhythms in Atlantic salmon in sea cages. Aquaculture, 92, 219–24. Kalle, K. (1971) General aspects of salinity. In: Marine Ecology: A Comprehensive, Integral Treatise on Life in Oceans and Coastal Waters. Vol. 1: The Physical and Chemical Environment (Ed. by I. O. Kinne), pp. 683–8. Wiley, New York. Kanehiro, H. & Kasu, J. E. (1988) Mechanism of photodegradation of nylon netting twines. Nippon Suisan Gakkaishi, 54, 1777–82. Karakassis, I. (2001) Ecological effects of ﬁsh farming in the Mediterranean. In: Environmental Impact Assessment of Mediterranean Aquaculture Farms. Proceedings of the Seminar of the CIHEAM Network TECAM, Zaragoza, Spain, 17–21 January 2000

References

333

(Ed. by A. Uriate & B. Bascurco), pp. 15–22. Options Méditerranéennes, Series Cahiers. Zaragoza: Vol. 55. CIHEAM/FAO, Rome. Karakassis, I., Tsapakis, M. & Hatziyanni, E. (1998) Seasonal variability in sediment proﬁles beneath ﬁsh farms cages in the Mediterranean. Marine Ecology Progress Series, 162, 243–52. Karakassis, I., Hatziyanni, E., Tsapakis, M. & Plaiti, W. (1999) Benthic recovery following cessation of ﬁsh farming: a series of successes and catastrophes. Marine Ecology Progress Series, 184, 205–18. Karl, H., Kuhlmann, H. & Oetjen, K. (2002) Transfer of toxaphene and chlordane into farmed rainbow trout, Oncorhynchus mykiss (Walbaum), via feed. Aquaculture Research, 33, 925–32. Karlsen, L. (1993) Developments in salmon aquaculture technology. In: Salmon Aquaculture (Ed. by K. Heen, R. L. Monaghan & F. Utter), pp. 59–82. Fishing News Books, Oxford. Kaspar, H. F., Hall, G. H. & Holland, A. J. (1988) Effects of sea cage salmon farming on sediment nitriﬁcation and dissimilatory nitrate reductions. Aquaculture, 70, 333–44. Kasu, J. E. (1988) The effect of additives on the photodegradation of nylon netting twines. Nippon Suisan Gakkaishi, 54, 1769–76. Kato, J., Noma, T. & Uekita, Y. (1979) Design of ﬂoating breakwaters. In: Advances in Aquaculture (Ed. by T. V. R. Pillay & W. A. Dill), pp. 458–66. Fishing News Books, Oxford. Kautsky, N., Berg, H., Folke, C., Larsson, J. & Troell, M. (1997) Ecological footprint for assessment of resource use and development limitations in shrimp and tilapia aquaculture. Aquaculture Research, 28, 753–66. Kawakami, T. (1964) The theory of designing and testing ﬁshing nets in model. In: Modern Fishing Gear of the World, Vol. 2 (Ed. by H. Kristjansson), pp. 471–82. Fishing News Books, Oxford. Kelly, A. M. & Kohler, C. C. (1996) Sunshine bass performance in ponds, cages and indoor tanks. The Progressive Fish-Culturist, 58, 55–8. Kelly, L. A. (1992) Dissolved reactive phosphorus release from sediments beneath a freshwater cage aquaculture development in West Scotland. Hydrobiologia, 235/236, 569–72. Kelly, L. A. (1995) Predicting the effect of cages on nutrient status of Scottish freshwater lochs using mass-balance models. Aquaculture Research, 26, 469–78. Kelly, L. A. & Elberizon, I. R. (2001) Freshwater ﬁnﬁsh cage culture. In: Environmental Impacts of Aquaculture (Ed. by K. D. Black), pp. 32–50. Shefﬁeld Academic Press, Shefﬁeld. Kelly, M. S., Brodie, C. C. & Mckenzie, J. D. (1998) Somatic and gonadal growth of the sea urchin Psammechinus miliaris (Gmelin) maintained in polyculture with Atlantic salmon. Journal of Shellﬁsh Research, 17, 1557–62. Kennedy, C. M. (1994) An integrated approach to control systems and risk management practical observations. In: Measures for Success (Ed. by P. Kestemont, J. F. Muir, F. Sevila & P. Williot), pp. 37–41. CEMAGREF, Paris. Kennedy, R. & Kreiberg, H. (2001) Heterosigma akashiwo blooms: warning signs for cage sites. Canadian Technical Report on Fisheries and Aquatic Sciences, 2386, 108–11. Kent, M. L. (2000) Marine netpen farming leads to infections with some unusual parasites. International Journal of Parasitology, 30, 321–6.

334

References

Kent, M. L. & Poppe, T. T. (2002) Infectious diseases of coldwater ﬁsh in salt water. In: Diseases and Disorders of Finﬁsh in Cage Culture (Ed. by P. T. K. Woo, D. W. Bruno & L. H. S. Lim), pp. 61–105. CABI Publishing, Wallingford. Kent, M. L., Dawe, S. C., St Hilaire, S. & Andersen, R. J. (1996) Effects of feeding rate, seawater entry and exposure to natural biota on the severity of net-pen liver disease among pen-reared Atlantic salmon. The Progressive Fish-Culturist, 58, 43–6. Kerr, N. M. (1981) Design of equipment and selection of materials – an engineer’s assessment. Schriften der Bundesforschungsanstalt für Fischerei, 16/17, 151–81. Kerr, N. M., Gillespie, M. J., Hull, S. T. & Kingwell, S. (1980) The design construction and location of marine ﬂoating cages. In: Proceedings of the Institute of Fisheries Management Cage Fish Rearing Symposium, University of Reading, 26–27 March 1980, pp. 23–49. Janssen Services, London. Kerry, J., Hiney, M., Coyne, R., Cazabon, D., NicGabhainn, S. & Smith, P. (1994) Frequency and distribution of resistance to oxytetracycline in micro-organisms isolated from marine ﬁsh farm sediments following therapeutic use of oxytetracycline. Aquaculture, 123, 43–54. Kerry, J., Hiney, M., Coyne, R., Gilroy, D., Hiney, M. & Smith, P. (1996) Spatial distribution of oxytetracycline and elevated frequencies of oxytetracycline resistance in sediments beneath a marine salmon farm following oxytetracycline therapy. Aquaculture, 145, 31–9. Kery, S. (1996) Mooring systems common in most types of open ocean aquaculture. In: Open Ocean Aquaculture. Proceedings of an International Conference, Portland, Maine, 8–10 May 1996 (Ed. by M. Polk), pp. 297–325. Report No. UNHMP-CP-SG96-9. New Hampshire/Maine Sea Grant College Program, Portland, ME. Keshavanath, P., Beveridge, M. C. M., Baird, D. J., Lawton, L. A., Nimmo, E. A. & Codd, G. A. (1994) The functional grazing response of a phytoplanktivorous ﬁsh Oreochromis niloticus to mixtures of toxic and non-toxic strains of the cyanobacterium Microcystis aeruginosa. Journal of Fish Biology, 45, 123–9. Ketchum, B. H. (Ed.) (1983) Estuaries and Enclosed Seas. Ecosystems of the World, Vol. 26. Elsevier, Amsterdam. Ketola, H. G. (1975) Requirements of Atlantic salmon for dietary phosphorus. Transactions of the American Fisheries Society, 104, 548–51. Kibria, G., Nugegoda, D., Fairclough, R., Lam, P. & Bradley, A. (1999) Utilisation of wastewater-grown zooplankton: nutritional quality of zooplankton and performance of silver perch Bidyanus bidyanus (Mitchell 1938) (Teraponidae) fed on wastewatergrown zooplankton. Aquaculture Nutrition, 5, 2221–7. Kiekhafer, H. (1983) Fishzuchtin Gehegen. Paul Parey, Hamburg. Kilambi, R. V., Hoffmann, E. E., Brown, A. V., Adams, T. C. & Wickizer, W. A. (1976) Effects of Cage Culture Fish Production Upon Biotic and Abiotic Environment of Crystal Lake, Arkansas. Arkansas University, Fayetteville, AR. Kils, U. (1977) The salinity effect on aeration in mariculture. Meeresforschung, 25, 201–6. Kils, U. (1979) Oxygen regime and artiﬁcial aeration of net cages. Meeresforschung, 27, 236–43. Kim, C. S., Lee, S. G., Kim, H. G. & Jung, J. (1991) Reactive oxygen species as causative agents in the ichthyotoxicity of the red tide dinoﬂagellate Cochlidium polykrikoides. Journal of Plankton Research, 21, 2105–15. Kim, I.-B. (2000) Cage aquaculture in Korea. In: Cage Aquaculture in Asia. Proceedings of the First International Symposium on Cage Aquaculture in Asia, Tungkang, Taiwan,

References

335

2–6 November 1999 (Ed. by I. C. Liao & C. K. Lin), pp. 59–73. Asian Fisheries Society, Manila and World Aquaculture Society – Asia Branch, Bangkok. Kinne, O. (1976) Marine Ecology: A Comprehensive, Integral Treatise on Life in Oceans and Coastal Waters. Vol. 3: Cultivation. Wiley, New York. Kinne, O. (Ed.) (1980) Diseases of Marine Animals, Vol. 1: General Aspects, Protozoa to Gastropoda. Wiley, New York. Kirchner, W. B. & Dillon, P. (1975) An empirical method of estimating the retention of phosphorus in lakes. Water Resources Research, 11, 182–3. Kirk, J. T. O. (1994) Light and Photosynthesis in Aquatic Ecosystems. Cambridge University Press, Cambridge. Kite-Powell, H. L. & Hoagland, P. (2001) The economics of New England ﬁnﬁsh growout: aquaculture at an offshore site. Journal of Shellﬁsh Research, 20, 521. Klaver, A. L. & Matthews, R. A. (1994) Effects of oxytetracycline on nitriﬁcation in a model aquatic system. Aquaculture, 123, 237–47. Kleinholz, C. W. & Luker, G. W. (1994) Evaluation of an enclosure for bath treatment of caged ﬁsh. Journal of Applied Aquaculture, 4, 71–8. Klust, G. (1982) Netting Materials for Fishing Gear. Fishing News Books, Oxford. Klust, G. (1983) Fibre Ropes for Fishing Gear. Fishing News Books, Oxford. Kogan, R. F. & Romanycheva, O. D. (1979) The use of ﬂoating breakwaters in ﬁsh culture. In: Marine Fish Culture (Ed. by N. E. Sal’nikov), pp. 19–30. Transl. VNIRO. Pishchevaya Promyshlennost, Moscow. Koivisto, V. & Blomqvist, E. M. (1988) Does ﬁsh farming affect natural Baltic ﬁsh communities. Kieler Meeresforschung Sonderheim, 6, 301–11. Konikoff, M. & Lewis, W. M. (1974) Variations in weight of cage-reared channel catﬁsh. The Progressive Fish-Culturist, 36, 138–44. Korycka, A. & Zdanowski, B. (1980) Some aspects of the effects of cage ﬁsh culture on lakes, with special reference to heated lakes. Schriften der Bundesforschungsanstalt für Fischerei, 16/17, 131–8. Korzeniewski, K. & Korzeniewska, J. (1982) Changes in the composition and physiological properties of the bacterial ﬂora of water and bottom sediments in Lake Letowo, caused by intensive trout culture. Polish Archives of Hydrobiology, 29, 671–82. Kowalski, T. (Ed.) (1974) Floating Tyre Breakwater Conference Papers. University of Rhode Island Marine Technical Report 24. University of Rhode Island, Narrangansett, RI. Kowalski, T. (1976) Scrap tyre ﬂoating breakwaters for North Sea operations. Naval Architecture, 6, 186–7. Kramer, D. L. (1983) The evolutionary ecology of respiratory mode in ﬁshes: an analysis based on the costs of breathing. Environmental Biology of Fishes, 9, 145–58. Kreiberg, H. & Solmie, A. (1987) A production-scale towable net-pen for efﬁcient high volume transport of Paciﬁc herring: design and comparative performance. Aquacultural Engineering, 6, 289–99. Kryvi, H., Ibrekk, H. O. & Elvestad, S. (1991) LENKA, a method for a nation-wide analysis of the suitability of the Norwegian coast for aquaculture. Marine Pollution Bulletin, 23, 785–8. Kulikovsky, Z., Wohlfarth, G. W. & Avtalion, R. R. (1994) The association between weight and growth in tilapias during communal testing. II. Cage testing. Israeli Journal of Aquaculture, 46, 89–94.

336

References

Kumazawa, T., Kutsuwa, S., Ishizaki, M., Fuwa, S. & Imai, T. (1994) An analysis of drag coefﬁcient of plane nets made from twisted knotless nettings and Reynolds number. Nippon Suisan Gakkaishi, 60, 167–71. Kuo, C. & Beveridge, M. C. M. (1990) Mariculture: biological and management problems and possible engineering solutions. In: Engineering for Offshore Fish Farming, pp. 171–84. Thomas Telford, London. Kupka-Hansen, P., Pittman, K. & Ervik, A. (1991) Organic waste from marine ﬁsh farms – effects on the seabed. In: Marine Aquaculture and Environment (Ed. by. T. Mäkinen), pp. 105–19. Nordic Council of Ministers, Copenhagen. Kuwa, M. (1983) Corrosion and protection of ﬁsh culturing ﬂoating cage made of wire netting. Bulletin of the Japanese Society of Scientiﬁc Fisheries, 49, 165–75. Kuwa, M. (1984) Fouling organisms on ﬂoating cage of wire netting and the removal by Oplegnathus sp. cultured with other marine ﬁsh. Bulletin of the Japanese Society of Scientiﬁc Fisheries, 50, 1635–40. Lafont, R. & Savoeun, D. (1951) Notes sur la pisiculture au Cambodge. Cybium, 6, 54–61. Lai, H.-C., Kessler, A. O. & Khoo, L.-E. (1993) Biofouling and its possible modes of control at ﬁsh farms in Penang, Malaysia. Asian Fisheries Science, 6, 93–116. Lall, S. P. (1979) Minerals in ﬁnﬁsh nutrition. Schriften der Bundesforschungsanstalt für Fischerei, 14/15, 85–98. Landless, P. J. (1985) Aeration in ﬂoating cages. Fish Farmer, 8 (6), 12–14. Landless, P. J. & Edwards, A. (1976) Economical ways of assessing hydrography for ﬁsh farms. Aquaculture, 8, 29–43. Landless, P. J. & Edwards, A. (1977) Some simple methods for surveying a marine farm site. Fish Farming International, 4 (1), 32–4. Langford, T. E. (1983) Electricity Generation and the Ecology of Natural Waters. Liverpool University Press, Liverpool. Larsen, D. P. & Mercier, H. T. (1976) Phosphorus retention capacity of lakes. Journal of the Fisheries Research Board of Canada, 33, 1742–50. Laurén-Määttä, C., Granlid, M., Henriksson, S.-H. & Koivisto, V. (1991) Effects of ﬁsh farming on the macrobenthos of different bottom types. In: Marine Aquaculture and Environment (Ed. by T. Mäkinen), pp. 57–83. Nordic Council of Ministers, Copenhagen. Lawson, T. B. (1995) Fundamentals of Aquacultural Engineering. Chapman & Hall, New York. Lazur, A. M. (1996) The effects of periodic grading on production of channel catﬁsh cultured in cages. Journal of Applied Aquaculture, 6 (4), 17–24. Le Cren, E. D. & Lowe-McConnell, R. H. (Eds) (1980) The Functioning of Freshwater Ecosystems. Cambridge University Press, Cambridge. Leon, K. A. (1986) Effect of exercise on feed consumption, growth, food conversion and stamina of brood trout. The Progressive Fish-Culturist, 48, 43–6. Leong, T.-S. & Colorni, A. (2002) Infectious diseases of coldwater ﬁsh in salt water. In: Diseases and Disorders of Finﬁsh in Cage Culture (Ed. by P. T. K. Woo, D. W. Bruno & L. H. S. Lim), pp. 193–230. CABI Publishing, Wallingford, Oxon. Leong, T.-S. & Wong, S.-Y. (1988) A comparative study of the parasite fauna of wild and cultured grouper (Epinephelus malabaricus Bloch et Schneider) in Malaysia. Aquaculture, 68, 203–7. Leung, K. M. Y., Chu, J. C. W. & Wu, R. S. S. (1999) Nitrogen budgets for the aerolated grouper Epinephelus cultured under laboratory conditions and in open-sea cages. Marine Ecology Progress Series, 186, 271–81.

References

337

Lewis, A. G. & Metaxas, A. (1991) Concentrations of total dissolved copper in and near a copper-treated salmon netpen. Aquaculture, 99, 81–90. Lewis, W. M. & Morris, D. P. (1986) Toxicity of nitrite to ﬁsh: a review. Transactions of the American Fisheries Society, 115, 183–95. Li, S. F. (1994) Fish culture in cages and pens. In: Freshwater Fish Culture in China: Principles and Practice (Ed. by S. Li & J. Mathias), pp. 305–46. Elsevier, Amsterdam. Liao, I. C. & Lin, C. K. (Eds) (2000) Cage Aquaculture in Asia: Proceedings of the First International Symposium on Cage Aquaculture in Asia, Tungkang, Taiwan, 2–6 November 1999. Asian Fisheries Society, Manila and World Aquaculture Society – Asia Branch, Bangkok. Liao, P. B. (1971) Water requirements of salmonids. The Progressive Fish-Culturist, 33, 210–15. Libert, L., Maucorps, A. & Innes, L. (1987) Mending of Fishing Nets, 2nd edition. Fishing News Books, Oxford. Lien, E. (1993) Tension leg cage – a new net pen cage for ﬁsh farming. In: Fish Farming Technology (Ed. by H. Reinertsen, L. A. Dahle, L. Jørgensen & K. Tvinnereim), pp. 251–8. A. A. Balkema, Rotterdam. Lien, E. (2000) Offshore cage systems. In: Cage Aquaculture in Asia. Proceedings of the First International Symposium on Cage Aquaculture in Asia, Tungkang, Taiwan, 2–6 November 1999 (Ed. by I. C. Liao & C. K. Lin), pp. 141–9. Asian Fisheries Society, Manila and World Aquaculture Society – Asia Branch, Bangkok. Lien, E., Rudi, H., Slaatellid, O. H. & Kolberg, D. (1996) Flexible mooring with multiple buoys. In: Open Ocean Aquaculture. Proceedings of an International Conference, Portland, Maine, 8–10 May 1996 (Ed. by M. Polk), pp. 93–105. Report No. UNHMPCP-SG-96-9. New Hampshire/Maine Sea Grant College Program, Portland, ME. Lindbergh, J. M. (1979) The development of a commercial Paciﬁc salmon culture business. In: Advances in Aquaculture (Ed. by T. V. R. Pillay & W. A. Dill), pp. 441–7. Fishing News Books, Oxford. Linfoot, B. T. & Hall, M. S. (1987) Analysis of the motions of scale-model sea-cage systems. In: Modelling and Simulation in Aquaculture (Ed. by J. G. Balchen), pp. 31–46. Pergamon Press, Oxford. Linfoot, B. T., Cairns, J. & Poxton, M. G. (1990) Hydrodynamic and biological factors in the design of sea-cages for ﬁsh culture. In: Engineering for Offshore Fish Farming, pp. 197–210. Thomas Telford, London. Ling, S. W. (1977) Aquaculture in Southeast Asia: A Historical Review. University of Washington, Seattle, WA. Lio-Po, G. D. & Lim, L. H. S. (2002) Infectious diseases of warmwater ﬁsh in fresh water. In: Diseases and Disorders of Finﬁsh in Cage Culture (Ed. by P. T. K. Woo, D. W. Bruno & L. H. S. Lim), pp. 231–81. CABI Publishing, Wallingford. Lipton, A. P. (1984) Studies on the culture of Heteropneustes fossilis in cages. In: Proceedings of the National Seminar on Cage and Pen Culture, Fisheries College, Tamil Nadu Agricultural University, Tuticorin, 18–19 March 1984, pp. 51–4. Tamil Nadu Agricultural University, Tuticorin, India. Lisac, D. & Muir, J. (2000) Comparative economics of offshore and mariculture facilities. In: Mediterranean Offshore Mariculture (Ed. by J. F. Muir & B. Basurco), pp. 203–11. Options Méditerranéenes, Series B, No. 30. CIHEAM, Zaragoza. Lister, A. R. & Boon, A. G. (1973) Aeration in deep tanks: an evaluation of ﬁne-bubble diffused air system. Journal of the International Water Pollution Conference, 5, 3–18. Little, D. C. & Hulata, G. (2000) Strategies for tilapia seed production. In: Tilapias:

338

References

Biology and Exploitation (Ed. by M. C. M. Beveridge & B. J. McAndrew), pp. 267–326. Kluwer, Dordrecht. Little, D. C. & Muir, J. F. (1987) A Guide to Integrated Warmwater Aquaculture. Institute of Aquaculture, University of Stirling, Stirling. Liu, C. Y. (2000) Comparison of various aspects of marine cage farming between Norway and Taiwan. In: Cage Aquaculture in Asia. Proceedings of the First International Symposium on Cage Aquaculture in Asia, Tungkang, Taiwan, 2–6 November 1999 (Ed. by I. C. Liao & C. K. Lin), p. 293. Asian Fisheries Society, Manila and World Aquaculture Society – Asia Branch, Bangkok. Lloyd, R. (1992) Pollution and Freshwater Fish. Fishing News Books, Oxford. Løland, G. (1993a) Current forces on, and water ﬂow through and around, ﬂoating ﬁsh farms. Aquaculture International, 1, 72–89. Løland, G. (1993b) Water ﬂow through and around net pens. In: Fish Farming Technology (Ed. by H. Reinersten, L. A. Dahle, L. Jørgensen & K. Tvinnereim), pp. 177–83. A. A. Balkema, Rotterdam. Lombardi, J. V., Marques, H. L. A. & Barreto, O. J. S. (2001) Floating cages in open sea water: an alternative for promoting integrated aquaculture in Brazil. World Aquaculture, 32, 47, 49–50. Lovegrove, T. (1979) Control of fouling in farm cages. Fish Farming International, 6 (1), 33–7. Lovell, R. T. (1978) Dietary requirements of channel catﬁsh. Transactions of the American Fisheries Society, 107, 617–21. Lovell, T. (1989) Nutrition and Feeding of Fish. Van Nostrand Reinhold, New York. Loverich, G. F. & Croker, T. R. (1993) OCEAN SPAR® net pen system: 32 months of offshore operations. In: Fish Farming Technology (Ed. by H. Reinertsen, L. A. Dahle, L. Jørgensen & K. Tvinnereim), pp. 243–9. A. A. Balkema, Rotterdam. Loverich, G. F. & Swanson, K. T. (1993) Offshore sea farms: 25 months of experience. In: Techniques for Modern Aquaculture (Ed. by J.-K. Wang), pp. 241–53. American Society of Agricultural Engineers, St Joseph, MI. Lowell, F., Newton, S. H. & Torrans, E. L. (1982) Caged catﬁsh production in private farm ponds. Arkansas Farming Research, 31 (2), 14–15. Loyacano, H. A. & Smith, D. C. (1976) Attraction of native ﬁshes to catﬁsh culture cages in reservoirs. In: Proceedings of the 29th Annual Conference of the Southeast Association of Game Fish Commissioners, 2–4 March 1976, Columbia, SC, pp. 63–73. South Carolina Agricultural Experimental Station, Columbia, SC. Lumb, C. M. (1989) Self-pollution by Scottish ﬁsh-farms? Marine Pollution Bulletin, 20, 375–9. Lunestad, B. T. (1992) Fate and effects of antibacterial agents in aquatic environments. In: Problems of Chemotherapy in Aquaculture: From Theory to Reality (Ed. by C. M. Michel & D. J. Alderman), pp. 152–61. Ofﬁce International des Epizooties, Paris. McAllister, P. E. & Owens, W. J. (1992) Recovery of infectious pancreatic necrosis virus from the faeces of wild piscivorous birds. Aquaculture, 106, 227–32. McAndrew, K. I., Little, D. & Beveridge, M. C. M. (2000) Entry points and low risk strategies appropriate for the resources poor to participate in cage aquaculture: experiences from CARE-CAGES project, Bangladesh. In: Cage Aquaculture in Asia. Proceedings of the First International Symposium on Cage Aquaculture in Asia, Tungkang, Taiwan, 2–6 November 1999 (Ed. by I. C. Liao & C. K. Lin), pp. 225–31. Asian Fisheries Society, Manila and World Aquaculture Society – Asia Branch, Bangkok.

References

339

McAndrew, K. I., Brugere, C., Beveridge, M. C. M., Ireland, M. J., Roy, T. K. & Yesmin, K. (2002) Improved management of small-scale tropical cage culture systems in Bangladesh: potential beneﬁts of an alliance between an NGO and a western research institute. In: Rural Aquaculture (Ed. by P. Edwards, D. C. Little & H. Demaine), pp. 129–41. CABI, Cambridge. McCaig, A. E., Phillips, C. J., Stephen, J. R., Kowalchuk, J. R., Harvey, S. M., Herbert, R. A., Embley, T. M. & Prosser, J. I. (1999) Nitrogen cycling and community structure of proteobacterial beta-sub groups ammonia-oxidising bacteria within polluted marine ﬁsh farm sediments. Applied and Environmental Microbiology, 65, 213–20. McCartney, T. H. (1969) The effect of dietary inorganic phosphate and Vitamin D supplementation on liver glycogen of ﬁngerling brown trout. Fisheries Research Bulletin of the New York State Conservation Department, 31, 5–7. McCave, I. N. (1979) Suspended sediment. In: Estuarine Hydrography and Sedimentation (Ed. by K. R. Dyer), pp. 131–85. Cambridge University Press, Cambridge. McCoy, H. D., II (2000) American and International Aquaculture Law. Blackwell, Oxford. McDonald, D. G. (1983) The effect of H+ upon the gills of freshwater ﬁsh. Canadian Journal of Zoology, 61, 691–703. McGhie, T. K., Crawford, C. M., Mitchell, I. M. & O’Brien, D. (2000) The degradation of ﬁsh-cage waste in sediments during fallowing. Aquaculture, 187, 351–66. McGinty, A. S. (1991) Tilapia production in cages: effects of cage size and number of non-caged ﬁsh. The Progressive Fish-Culturist, 53, 246–9. McGregor, R. C. (1978) Design of scrap tyre ﬂoating breakwaters with special reference to ﬁsh farms. Proceedings of the Royal Society of Edinburgh, 76B, 115– 33. McGregor, R. C. & Gilbert, C. H. G. (1983) Floating tyre breakwaters: a case history. In: Proceedings of the 18th International Conference on Coastal Engineering, Capetown, South Africa, November 1982 (Ed. by B. L. Edge), pp. 183–91. American Society of Civil Engineers, Reston, VA. McGuigan, J. B. & Sommerville, C. (1985) Studies on the effects of cage culture of ﬁsh on the parasite fauna in a lowland freshwater loch in the west of Scotland. Zeitschrift für Parisitenkunde, 71, 673–82. Mackereth, F. J. H., Heron, J. & Talling, J. F. (1978) Water Analysis: Some Revised Methods for Limnologists. Freshwater Biological Association, Windermere. McKibben, M. A. & Hay, D. W. (in press) Distributions of planktonic sea lice larvae Lepeophtheirus salmonis in the inter-tidal zone in Loch Torridon, Wester Ross in relation to ﬁsh farm production cycles. Aquaculture Research. McLarney, W. (1984) The Freshwater Aquaculture Book. Hartley & Marks, Washington, DC. McVey, J. P. (1991) Finﬁsh Aquaculture. Handbook of Mariculture, Vol. II. CRC Press, Boca Raton, FL. McVey, J. P., Stickney, R. R., Yarish, C. & Chopin, T. (2002) Aquatic polyculture and balanced ecosystem management: new paradigms for seafood production. In: Responsible Marine Aquaculture (Ed. by R. R. Stickney & J. P. McVey), pp. 91–104. CAB International, Cambridge. McVicar, A. H. (1997) Disease and parasite implications of the coexistence of wild and cultured salmon populations. ICES Journal of Marine Science, 54, 998– 1008.

340

References

Mahnken, C. V. W. (1975) Status of commercial net pen farming of Paciﬁc salmon in Puget Sound. Proceedings of the Annual Meeting of the World Mariculture Society, 6, 285–98. Mäkinen, T. (1991) Nutrient load from marine aquaculture. In: Marine Aquaculture and Environment (Ed. by T. Mäkinen), pp. 1–8. Nordic Council of Ministers, Copenhagen. Mäkinen, T. (1993) Fish farm efﬂuents and their control in Finland. In: Workshop on Fish Farm Efﬂuents and Their Controls in EC Countries (Ed. by H. Rosenthal, V. Hilge & A. Kamstra), pp. 81–7. University of Kiel, Germany. Mamcarz, A. & Kozlowski, J. (1989) An automatic dry food dispenser for coregonid larvae reared in illuminated cages. Aquacultural Engineering, 8, 73–7. Mamcarz, A. & Nowak, M. (1987) New version of an illuminated cage for coregonid rearing. Aquaculture, 65, 183–8. Mamcarz, A. & Wornial´´l o, E. (1986) Effects of diatom blooms on the gills of whiteﬁsh (Coregonus lavaretus L.) reared in illuminated cages. Aquaculture, 53, 1–5. Manning, R. (1891) On the ﬂow of water in open channels and pipes. Transactions of the Institute of Civil Engineers of Ireland, 20, 161–207. Marcello, R. A., Jr & Strawn, K. (1973) Cage culture of some marine ﬁshes in the intake and discharge canals of a steam-electric generating station, Galveston Bay, Texas. Proceedings of the Annual Meeting of the World Mariculture Society, 4, 97–112. Marte, C. L. (1988) An improved method for collecting naturally spawned milkﬁsh eggs from ﬂoating cages. Aquaculture, 71, 387–92. Marte, C. L., Cruz, P. & Flores, E. E. E. (2000) Recent developments in freshwater and marine cage aquaculture in the Philippines. In: Cage Aquaculture in Asia. Proceedings of the First International Symposium on Cage Aquaculture in Asia, Tungkang, Taiwan, 2–6 November 1999 (Ed. by I. C. Liao & C. K. Lin), pp. 83–96. Asian Fisheries Society, Manila and World Aquaculture Society – Asia Branch, Bangkok. Martin, R. M. & Heard, W. R. (1987) Floating vertical raceway to culture salmon (Oncorhynchus spp.). Aquaculture, 61, 295–302. Martinez Cordero, F. J., Beveridge, M. C. M., Muir, J. F., Mitchell, D. & Gillespie, M. (1994) A note on the behaviour of adult Atlantic halibut, Hippoglossus hippoglossus (L.), in cages. Aquaculture and Fisheries Management, 25, 475–81. Martyshev, F. G. (1983) Pond Fisheries. A. A. Balkema, Rotterdam. Maruyama, T. & Ishida, R. (1976) Effect of water depth in net cages on the growth and body shape of Tilapia mossambica. Bulletin of the Freshwater Fisheries Research Laboratory of Tokyo, 26, 11–19. Maruyama, T. & Ishida, R. (1977) Effects of water depth on the growth and body shape of common carp raised in net cages. Bulletin of the Freshwater Fisheries Research Laboratory of Tokyo, 27, 95–102. Mason, C. (2001) Biology of Freshwater Pollution. Prentice Hall, New York. Mate, B. R. & Harvey, J. T. (Eds) (1986) Acoustical Deterrents in Marine Mammal Conﬂicts with Fisheries. Oregon State University, Corvallis, OR. Matheson, A. (1979) An Investigation of the Occurrence and Control of Pleurocercoids of Diphyllobothium spp. in Farmed Atlantic Salmon (Salmo salar). MSc thesis, University of Stirling, Stirling. Mayer, I. & McLean, E. (1995) Bioengineering and biotechnical strategies for reduced waste aquaculture. Water Science and Technology, 31, 85–102. Mazhari, K. & Petrell, R. J. (1991) Integrated salmon and kelp production. In: Engineering Aspects of Aquaculture, pp. 102–17. Northeast Regional Agricultural Engineering Service, Ithaca, NY.

References

341

Mazzola, A. & Sarà, G. (2001) The effect of ﬁsh farming organic waste on food availability for bivalve molluscs (Gaeta Gulf, Central Tyrrhenian, MED): stable carbon isotopic analysis. Aquaculture, 192, 361–79. Meaden, G. J. & Kapetsky, J. M. (1991) Geographical Information Systems and Remote Sensing in Inland Fisheries and Aquaculture. FAO Fisheries Technical Paper 318. FAO, Rome. Melack, J. M. (1976) Primary productivity and ﬁsh yields in tropical lakes. Transactions of the American Fisheries Society, 105, 575–80. Meng, L., Gray, C., Taplin, B. & Kupcha, E. (2000) Using winter ﬂounder growth rates to assess habitat quality in Rhode Island’s coastal lagoons. Marine Ecology Progress Series, 201, 287–99. Merican, Z. O. & Phillips, M. J. (1985) Solid waste production from rainbow trout, Salmo gairdneri Richardson, cage culture. Aquaculture and Fisheries Management, 16, 55–70. Meske, C. H. & Manthey, M. (1983) Sarotherodon niloticum – tropical cichlids as food ﬁsh. International Fischwirt, 30, 30–4. Meyer, J. (1980) Herons at ﬁsh farms. In: Proceedings of the Institute of Fisheries Management 11th Annual Study Course, University of Sussex, 23–25 September 1980, pp. 58–63. Janssen Services, London. Michel, C. M. & Alderman, D. J. (Eds) (1992) Problems of Chemotherapy in Aquaculture: From Theory to Reality. Ofﬁce International des Epizooties, Paris. Mikheyev, P. V., Meysner, Y. V. & Mikheyev, V. P. (1970) The importance of contact with the air in the life of the rainbow trout. Journal of Ichthyology, 10, 691–3. Millie, D. F., Schoﬁeld, O. M., Diogigi, C. P. & Johnsen, P. B. (1995) Assessing noxious phytoplankton in aquaculture systems using bio-optical methodologies: a review. Journal of the World Aquaculture Society, 26, 329–45. Mills, D. (1965) The distribution and food of the cormorant in Scottish inland waters. Freshwater and Salmon Fisheries Research, 35, 31–7. Mills, D. (1979) Bird predation – current views. In: Proceedings of the Institute of Fisheries Management, 10th Annual Study Course, Nottingham University, 18–20 September 1979, pp. 264–71. Janssen Services, London. Milne, P. H. (1970) Fish Farming: A Guide to the Design and Construction of Net Enclosures. HMSO, Edinburgh. Milne, P. H. (1972) Fish and Shellﬁsh Farming in Coastal Waters. Fishing News Books, Oxford. Milne, P. H. (1974) A visit to Japan’s ﬁsh farming industry. Fish Farming International, 1 (2), 38–55. Milne, P. H. (1976) Engineering and the economics of aquaculture. Journal of the Fisheries Research Board of Canada, 33, 288–98. Milne, P. H. (1979) Selection of sites and design of cages, ﬁsh pens and enclosures for aquaculture. In: Advances in Aquaculture (Ed. by T. V. R. Pillay & W. A. Dill), pp. 416–23. Fishing News Books, Oxford. Miranda, C. D. & Zemelman, R. (2002) Bacterial resistance to oxytetracycline in Chilean salmon farming. Aquaculture, 212, 31–47. Møller, D. (1979) Recent developments in cage and enclosure aquaculture in Norway. In: Advances in Aquaculture (Ed. by T. V. R. Pillay & W. A. Dill), pp. 447–53. Fishing News Books, Oxford. Moon, T. S., Choi, H. S. & Park, S. Y. (1993) Study on rearing density and cutting teeth effects of culture of tiger puffer, Takifufgu rubripes, in marine ﬂoating net cage culture.

342

References

Bulletin of the National Fisheries Research and Development Agency Korea, 47, 25–44. Moore, R. E. (1982) Toxins from marine blue-green algae. In: The Water Environment (Ed. by W. W. Carmichael), pp. 15–24. Plenum Press, New York. Moring, J. R. (1982) Fin erosion and culture-related injuries of chinook salmon raised in ﬂoating net pens. The Progressive Fish-Culturist, 44, 189–91. Moring, J. R. & Moring, K. A. (1975) Succession of net biofouling material and its role in the diet of pen-cultured chinook salmon. The Progressive Fish-Culturist, 37, 27–30. Mortensen, A. (1985) Massive ﬁsh mortalities in Tjaldavik cove on the Faroe Islands, on 10–20 July, 1984. In: Proceedings of the Third International Conference on Toxic Dinoﬂagellates, St Andrews, New Brunswick, 8–12 June 1985 (Ed. by D. M. Anderson, A. W. White & D. G. Baden), pp. 121–34. Elsevier, New York. Moyle, P. B. & Light, T. L. (1996) Biological invasions of freshwater: empirical rules and assembly theory. Biological Conservation, 78, 149–61. Muir, J. F. & Basurco, B. (Eds) (2000) Mediterranean Offshore Mariculture. Options Méditerranéennes, Series B, No. 30. CIHEAM, Zaragoza. Muir, J. F. & Kapetsky, J. M. (1988) Site selection decision and project cost: the case of brackish water pond systems. In: Aquaculture Engineering Technologies for the Future. Institute of Chemical Engineers Symposium Series, Vol. 111, pp. 45–63. Institute of Chemical Engineers, Rugby. Muir Wood, A. M. & Fleming, C. A. (1981) Coastal Hydraulics, 2nd edition. Macmillan, London. Muller, F. & Varadi, L. (1980) The results of cage ﬁsh culture in Hungary. Aquaculture Hungarica, 2, 154–67. Muller-Fuega, A., Petit, J. & Sabaut, T. J. (1971) The inﬂuence of temperature and wet weight on the oxygen demand of rainbow trout (Salmo gairdneri R.) in freshwater. Aquaculture, 14, 355–63. Müller-Haekel, A. (1986) Control of water quality around a cage ﬁsh farm in the Norrby Archipelago (Northern Bothnian sea). Vatten, 42, 205–9. Murugesan, V. K. & Parameswaran, S. (1984) Culture of air-breathing ﬁshes in cages and pens. In: Proceedings of the National Seminar on Cage and Pen Culture, Fisheries College, Tamil Nadu Agricultural University, Tuticorin, 18–19 March 1984, pp. 59–62. Tamil Nadu Agricultural University, Tuticorin, India. Myrseth, B. (2000) Automation of feeding management in cage culture. In: Cage Aquaculture in Asia. Proceedings of the First International Symposium on Cage Aquaculture in Asia, Tungkang, Taiwan, 2–6 November 1999 (Ed. by I. C. Liao & C. K. Lin), pp. 151–5. Asian Fisheries Society, Manila and World Aquaculture Society – Asia Branch, Bangkok. Naiberg, A., Petrell, R. J., Savage, C. R. & Neufeld, T. P. (1993) A non-invasive ﬁsh size assessment method for tanks and sea cages using stereo video. In: Techniques for Modern Aquaculture (Ed. by J.-K. Wang), pp. 372–81. American Society of Agricultural Engineers, St Joseph, MI. Nakada, M. & Murai, T. (1991) Yellowtail aquaculture in Japan. In: Finﬁsh Aquaculture. Handbook of Mariculture, Vol. II (Ed. by J. P. McVey), pp. 55–72. CRC Press, Boca Raton, FL. Nakao, S., Shazili, N. A., Md. & Salleh, H. U. (1989) Benthic communities in the areas under and around the ﬁsh-culture rafts at the Kuala Trengganu River estuary, Malaysia. Bulletin of the Faculty of Fisheries of Hokkaido University, 40, 154–8.

References

343

Nash, C. E., Iwamoto, R. N. & Mahnken, C. V. W. (2000) Aquaculture risk management and marine mammal interactions in the Paciﬁc Northwest. Aquaculture, 183, 307–23. Natarajan, P., Ramanthanam, N. & Venkataramani, V. K. (1984) Cage and pen culture of carps and prawns. In: Proceedings of the National Seminar on Cage and Pen Culture, Fisheries College, Tamil Nadu Agricultural University, Tuticorin. 18–19 March 1984, pp. 33–6. Tamil Nadu Agricultural University, Tuticorin, India. Nath, S. S., Bolte, J. P., Ross, L. G. & Aguilar-Manjarrez, J. (2000) Applications of geographical information systems (GIS) for spatial decision support in aquaculture. Aquacultural Engineering, 23, 233–78. Naylor, R. L., Goldburg, R. J., Mooney, H., Beveridge, M. C. M., Clay, J., Folke, C., Kautsky, N., Lubchenko, J., Primavera, J. & Williams, M. (1998) Nature’s subsidies to shrimp and salmon farming. Science, 282, 883–4. Naylor, R. L., Goldburg, R. J., Primavera, J., Kautsky, N., Beveridge, M. C. M., Clay, J., Folke, C., Lubchenko, J., Mooney, H. & Troell, M. (2000) Effects of aquaculture on world ﬁsh supplies. Nature, 405, 1017–24. NCC (1989) Fish Farming and the Safeguard of the Marine Environment of Scotland. Report prepared for the Nature Conservancy Council by the Institute of Aquaculture, University of Stirling. NCC, Edinburgh. NCC (1990) Fish Farming and the Scottish Freshwater Environment. Report prepared for the Nature Conservancy Council by the Institute of Aquaculture, University of Stirling, Institute of Freshwater Ecology, Penicuik and the Institute of Terrestial Ecology, Banchory. NCC, Edinburgh. Nelson, R. G., Duarte, S. A. & Masser, M. P. (2001a) Financial risk analysis of three aeration regimes in catﬁsh cage culture. Aquaculture Economics and Management, 5, 171–8. Nelson, S. G., Glenn, E. P., Conn, J., Moore, D., Walsh, T. & Akutagawa, M. (2001b) Cultivation of Gracilaria parvispora (Rhodophyta) in shrimp-farm efﬂuent ditches and ﬂoating cages in Hawaii: a two phase polyculture system. Aquaculture, 193, 239–48. New, M. B., Tacon, A. G. J. & Csavas, I. (Eds) (1993) Farm-Made Aquafeeds. Proceedings of the FAO/AADCP Regional Expert Consultation on Farm-Made Aquafeeds, 14–18 December 1992, Bangkok, Thailand. FAO-RAPA/AADCP, Bangkok. Newton, S. H. (1980) Review of cage culture activity indicates continuing interest. Aquaculture Magazine, 7 (1), 32–6. Nishimura, A. (1982) Effect of organic matter produced in ﬁsh farms on the growth of red tide algae Gymnodinium type-65 and Chattonella antiqua. Bulletin of the Plankton Society of Japan, 29, 1–7. Nishimura, A. (1983) The process of decomposition and sedimentation of deposits in the pearl oyster and yellowtail culture grounds. Bulletin of the Japanese Society of Oceanography, 44, 44–50. Noble, R. (1975) Growing ﬁsh in sewage. New Scientist, 67, 259–61. Norberg, J. (1999) Periphyton fouling as a marginal energy source in tropical tilapia cage farming. Aquaculture Research, 30, 427–30. Nordvarg, L. & Johansson, T. (2002) The effects of ﬁsh farm efﬂuents on the water quality in the Åland archipelago, Baltic Sea. Aquacultural Engineering, 25, 253–80. Nose, T. (1985) Recent advances in aquaculture in Japan. Geojournal, 10, 261–76. Nose, T. & Arai, S. (1979) Recent advances in studies on mineral nutrition of ﬁsh in Japan. In: Advances in Aquaculture (Ed. by T. V. R. Pillay & W. A. Dill), pp. 584–9. Fishing News Books, Oxford.

344

References

NRC (1993) Nutrient Requirements of Fish. National Academy Press, Washington, DC. O’Connor, B., Costelloe, J., Dineen, P. & Faull, J. (1994) The effect of harrowing and fallowing on sediment quality under a salmon farm on the west coast of Ireland. In: Measures for Success (Ed. by P. Kestemont, J. F. Muir, F. Sevilla & P. Williot), pp. 129–32. CEMAGREF, Paris. OECD (1982) Eutrophication of Waters. Monitoring, Assessment and Control. OECD, Paris. Ogino, C. & Takeda, T. (1976) Mineral requirements in ﬁsh. 3. Calcium and phosphorus. Bulletin of the Japanese Society of Scientiﬁc Fisheries, 42, 793–9. Ogino, C. & Takeda, T. (1978) Requirements of rainbow trout for dietary calcium and phosphorus. Bulletin of the Japanese Society of Scientiﬁc Fisheries, 44, 1019–22. Ojeda, G. M. & Strawn, K. (1980) Comparison of wire cages and net cages for the culture of black drum (Pogonius cromis). Proceedings of the Annual Meeting of the World Mariculture Society, 11, 185–91. Okaichi, T. (1989) Red tide problems in the Seto Inland Sea, Japan. In: Red Tides, Biology, Environmental Science and Toxicology (Ed. by T. Okaichi, D. M. Anderson & T. Nemoto), pp. 137–42. Elsevier, New York. Okaichi, T., Anderson, D. M. & Nemoto, T. (Eds) (1989) Red Tides, Biology, Environmental Science and Toxicology. Elsevier, New York. Oltedal, G., Lien, E. & Aarsnes, J. V. (1988) Simulation of ﬁsh cage response to wave and currents. In: Aquaculture Engineering Technologies for the Future. Institute of Chemical Engineers Symposium Series, Vol. 111, pp. 123–32. Institute of Chemical Engineers, Rugby. Oppedal, F., Juell, J., Taranger, G. L. & Hansen, T. (2001) Artiﬁcial light and season affects vertical distribution and swimming behaviour of post-smolt Atlantic salmon in sea cages. Journal of Fish Biology, 58, 1570–84. Ormberg, H. & Slaattelid, O. H. (1993) Response analysis of ﬂoating ﬁsh farms: construction and veriﬁcation. In: Fish Farming Technology (Ed. by H. Reinertsen, L. A. Dahle, L. Jørgensen & K. Tvinnereim), pp. 185–92. A. A. Balkema, Rotterdam. Ostrofsky, M. L. (1978) Modiﬁcation of phosphorus retention models for use with lakes with low areal loading. Journal of the Fisheries Research Board of Canada, 35, 1532–6. Ottera, H., Roth, B. & Torrissen, O. J. (2001) Do killing methods affect the quality of Atlantic salmon. In: Farmed Fish Quality (Ed. by S. C. Kestin & P. D. Warriss), pp. 398–9. Blackwell, Oxford. Paclibare, J. O., Evelyn, T. P. T., Albright, L. J. & Prosperi-Porta, L. (1994) Clearing of the kidney disease bacterium Renibacterium salmoninarum from seawater by the blue mussel Mytilus edulis, and the status of the mussel as a reservoir of the bacterium. Diseases of Aquatic Organisms, 18, 129–33. Paetsch, B. (1977) Untersuchungen uber die ﬁschhalterung von karpfen in netzkaeﬁgen. Zeitschrift für Binnenﬁscherei, 24, 364–5. Page, F. & Robinson, S. (1992) Salmon farming in the Bay of Fundy: a chilling reminder. World Aquaculture, 23 (4), 31–4. Panchang, V. G., Cheng, G. & Newell, C. R. (1993) Application of Mathematical Models in the Environmental Regulation of Net-Pen Aquaculture. Final Report submitted to the National Marine Fisheries Service. NOAA, Washington, DC. Pankhurst, R. C. & Holder, D. W. (1952) Wind-Tunnel Techniques. Pitmans, London. Pantastico, J. B. & Baldia, J. P. (1981) An assessment of algal growth in net cages in Laguna Lake. Fisheries Research Journal of the Philippines, 6, 19–25.

References

345

Pantulu, V. R. (1979) Floating cage culture of ﬁsh in the lower Mekong River Basin. In: Advances in Aquaculture (Ed. by T. V. R. Pillay & W. A. Dill), pp. 423–7. Fishing News Books, Oxford. Parker, M. (Ed.) (1981) Red Tides. Fisheries Research Centre, Dublin. Parsons, G. J., Shumway, S. E., Kuenstner, S. & Gryska, A. (1999) Polyculture of sea scallops (Placopecten magellanicus) suspended from salmon cages. In: Book of Abstracts from the 12th International Pectinid Workshop, Bergen (Norway), 5–11 May 1999, p. 73. Institute of Marine Research, Bergen. Parsons, T. R., Maita, Y. & Lalli, C. M. (1984) A Manual of Chemical and Biological Methods for Seawater Analysis. Pergamon Press, Oxford. Pauly, D., Christensen, V., Dalsgaard, J., Froese, R. & Torres, F., Jr (1998) Fishing down marine food webs. Science, 279, 860–3. Pawar, V., Matsuda, O. & Fujisaki, N. (2002) Relationship between feed input and sediment quality of the ﬁsh cage farms. Fisheries Science, 68, 894–903. Pearson, T. H. & Black, K. D. (2001) The environmental impacts of marine ﬁsh cage culture In: Environmental Impacts of Aquaculture (Ed. by K. D. Black), pp. 1–30. Shefﬁeld Academic Press, Shefﬁeld. Pemberton, D. & Shaughnessy, P. D. (1993) Interaction between seals and marine ﬁshfarms in Tasmania, and management of the problem. Aquatic Conservation, 3, 149–58. Peña, R. & Prospero, O. (1984) Floating nursery cage for higher sugpo surival. Asian Aquaculture, 6 (3), 6–8. Penczak, T., Galicka, W., Molinski, M., Kusto, E. & Zalewski, M. (1982) The enrichment of a mesotrophic lake by carbon, phosphorus and nitrogen from the cage aquaculture of rainbow trout Salmo gairdneri. Journal of Applied Ecology, 19, 371–93. Pérez, O. M., Telfer, T. C., Beveridge, M. C. M. & Ross, L. G. (2002) Geographical Information Systems (GIS) as a simple tool to aid modeling of particulate waste distribution at marine ﬁsh cage sites. Estuarine, Coastal and Shelf Science, 54, 761–8. Pérez, O. M., Ross, L. G., Telfer, T. C. & Campo Barquin, L. M. del (2003a) Water quality requirements for marine ﬁsh cage site selection in Tenerife (Canary Islands): predictive modeling and analysis using GIS. Aquaculture, 224, 51–68. Pérez, O. M., Telfer, T. C. & Ross, L. G. (2003b) Use of GIS-based models for integrating and developing marine ﬁsh cages within the tourism industry in Tenerife (Canary Islands). Coastal Management, 31, 355–66. Persson, P. E. (1980) Muddy odour in ﬁsh from hypereutrophic waters. Developments in Hydrobiology, 2, 203–8. Petrell, R. J. & Alie, S. Y. (1996) Integrated cultivation of salmonids and seaweeds in open systems. Hydrobiologia, 326–7, 67–73. Petrell, R. J. & Ang, K. P. (2001) Effects of pellet contrast and light intensity on salmonid feeding behaviours. Aquacultural Engineering, 25, 175–86. Petrell, R. J., Neufeld, T. P. & Savage, C. R. (1993) A video method for noninvasively counting ﬁsh in sea cages. In: Techniques for Modern Aquaculture (Ed. J.-K. Wang), pp. 352–61. American Society of Agricultural Engineers, St Joseph, MI. Petrusevics, P. M. & Clarke, S. (1998) Optimisation of southern blueﬁn tuna resource in semi-open ocean farming, Boston Bay, South Australia, using numerical simulation. In: Charting the Future of Ocean Farming. Proceedings of the International Conference on Open Ocean Aquaculture ‘97, Maui, HI (USA), 23–25 April 1997, pp. 98–108. Hawaii University, Manoa, HI. Pettersson, K. (1988) The mobility of phosphorus in ﬁsh-food and fecals. Proceedings of the International Association for Theoretical and Applied Limnology, 23, 200–206.

346

References

Phillips, M. J., Beveridge, M. C. M. & Ross, L. G. (1985a) The environmental impact of salmonid cage culture on inland ﬁsheries: present status and future trends. Journal of Fish Biology, 27 (Suppl. A), 123–37. Phillips, M. J., Roberts, R. J., Stewart, J. A. & Codd, G. A. (1985b) The toxicity of the cyanobacterium Microcystis aeruginosa to rainbow trout Salmo gairdneri Richardson. Journal of Fish Diseases, 8, 339–44. Phillips, M. J., Clarke, R. & Mowat, A. (1993) Phosphorus leaching from Atlantic salmon diets. Aquacultural Engineering, 12, 47–54. Piccinetti, C., Piccinetti, W. & Manfrin, G. (1980) The reproduction of blue ﬁn tuna (Thunnus thynnus L.). Memoirs in Biology and Marine Oceanography, 10 (6), 255–8. Pickering, A. D. (Ed.) (1981) Stress and Fish. Academic Press, London. Pickering, A. D. (1993) Husbandry and stress. In: Recent Advances in Aquaculture, Vol. 4 (Ed. by J. F Muir & R. J. Roberts), pp. 155–69. Blackwell, Oxford. Pickering, A. D. (1998) Stress response of farmed ﬁsh. In: Biology of Farmed Fish (Ed. by K. D. Black & A. D. Pickering), pp. 222–55. Shefﬁeld Academic Press, Shefﬁeld. Pike, I. H. & Barlow, S. M. (2003) Impact of ﬁsh farming on ﬁsh stocks. Fish Farmer January/February, 14–16. Pingree, R. D., Holligan, P. M. & Mardell, G. T. (1978) The effects of vertical stability on phytoplankton distribution in the summer on the northwest European shelf. DeepSea Research, 25, 1011–28. Piper, R. G., McElwain, I. B., Orme, L. E., McCracken, J. P., Fowler, L. G. & Leonard, J. R. (1983) Fish Hatchery Management. US Department of the Interior Fish and Wildlife Service, Washington, DC. Pitcher, T. J. & Hart, P. J. B. (1982) Fisheries Ecology. Croom Helm, London. Pitt, R., Tsur, O. & Gordin, H. (1977) Cage culture of Sparus aurata. Aquaculture, 11, 285–96. Pitta, P., Karakssis, I., Tsapakis, M. & Zivanovic, S. (1999) Natural vs maricultureinduced variability in nutrients and plankton in the eastern Mediterranean. Hydrobiologia, 391, 181–94. Pond, S. & Pickard, G. L. (1983) Introductory Dynamic Oceanography, 2nd edition. Pergamon Press, Oxford. Poe, W. E. & Wilson, R. P. (1989) Palatability of diets containing sulfadimethoxine, ormetoprim and Romet 30 to channel catﬁsh ﬁngerlings. The Progressive FishCulturist, 51, 226–8. Polk, M. (Ed.) (1996) Open Ocean Aquaculture. Proceedings of an International Conference, Portland, Maine, 8–10 May, 1996. Report No. UNHMP-CP-SG-96-9. New Hampshire/Maine Sea Grant College Program, Portland, ME. Porter, C. (1981) Cage culture of gilthead bream (Sparus aurata) at an exposed site in the Red Sea. Special Publication of the European Mariculture Society, 6, 15–24. Postma, H. (1967) Sediment transport and sedimentation in the estuarine environment. In: Estuaries (Ed. by G. H. Lauff), pp. 158–79. American Association for the Advancement of Science, Washington, DC. Powell, M. R. (1976) The resistance of copper-nickel expanded metal to fouling and corrosion in mariculture operations. The Progressive Fish-Culturist, 38, 58–9. Poxton, M. (1996) Water Quality for Fish Culture. Kluwer, Amsterdam. PPD (1986) Manual on Floating Netcage Fish Farming in Singapore’s Coastal Waters. Fisheries Handbook, Primary Production Department, Singapore. Pritchard, D. W. (1967) Observations of circulation in coastal plain estuaries. In: Estuaries (Ed. by G. H. Lauff), pp. 37–44. American Association for the Advancement of Science, Washington, DC.

References

347

Privol’nev, T. I. (1975) The prospects for raising rainbow trout in cages. Translation Series of Fish and Marine Services, Canada, 3577. Pullin, R. S. V. & Lowe-McConnell, R. H. (Eds) (1982) The Biology and Culture of Tilapias. ICLARM, Manila. Pullin, R. S. V., Rosenthal, H. & Maclean, J. L. (Eds) (1993) Environment and Aquaculture in Developing Countries. ICLARM Conference Proceedings 31. ICLARM, Manila. Quick, N. J., Middlemas, S. J. & Armstrong, J. D. (2004) A survey of anti-predator controls at marine salmon farms in Scotland. Aquaculture, 230, 169–80. Raa, J. & Gildberg, A. (1982) Fish silage: a review. CRC Critical Review Series in Food Science and Nutrition, 16, 383–420. Raa, J., Gildberg, A. & Strom, T. (1983) Silage production – theory and practice. In: Upgrading Waste for Feeds and Food (Ed. by D. A. Ledward, A. J. Taylor & R. A. Lawrie), pp. 117–32. Butterworths, London. Ranson, K. & Beveridge, M. C. M. (1983) Raiders from the skies. Fish Farmer, 6 (1), 22–3. Ravindran, K. (1983) Copper-nickel alloys for marine aquacultural engineering. Fisheries Technology Society, Cochin, 20, 90–96. Rawson, M. V., Chen, C., Ji, R., Zhu, M., Wang, D., Wang, L., Yarish, C., Sullivan, J. B., Chopin, T. & Carmona, R. (2002) Understanding the interaction of extractive and fed aquaculture using ecosystem modeling. In: Responsible Marine Aquaculture (Ed. by R. R. Stickney & J. P. McVey), pp. 263–96. CAB International, Cambridge. Reay, P. (1979) Aquaculture. Edward Arnold, London. Reckow, K. H. (1983) A method for the reduction of lake model prediction error. Water Research, 17, 911–16. Reksalegora, O. (1979) Fish cage culture in the town of Jambi, Indonesia. In: Proceedings of the IDRC/SEAFDEC International Workshop on Pen and Cage Culture of Fish, Tigbauan, Iloilo, Philippines, 11–22 February 1979, pp. 51–3. SEAFDEC, Iloilo, Philippines. René, F. (1984) Essais d’elevage du loup (Dicentrarchus labrax), de la daurade (Sparus auratus) et du sar (Diplodus sargus) en Marinique. In: L’Aquaculture du Bar et des Sparides (Ed. by G. Barnabé & R. Billard), pp. 403–18. INRA, Paris. Rengaswamy, V. S., Marichamy, R., Rajapackiam, S. & Sundararajan, D. (1999) Collection and transportation of groupers for farming. In: Proceedings of the Fourth Indian Fisheries Forum, 24–28 November 1996, Kochi, Kerala (Ed. by M. M. Joseph, N. R. Menon & N. U. Nair), pp. 401–3. Asian Fisheries Society, Indian Branch, Mangalore. Rensel, J. E. & Prentice, E. F. (1979) Growth of juvenile spot prawn, Pandalus platyceros, in the laboratory and in net pens using different diets. Fisheries Bulletin, 78, 886–90. Rensel, J. E., Horner, R. A. & Postel, J. R. (1989) Effects of phytoplankton blooms on salmon aquaculture in Puget Sound, Washington: initial research. Northwest Environment Journal, 5, 53–69. Reynolds, C. S. (1984) The Ecology of Freshwater Phytoplankton. Cambridge University Press, Cambridge. Reynolds, C. S. & Walsby, A. E. (1975) Water blooms. Biological Reviews, 50, 437–81. Reynolds, F. A. & Haines, T. A. (1980) Effects of chronic exposure to hydrogen sulphide on newly hatched brown trout, Salmo trutta L. Environmental Pollution (Series A), 22, 11–17.

348

References

Richards, K. (1982) Rivers: Form and Process in Alluvial Channels. Methuen, London. Richards, R. H., Inglis, V., Frerichs, G. N. & Millar, S. D. (1992) Variation in antibiotic resistance patterns of Aeromonas salmonicida isolated from Atlantic salmon Salmo salar L. in Scotland. In: Problems of Chemotherapy in Aquaculture: From Theory to Reality (Ed. by C. M. Michel & D. J. Alderman), pp. 276–87. Ofﬁce International des Epizooties, Paris. Riddell, B. E. (1993). Salmonid enhancement: lessons from the past and a role for the future. In: Salmon in the Sea and New Enhancement Strategies (Ed. by D. Mills), pp. 228–55. Fishing News Books, Oxford. Rippey, S. R. & Cabelli, V. J. (1980) Occurrence of Aeromonas hydrophila in the limnotic environment: relationship of organism to trophic state. Microbial Ecology, 6, 45–54. Ritz, D., Lewis, M. E. & Shen, M. (1989) Response to organic enrichment of infaunal macrobenthic communities under salmonid sea cages. Marine Biology, 103, 211–14. Rivas, C., Cepeda, C., Dopazo, C. P., Novoa, B., Noya, B. & Barja, J. L. (1993) Marine environment as a reservoir of birnaviruses from poikilothermic animals. Aquaculture, 115, 183–94. Robb, D. H. & Roth, B. (2003) Brain activity of Atlantic salmon (Salmo salar) following electrical stunning using various ﬁeld strengths and pulse durations. Aquaculture, 216, 363–9. Roberts, R. J. (Ed.) (2001) Fish Pathology, 3rd edition. Baillière Tindall, London. Roberts, R. J. & Shepherd, C. J. (1997) Handbook of Trout and Salmon Diseases, 3rd edition. Fishing News Books, Oxford. Roberts, R. J. & Sommerville, C. (1982) Diseases of tilapias. In: Biology and Culture of Tilapias (Ed. by R. S. V. Pullin & R. H. Lowe-McConnell), pp. 247–64. ICLARM, Manila. Roberts, R. J., Bullock, A. M., Turner, M., Jones, K. & Tett, P. (1983) Mortalities of Salmo gairdneri exposed to cultures of Gyrodinium aureolum. Journal of the Marine Biological Association of the UK, 63, 741–3. Rodger, H. D., Turnbull, T., Edwards, C. & Codd, G. A. (1994) Cyanobacterial (bluegreen algal) bloom associated pathology in brown trout Salmo trutta L. in Loch Leven, Scotland. Journal of Fish Diseases, 17, 177–81. Romanycheva, O. D. & Barybina, I. A. (1979) A study of behaviour of bester (a hybrid between beluga and sterlet sturgeons) and a mathematical model for ﬁsh feeding in cages. In: Proceedings of the 7th Japan–Soviet Joint Symposium on Aquaculture, pp. 19–24. Tokai University, Tokyo and Ministry of Fisheries, Moscow. Rönnberg, O., Adjers, K., Ruokolathi, C. & Bondestam, M. (1992) Effects of ﬁsh farming on growth, epiphytes and nutrient content of Fucus vesiculosus L. in the Aland archipelago, northern Baltic sea. Aquatic Botany, 42, 109–20. Rosario, W. R. (1985) Transfer of ﬁsh culture technology in Central Luzon, Philippines. In: Proceedings of the PCARRD/ICLARM Tilapia Economics Workshop, U. P. Los Banos, Laguna, Philippines, 10–13 August 1983, (Ed. by I. R. Smith, E. B. Torres & E. O. Tan), pp. 174–9. ICLARM, Manila. Rosenthal, H. & Hilge, V. (2000) Aquaculture production and environmental regulations in Germany. Journal of Applied Ichthyology, 16, 163–6. Ross, A. (1988) Controlling Nature’s Predators on Fish Farms. Marine Conservation Society, Ross on Wye. Ross, A. J. & Johnson, H. E. (1962) Studies of the transmission of mycobacterial infections by chinook salmon. The Progressive Fish-Culturist, 24, 147–9.

References

349

Ross, L. G. (2000) Environmental physiology and energetics. In: Tilapias: Biology and Exploitation (Ed. by M. C. M. Beveridge & B. J. McAndrew), pp. 89–128. Kluwer, Dordrecht. Ross, L. G. & Ross, B. (1999) Anaesthetics and Sedative Techniques for Aquatic Animals. Blackwell, Oxford. Ross, L. G., Mendoza, E. A. Q. M. & Beveridge, M. C. M. (1993) The application of geographical information systems to site selection for coastal aquaculture: an example based on salmonid cage culture. Aquaculture, 112, 165–78. Rossbach, M. (1993) Dynamic investigations to describe the fatigue behaviour of VESTOLEN® A 5041 R (PE - HD) for ﬁsh farming cages. In: Fish Farming Technology (Ed. by H. Reinertsen, L. A. Dahle, L. Jørgensen & K. Tvinnereim), pp. 193–8. A. A. Balkema, Rotterdam. Roth, M., Richards, R. H. & Sommerville, C. (1993) Current practices in the chemotherapeutic control of sea lice infestations in aquaculture. Journal of Fish Diseases, 16, 1–16. Rothbard, S. (1988) The use of chemical tranquilizers in ﬁsh transport. Fisheries and Fishbreeding in Israel, 21, 28–34. Rottiers, D. V. (1991) Towable cage for studies of smoltiﬁcation in Atlantic salmon. The Progressive Fish-Culturist, 53, 124–7. Round, F. E. (1981) The Ecology of Algae. Cambridge University Press, Cambridge. Rubach, S. & Svendsen, Y. S. (1993) Salmon farming in submersible cages? In: Fish Farming Technology (Ed. by H. Reinertsen, L. A. Dahle, L. Jørgensen & K. Tvinnereim), pp. 237–9. A. A. Balkema, Rotterdam. Rudi, H. & Dragsund, E. (1993) Localization strategies. In: Fish Farming Technology (Ed. by H. Reinertsen, L. A. Dahle, L. Jørgensen & K. Tvinnereim), pp. 169–76. A. A. Balkema, Rotterdam. Rudi, H., Aarsnes, J. V. & Dahle, L. A. (1988) Environmental forces on a ﬂoating cage system, mooring considerations. In: Aquaculture Engineering Technologies for the Future. Institute of Chemical Engineers Symposium Series, Vol. 111, pp. 97–122. Institute of Chemical Engineers, Rugby. Rudi, H., Aarsnes, J. V., Lien, E. & Jørgensen, R. (1996) Operational regularity, exposed locality. In: Open Ocean Aquaculture. Proceedings of an International Conference, Portland, Maine, 8–10 May 1996 (Ed. by M. Polk), pp. 203–16. Report No. UNHMPCP-SG-96-9. New Hampshire/Maine Sea Grant College Program, Portland, ME. Rueggeberg, H. & Booth, J. (1989) Interactions between Wildlife and Salmon Farms in British Columbia: Results of a Survey. Technical Report Series Vol. 67, Paciﬁc and Yukon Region, Canadian Wildlife Service. Canadian Wildlife Service, Ottawa. Ruff, B. P., Marchant, J. A. & Frost, A. R. (1994) A stereo image analysis system for monitoring size distributions and biomass of farmed ﬁsh. In: Measures for Success (Ed. by P. Kestemont, J. F. Muir, F. Sevilla & P. Williot), pp. 157–62. CEMAGREF, Paris. Ruiz, J. M. (2001) Effects of ﬁsh farm loadings on seagrass (Posidonia oceanica) distribution, growth and photosynthesis. Marine Pollution Bulletin, 42, 749–60. Ruokolathi, C. (1988) Effects of cage ﬁsh farming on growth and chlorophyll a content of Cladophora. Marine Pollution Bulletin, 19, 166–9. Russell, E. S. (1931) Some theoretical consideration on the overﬁshing problem. Journal de Conseil International d’Exploration de la Mer, 6, 3–20. Ruzzante, D. E. (1994) Domestication effects on aggressive and schooling behaviour in ﬁsh. Aquaculture, 120, 1–24.

350

References

Sakiyama, T. (1979) The impact of expanding environmental pollution on Japanese ﬁsheries in the coastal zone. In: Economics of Aquaculture, Sea Fishing and Coastal Resource Use in Asia (Ed. by A. R. Librero & W. L. Collier), pp. 345–55. Agricultural Development Council and Philippine Council for Agriculture and Resources Research, Los Baños, Philippines. Sakomoto, S. & Yone, Y. (1980) A principal source of deposited lipid in phosphorus deﬁcient red sea bream. Bulletin of the Japanese Society of Scientiﬁc Fisheries, 46, 1227–30. Samuelsen, O. B. (1989) Degradation of oxytetracycline in seawater at two different temperatures and light intensities and the persistence of oxytetracycline in the sediment from a ﬁsh farm. Aquaculture, 83, 7–16. Samuelsen, O. B. (1992) The fate of antibiotics/chemotherapeutics in marine aquaculture sediments. In: Problems of Chemotherapy in Aquaculture: From Theory to Reality (Ed. by C. M. Michel & D. J. Alderman), pp. 162–73. Ofﬁce International des Epizooties, Paris. Samuelsen, O. B., Lunestad, B. T., Husevåg, B., Hølleland, T. & Ervik, A. (1992) Residues of oxilinic acid in wild fauna following medication in ﬁsh farms. Diseases of Aquatic Organisms, 12, 111–19. Sandodden, R., Finstad, B. & Iversen, M. (2001) Transport stress in Atlantic salmon (Salmo salar L.): anaesthesia and recovery. Aquaculture Research, 32, 87–90. Santhanam, R., Natarajan, P. & Kuthalingam, M. D. K. (1984) Fouling problems in cages and pens. In: Proceedings of the National Seminar on Cage and Pen Culture, Fisheries College, Tamil Nadu Agricultural University, Tuticorin, 18–19 March 1984, pp. 143–7. Tamil Nadu Agricultural University, Tuticorin, India. Santiago, A. (1994) The ecological impact of tilapia cage culture in Sampaloc lake, Philippines. In: Proceedings of the Third Asian Fisheries Forum, Singapore, 26–30 October (Ed. by L. M. Chou, A. D. Munro, T. J. Lum, T. W. Chen, L. K. K. Cheong, J. K. Ding, K. K. Hooi, H. W. Khoo, V. P. E. Phang, K. F. Shun & C. H. Tan), pp. 214–22. Asian Fisheries Society, Singapore. Santiago, C. B., Coloso, R. M., Millamena, M. N. & Borlongan, I. G. (1996) Feeds for Small-Scale Aquaculture. SEAFDEC, Iloilo, Philippines. Saunders, R. L. (1995) Salmon aquaculture: present status and prospects for the future. In: Cold-Water Aquaculture in Atlantic Canada (Ed. by A. D. Boghen), pp. 35–81. Tribune Press, Sackville, NB. Sayer, M. D. J., Treasurer, J. W. & Costello, M. J. (Eds) (1996) Wrasse: Biology and Use in Aquaculture. Blackwell, Oxford. Schmittou, H. R. (1969) Cage culture of channel catﬁsh. In: Proceedings of the Fish Farming Conference and Annual Convention of Catﬁsh Farmers, Texas, 1969, pp. 72–5. Schmittou, H. R. (1993) High Density Fish Culture in Low Volume Cages. American Soybean Association, Singapore. Schneider, G., Jäger, T. & Rosenthal, H. (1990) Growing trout in an efﬂuent plume of a coastal power station. Proceedings of ICES CM 1990/F: 68. ICES, Copenhagen. Schnick, R. A. (1991) Chemicals for worldwide aquaculture. In: Fish Health Management in Asia–Paciﬁc. Report on a Regional Study and Workshop on Fish Disease and Fish Health Management, pp. 441–67. NACA, Bangkok. Schreck, C. B. (1990) Physiological, behavioural and performance indicators of stress. American Fisheries Society Symposium, 8, 29–37.

References

351

Schreck, C. B., Maule, A. G. & Kaattari, S. L. (1993) Stress and disease resistance. In: Recent Advances in Aquaculture, Vol. 4 (Ed. by J. F Muir & R. J. Roberts), pp. 170–75. Blackwell, Oxford. Scott, D. B. C. & Muir, J. F. (2000) Offshore cage systems – a practical overview. In: Mediterranean Offshore Mariculture (Ed. by J. F. Muir & B. Basurco), pp. 79–89. Options Méditerranéenes, Series B, No. 30. CIHEAM, Zaragoza. Secretan, P. A. D. (1979) Too much stock escapes from nets and cages. Fish Farming International, 6 (3), 23. Secretan, P. A. D. (1980) Insurance and risk management as related to cage culture. In: Proceedings of the Institute of Fisheries Management Cage Fish Rearing Symposium, University of Reading, 26–27 March 1980, pp. 126–36. Janssen Services, London. SEERAD (2002) Scottish Fish Farms. Annual Production Survey 2000. Fisheries Research Services, Aberdeen. Shariff, M., Subasinghe, R. P. & Arthur, J. R. (Eds) (1992) Diseases in Asian Aquaculture. Asian Fisheries Society, Manila. Sharma, R. N. (1979) Cage ﬁsh culture in Nepal. In: Proceedings of the IDRC/SEAFDEC International Workshop on Pen and Cage Culture of Fish, Tigbauan, Iloilo, Philippines, 11–22 February 1979, pp. 93–6. SEAFDEC, Iloilo, Philippines. Shaw, S. & Muir, J. F. (1986) Salmon Economics and Marketing. Croom Helm, London. Shellard, H. C. (1965) Extreme wind speeds over the United Kingdom for periods ending 1963. Climatology Memo 50. Meteorological Ofﬁce, London. Shenoda, B. Z. & Naguib, M. (2000) Net cage culture of silver carp, Hypophthalmichthys molitrix Valenciennes, without artiﬁcal feeds. In: Sustainable Fish Production in Lake Nasser: Ecological Basis and Management Policy. ICLARM Conference Proceedings No. 61 (Ed. by J. F. Craig), p. 91. ICLARM, Penang. Shepherd, C. J. (2001) Husbandry and management in relation to disease. In: Fish Pathology, 3rd edition (Ed. by R. J. Roberts), pp. 276–82. Baillière Tindall, London. Sigler, J. W., Bjornn, T. C. & Everest, F. H. (1984) Effects of chronic turbidity on density and growth of steelheads and coho salmon. Transactions of the American Fisheries Society, 113, 142–50. Silvert, W. (1992) Assessing environmental impacts of ﬁnﬁsh aquaculture in marine waters. Aquaculture, 107, 67–79. Silvert, W. (1994) Modelling benthic deposition and impacts of organic matter loading. In: Modelling Benthic Impacts of Organic Enrichment from Marine Aquaculture (Ed. by B. T. Hargrave), pp. 1–18. Canadian Technical Report on Fisheries and Aquatic Sciences No. 1949. Department of Fisheries and Oceans, Ottawa. Silvert, W. & Sowles, J. W. (1996) Modelling environmental impacts of marine ﬁnﬁsh aquaculture. Journal of Applied Ichthyology, 12, 75–81. Simpson, J. H. (1981) The shelf-sea fronts: implications of their existence and behaviour. Philosophical Transactions of the Royal Society, London, Series A, 302, 531–46. Simmonds, S. (1996) Insurance and risk management for aquaculture. In: Open Ocean Aquaculture. Proceedings of an International Conference, Portland, Maine, 8–10 May 1996 (Ed. by M. Polk), pp. 217–19, Report No. UNHMP-CP-SG-96-9. New Hampshire/Maine Sea Grant College Program, Portland, ME. Sindermann, C. J. (1993) Disease risks associated with importation of non-indigenous marine animals. Marine Fisheries Review, 54, 1–10. Skjervold, P. O., Fjaera, P. V., Ostby, P. B. & Einen, O. (2001) Live-chilling and crowding stress before slaughter of Atlantic salmon (Salmo salar). Aquaculture, 192, 267–82.

352

References

Skulberg, O. M., Codd, G. A. & Carmichael, W. W. (1984) Toxic blue-green algal blooms in Europe: a growing problem. Ambio, 13, 244–7. Slaattelid, O. H. (1990) Model tests with ﬂexible, circular ﬂoats for ﬁsh farming. In: Engineering for Offshore Fish Farming, pp. 93–106. Thomas Telford, London. Slinger, S. J., Razzaque, A. & Cho, C. Y. (1979) Effect of feed processing and leaching on the losses of certain vitamins in ﬁsh diets. Schriften der Bundesforschungsanstalt für Fischerei, 14/15, 425–34. Smith, I. P., Metcalfe, N. B., Huntingford, F. A. & Kadri, S. (1993) Daily and seasonal patterns in the feeding behaviour of Atlantic salmon (Salmo salar L.) in a sea cage. Aquaculture, 117, 165–78. Smith, I. P., Metcalfe, N. B. & Huntingford, F. A. (1995) The effects of food pellet dimensions on feeding responses by Atlantic salmon (Salmo salar L.) in a marine net pen. Aquaculture, 130, 165–75. Smith, I. R. (1975) Turbulence in Lakes and Rivers. Freshwater Biological Association Publication No. 29. FBA, Windermere. Smith, I. R., Torres, E. B. & Tan, E. O. (Eds) (1985) Proceedings of the PCARRD/ICLARM Tilapia Economics Workshop, U.P. Los Banos, Laguna, Philippines, 10–13 August 1983. ICLARM, Manila. Soderberg, R. W. & Krise, W. F. (1987) Fin condition of lake trout, Salvelinus namaycush Walbaum, reared at different densities. Journal of Fish Diseases, 10, 233–5. Sodikin, D. (1978) Fish Culture in Indonesia: Its Construction and Management. SCS/GEN/77/15. South China Sea Fisheries Development and Co-ordinating Programme, Manila. Solomon, D. J. & Hawkins, A. D. (1981) Fish capture and transport. In: Aquarium Systems (Ed. by A. D. Hawkins), pp. 197–221. Academic Press, London. Sommerville, C. S. (1998) Parasites of farmed ﬁsh. In: Biology of Farmed Fish (Ed. by K. D. Black & A. D. Pickering), pp. 146–79. Shefﬁeld Academic Press, Shefﬁeld. Soto, D. & Mena, G. (1999) Filter feeding by the freshwater mussel, Diplodon chilensis, as a biocontrol of salmon farming eutrophication. Aquaculture, 171, 63–81. Sowerbutts, B. J. & Forster, J. R. M. (1981) Gaseous exchange and reoxygenation. Schriften der Bundesforschungsanstalt für Fischerei, 16/17, 199–218. Sowles, J. W., Churchill, L. & Silvert, W. (1994) The effect of benthic carbon loading on the degradation of bottom conditions under farm sites. In: Modelling Benthic Impacts of Organic Enrichment from Marine Aquaculture (Ed. by B. T. Hargrave), pp. 31–46. Canadian Technical Report on Fisheries and Aquatic Science No. 1949. Department of Fisheries and Oceans, Ottawa. Speare, D. J. (2002) Non-infectious disorders of coldwater ﬁsh. In: Diseases and Disorders of Finﬁsh in Cage Culture (Ed. by P. T. K. Woo, D. W. Bruno & L. H. S. Lim), pp. 171–91. CABI Publishing, Wallingford. Srivastava, R. K., Brown, J. A. & Allen, J. (1991) The inﬂuence of wave frequency and wave height on the behaviour of rainbow trout (Oncorhynchus mykiss) in cages. Aquaculture, 97, 143–53. Starling, F., Beveridge, M. C. M., Lazzaro, X. & Baird, D. J. (1998) Silver carp biomass effects on the plankton community in Paranoá Reservoir (Brazil) and an assessment of its potential for improving water quality in lacustrine environments. International Review of Hydrobiology, 83, 499–508. Stephen, C. & Ribble, C. S. (1995) An evaluation of surface moribund salmon as indicators of netpen disease status. Aquaculture, 133, 1–8.

References

353

Stewart, A. R. J. & Grant, J. (2002) Disaggregation rates of extruded salmon feed pellets: inﬂuence of physical and biological variables. Aquaculture Research, 33, 799–810. Stirling, H. P. (Ed.) (1985) Chemical and Biological Methods of Water Analysis for Aquaculturists. Institute of Aquaculture, University of Stirling, Stirling. Stirling, H. P. & Dey, T. (1990) Impact of intensive cage ﬁsh farming on the phytoplankton and periphyton of a Scottish freshwater loch. Hydrobiologia, 190, 193–214. Stirling, H. P. & Okomus, I. (1995) Growth and production of mussels (Mytilus edulis L.) suspended at salmon cages and shellﬁsh farms in two Scottish sea lochs. Aquaculture, 134, 193–210. Stokoe, A. (1973) Reed’s Naval Architecture for Marine Engineers. Thomas Reed, Glasgow. Strasˇkraba, M. (1980) The effect of physical variables on freshwater production. In: The Functioning of Freshwater Ecosystems (Ed. by E. D. Le Cren & R. H. LoweMcConnell), pp. 13–84. Cambridge University Press, Cambridge. Struve, M. R. & Bayne, D. R. (1991) Isolation columns as an alternative to replicate ponds in aquaculture research. Journal of Applied Aquaculture, 1 (4), 73–81. Sutherland, T. F., Martin, A. J. & Levings, A. D. (2001) Characterization of suspended particulate matter surrounding salmonid net-pen in the Broughton Archipelago, British Columbia. ICES Journal of Marine Science, 58, 404–10. Sutterlin, A. M. & Merrill, S. P. (1978) Norwegian Salmonid Farming. Technical Report of the Fisheries and Marine Services, Canada, No. 779. Department of Fisheries and Oceans, Ottawa. Sutterlin, A. M., Jokola, K. J. & Holte, B. (1979) Swimming behaviour of salmonid ﬁsh in ocean pens. Journal of the Fisheries Research Board of Canada, 36, 948–54. Svealv, T. L. (1988) Inshore versus offshore farming. Aquacultural Engineering, 7, 287–97. Svobodová, Z., Lloyd, R., Máchová, J. & Vykusová, B. (1993) Water Quality and Fish Health. EIFAC Technical Paper 54. FAO, Rome. Svobodová, Z., Kalab, P., Dusek, L., Vykusová, B., Kolarova, J. & Janouskova, D. (1999) The effect of handling and transport on the concentration of glucose and cortisol in blood plasma of common carp. Acta Veterinaria (Brno), 68, 265–74. SWCL (1988) Marine Fish Farming in Scotland. A Discussion Paper. SWCL, Perth, Scotland. Swingle, W. E. (1971) A marine cage design. The Progressive Fish-Culturist, 33, 102. Sylvester, J. R. (1975) Biological considerations on the use of thermal efﬂuents for ﬁnﬁsh aquaculture. Aquaculture, 6, 1–10. Tabira, N. (1980) The ﬁsh culture under the optimum density and the ﬁshery right administered by ﬁshermen’s cooperative association. Memoirs of the Faculty of Fisheries of Kagoshima University, 29, 157–67. Tacon, A. G. J. (1992) Nutritional Fish Pathology. Morphological Signs of Nutrient Deﬁciency and Toxicity in Farmed Fish. FAO Fisheries Technical Paper 330. FAO, Rome. Tahil, A. S. (1978) Experiments in rearing Siganus guttatus (pisces: Osteichyes, Siganidae) in a sea cage and ﬁshpen in the Philippines. Philippine Science Journal, 15, 50–66. Takashima, F. & Arimoto, T. (2000) Cage culture in Japan toward the new millennium. In: Cage Aquaculture in Asia: Proceedings of the First International Symposium on Cage Aquaculture in Asia (Ed. by I. C. Liao & C. K. Lin), pp. 53–8. Asian Fisheries Society, Manila, and World Aquaculture Society – Asia Branch, Bangkok.

354

References

Takeuchi, M. & Nakazoe, J. (1981) Effects of dietary phosphorus on lipid content and its composition in carp. Bulletin of the Japanese Society of Scientiﬁc Fisheries, 47, 347–52. Tan, C. K. F., Nowak, B. F. & Hodson, S. L. (2002) Biofouling as a reservoir of Neoparamoeba pemaquidensis (Page, 1970), the causative agent of amoebic gill disease in Atlantic salmon. Aquaculture, 210, 49–58. Tangen, K. (1977) Blooms of Gyrodinium aureolum in North European waters, accompanied by mortalities of marine organisms. Sarsia, 63, 123–33. Taniguchi, M. (1983) Feeding experiment of yellowtail using frozen diet. Bulletin of the Japanese Society of Scientiﬁc Fisheries, 49, 611–15. Tattanon, T. & Maneewongsa, S. (1982) Distribution and transport of seabass fry. In: Report of Training Course on Seabass Spawning and Larval Rearing, Songkhla, Thailand, 1–20 June 1982, p. 33. SCSIGEN/82/39. South China Sea Development and Coordination Programme, Manila. Tatum, W. M. (1974) Experiments in the culture of marine species in ﬂoating baskets. Proceedings of the Annual Meeting of the World Mariculture Society, 5, 411–18. Taylor, B. E., Jamieson, G. & Carefoot, T. H. (1992) Mussel culture in British Columbia: the inﬂuence of salmon farms on growth of Mytilus edulis. Aquaculture, 108, 51–66. Taylor, D. L. & Seliger, H. H. (Eds) (1979) Toxic Dinoﬂagellate Blooms. Elsevier, New York. Tazaki, S., Ishida, Y., Urakawa, T. & Takamatsu, T. (1975) Development of a ﬂoating breakwater. Oceanology International, 75, 191–4. Telfer, T. C. & Beveridge, M. C. M. (2001) Practical experiences: the Northern European perspective. In: Environmental Impact Assessment of Mediterranean Aquaculture Farms. Proceedings of the Seminar of the CIHEAM Network TECAM, Zaragoza, Spain, 17–21 January 2000 (Ed. by A. Uriante & B. Basurco), pp. 85–90. Options Méditerranéennes, Series Cahiers. Zaragoza: Vol. 55. CIHEAM/FAO, Rome. Tencalla, F. G., Dietrich, D. R. & Schlatter, C. (1994) Toxicity of Microcystis aeruginosa peptide toxin to yearling rainbow trout (Oncorhynchus mykiss). Aquatic Toxicology, 30, 215–24. Teng, S. K. & Chua, T. E. (1979) Effect of stocking density on the growth of estuary grouper Epinephelus salmoides Maxwell cultured in ﬂoating net cages. Aquaculture, 15, 273–87. Thomassen, J. M. & Lekang, O. I. (1993) Optimal distribution of feed in sea cages. In: Fish Farming Technology (Ed. by H. Reinertsen, L. A. Dahle, L. Jørgensen & K. Tvinnereim), pp. 439–42. A. A. Balkema, Rotterdam. Thoms, A. (1989) Pointers to safer moorings. Fish Farmer, 12 (3), 27–8. Thorpe, J. E. & Huntingford, F. A. (Eds) (1992) The Importance of Feeding Behaviour for the Efﬁcient Culture of Salmonid Fishes. World Aquaculture Society, Baton Rouge, LA. Thorpe, J. E., Talbot, M. S., Miles, M. S., Rawlings, C. E. & Keay, D. S. (1990) Food consumption in 24 hours by Atlantic salmon (Salmo salar L.) in a sea cage. Aquaculture, 90, 41–7. Tomi, W., Naiki, K. & Yamada, Y. (1979) Investigations into technical development of mariculture on commercial scale applied to offshore region. Proceedings of the Japanese–Soviet Joint Symposium on Aquaculture, 7, 111–20. Toor, H. S., Sehgal, H. S. & Sehdev, R. S. (1983) A case study of acute ﬁsh diseases in tanks loaded with high levels of organic manures. Aquaculture, 35, 277–82.

References

355

Toranzo, A. E., Nieto, F. & Barja, J. L. (1990) Mortality associated with cyanobacterial bloom in farmed rainbow trout in Galicia (Northwest Spain). Bulletin of the European Association of Fish Pathologists, 10, 106–7. Treasurer, J. W. (2002a) A review of potential pathogens of sea lice and the application of cleaner ﬁsh in biological control. Pest Management Science, 58, 546–58. Treasurer, J. W. (2002b) Application of sonar systems in aquaculture. Fish Farmer, 25 (4), 37–9. Troell, M. & Berg, H. (1997) Cage ﬁsh farming in the tropical Lake Kariba, Zimbabwe: impact and biogeochemical changes in sediment. Aquaculture Research, 28, 527–44. Troell, M. & Norberg, J. (1998) Modelling output and retention of suspended solids in an integrated salmon-mussel culture. Ecological Modelling, 110, 65–77. Troell, M., Kautsky, N. & Folke, C. (1999) Applicability of integrated coastal aquaculture systems. Ocean and Coastal Management, 42, 63–9. Trojanowski, J., Trojanowska, C. & Ratajczyk, H. (1982) Effect of intensive trout culture in Lake Letowo on its bottom sediments. Polskie Archiv für Hydrobiologie, 29, 659–70. Tseng, W. Y. (1983) Prospects for commercial netcage culture of red grouper (Epinephelus akaara T & S) in Hong Kong. Journal of the World Mariculture Society, 14, 650–60. Tseng, W. Y. & Yuen, K. H. (1979) Studies on fouling organisms on mariculture nets and cages in Hong Kong. In: Proceeding of the Conference on the Aquatic Environment in the Paciﬁc, pp. 151–9. SCOPE, Academia Sinica, Taipei. Tsutsumi, H. & Kikuchi, T. (1983) Benthic ecology of a small cove with essential oxygen depletion caused by organic pollution. Memoirs of the Marine Biological Laboratory, Kyushu University, 7, 17–40. Tsutsumi, H., Kikuchi, T., Tanka, M., Higashi, T., Imasaka, K. & Miyazaki, M. (1991) Benthic faunal succession in a cove organically polluted by ﬁsh farming. Marine Pollution Bulletin, 23, 233–8. Tuan, L. A., Nho, N. T. & Hambrey, J. (2000) Status of cage mariculture in Vietnam. In: Cage Aquaculture in Asia: Proceedings of the First International Symposium on Cage Aquaculture in Asia (Ed. by I. C. Liao & C. K. Lin), pp. 111–23. Asian Fisheries Society, Manila and World Aquaculture Society, Asia Branch, Bangkok. Tucholski, S. & Wojno, T. (1980) Studies on removal of wastes produced during cage-rearing of rainbow trout (Salmo gairdneri Richardson) in lakes III. Budgets of material and some nutrient elements. Rocznick Nauk Rolniczynch, 82, 31–50. Tucholski, S., Kok, J. & Wojno, T. (1980a) Studies on removal of wastes produced during cage-rearing of rainbow trout (Salmo gairdneri Richardson) in lakes I. Chemical composition of wastes. Rocznick Nauk Rolniczynch, 82, 3–15. Tucholski, S., Wieclawski, F. & Wojno, T. (1980b) Studies on removal of wastes produced during cage-rearing of rainbow trout (Salmo gairdneri Richardson) in lakes II. Chemical regime of water and bottom sediments. Rocznick Nauk Rolniczynch, 82, 17–30. Tucker, C. S. (2000) Off-ﬂavour problems in aquaculture. Reviews in Fisheries Science, 8, 45–88. Tucker, C. S. & Martin, J. F. (1991) Environment-related off-ﬂavours in ﬁsh. In: Aquaculture and Water Quality. Advances in World Aquaculture, Vol. 3 (Ed. by D. E. Brune & J. R. Tomasso), pp. 133–79. Turnbull, J. F., Adams, C. E., Richards, R. H. & Robertson, D. A. (1995) Attack site and resultant damage during aggressive encounters in Atlantic salmon Salmo salar L. parr. Aquaculture, 159, 343–53.

356

References

Turnbull, J. F., Richards, R. H. & Robertson, D. A. (1996) Gross, histological and scanning electron microscopic appearances of dorsal ﬁn rot in farmed Atlantic salmon Salmo salar L. parr. Journal of Fish Diseases, 19, 415–27. Turner, G. E. (1988) Codes of Practice and Manual of Procedures for Consideration of Introductions and Transfers of Marine and Freshwater Organisms. EIFAC Occasional Paper 23. FAO, Rome. Turner, R. (2000a). Offshore mariculture: mooring system design. In: Mediterranean Offshore Mariculture (Ed. by J. F. Muir & B. Basurco), pp. 159–72. Options Méditerranéennes, Series B, No. 30. CIHEAM, Zaragoza. Turner, R. (2000b) Offshore mariculture: site evaluation. In: Mediterranean Offshore Mariculture (Ed. by J. F. Muir & B. Basurco), pp. 141–57. Options Méditerranéennes, Series B, No. 30. CIHEAM, Zaragoza. Twu, S. W., Kao, R. C. & Hwung, H. H. (1986) On a wave dissipation method for offshore aquaculture areas. Aquacultural Engineering, 5, 271–86. Uotila, J. (1991) Metal contents and spread of ﬁsh farming sludge in southwestern Finland. In: Marine Aquaculture and Environment (Ed. by T. Mäkinen), pp. 121–6. Nordic Council of Ministers, Copenhagen. Uriarte, A. & Basurco, B. (2001) (Eds) Environmental Impact Assessment of Mediterranean Aquaculture Farms. Proceedings of the Seminar of the CIHEAM Network TECAM, Zaragoza, Spain, 17–21 January 2000. Options Méditerranéennes, Series Cahiers. Zaragoza: Vol. 55. CIHEAM/FAO, Rome. Uryn, B. A. (1979) Farming of juvenile whiteﬁsh Coregonus lavaretus (L.) in submerged, illuminated cages. Special Publication of the European Mariculture Society, 4, 289–97. US Army Corps of Engineers (1984) Shore Protection Manual, Vol I and II. US Government Printing Ofﬁce, Washington, DC. Vallentyne, J. R. (1974) The algal bowl: lakes and man. Miscellaneous Publications of the Department of the Environment, Fisheries and Marine Services, Canada, No. 22. Department of Fisheries and Oceans, Ottawa. Van Houtte, A. (1993) Preliminary draft survey of EIFAC members national legislation regulating control of efﬂuent discharges from ﬁsh farms. In: Workshop on Fish Farm Efﬂuents and their Control in EC Countries (Ed. by H. Rosenthal, V. Hilge & A. Kamstra), pp. 43–52. University of Kiel, Germany. Van Houtte, A. R., Bonucci, N. & Edeson, W. R. (1989). A Preliminary Review of Selected Legislation Governing Aquaculture. ADCP/REP/89/42. FAO, Rome. Vass, K. R. & Sachlan, M. (1957) Cultivation of common carp in running water in West Java. Proceedings of the IndoPaciﬁc Fisheries Commission, 6 (1–2), 187–96. Vincent, A. C. J. & Pajaro, M. J. (1997) Community-based management for a sustainable seahorse ﬁshery. In: Developing and Sustaining World Fisheries Resources. The State of Science and Management (Ed. by D. A. Hancock, D. C. Smith, A. Grant & J.-P. Beumer), pp. 716–66. CSIRO, Collingwood, Australia. Vines, C. R., Braid, M. R., Raymond, C. B. & Sanders, W. S. (1996) A collapsible cage design for aquatic culture. Journal of Applied Aquaculture, 6 (2), 73–9. Vollenweider, R. A. (1968) Scientiﬁc Fundamentals of the Eutrophication of Lakes and Flowing Water With Particular Reference to Nitrogen and Phosphorus as Factors in Eutrophication. Technical Report DASISU/68-27. OECD, Paris. Vollenweider, R. A. (1975) Input–output models with special reference to the phosphorus loading concept in limnology. Schweizerische Zeitschrift für Hydrologie, 37, 455–72.

References

357

Wagner, E. J. & Driscoll, D. M. (1994) Physiological stress responses of cutthroat trout to loading by ﬁsh pump, conveyor or dip net. Journal of Applied Aquaculture, 4 (1), 19–27. Walford, J. & Lam, T. J. (1986) Floating hatchery and net cage culture of Penaeus indicus in the Straits of Johore, Singapore. Aquaculture, 62, 11–32. Wallin, M. & Håkanson, L. (1991a) Nutrient loading models for estimating the environmental effects of marine ﬁsh farms. In: Marine Aquaculture and Environment (Ed. by T. Mäkinen), pp. 39–55. Nordic Council of Ministers, Copenhagen. Wallin, M. & Håkanson, L. (1991b) The importance of inherent properties of coastal areas. Marine Pollution Bulletin, 22, 381–8. Walmsley, R. D. & Thornton, I. A. (1984) An evaluation of OECD-type phosphorus eutrophication models for predicting the trophic state of southern African man-made lakes. South African Journal of Science, 16, 12–21. Watanabe, T., Murukami, A., Takeuchi, L., Nose, T. & Ogino, C. (1980a) Requirement of churn salmon held in freshwater for dietary phosphorus. Bulletin of the Japanese Society of Scientiﬁc Fisheries, 46, 361–7. Watanabe, T., Takeuchi, L., Murukami, A. & Ogino, C. (1980b) The availability to Tilapia nilotica of phosphorus in white ﬁsh meal. Bulletin of the Japanese Society of Scientiﬁc Fisheries, 46, 897–900. Watson, D. J. & Raja, P. T. (1979) A review of the present state of cage culture in the Baram district of Sarawak. In: Proceedings of the IDRC/SEAFDEC International Workshop on Pen and Cage Culture of Fish, Tigbauan, Iloilo, Philippines, 11–22 February 1979, pp. 89–92. SEAFDEC, Iloilo, Philippines. Watson, M., Power, R. & Munro, J. L. (2001) Use of light-attracted zooplankton for rearing post-settlement coral reef ﬁsh. Proceedings of the Gulf and Caribbean Fisheries Institute, 52, 340–51. Webb, J. H. & Youngson, A. F. (1992) Reared Atlantic salmon Salmo salar L. in the catches of a salmon ﬁshery on the western coast of Scotland. Aquaculture and Fisheries Management, 23, 393–7. Wee, K. L. (1979) Ventilation of Floating Cages. MSc thesis, University of Stirling, Stirling. Wee, K. L. (1982) Snakeheads – their biology and culture. In: Recent Advances in Aquaculture, Vol. 1 (Ed. by J. F. Muir & R. J. Roberts), pp. 179–213. Croom Helm, London. Welcomme, R. L. (1988) International Introductions of Inland Aquatic Species. FAO Fisheries Technical Paper 294. FAO, Rome. Weston, D. P. (1991) The effects of aquaculture on indigenous biota. In: Aquaculture and Water Quality. Advances in World Aquaculture, Vol. 3 (Ed. by D. E. Brune & J. R. Tomasso), pp. 534–67. World Aquaculture Society, Baton Rouge, LA. Weston, D. P. (1996) Environmental considerations in the use of antibacterial drugs in aquaculture. In: Aquaculture and Water Resources Management (Ed. by D. J. Baird, M. C. M. Beveridge, L. A. Kelly & J. F. Muir), pp. 140–65. Blackwell, Oxford. Wheatley, S. B., McLoughlin, M. F., Menzies, F. D. & Goodall, E. A. (1995) Site management factors inﬂuencing moratility rates in Atlantic salmon (Salmo salar L.) during marine production. Aquaculture, 136, 195–207. Whittaker, T. J. T., Bell, A. K. & Shaw, M. R. (1990) Model testing of a cage system for the offshore environment. In: Engineering for Offshore Fish Farming, pp. 79–92. Thomas Telford, London. WHO (1999) Food Safety Issues Associated with Products from Aquaculture. Report of a Joint FAO/NACA/WHO Study Group. WHO Technical Report Series 883. WHO, Geneva.

358

References

Whyte, J. N. C., Davis, J. C. & Forbes, J. R. (1997) Harmful algae in Canadian waters and management strategies. Ocean Research, 19, 161–71. Wiegel, R. L. (1964) Oceanographical Engineering. Prentice Hall, New Jersey. Willett, I. R. & Jacobsen, P. (1986) Fertilizing properties of trout farm waste. Agricultural Wastes, 17, 7–13. Williams, K., Gebhart, G. E. & Maughan, O. E. (1987) Enhanced growth of cage cultured channel catﬁsh through polyculture with blue tilapia. Aquaculture, 62, 207–14. Williamson, R. B. & Beveridge, M. C. M. (1994) Fisheries and aquaculture. In: The Freshwater Resources of Scotland: A National Resource of International Signiﬁcance (Ed. by P. S. Maitland, P. J. Boon & D. S. McLusky), pp. 317–32. John Wiley, Chichester. Willinsky, M. D., Robson, D. R., Vangool, W. J., Fournier, R. A. & Allen, J. H. (1991) Design of a spherical submersible, self-cleaning aquaculture system for exposed sites. In: Proceedings of the National Science Foundation Workshop on Offshore Engineering and Mariculture, 25–30 September 1991 (Ed. by G. N. Hirata), pp. 317–36. Honolulu, HA. Willinsky, M. D., Berteaux H., Huguenin, J. E. & Paul, W. (1995a) Design and development of a spar buoy and support system for autonomous operation of offshore ﬁsh farms. Unpublished paper presented at World Aquaculture Society ’95, 1–4 February 1995, San Diego, CA. Willinsky, M. D., Jäger, T. & Huguenin, J. E. (1995b) Illuminated in-situ incubators for the mass rearing of larval ﬁsh. Unpublished paper presented at World Aquaculture Society ’95, 1–4 February 1995, San Diego, CA. Willis, K. J. & Ling, N. (2003) The toxicity of emamectin benzoate, an aquaculture pesticide, to planktonic marine copepods. Aquaculture, 221, 289–97. Willumsen, B. (1989) Birds and wild ﬁsh as potential vectors of Yersinia ruckeri. Journal of Fish Diseases, 12, 275–7. Willumsen, F. V., Berge, F. S., Dragsund, E., Tangen, K. & Winther, U. (1993) SEAWATCH – An operational marine surveillance and information system. In: Fish Farming Technology (Ed. by H. Reinertsen, L. A. Dahle, L. Jørgensen & K. Tvinnereim), pp. 163–8. A. A. Balkema, Rotterdam. Wilson, R. P. (Ed.) (1991) Handbook of Nutrient Requirements of Finﬁsh. CRC Press, Boca Raton, FL. Wilton, S. R. (1979) An Investigation into the Cage Culture of Rainbow Trout. MSc thesis, University of Stirling, Stirling. Wong, C. (1982) Harvesting and Marketing of Cultured Marine Fish in Hong Kong. SCS/GEN/82/34. South China Sea Fisheries Development and Coordinating Programme, Manila. Wong, P. S. & Wu, R. S. S. (1987) Red tides in Hong Kong: problems and management strategy with special reference to mariculture industry. Journal of Shoreline Management, 3, 1–21. Wong, S.-Y., Ong, B. & Chua, T.-E. (1979) Isolation, identiﬁcation of causative agent of ‘red-boil disease’ in grouper (Epinephelus salmoides) and its possible control by vaccination. In: Proceedings of the IDRC/SEAFDEC International Workshop on Pen and Cage Culture of Fish, Tigbauan, Iloilo, Philippines, 11–22 February 1979, pp. 81–7. SEAFDEC, Iloilo, Philippines. Woo, P. T. K., Bruno, D. W. & Lim, L. H. S. (Eds) (2002) Diseases and Disorders of Finﬁsh in Cage Culture. CAB Publishing, Wallingford. Woods Hole Engineering Associates (1984) Design Guide for Use of Copper Alloy Expanded Metal Mesh in Marine Aquaculture. Contract Report Technology for the

References

359

Copper Industry. INCRA Proj. 268B. Woods Hole Engineering Associates, Woods Hole, MA. Wootten, R. (1979) Tapeworm threat to trout in ﬂoating freshwater cages. Fish Farmer, 2 (3), 5. Wu, R. S. S., Lam, K. S., MacKay, D. W., Lau, T. C. & Yam, V. (1994) Impact of marine ﬁsh farming on water quality and bottom sediment: a case study in the sub-tropical environment. Marine Environmental Research, 38, 115–45. Würsig, B. & Gailey, G. A. (2002) Marine mammals and aquaculture: conﬂicts and potential resolutions. In: Responsible Marine Aquaculture (Ed. by R. R. Stickney & J. P. McVey), pp. 45–59. CAB International, Cambridge. Yambot, A. V. (2000) Problems and issues of Nile tilapia cage farming in Taal Lake, Philippines. In: Cage Aquaculture in Asia. Proceedings of the First International Symposium on Cage Aquaculture in Asia (Ed. by I. C. Liao & C. K. Lin), pp. 241–52. Asian Fisheries Society, Manila and World Aquaculture Society – Asia Branch, Bangkok. Yang, S.-L. (1982) Fish culture and reservoir management in the Republic of Singapore. In: Proceedings of Seminar on Production and Exploitation of Open Waters, 15–18 June 1982, Bogor, Indonesia. Unpublished ms. Yang, Y. & Lin, C. K. (2000) Integrated cage culture in ponds: concepts, practices and perspectives. In: Cage Aquaculture in Asia. Proceedings of the First International Symposium on Cage Aquaculture in Asia (Ed. by I. C. Liao & C. K. Lin), pp. 233–40. Asian Fisheries Society, Manila and World Aquaculture Society – Asia Branch, Bangkok. Yang, Z. B., Takayama, H., Matsuoka, K. & Hodgkiss, I. J. (2000) Karenia digitata sp. nov. (Gymnodiniales, Dinophyceae), a new harmful algal bloom species from the coastal waters of west Japan and Hong Kong. Phycologia, 39, 463–70. Ye, L.-X., Ritz, D. A., Fenton, G. E. & Lewis, M. E. (1991) Tracing the inﬂuence of organic waste from a salmonid farm using stable isotope analysis. Journal of Experimental Marine Biology and Ecology, 145, 161–74. Yin, S. K., Harrison, P. J., Chen, J., Huang, W. & Qian, P. Y. (1999) Red tides during spring 1998 in Hong Kong: is El Niño responsible? Marine Ecology Progress Series, 187, 289–94. Yndestad, M. (1992) Public health aspects of residues in animal products: fundamental considerations. In: Problems of Chemotherapy in Aquaculture: From Theory to Reality (Ed. by C. M. Michel & D. J. Alderman), pp. 494–511. Ofﬁce International des Epizooties, Paris. Yokom, S., Axler, R., McDonald, M. & Wilcox, D. (1997) Recovery of a mine pit lake from aquacultural phosphorus enrichment: model predictions and mechanisms. Ecological Engineering, 8, 195–218. Yokoyama, H. (2000) Environmental quality criteria for aquaculture farms in Japanese coastal area – a new policy and its potential problems. Bulletin of the National Research Institute of Aquaculture, 29, 123–34. Youngson, A. F., Hansen, L. P. & Windsor, M. L. (1998) Interaction between Salmon Culture and Wild Stocks of Atlantic Salmon: Scientiﬁc and Management Issues. NINA, Oslo. Youngson, A. F., Dosdat, A., Saroglia, M. & Jordan, W. C. (2001) Genetic interactions between marine ﬁnﬁsh species in European aquaculture and wild conspeciﬁcs. Journal of Applied Ichthyology, 17, 153–62. Yu, O. K., Vizcarra, A. T. & Sitoy, H. S. (1979) Development of circular ﬂoating cages for milkﬁsh broodstock at the SEAFDEC Aquaculture Department. In: Proceedings of

360

References

the IDRC/SEAFDEC International Workshop on Pen and Cage Culture of Fish, Tigbauan, Iloilo, Philippines, 11–22 February 1979, pp. 107–17. SEAFDEC, Iloilo, Philippines. Zerner, C. (1992) Development of small scale freshwater cage culture ﬁshery in reservoirs in Java: legal, environmental and socio-economic issues. In: Contributions to Fishery Development Policy in Indonesia (Ed. by R. B. Pollnac, C. Bailey & A. Poernomo), pp. 133–9. Central Research Institute for Fisheries, Jakarta. Zimba, P. V., Khoo, L., Gaunt, P. & Carmichael, W. W. (2001) Conﬁrmation of catﬁsh Ictalurus punctatus (Raﬁnesque) mortality from Microcystis blooms. Journal of Fish Diseases, 24, 41–7. Zitko, V. (2001) Analytical chemistry in monitoring the effects of aquaculture: one laboratory’s perspective. ICES Journal of Marine Sciences, 58, 486–91. Zmyslowska, I., Harisz, M. & Lewandowska, D. (2001) Sanitary and bacteriological evaluation of water quality during cage culture of wells (Siluris glanis L.) in cooling water. Archives of Polish Fisheries, 9, 191–9. Zoran, M., Milstein, A. & Krambeck, H. J. (1994) Limnological aspects of dual purpose reservoirs for irrigation and ﬁsh culture in the coastal area of the Jordan Valley. Israeli Journal of Aquaculture, 46, 64–75.

Cage Aquaculture, Third Edition Malcolm C. M. Beveridge Copyright © 1996, 2004 by Blackwell Publishing Ltd

Index

abrasion, 24, 44, 50, 72, 93, 100, 102, 132, 244 access, 25–6, 31, 94, 111, 112, 151, 155 acidic waters, 120–21 aeration, 202, 257–64, 307 Aeromonas sp., 129, 175, 248 aﬂatoxins, 217 Africa, 40, 133, 223, 246 agriculture, 1 alarms, 265 algae (see phytoplankton) altitude, 118, 300, 306 aluminium alloy, 61, 63, 75, 91, 297 ammonia, 112, 120–21, 166, 167, 168, 197, 206, 227, 276 Anabaena spp., 123 anaesthetics, 227, 234 anchor, 94–107, 201, 282, 285–7 Anguilla spp., 23 antifoulants, 44, 172, 239, 252–4 antimicrobial(s), 174–5, 250 aquaculture extensive, 15–18 intensive, 15, 19–20 law, 151–4 origins, 1–2 semi-intensive, 15, 18–19 Arctic charr (see Salvelinus alpinus) Ardea cinerea, 164, 266–75 Arenicola marina, 178 Aristichthys nobilis, 16, 17 Atlantic salmon (see Salmo salar) Australia, 21, 50–51, 63, 132, 133, 152, 205, 208, 211, 218, 254, 266, 268, 271, 273, 288 azamethiphos, 176 bacteria, 126, 132, 170, 204, 206, 212, 218, 232, 243, 245, 254, 256, 270, 276 Baltic Sea, 136, 145, 169, 250 bamboo, 6, 37–40, 69–74, 76, 83–6, 250 Bangladesh, 25, 30, 210

Bay of Bengal, 135 of Fundy, 114, 150, 296 Beggiatoa, 170 behaviour, caged ﬁsh, 26, 32, 33, 37, 145, 220, 225, 229, 244, 245, 264, 269, 298, 306 benthos, 120, 169–72, 192, 194 bighead carp (see Aristichthys nobilis) biodiversity, 170 birds, 163–4, 214, 269–75 boats, 6, 9, 15, 91, 107, 145, 204–6, 214, 232, 261, 280 Bolivia, 242, 306 boring organisms, 40 breakwaters, 136, 281–8 buoyancy, 9, 12, 74, 75–8, 83, 89, 109–10, 283, 287, 294, 297 buoys, 9, 107, 294 cage(s) bag, 9–11, 26, 33–7, 40–68, 69, 71, 109, 132, 141, 147, 150, 160, 234–9, 241–2, 246, 268–9, 288 classiﬁcation, 9–14 collar, 9, 37, 74–91, 94, 97, 100, 110, 132, 160, 288, 290, 292, 294, 297, 301 construction, 40–91, 160, 172 deﬁnition, 4–5 depth, 36–7, 260 design(s), 9, 32–110 ﬁxed, 9, 10, 69–73, 150, 151, 250–51, 294 ﬂexible collar designs, 10, 12, 13, 89, 288, 290–91, 295–6 ﬂoating, 9, 74–91, 150, 294 ﬂotation system 75–8, 109–10 fouling-resistant designs, 254–5 handrail, 84, 87, 89, 91 hapa, 15, 63–4 illuminated, 14, 17 joints, 84–6, 88, 89 licensing, 153 361

362

Index

linkages, 81, 82, 92–4, 96 maintainance, 234–9 motion, 77, 82–3, 93, 96, 141, 243, 281, 292, 297 offshore, 24, 29, 43, 45, 66, 96, 281, 288–98 rigging, 9, 55, 67, 72, 294 rotating, 9–12, 232–3, 235, 236, 246–7, 252, 255, 256, 294 shading, 306–7 shape, 9, 32, 33, 58, 61, 87, 97 size, 5, 26, 33–7 submerged, 9, 11, 12, 14, 241–2, 291, 297–8 submersible, 9, 11–2, 136, 255, 291–7 top nets, 294, 298, 301, 307 towable, 205, 207 traditional designs, 6–8, 37–40 treatment, 246 uses, 14–15 walkways, 72, 83, 87, 91 wooden, 6, 7, 40, 241–2 cage culture advantages and disadvantages, 24–31 economics, 26–30, 32, 40 environmental issues, 25–6, 159–200 extensive, 15–18, 20, 26, 132, 145, 148, 163, 184, 190–91, 196, 199, 209, 256 freshwater, 26, 96, 114, 123–5, 129, 132–4, 152–3, 167–8, 169, 179–81, 184–92, 269, 276 integrated, 21–2 intensive, 19–20, 26, 27–9, 30, 130, 161, 163, 169, 174–5, 182, 184–90, 192, 195, 198–9, 210–14, 220, 226, 227, 228 marine, 18–19, 21–2, 24, 26, 30, 88–9, 115, 120, 126–8, 132, 136, 147, 167–9, 170–72, 181, 192–5, 196–8, 252 monoculture, 20–21 offshore, 24, 27, 29, 32–3, 45, 66, 82, 115, 128, 145, 281, 297 origins and history, 6–8 polyculture, 20–21 production, 22–4 semi-intensive, 18–19, 20, 26, 27, 132, 145, 191–2, 196, 199–200, 209–10, 218 social issues, 25–6, 31, 159–60, 254, 159–60

Calamus spp., 37 Caligus, 26 elongatus, 26, 129, 175, 249 Cambodia, 6, 7, 148, 160, 250 Canada, 15, 114, 128, 132, 136, 173, 182, 266–7, 271, 296–7 Capitella spp., 170, 176 carbon content of ﬁsh, 191 wastes, 22, 169, 193, 194–5, 278 carp (see Cyprinus carpio, etc.) catﬁsh (see Pangasius, Clarias, Ictalurus) Chaetoceros spp., 126 chain, 93, 96, 98–104 Channa spp., 6 Chanos chanos, 12, 19, 33, 87, 161, 201, 268 chemicals, 172–8, 246–9, 252–4 chemotherapeutants, 172–8, 182 chemotherapy, 178, 246–9 bath treatments, 176, 246–8 in-feed medication, 174–5, 248–9 Chile, 32, 160, 181 China, 1, 3, 6, 14, 16, 23, 24, 161–3 clams (see Tridacna, Tapes) Clarias spp., 6 Clupea harengus harengus, 15 cobia (see Rachycentron canadum) Cochlodinium sp., 127 conservation, 15 copper, 253–5 cormorant (see Phalacracorax) corrosion, 37, 45–52, 61, 102, 255 Crenilabrus melops, 254 crustacean(s), 15, 22–3, 24, 176, 254 Cryptoctyle lingua, 129 Ctenolabrus rupestris, 254 currents, 1, 37, 44, 45, 51–2, 55–61, 71, 78–80, 92, 96, 97, 111, 130–32, 133, 145–9, 155, 164, 167, 169, 193–4, 213, 224, 226, 241–2, 251, 256, 275, 279, 282, 294, 297 cyanobacteria, 123–5, 126, 167 cyclones, 134–6, 141 cypermethrin, 176, 249 Cyprinus carpio, 6, 14, 18, 213, 217, 254, 261 depth, 150–51 diatoms, 123, 126 Dicentrarchus labrax, 19, 22

Index

dichlorvos, 176, 249 digestion, 118, 164, 224 dinoﬂagellates, 126–8, 182 Diphyllobothrium spp., 129, 269 Diplodus sargus, 294 Diplostomum spathaceum, 129 disease, 3, 4, 21, 26, 91, 112, 121, 128–30, 132, 150, 153, 172, 178, 204, 208, 229–30, 243–5, 269–70, 276–8, 306 disinfectants, 172 dogﬁsh (see Squalus) dogs, 265 dolphins, 269, 271 drag, 37, 45, 55, 58, 76, 78–9, 80, 96, 252 drifting objects, 250–51 ecological footprint, 162–3 eel (see Anguilla) Egypt, 152 Eichhornia crassipes, 250 emamectin benzoate, 176 enclosure, 4–5, 24 environmental capacity, 159–200, 245 impacts, 18, 159–200, 217 quality objectives (EQO), 183 quality standards (EQS), 183 risk assessments (ERA), 183 services, 159 Epinephelus spp., 19, 22, 26, 30, 128, 161, 204, 211, 232, 244, 254, 268–9, 278 escapes (see feral animals) estuaries, 116–17 Eumetopias jubatus, 266, 271 Europe, 9, 18, 19, 22, 32, 33, 122, 127, 128, 151–2, 173, 174, 175, 178, 214–15, 218, 240, 253, 281, 288, 292, 294–6, 298 Euthynnus pelamis, 22 eutrophication, 15, 16–17, 22, 167–9 exotic species, 178–82 experimentation, 15 fallowing, 175, 245, 250, 280–81 Farm Ocean Cage, 255, 292–4 feeders, 220–25 automatic, 220–22, 225–6 demand, 222–4 interactive, 224–5

363

feed(ing), 217–26 conversion ratio (FCR), 161, 165 dry, pelleted, 212–14, 220–24 extensive, 15–18, 190–91, 209 ﬁshmeal, 1, 4, 15, 161–3 ﬁsh oil, 161–3, 212 ﬂoating, 213–14 hand, 218–20, 225–6 intensive, 19–20, 210–14 Lift-Up system, 279 losses, 20, 35–6, 149, 226, 241, 278 methods, 217–26, 281 moist, 209–210, 211–12, 218 pigments, 172, 212 rings, 226 semi-intensive, 15, 18–19, 26–7, 132, 145, 148, 162, 163, 184, 191–2, 196, 199–200, 209–210, 218–20, 243 silage, 212, 215 sinking, 213 storage, 211, 214–17, 243 trash ﬁsh, 6, 19–20, 165, 211–12, 215, 218, 248 trays, 216 feral animals, 164, 178–82, 244 fetch, 79, 81, 82, 137–45, 286, 287 Finland, 153 ﬁsh aggregation devices (FADs), 267 air-breathing, 36 consumption, 1 pelagic, 22, 35 pump, 207, 228–9 sunburn and UV effects, 304–7 Fishbase, 179 ﬁsheries, 1–4, 15, 25, 161–3, 181–2 fjords, 94, 117, 126, 128, 130–31, 147 fouling, 9, 15–16, 20–21, 26, 37, 44, 45, 49–50, 54–61, 76, 80, 132–4, 164, 236–9, 240, 251–6, 302 copper–nickel mesh, 49–50 control with herbivores, 15–16, 20–21, 254 effect on drag, 54–61 effect on ﬂotation, 55, 76, 77 pollution, 133, 252 removal, 236–9, 251–6 water exchange, 58–61 France, 242, 294 fungi, 129, 214, 216, 243, 244, 245, 250, 269–70

364

Index

gas supersaturation, 120 Geographical Information Systems (GIS), 157–8, 193–4 gills, 124, 126–7, 268, 278, 280 global positioning system (GPS), 107 goby (see Oxyeleotris) grading, 34, 91, 179, 228–9 grouper (see Epinephelus) growth, 2–3, 35–6, 38, 112, 118, 120, 121, 132, 145, 184, 196, 201, 208–10, 217, 220–26, 227, 228, 267, 298, 306 grunt (see Pomadasys kaakan) guillemots (see Uria) Gulf of Aqaba, 197 Maine, 267 gulls (see Larus) Gyrodactylus salaris, 178 Gyrodinium aureolum, 127 habitat destruction, 181 HABs (see phytoplankton, harmful algal blooms) halibut (see Hippoglossus hippoglossus) handling, 128, 180, 202, 207, 208, 243, 244, 245, 256 hardwood, 37, 40, 70–71, 85–6, 99, 160 harrowing, 280 harvesting, 179, 227, 231–4 herons (see Ardea) herring (see Clupea harengus harengus) Heterosigma akashiwo, 128 Hippocampus, 15 Hippoglossus hippoglossus, 22, 36, 209, 213 Hong Kong, 126–7, 133, 152, 164, 168–9, 204, 232–4, 278 Hungary, 20 hurricanes, 134–6 hydroacoustics, 32, 165, 227 hydrogen peroxide, 176–8 hydrogen sulphide, 169, 276 Hypophthalmichthys molitrix, 16–17, 63, 125, 196, 202, 203, 261 ice, 74, 115, 136, 217–18, 287–8, 291, 296–7, 298–300 Iceland, 136 Ictalurus punctatus, 19, 20, 28–9, 125, 133, 213, 223, 261

impact (see also environment) noise, 163 social, 25, 265 visual, 160 wildlife, 163–4 India, 49–50, 69, 133 Indonesia, 6, 14, 17, 25, 148, 152, 161, 241, 278 insurance, 33, 136, 230, 240, 267, 299 Inter-tropical Convergence Zone, 135 Ireland, 155, 176, 178, 249 irrigation canals, 21, 33, 148, 150, 152, 242 Israel, 21, 120, 266, 276 Italy, 242 ivermectin, 176 Japan, 6–7, 15, 22, 30, 32, 42–3, 58, 63, 122, 126–7, 128, 136, 173–4, 194, 211, 218, 252, 254, 255, 276–7, 283, 288, 291, 294 jellyﬁsh, 250 Kames cages, 87, 88, 91 Kenya, 306 knifejaws (see Oplegnathus) Labeo rohita, 254 lakes, 6, 16, 18, 26, 30–31, 114–15, 120–21, 122, 123–5, 131–2, 133, 136, 149, 150, 152, 160, 163, 167, 169, 184–92, 195–6, 198–200, 242, 250–51, 264, 265, 276, 279, 280, 298–300 Great Lake, Cambodia, 6 Kariba, Zimbabwe, 162–3, 269 Laguna de Bay, Philippines, 10, 133, 152, 159, 160, 250–51 landscape, 26, 160 larch (see Larix) Larix spp., 87 Larus spp., 266 Lates calcarifer, 19, 50–52 legislation, 151–4, 173 Lepeophtheirus salmonis, 129, 175–8, 183, 197, 229–30, 244, 245, 246, 249, 307 Leptobarbus hoeveni, 6 lice (see Lepeophtheirus, Caligus) light, 16, 76, 133, 213, 244, 300–307 photodegradation, 76, 255, 300–304

Index

ultraviolet (UV), 76, 300, 301, 304–7 Livingstonia rotundifolia, 70–71, 250–51 loads, 43, 55–61, 67, 69, 74–5, 93, 95, 96, 97–8, 100, 104, 145, 286–7, 297 Lutjanus johni, 20 Lutra spp., 266, 268 Malaysia, 18, 25, 133, 160, 266 management, 24, 91, 151–5, 201–39, 252, 275, 280–81 mangrove timber (see Rhizophora, Avicenia) Martesia striata, 132 Martinique, 294 mesh, 6, 38, 41, 61–7, 78, 254–5 bamboo, 38 copper alloy (Cu–Ni), 26, 49–50, 64, 238–9, 254–5 ﬂexible (see netting) galvanized steel, 50–52, 63, 271 metal, 26–7, 239 plastic-coated metal, 52, 242, 255 polymer, 45 rigid, 46–52, 61–2, 239, 242, 265, 275 shape, 65–6 size, 78, 132, 235–6, 241, 243, 252, 268, 269 methane, 276–8 Mexico, 86, 307 Microcystis spp., 123–5 microorganisms, 164, 178, 243 milkﬁsh (see Chanos chanos) mink (see Mustela) models environmental capacity, 183–200 moorings, 95–7 stratiﬁcation, 117 waste dispersion, 193–4, 195 wave prediction, 139–45 modulus of elasticity, 83 molluscs, 22, 23–4, 132–3, 176 monitoring, 34, 112, 128, 196, 199, 201, 226–7, 229–30 moorings, 32, 43, 55, 68, 74, 76, 82, 91, 94–107, 132, 145, 147, 150, 151, 201, 234, 239, 240, 241, 250, 276, 278, 281, 285, 286–7, 288–91, 294, 298 mortalities, 27, 34, 35, 52, 111, 120, 121, 124, 125, 127, 128–30, 141, 164, 196, 201, 205, 206, 208, 209,

365

221, 227, 230–31, 240, 244, 245, 246, 250, 275, 276–7, 281 Mustela lutreola, 268 Neoparamoeba pemaquidensis, 132, 254 Nepal, 254 Nerocilia sp., 129 net changing, 235–6 Netherlands, 242 Netlon®, 46 netpen liver disease, 132 netting, 6, 40–45, 54–67, 78–9, 235–8, 254–6, 271–4 antipredator, 63, 67, 271–4 cleaning, 235–8, 252, 256 colour, 302–3 fouling resistant, 55–7, 254–5 knotless, 43, 252 knotted, 42–3, 252 nylon (also polyamide, PA), 41, 43, 46, 256, 301, 302, 303 polyester (PES), 41, 43, 46, 301, 303 polyethylene (PE), 41, 42, 43, 46, 254–5, 256, 302, 303 polypropylene (PP), 41, 46, 302, 303 yarns, 40, 41–4, 46 New Zealand, 128, 131, 288, 302 nitrogen, 16, 165–6, 167, 169, 192, 227, 230, 277–8, 279 North America, 4, 19, 22, 32, 122, 126, 131, 160, 174, 246, 253, 298 Norway, 7, 15, 25, 27, 32–3, 35, 36–7, 94, 96, 97, 126, 128, 130, 136, 152, 153, 174, 176, 178, 181, 182, 198, 204, 208, 224, 229–30, 236, 249, 250, 265, 266, 267, 306 Ocean Spar®, 67, 297 oil drums, 40, 85–7, 160, 303–4 spills, 122 oligochaetes, 169 Oncorhynchus mykiss, 20, 24, 59–60, 124, 127, 179, 182, 185, 188, 195, 198, 223, 234, 242, 260, 261, 269, 278–9, 294, 298, 307 Oplegnathus spp., 53, 254, 255 Oreochromis spp., 7–8, 15–16, 18, 20, 26–7, 29–31, 33, 35–6, 63–5, 72, 124–5, 132, 133, 159, 161, 163, 165–6, 179, 185, 188, 190, 191,

366

Index

196, 199–200, 201, 208, 210–11, 213, 217, 223, 224, 226, 232, 240, 245, 249, 251, 252, 254, 261–2, 269, 278, 294 O. aureus, 18 O. mossambicus, 18 O. niloticus, 18, 261 Oscillatoria spp., 123, 124, 125, 126 otters (see Lutra) oxolinic acid, 174–5 Oxyeleotris marmorata, 6 oxygen, 52, 92, 112, 118–20, 126, 127, 145, 167, 168–9, 177, 202, 204, 206, 227, 234, 244, 247–8, 256–64, 276, 277, 278, 307 oxytetracycline, 174, 175 Pagrus major, 252, 278 Pangasius, 6 parasites (see also sea lice), 20–21, 26, 27, 129, 173, 175–8, 183, 202, 243–50, 269–70 pen, 4, 5, 24, 27, 30, 111, 159, 160, 265 Penaeus spp., 23, 24, 201 Penang, Straits of, 128–9, 132–3 pH, 120–21, 122, 227, 243 Phalacracorax carbo, 164, 266, 268, 269, 270, 271 P. aristotelis, 164 P. auritus, 266 P. penicillatus, 266 Philippines, 9, 10, 12, 15, 16, 18, 19–20, 22, 23, 26, 27, 29–31, 70–71, 73, 83–5, 120–21, 152, 155, 159–60, 214, 215, 250–51, 265, 268, 276, 294 Phoca spp., 265–75 phosphorus, 16, 165, 167, 169, 184–90, 192, 197, 198–9, 200, 227, 230, 277–8, 279 Phragmites spp., 37 phytoplankton blooms, 15, 118, 122–8, 167–9, 184, 256, 277–8 food, 15–19 freshwater, 121, 123–5, 167–8, 196, 256, 258 harmful algal blooms, 128 marine, 19, 126–8, 168–9 off-ﬂavours, 123 toxins, 123–5, 126–8

pike (see Esox lucius) plankton, 15, 17, 18, 20, 167–9 planning, 151–4, 160, 228, 231, 265, 280 pneumatic barriers, 288, 289 poaching, 31, 111, 265 Poland, 149, 151–2, 278–9 Pollarchius pollarchius, 15 virens, 129 pollution, 30, 111, 112, 122, 209, 212, 240, 244 polychaetes, 170, 176 polystyrene, 75, 76, 86, 87, 303–4 Pomadasys kaakan, 20 pond, 3, 4, 6, 15, 17–18, 20, 21, 24, 25–6, 27, 30–32, 111, 118, 124, 133, 155, 162, 163, 225, 267 pontoon, 9, 91, 94, 151, 282, 292–4 Posidonia, 172, 280–81 Povilla adusta, 40, 133, 134 power stations, 15, 117–18, 120, 122, 149, 233, 242 prawns (see Penaeus) predators, 26, 36, 37, 43–4, 51–2, 63, 67, 129–30, 163–4, 180–81, 234, 235, 236, 240, 243, 245, 265–75 primary production, 16, 26, 132, 184, 190–92, 199–200 protozoa, 243, 248, 254 prymnesiophytes, 126 Prymnesium parvum, 126 Psammechinus miliaris, 20 raceways ﬂoating, 281 land-based, 4, 24, 27–9, 225 Rachycentron canadum, 8 rainbow trout (see Oncorhynchus mykiss) recreational ﬁshing, 15, 160 reduction–oxidation (redox) potential, 170, 194 Refa cage, 294–6, 297 remote sensing, 112, 113 Renibacterium salmoninarum, 132 reservoirs, 14, 16, 18, 25, 115, 120, 123, 131–2, 136, 149, 150, 159, 184, 186, 188, 196, 199, 276, 280, 298 Saguling, Indonesia, 17, 25, 278 Selatar, 16–17 resource consumption, 159–63 Rhizophora spp., 85–6, 253

Index

river, 9, 17–18, 20, 121, 132, 148, 149, 181–2, 241 Mekong, 148, 160 Po, 149 Tengi, Malaysia, 18 rohu (see Labeo rohita) rope, 9, 67, 83, 93, 97–9, 100–104, 284, 291 Russia, 129, 241, 299 safety, 33, 94, 95–6 saithe (see Pollachius virens) Salacca glabescens, 37 salinity, 47–8, 111, 112–18, 121, 128, 130, 131, 133, 156, 157, 158, 206, 243, 257–64, 288 Salmo spp., 6, 19, 33–7, 91–2, 126–7, 173, 175, 203, 208, 211, 213–14, 217, 220, 225, 245, 254, 264, 278, 294, 299 S. salar, 7, 14, 15, 20, 24–7, 114, 122, 126, 129, 132, 136, 152, 155, 158, 161, 165, 169, 171–2, 174, 175–8, 179–83, 204–7, 208, 209, 213, 218, 223, 225–6, 227, 230, 231, 232, 234, 246, 249, 254, 264, 266, 267, 268, 270, 274, 275, 276, 281, 296, 298, 306 S. trutta, 129, 305 Salmonella, 217 Salvelinus alpinus, 20, 209 sar (see Diplodus sargus) scavengers, 174, 265–75 Scolelepis spp., 170 Scophthalmus maximus, 22, 36, 213 Scotland, 7, 15, 20, 24, 25, 58, 74, 87, 117, 122, 126, 128, 130, 131, 133, 153–4, 158, 176, 178, 179, 181, 182, 192, 216, 229–30, 249, 250, 264, 266, 267, 268, 271 sea bass (see Lates, Dicentrarchus) sea bream (see Sparus) sea horse (see Hippocampus) sea lice (see Lepeophtheirus and Caligus) sea lion (see Eumetopias jubatus, Zalophus californianus) sea urchins (see Psammechinus miliaris) seagrass (see Posidonia) seal scarers, 274 seals (see Phoca spp.) seawater, 112–17

367

security, 30–31, 33, 111, 155, 265 sediment(s), 96–7, 150, 151, 173, 174, 194–5, 196, 197, 239, 282, 287 chemistry, 169–72, 173, 174–5, 176, 182, 186, 188, 190, 254, 276–8, 280–81 seed, 15, 26, 27, 161, 201 seiche, 149 Seriola quinqueradiata, 6–7, 19, 24, 30, 34, 35, 53, 136–7, 161, 209, 211–12, 218, 234, 235, 252, 254–5, 276–7, 294 sewage-fed systems, 16, 17 shag (see Phalacracorax) shelter, 9, 20, 22–3, 26, 32, 83, 87, 94, 96, 97, 111, 137–45 shooting, 275 Siganus spp., 16, 19, 161, 254 silver carp (see Hypophthalmichthys) Singapore, 16, 19, 24, 133, 152 site(s), 111–58 Skaggerak, 126 skeletal deformities, 145, 229 skin, 26, 128, 129, 229, 244, 246, 304–7 snakehead (see Channa) snapper (see Lutjanus johni) solar radiation, 114–18, 300–307 Southeast Asia, 15, 69, 148, 204, 232, 246, 265 Spain, 124 Sparus aurata, 22, 27 Squalus spp., 266, 271 squid, 265, 266, 268 Sri Lanka, 21 steel, 9, 45, 49, 50–52, 63, 74, 75, 76, 81, 83, 85, 87, 89, 91, 100, 102, 255, 256, 271, 273, 288, 295–6, 297 stock(ing), 20, 30, 33, 58, 147, 150, 159, 201, 207–9, 227, 244, 249 densities, 18, 207–9, 213, 214, 217, 226, 230, 241, 250, 254, 278, 299 storm damage, 31, 145, 179, 265 Straits of Penang, 128, 132–33 stratiﬁcation, 114–15, 116–17, 119, 120, 128, 256, 264, 276 stress, 52, 112, 196, 201, 202, 205, 207, 208, 226, 229, 230, 234, 236, 244, 246, 248, 249, 250, 267, 269–70, 306 substrate, 151 Sweden, 178, 186, 279

368

Index

swim bladders, 298 Tagifuku rubripes, 179 Taiwan, 19, 136 Tapes spp., 22 Tassie cage, 50 teﬂubenzuron, 176 telemetry, 32 temperature, 59, 111, 112–18, 119, 130, 133, 135, 136, 167, 206, 211, 216–18, 225, 227, 232, 243, 256, 257, 258–9, 298–300 tex, 40 Thailand, 22, 38–40, 153, 211 thermal discharges, 117–18, 120, 122, 149 Thunnus thynnus, 22, 35, 161, 205, 211, 219, 252 tides, 145–9, 150, 167, 246, 250 tilapia (see Oreochromis, Tilapia, Sarotherodon) Tilapia spp., 245 transport, 201–7 trapping and removal, 275 Triaenophorus nodulosus, 129 Tridacna spp., 15 Trident cages, 296–7 tuna (see Thunnus, Euthynnus) turbidity, 121, 167, 304 turbot (see Scophthalmus maximus) turtles, 265, 266 typhoons, 134–6 UK, 129, 138–9, 147, 160, 266, 269 United States, 18, 21, 22, 29, 91, 124, 148, 182, 266, 267, 275 Uria aalgae, 271 vaccines, 175, 245, 250 vandalism, 31, 111, 179, 265 Vibrio spp., 126, 128–9, 175, 245, 248 video monitoring, 32, 165, 207, 220, 227 Vietnam, 7, 8, 150, 160 viruses, 243, 244, 270

169–72, 194–5, 239, 275, 276, 278 chemical, 172–8 collection, 230, 278–80 dispersal, 29, 96–7, 145, 158, 167–9, 170–72, 241, 264, 280–81 excretory (see urinary) faecal, 164 fertilizers, use as, 281 loadings, 120, 164, 183, 230 origins, 164 quantifying, 164–6, 230 uneaten food, 20, 164, 212, 221, 224, 225, 241, 267 urinary, 164 water density, 55, 80, 114–16, 117, 299 exchange, 33, 34, 38, 52, 55, 58, 59, 60, 97, 111, 112, 128, 130–32, 145, 149, 194, 235, 240, 252, 264 hyacinth (see Eichhornia crassipes) quality, 111–22 use in aquaculture, 2, 159–60 wave(s), 55, 76, 79–80, 96, 137–45, 146, 224, 281–98 forces, 55, 71, 74, 77, 79, 80–81, 82, 83, 93, 95, 96, 100, 137, 138, 139, 281–2 height, 79, 80, 137, 138, 139, 140, 141, 145, 157, 281, 282, 286, 288, 291–2, 294, 296, 297 length, 79, 80, 82, 281, 282–3, 286, 291–2 models, 142–4, 145 period, 79, 80, 81, 82, 95, 96, 137, 138, 139, 286 weather, 11, 68, 94, 111, 134–6, 180–81, 205, 241, 281–98 weight/size determination, 227 welfare, 33, 111, 201, 209, 229, 234 White Fish Authority, Scotland, 7 wind, 63, 74, 77–8, 79, 97, 134–6, 137, 139–41, 294 wrasse (see Crenilabrus, Ctenolabrus) yellowtail (see Seriola quinqueradiata)

wastes (see also models), 15–18, 20–22, 30, 92, 96, 128, 159, 164–83, 275–81 accumulation, 120, 130, 131, 150, 151,

Zalophus californianus, 266 zinc, 173 zooplankton, 17, 167, 196