PRODUCTION SYSTEMS FOR COMMONLY CULTURED FRESHWATER FISHES OF SOUTHEAST ASIA Report on a 1983 Workshop by James S. Diana -^^ ^ William Y. B. Chang^ David R. Ottey-^ Wiang Chuapoehuk-^ with a preface by Karl F. Lagler^* 2 •^Great Lakes Research Division and ^School of Natural Resources The University of Michigan Ann Arbor, Michigan ^Department of Aquaculture Faculty of Fisheries Kasetsart University Bangkok, Thailand International Programs Report No. 7 Great Lakes and Marine Waters Center The University of Michigan Ann Arbor, Michigan 48109 1985 PREFACE During parts of 1980 and 1981 it was my good fortune to be associated with my friend Dr. V. R. Pantulu of the UN Mekong Secretariat while he was directing adaptive environmental simulation modeling of fifteen years of management actions and outcomes in the Nam Pong basin in Thailand. From this experience I was stimulated to try a similar study of the dozen most important and commonly propagated freshwater fishes (including the giant freshwater prawn) in Southeast Asia. Were one to be successful in creating working simulation models for these species, teaching of aquaculture could be greatly facilitated and improved, needed research could be identified, prioritized, and undertaken, and opportunities for transfer of key information to commercial producers could be pinpointed and extension thus enhanced. Working with colleagues of the Aquaculture Department, Faculty of Fisheries, Kasetsart University, Bangkok, Thailand, led by my long-time colleague Dean Mek Boonbrahm, the investigational project was formulated of which this document is a part. The project was funded in 1982 under the title "Strengthening of Southeast Asian Aquaculture Institutions." Because of my conflicting involvement in other projects, principally in Africa, various stages of the implementation leadership of this project fell upon the capable shoulders of Dr. James S. Diana of the School of Natural Resources, University of Michigan. His co-authors in this present report were his principal associates, but the entire Faculty of Aquaculture at Kasetsart University shared in the study. They had previously worked vigorously in 1982 with me in planning the workshop reported herein and in issuing the invitations to, and in making the local arrangements for it. Although some of the objectives of this overall study have not yet been realized, the present document contains much well-organized information of considerable value. It is presented herewith for its usefulness and for its promise toward the completed study. Banjul Karl F. Lagler The Gambia Principal Investigator 5 December 1983 iii CONTENTS Preface ill Workshop Agenda vili Chapter 1 . INTRODUCTION AND GENERAL DESCRIPTION 1 Rationale for Workshop 1 General Fish Culture 2 History of Fish Culture in Southeast Asia 2 Major Differences in Culture Systems 4 Production by Each Country • 7 Institutions Involved. 9 Chapter 2 . TROPHIC DYNAMICS 10 Background • 10 Model Types 11 Bioenergetics Models • 13 Extensive Systems • • 14 Intensive Systems • 16 Intensive with Oxygen 16 Chapter 3 • GRASS CARP (PLA CHOA) • 17 Introduction • 17 Feeding Habits 18 Growth Rates 22 Reproduction 23 Culture Systems 25 Limiting Factors 27 Model Networks 28 Chapter 4, SILVER CARP (PLA LIN) 31 Introduction • • • .••.••••• 31 Feeding Habits « 31 Growth Rates 32 Reproduction • 34 Culture Systems « 34 Limiting Factors 36 Model Networks • >••••• 37 Chapter 5 . BIGHEAD CARP (PLA SOONG) 40 Introduction 40 Feeding Habits 40 Growth Rates 40 Reproduction 41 Culture Systems 41 Limiting Factors 42 Model Network 42 Chapter 6. NILE TILAPIA (PLA NIN) «««•«• e ••• o c « o ««•« e « « « 45 Introduction • 45 Feeding Habits 45 Growth Rates » 46 Reproduction «•...* 47 Culture Systems • . « « . 47 Breeders and Fry . • • « « • 48 OrOW'^OUu JrOnCLS •oaeoeeee>oeeoeeeoeeae«.e».«oe«<»e*.9«e«c«eoeeeeee DvJ L/Uner V^OUnuneS •«oeoe.eoe«oe.»e»«..«..*ee««oeoee.««ee«e.ee««ee D JL Limiting Factors « •. .•..•••.... • « 51 Model Network • . « 53 Chapter 7 • GIANT FRESHWATER PRAWN (KUNG KAM-KRAM) c e . . 55 Introduction • o 55 Feeding Habits « 55 i^rOWrn lA.at6S •ee*o»e«eeeee.«...e.e<>oo.o.....«..eceee...Geee.«.»s.e.. DO Reproduction. . « . • o 57 Culture Systems « 58 Limiting Factors 59 Model Network 59 Chapter 8 . SNAKESKIN GOURAMI (PLA SALID) 62 -LUL i OQUC L ion • ••••«<>•«••••. •.«.e«ee..e.e9eoo.. .«....«.. ...... ....... O^ Feeding Habits 62 Growth Rates 62 Reproduction. 63 Culture Systems 63 Traditional System 63 New Methods • « 64 ■uimiLing f aCuOrS •ec.eee..see...e«...ee9.e«e...««c..e.ee.o. ««.«...•• OD Model Network • . « « 66 Chapter 9 . TAWES (PLA TAPIEN-KHAO) 68 Introduction 68 Feeding Habits « 68 wrOWUn .KaUeS ....... .««.«e««...e.ee.eaeeeee««e..eoe. ..«..«.. .e...... OO i^eprOuUC nOn .•eoe««oee.»e«.. •'. eee*eeeoeee.«.e.eeeeeoeeeei».eee.ee.«« O^ UUXrUre OyStemS ee>e.««>eeee«oee«eeeoee.eeoeeeo...ocoe«ecoeo. ......... /U i^r eeQ xng •«.... .......e....... ............ ...... ...... ......... /u Nursery 70 Grow~out Ponds • « • 70 Limiting Factors 71 Model Networks 72 Chapter 10. SNAKEHEAD (PLA CHON) 75 Introduction 75 Feeding Habits 75 Growth Rates 75 Reproduction 76 Culture Systems 76 Nursery • 76 Grow-out Ponds 77 Limiting Factors 78 Model Network 78 VI Chapter 11. SAND GOBY (PLA BU-SAI) 80 Introduction 80 Feeding Habits 80 Growth Rates 80 Reproduction • 81 Culture Systems > 82 Limiting Factors 84 Model Network , 85 Chapter 12. WALKING CATFISH (PLA DUK-DAN) 87 Introduction 87 Feeding Habits 87 Growth Rates 88 Reproduction • 88 Culture Systems 89 Brooding . . • « c 89 Nursery 90 Grow-out Ponds 91 Other Systems 92 Limiting Factors 92 Model Networks 94 Chapter 13. SUTCH'S CATFISH (PLA SAWAI) 96 Introduction 96 Feeding Habits 96 Growth Rates 96 Reproduction 97 Culture Systems 97 Brooding 97 Nursery. . . . « 97 Grow-out Systems 98 Limiting Factors 99 Model Network 100 Chapter 14 . SUMMARY 102 Introduction 102 Miajor Limiting Factors 102 Major Problems and Status of Solutions 104 Grass Carp 104 Silver Carp and Bighead Carp 104 Nile Tilapia 104 Giant Freshwater Prawn 105 Snakeskin Gourami 105 Tawes 105 Snakehead and Sand Goby 106 Walking Catfish. 106 Sutch's Catfish 106 Summary of Research Needs 107 References • > 109 Appendix 1. List of Workshop Participants 115 Appendix 2 . Workshop Agenda 118 vii WORKSHOP AGENDA The workshop involved scientists from the U.S.A., China, Malaysia, the Philippines, and Thailand. The scientists were chosen for their knowledge and experience with one or more of the selected species. Workshop participants and the addresses of their institutions are listed in Appendix 1. Workshop activities, which included discussions, subgroup meetings, demonstrations, exercises, and field observations, are indicated in Appendix 2. The goal of this text is to outline the state-of-the-art for culture of the commonly cultured freshwater species in Southeast Asia. While a goal of producing computer models of these systems is also a part of the original project, these results are not included in this text. The computer models and some simulations of these models will be reported in a later document. The present text contains a general description of the major components of culture systems, followed by descriptions of common species* culture techniques . Vlll CHAPTER I. INTRODUCTION AND GENERAL DESCRIPTION Rationale for Workshop It is well recognized that more than half of the world population suffers from malnutrition due to protein deficiency. This condition can be partially alleviated by increasing animal protein production from inland aquaculture. However, there are still many technological constraints which maintain low production in most inland aquaculture systems in Southeast Asia (Lam 1982). The Agency for International Development (AID), The University of Michigan, and Kasetsart University initiated research in 1982 to attempt the infusion of new knowledge and approaches into the solution of this problem. This text represents the outcome of the majority of that work. The text describes the results of a workshop, held 15-30 April 1983, on the state of the art for the commonly cultured freshwater fishes of Southeast Asia. Additionally, the text defines energy pathways which describe the major energy and chemical transfers through these ponds. The ultimate goal of this workshop was to devise means of reducing the principal technical deterrents to future inland aquaculture in Southeast Asia. The specific objectives were to complete the following goals for the 11 most commonly-cultured species in the area: (1) to identify key points of known technology that may be of particular value to fish farmers but which may not be reaching them; (2) to identify and establish priorities for the critically needed areas of research that are practical and production oriented; (3) to use simulation modeling techniques to help in achieving (1) and (2) above, and at the same time to provide new tools for teaching, learning, and research planning in aquaculture; (4) to cooperatively initiate the most valuable and practical research targeted at improvement of pond production of the selected 1 species; and (5) to publish and disseminate the results by both established and innovative means, including a document on the outcome of the workshop. General Fish Culture History of Fish Culture in Southeast Asia The culture of fishes in Southeast Asia has been practiced for at least 3,500 years (Ling 1977). While the region in question includes Indonesia, the Philippines, Malaysia, Singapore, Vietnam, Cambodia, Thailand, Burma, and southern China, our workshop included only members from China, Thailand, Malaysia, and the Philippines. Therefore, our major conclusions are drawn from work in those areas, and expanded to the region. Early culture techniques evolved through a series of steps from trapping of fish, to trapping and holding fish live to maintain freshness, to even- tually trapping, holding, and then growing the fish to a larger size. As the harvestable fish varied in species and sizes, the culture techniques often varied. However, the early Chinese culturists realized that using one dominant species in a pond increased their yield. The culture of common carp continued on this premise. Around 500 BC the culture of common carp was widespread, and the creation of semi-natural breeding conditions allowed the dependable production of fry (Ling 1977). Their technique was to stock ripe male and female carp into ponds, then allow the pond to develop its ox^n population. The control of stunting by predator introduction was also commonly practiced. The culture of common carp in China continued in much the same fashion for a thousand years. Around 600 AD, common carp culture was sharply curtailed by the Tang Dynasty. A search was begun for replacement species, and four additional native Chinese carps (silver, mud, black, and grass carp) were chosen as 2 substitutes • Original methods had always used common carp in monoculture, but now these four species were all grown together. Each species utilizes different foods in the ponds, and eventually polyculture proceeded to the point that each species was stocked in rough abundance related to the production of each type of food in a pond. The total production from ponds with polyculture systems was much higher than yields formerly obtained under monoculture systems with common carp. Application of waste materials to the ponds as a fertilizer or feed was also practiced, but to an unknown extent. This eventually gave way to integrated systems, where the wastes from chicken, pig, or duck rearing were added to the pond to increase primary productivity. Other agriculture wastes (vegetable or silkworm offal) were also added to ponds as feed. Thus, the carp culture systems in the region are very old, with long-established methods for stocking and fertilization of the ponds. The procurement of seed to stock grow-out ponds generally involved col- lection of eggs or fry from the wild. The major limitation to the widespread culture of many fishes, particularly the Chinese carps, was seed availability. The only suitable method of seed production involved the extensive transport of fry* This was particularly true for grass carp, which possess reproductive site requirements (running water with temperature conditions between 20-24°C), that are not available in many countries. Even today, the collection of fry from the wild remains one of the most common techniques available for many species (such as the sand goby, see Chapter 11, or the snakehead, see Chap- ter 10 )o The limited supply of seed in natural waters, and the unavailability of fry at certain times of year, led to the brooding of fish in semi-natural systems to produce fry. These methods usually involved creating an environ- ment suitable for reproduction, then stocking the ponds at a controlled density and collecting fry at regular intervals* The combination of this method with supplemental feeding has expanded the breeding season for several species, making fry availability more predictable and more dependable. While this system was used for common carp ca. 500 BC, it has not been ex- panded to many other species until fairly recently. Natural production of fry using brood ponds is commonly practiced for several species today, including Tilapia , and snakeskin gourami. A most important development which expanded the geographical extent of culture for several species has been the elaboration of artificial propagation techniques, particularly induced maturation. Hypophysation and hormonal manipulations have been important in the culture of the Chinese carps, par- ticularly the grass carp. In 1954, the induced maturation of carp by use of hormones was a major breakthrough (Bardach et al. 1972). This has allowed the use of hatcheries to produce fry, using flowing-water systems to simulate the early life requirements of grass carp fry. Current research on several species (such as the snakehead and sand goby) whose fry are now collected in the field may soon result in artificial propagation of these species by induced maturation. Major Differences in Culture Systems Before elaborating the details of the culture most commonly used in Southeast Asia, a general background and definition of terms is needed. Fish culture has been practiced in the area for a very long time, but low productivity systems with traditional methods are still in use today (Lam 1982). The amount of external inputs into a fish culture system can vary from 4 an extensive system, where natural production of fish foods through internal cycles predominates, to an intensive* system, where considerable amounts of fertilizers, feeds, and other external products are added to increase production levels. Of course, in most culture the intensity is related to many factors, including the income of the farmer, the value of the product, the cost of the supplemental materials, and the traditional values of the farmer • While the reality of economic incentives cannot be ignored (see Chapter 14), there often is the potential to improve culture systems without disturbing religious or ethnic customs. It is within these boundaries that we must set our production goals. In addition to the input of materials, there are several other levels of intensity which are commonly encountered in aquaculture systems. Fishes may be groTim in monoculture or polyculture. Tlie combination of species by chance processes, generally due to mixture of seed or flooding with some native fish still in the pond, probably results in little change in overall production of the system. However, the intentional combination of several species, with the purpose to produce a forage and predator fish population, to use different components of the energy available in the pond ecosystem, or to improve water quality, may result in substantial improvements in overall yield. Examples of this sort of manipulation include the culture of Chinese carps, walking catfish and Tilapia , and many other combinations. The widespread polyculture of carps has made it difficult to find information on adult monoculture of the three carp species examined in this text. Indeed, the data included in Chapters 3 to 5 are really considered for those fishes in polyculture, except in the case of fry culture. For most of the other species, monoculture was the most common technique used. While the basic problems for most fish culture systems are fairly similar, there are species-related differences in culture dependent on eco- logical differences. One example we have already mentioned is the method for fry production: either production by artificial means, or semi-natural sys- tems, or collection in the wild. Obviously, the availability and abundance of seed sets upper limits to fish culture for any species. However, many other factors are also involved in this limit. The production of fry suitable for stocking in a grow-out pond is also often a limit to production. An excellent example of this is the sand goby. Although semi-natural methods are available for producing fertilized eggs (see Chapter 11), the rearing of these young from hatching to a suitable stocking size is very problematic. The production of suitable food for the minute fry, without also producing other sizes of zooplankton which may eat the sand goby fry, is still highly experimental. Even for commonly reared and cultured fishes, the fry stage may often require different techniques and different systems than the growing of adults . A cursory examination of the information in each chapter indicates that there are separate fry culture systems for eight of the eleven species examined in this text, and for those without distinct fry culture systems, two species (snakehead and sand goby) are collected in the wild as advanced fry or adults. Economic incentives for increased production vary considerably from re- gion to region. For example, in Thailand a major cultured and sold fish is the walking catfish, and improvements in culture could have tremendous economic benefits. The same is true for Tilapia in the Philippines, and carp in China and Malaysia. The widespread application of improved techniques is unlikely for the species that are of local concern. However, several species 6 which are cultured in the region (sand goby and Macrobrachium ) have high export value, as well as value for local consumption, which makes the wider application of improved technology feasible and profitable. Production by Each Country As mentioned previously, there are very different factors influencing the local consumption and production of fish species in Southeast Asia. Generally, the popularity of culture of a species is related to its local value or consumption. These local favorites vary by country in the region. The walking catfish, snakehead, and snakeskin gourami are predominant in Thai fish culture, for example, while the carps are most important in Chinese culture. Local variations in the region (Table 1-1) are often hard to evaluate statistically, since the data collection effort is uneven between sites. Sources documenting world production, such as FAO, do not always cover each country adequately. However, it is clear from Table 1-1 that Indonesia ranks highest overall in total fish production, and that the culture there is dominated by carp, snakehead, snakeskin gourami, and Puntius . Culture in the Philippines is much lower overall, and is dominated by Tilapia . This table is very limited, as data for China are unavailable, and China probably dominates the regional aquaculture production. Data for Malaysia are also limited. It is also important to note that this is for freshwater fishes only, and the culture of milkfish and other marine species could largely affect this production ranking. Finally, the data are compiled for cultured fish as well as wild-caught fish, and the relative importance of each is unknown. TABLE 1-1. Annual production statistics for various freshwater fishes in sev- eral countries in Southeast Asia. Data (FAO 1983a) are rounded to the nearest ten metric tons, and refer to landings on a wet weight basis. Fish Country 1977 1979 1981 Clarias species Indonesia Philippines Thailand 320 1,620 19,100 420 1,020 21,450 620 2,930 22.260 Cyprinus carpio Channa striatus Indonesia Philippines^ Thailand Indonesia'^ Philippines Thailand 39,950 5,740 1,100 35,210 3,250 22,060 35,820 9,710 2,040 37,160 3 , 040 23,760 53,300 13,600 2.200 Oxyeleotris^ marmoratus Indonesia 510 830 780 Macrobrachium rosenbergii Indonesia 2,760 3,690 3,740 37,630 8,040 28,000 Puntius gonionotus*^ Indonesia Thailand 29,420 11 , 000 39,140 12,840 42,940 15,980 Tilapia nilotica Indonesia Philippines® 3,610 12,990 4,830 6,120 5,960 27,850 Trichogaster pectoral is Indonesia Thailand 24,040 16,860 22,270 19,190 23,140 20,520 ^ Philippines landings refer to various cyprinids. ^ Reported as Eleotridae. ^ Indonesian landings refer to Channa species other than £. micropeltus < ^ Reported as Puntius species • ^ Philippines landings refer to Tilapia species. Institutions Involved Fish culture originated as a local farm initiative, and government agencies and other groups have only recently become involved (Ling 1977) • As such, the culture progressed mainly by tradition and trial and error. Few attempts were made to adopt culture techniques or methods developed in other countries. Rather, local species and culture techniques often differed considerably. Only recently has the spread of information from other countries been attempted, and culture techniques in each country still lag behind in incorporating current knowledge. Local governmental agencies have often developed to extend this new information to the farmers they serve. Similarly, the production of fish fry often fell into the hands of the farmers themselves. With the advent of governmental agencies, development of new culture techniques for fry, as well as production of fry which formerly were limited or difficult to produce, have become their major responsibility. Currently the Thai Department of Fisheries, as an example, runs at least 20 fishery stations which produce fry for purchase by local farmers in freshwater areas (Tarnchalanukit , personal communication). These stations also research production difficulties, such as rearing of sand goby fry. The intervention of government agencies into the culture practices of several countries has definitely improved the availability and the quality of fish seed. The ex- ample of Thailand is probably typical of fish culture stations in most of the region. CHAPTER 2. TROPHIC DYNAMICS Background The general approach used in this text to evaluate, interpret, and ulti- mately model the pond culture systems of Southeast Asia is one of trophic- dynamics. Trophic-dynamics was first formulated by Lindeman (1942) as the study of energy transfer through communities. His pioneering work synthesized and organized basic energetic concepts into an ecological perspective. He realized that energy transfer through trophic levels has an inefficient nature, and this idea has been supported by many later studies. Lindeman' s research led to many studies of animal conversion efficiency, as well as trophic level transfer efficiencies. These studies become progressively more complex as one moves from individual to population to community levels. On one end of the spectrum, many studies have attempted to relate fish production in lakes to primary productivity (see Adams et al. 1983 for review). These studies are generally undertaken to determine the relative importance of allochthonous (externally produced) or autochthonous (internally produced) energy in the function of aquatic systems. The importance of energy sources differs with each system. Indeed, energy sources are also major differences between the intensive and extensive pond types described in Chapter 1. The transfer of autochthonous energy between trophic levels has not been estimated with much certainty. This transfer involves many steps, particu- larly the cropping efficiency of an animal and its growth conversion effi- ciency. These steps are in turn under the influence of many environmental factors, such as season, temperature, and maturation stage. Therefore, one would not expect the trophic efficiency to be similar over time or within 10 different ecosystems (Slobodkin 1972). There is presently little empirical evidence available on ponds or other aquatic systems, yet it has become common to model trophic dynamics using a 10-15% efficiency of transfer between trophic levels (Kozlovsky 1968, Adams et al. 1983). This area of trophic efficiency needs much future effort, particularly in aquaculture, if we are to truly understand energy flow in extensive culture systems. Tbiis chapter will review the approach used in the text to model the three main types of aquaculture systems found in Southeast Asia. Model Types Models of fish production vary tremendously in their complexity and general applicability. Probably the simplest general model of fish production in inland lakes is the morphoedaphic index (Ryder 1982). This index uses mean depth and total dissolved solids to predict fish yield. While it has been useful for some first applications to new or unknown systems, it is not very useful in pond culture systems where these parameters (depth and dissolved solids) vary little between ponds. Similarly, many of the main models used in capture fisheries are not of much use in aquaculture. The dynamic pool, surplus production, and stock- recruitment models (Ricker 1975) utilize the tradeoff in fish growth and either reproduction or mortality to determine optimal fishing strategies. These strategies seldom are complete harvest at one point in time, yet this is the most common practice in aquaculture. Therefore, the models consider many factors which are of little concern in aquaculture production. An empirical model which is widely used in aquaculture is feed conver- sion. This model estimates fish growth based on a conversion ratio (grams of food eaten per grams of fish growth), and a known feeding level. This is 11 useful in supplemental feeding situations in a local area. However, it is not widely applicable; values determined in one area for one food type do not transfer to other areas or foods. Factors such as temperature, fish metabolism, assimilation efficiency, fish size, and water quality all strongly affect gross conversion (Brett and Groves 1979). Thus it becomes necessary to evaluate these physiological parameters to relate conversion efficiency information to other sites. Models which include these physiological processes are termed bioener- getic models. Their development traces back to Brody (1945), Kleiber (1961), and Winberg (1956). Computer simulations of bioenergetics for fish are much more recent, and generally follow the work of Kitchell (Kitchell et al. 1974, 1977). Kitchen's models use scaling processes for temperature, body size, assimilation, activity, and feeding rate to estimate a growth rateo These models can be linked with population processes (birth rate, mortality) to predict yield. They can also be freely transferred between regions providing basic environmental parameters are known. We have chosen to use Kitchell' s models for our evaluation of fish pro- duction in aquaculture ponds. The advantages of these models also include the ability to model oxygen levels in ponds, based on fish metabolism, oxygen production by aquatic plants, and respiration by plants and other animals. The primary production and other respiration components of these models are not from Kitchen's models, but rather from the work of Boyd (1981) and others. 12 Bioenergetlcs Models Tlie major portion of a bioenergetic model is the process of fish metabolism. Fish metabolism can be estimated by the following equation (Rice et al. 1983): Qs = a . W^ e mT where Qs = standard metabolism (mg 02/kg/hr) W = body weight (kg) a = constant for standard metabolism-weight equation at 1°C b = exponent for standard metabolism-weight equation m = temperature coefficient T = temperature CO This metabolism can be converted to kcal/day using an oxycalorific coefficient (3,22 kcal/mg O2 consumed, Brafield and Solomon (1972)). Fish growth in the bioenergetic model is the difference between energy ingested and energy used in metabolissm. It is estimated by the balanced energy equation (Webb 1978): Qg = Qr - (Qs + QsDA + Of + On ^' Ql) where Qq = growth (all units below in kcal/day) Qr = ration ingested Qp = feces Qjj = non-fecal energy loss QsDA ~ apparent specific dynamic action Ql = cost of locomotor activity. 13 Qf> Qn> ^^*^ QsDA ^^^ ^^^ considered to be a constant fraction of the ingested ration. The feeding component of this model compares a measured available ration (either food fed to the fish or natural food produced) to the maximum total consumption for the fish. Maximum consumption is in turn calculated from the maximum growth rate (which is commonly available for most species) by the formula: Qr^^^ = QCmax + Qs + QsDA + Qf + Qn + Ql where QRmax ~ maximum ration ingested (kcal/day) QCmax ~ maximum growth rate (kcal/day) and the other parameters were previously defined. These models were applied, with species-specific values when possible, to three different classes of pond systems: (1) Extensive systems, (2) Intensive systems with no concern for oxygen levels, and (3) Intensive systems with oxygen as an important parameter. These systems models are described below. A general production model for pond systems is given in Figure 2-1. Extensive Systems These culture systems occur for fish which are not directly given supple- mental food, such as bighead carp, silver carp, Tilapia , and snakeskin gourami. Natural food production dominates these systems, and due to the limited food available, stocking densities seldom reach levels sufficient to cause low oxygen problems. The natural food pathways may differ depending on the food habits of the cultured fish. Our models of this class are for the two carps mentioned above. 14 o 2 I P 2 Z o ■"^ "o" H- 2 U) cn O § 8 q: u. (T GL Ui 2 UI 1 _J O h- O qE Q. S ^ i C ^ 2 2 O .f=^ o cr o ^ s\NN\\\N\\\\SS\\ \\\\\\^ nX^ ^ \^v^ ^\\\\\\\V X^ O O o r-? ,lii -^LiL "N 2 o O z < o -J H- ^ UJ UL -J UJ a. 5 o 2 o ^ 3 2 X w ^ O z o I z I I o I 1 J2 ^ O UJ -J < z UJ (/} s a UJ UJ «J UJ a. u. a. 3 en 2 a )i ii z g 2 >- I U ^ (o o S iT >■ 3 **• X cn o z o o < O Z Q 5 i*i z J:? O y ^ CO S 0) 4J w Cf3 o CU u o 0) O o T3 O J-i a CO a; c O S I CM M 15 Intensive Systems These culture systems occur for fish that are cage cultured in flowing water (sand goby, Sutch*s catfish) or fishes that are air breathers (walking catfish). The models are usually very simple, with supplemental food being used and fish energetic processes converting that food to body growth. Mortality functions may be constant or density-dependent. These models are probably the best predictors of all that we have developed, and are exemplified by the walking catfish. Intensive with Oxygen Most intensive systems fall into this category, including those for grass carp and giant prawns. These models include natural oxygen production by phytoplankton and respiration by mud, plankton, and fish. A minimum oxygen level is predicted for dawn of each day. This level sets the mortality function, and itself is set by stocking density and size. The minimum oxygen levels are highly dependent on wind velocity in nature, while our models do not include these local events. Also, uneaten supplemental feed decays at a rate which consumes oxygen, yet is not well understood or quantified. Thus our models give conservative stocking densities. We have computed models of this class for grass carp and giant prawns. 16 CHAPTER 3. GRASS CMP (PLA CHOA) Introduction Culture of the grass carp, Ctenopharyngodon Idella , was first recorded in China during the Tang Dynasty (618-907 A.D). Fish culturists were historically limited to naturally produced grass carp, as fry collected from rivers and streams were the only source of seed for culture. The culture of grass carp has also long been limited by means for transporting live fry. Chinese farmers who settled on Taiwan 300 to 400 years ago brought with them the practice of pond culture, based on the annual importation of carp fry from the Chinese mainland. This method of propagating the grass carp was used throughout Southeast Asia, Japan, and parts of the Soviet Union before the introduction of artificial induction of spawning. Along with other Chinese carps ( Aristichthys nobilis , Hypophthalmichthys molitrix , and Cyprinus carpio ) , the grass carp was thus an exotic introduction which became established on the Malay Peninsula (Merican and Soong 1966). In the early 1960s in China, the development of techniques to induce spawning of Chinese carp by hypophysation eliminated dependence on wild stocks (Lin et al . 1980, FAO 1983b). In Malaysia, attempts to induce breeding of Chinese and Indian major carps were being evaluated by the mid-sixties (Merican and Soong 1966, Chen et al. 1969). Within a decade, the grass carp became the subject of experimentation and acclimatization in a number of western countries, including Bulgaria, Czechoslovakia, France, Hungary, Iraq, Israel. » Poland, Romania, the United Arab Republic, the United States, West Germany, and Yugoslavia. 17 Feeding Habits The adult grass carp is a facultative herbivore. It will eat nearly any organic matter, but prefers green vegetable food. The feeding habits during early life stages are considerably different from those characterizing the adult. Three days after hatching, at about 0,7 cm in length, the fry are zoo- plankton feeders, with a diet similar to young silver carp and bighead carp (Table 3-1). When fry reach a size of 1 cm, the diet consists mainly of water fleas ( Daphnia) , copepods, and rotifers. They maintain this component, but as they grow from 1 to 3 cm in length, benthic insect larvae such as Chironomus and fragments of plants are included. Above 3 cm, the intestine is about twice as long as the body, and pharyngeal teeth suitable for cutting higher plants are developed. At this point their feeding habits change, and they begin to feed more on leaves and sprouts of tender aquatic plants such as Lemna , Wolff ia , and Hydrilla . Grass carp more than 10 cm in length are capable of grinding and cutting grasses, and feed mainly on aquatic vegetation and tender land plants that reach the water. During intensive culture in captivity, many kinds of supplementary foods are added. These include various wild and cultivated plants such as potatoes, peanuts, rice and other cereals, bran of beans, soy dregs, brewer's grain, fruits and vegetables including the leaves and stems, plus silk worm pupae, earthworms, and entrails of animals and poultry. In China, supplementary food amounting to about 1 to 4% of body weight is applied twice a day, about mid- morning and again about mid-afternoon. There is considerable variation in both feeding rates and feed stuffs that are used (Table 3-2). Food conversion ratios differ accordingly (Table 3-3). 18 TABLE 3-1. Principal natural foods of Chinese carp fry. X indicates primary foods; + indicates secondary foods. Based on Bardach et al . (1972), Lin et al- (1980), Lin and Kangnian (1983), DeSilva and Weerakoon (1981), and FAO (1983b), Food Size of Fry (mm) 10- 12 13- 17 18- 23 24- 30 Protozoa Copepod nauplii Rotifers Copepods Cladocera Insects Detritus Algae Phytoplankton X X X + + X X X + X X + + -h X X X + X X X + + GRASS CARP Food Size of Fry (mm) 10- 12 13- 17 18- 23 24- 30 Protozoa Copepod nauplii Rotifers Copepods Cladocera Phytoplankton X X X + + X X X X X X + X X + + + X SILVER CARP Food Size of Fry (mm) 10- 12 13- 17 18- 23 24- 30 Protozoa Copepod nauplii Rotifers Copepods Cladocera Phytoplankton X + X + X X X X X X X X X X X + X + BIGHEAD CARP 19 d i cd rH O *J 9 C « o ^ o c o O J s o • Cd r^ 9i U ^ Vi O N-^ (d Q) ^ • -H «d fH c « (d o 9 s M a »^ 0) o to Kk4 U OD CO 4) W) M (d td td -u o u 00 « 41 0) OB « (H 4) OJ O. CCS O U M O 4) • CB 4J 00 ao o 6 C -"H •H JS fH V4 60 dO c C 4i -H e 9 Cd Cd o >MH V4 1^ O M^ • «*^ a 00 ^ )^ o o O O -H »W 60 OB I e 00 CD O «H 9 8 e j^ o 4) O U 00 JJ S O -H 00 *J T3 00 •H a 00 o 60T3 « x: C 4) «t3 00 •H W 4i ^ ^ Cd o t-t (d 4) 4) O « Wl k4 ^ 4) T3 M-J ^ >» C o i-i (d 00 Cd cd ** 4* CJ ^-^ Glc 00 o«cn a 00 >sao Cd C > fH 00 • • "^ W» -H ^ o o 00 tM a o o 000 v« v|<. Cd ^^ Cd 4J O O 00 •a 00 t3 o a c^ ^ -^ ^ • >* o Cd >» >» 1 O Cd )-i 4) Vi G U o — *^ o -^ O 00 >% O 00 O 00 U « 00 ^ O C ^^ M^ o c "^ S i5 'yS td r-< w. o 00 u 0*0 ^0'«^>.4) OO ^4) eva **^ •-^ 00 00 ** • 00 ri ^ ^ lOM •Hcdoocoo'Hcd 3 41 a o . f-i -H CO « "H c OS 5J ^ Cd p-» H ^ 00^' >k « u M 00 % »H 4J Cd 4) Cd Cd o a d 00 ^ •t3 « 2 (d 3 Cd a a ^ « ^ o o « iH *o a f-H d kH o >» u Cd a-TS ►k o « 9 Wl S a 4) 00 ^ d 4) ua 9 4) 00 f-i Cd 4J >« en to 00 u •H 4) M^ o u o a a (d en ^^ Cv* a 00 o C G o •*-< v-^ o 1 ^ 1^ U 41 O *J 4J (d en 0tf /-s 00 >t Cd •o o M-» 'w^ 1 en o V JC 4) CD 00 s s ^ O O d (d 00 o o u V 9 • O w o d o 9 iH Cd d M-t 1^ Wi •2 ^ o d 4) 4> 9 N 4) 4) ^ 00 H Cd •^ Cd (d d Cd "M Cd 4) « u M 4J O 9 d 0) d cd u Cd 4» "H M d 4) Cd ^ u Cd ,fi u 4) ^ u 9 "H d M Cd oa^uatot-)^ cuo I O o en V I o I o I o O o o I o *t3 OB d 00 Cd Cd u *« 00 4J Qi U ^ m o d 9 4J Q 4J f %0 O «-« I o so en to o a a d 4) o O to Cd •« « d ^ o •< to 00 4) •H O 4) a en >> d 9 o o I I a u •» M Cd •o Cd a 00 Cd .H •H M ^ 0Q O cn I e>j V I o o o 0» U cu M * i-i « V4 Cd T3 (d O TJ Cd o u 9i u x: 00 4) .d cB 4) 00 > 00 > 00 Cd 1-* 00 Cd r-^ pQ o cn A o cn ^ f-4 1 J t-4 1 m m SO o o o 0k fs| I o o o o o m I o o o o o o fn * a a a u u a u Cd • Cd u •T3 cd O T3 O Cd Cd o Cd o 4) M 4) U ad 00 4) X 4) 00 00 > > 00 00 « »H 00 »H Cd 'H U 'H •H -H u M O en A cn o o o o o m a u Cd u 00 00 Cd u o 08 d us o CO d ^ o 4) 9 d 20 00 o G 0) C/3 1^ CM s CO o c c M O 4) O 4J *J « CO ptf OS !k cd o 0) CD 60 •H « x: c ri ed O 4J ^** o ^ — 50 3 >» O « 00 4) (Q O "H 00O>^P.4)O r-liH 60*^ na 00 O 60 « U Q) ^^>jda)«j^oo s c V iH iH Cd •H C 4-1 Tl u 9t (d 0) c 4) -H jj M-l ^ **^ 3 CO •H ^ M-l C CD TJ *J iH Cd Cd c (d sa <£ N CD Vi CN C^J > »n •H OJ 1 1 > a > •-< i-( M to 3 ^ ua CO CO I a >-/ x: 00 *J TJ • a c « Q a« /-> ^ -• 8 «d T3 m QJ c •-4 fr4 < cu Oi 3 (3 a 5 CD o CO o 0U la « i-i Cd "C5 Cd o Cd o V u x: to Q> 00 > 00 Cd rH •H W "H 03 U CO a" 00 e o 8 a « u T> Cd Cd o us OB 00 00 Cd PQ O O 00 o o a 00 V o in • a a u u td U 73 Cd U Cd o V4 4) 1U Hi US 00 u > 00 > fH 00 Cd r-4 CO PQ C9 Cf) m in o CM o V CM /\ en ' o 00 »n mo u s Cd Cd 9C • a o. a M ta »-• Cd • Cd •a Cd o T3 o cd o Cd x: 00 4) j: 4) CO 00 CD > > 00 00 00 Cd iH oo»H (d Cd •H V- ^ -Hi-* |4 M 09 O CO A CO O CD 00 CO OS X (B ?!. JS Cd S s S ^ 21 TABLE 3-3. Food conversion ratios for supplementary feeds commonly used in culture of adult Chinese carp. Data presented during workshop discussions were augmented with values from Hora and Pillay (1962) and Lin et al. (1980), Type of Feed Food Conversion Ratio Silkworm pupae (dried) Pellets (20-25% C«P) Manures plus inorganic fertilizer (N:P:K (1:0.5:0.25) or NH4SO4) Peanut and soybean cake Barley and oats Rice sprouts Rice grain, rice and wheat bran Silkworm pupae (fresh) Silkworm feces Mixed vegetables, tender grasses Duckweed, Wolff ia Sugar cane leaves Pig and duck dung Vallisneria 1.1-2.1 1.5- 2.2 2 2.2- 4.5 2.6 3.9 4-6 5 - 5.5 17 33 -36 37 -41 40 43 -45 101 Growth Rates Grass carp fry grow at a remarkable rate. When reared in nursery ponds, they double their weight every other day during the first 10 days of life with daily increments of 0.01 to 0.02 g (FAO 1983b). Just after hatching, the fry are about 0.7 cm long, and average about 0.002 g. They attain a length of 1.3 cm and a weight of 0.1 g in about 12 days, and reach 5.7 cm and 1.5 g in 40 to 50 days . As growth continues , the relative growth rate decreases considerably compared to the early fry stage. Weight increases average about one-fold each 5 to 10 days during the early fingerling stage, but absolute weight increases rapidly at about 3-9 g per day during the period. For adults, the maximum annual growth in length occurs during the second year, whereas the maximum weight increment occurs during the third year. Under ideal rearing conditions, grass carp can gain 1 to 2 kg in the first year, 2 to 3 kg in the second year, and 5 to 6 kg in the third year. After the third year, growth in both length and weight decreases sharply. 22 Temperature is an important factor in the growth of grass carp. They usually prefer temperatures greater than 20*^0, and grow best between 22 and 25*^C. Below 15 °C, poor appetite occurs, and if temperatures fall below 8 to 10*^0, the fish stop feeding. As may be expected, growth rates decline sharply as temperatures fall below 20 ''C. Reproduction Age at which sexual maturity is attained varies greatly with climate and environmental factors. For example, grass carp attain maturity in about 1 to 2 years in Malaysia, but take about 10 years in the vicinity of Moscow (Bar- dach et al . 1972) (Table 3-4). Among the factors responsible, temperature has the most profound effects on maturation. A formula for estimating the age of first maturation of Chinese carps has been developed in China using accumulated temperature exposure. Expected age-at-maturity is equal to the number of annual growing seasons required to accumulate 5,000-15,000 degree- days, where degree-day accumulation is defined as the summed product of the number of days during which mean water temperature exceeds 15 °C, times the number of degrees by which these mean daily temperatures exceed 15 °C. The accumulated degree-day concept is similarly applied in central European carp culture to estimate the dates of annual spawning of breeder stocks (FAO 1976). There, a sum of daily mean water temperatures between 1, 300-1, 400 °C (silver carp), 1, 350-1,450 °C (grass carp), or 1,400-1, 500 °C (big head carp) within a calendar year indicates the attainment of maturity. The age of first maturity is thus closely tied to both the number of days annually when the temperature exceeds 15 °C and to average water temperatures during growing seasons. This relationship is currently applied in fish culture in Northern China, where fish are induced to mature at an earlier 23 TABLE 3-4. Age at maturation of grass carp in different geographical regions. Data were taken from Bardach et al. (1972) • Region Age (In Years) Males Females Malaysia India U.ScS.R. (Turkmenia) China (South) Taiwan China (Central) U.S.S.R. (Krasnodar area) China (Northeast) U.S.S.R. (Ukraine) U.S.S.R. (Siberia) U.S.S.R. (Moscow area) 1-2 <1 2-3 3-4 3-4 3-4 4 5-6 7-8 8-9 9 1-2 2 3-4 4-5 4-5 4-5 5 6-7 8-9 8-9 10 age by increasing the water temperature. Once they reach maturity, the reproductive life span can continue for the next 15 years. Average fecundity is about 100,000-140,000 eggs per kilogram of total body weight (Lin et al. 1980, Zainuddin et al. 1982). Absolute fecundity was estimated by the relationship : F = 41.4 W^*^^, where F is in thousands of eggs and W is body weight in kg, for grass carp stocks in Malacca, but these fecundities are roughly two to three times those observed in other Asian culture (Zainuddin et al. 1982). The rate of hatching varies depending largely on fertilization success (which ranges from 50-90% with an average of 85%). As a further limit to repro- duction, only some 90% of brooders are capable of spawning, and only about 85% of the females ovulate under typical pond culture conditions. 24 Culture Systems Culture systems for the grass carp can be described according to four life stages of the fish: brooder, fry, fingerling, and adult. For rearing and maintaining brood stock of the grass carp in China, ponds with a water depth of 1.5-2.5 m and an area of 0.20-0.45 hectares are recommended. Ponds with firm, flat sandy-soil bottoms are most suitable, as these permit easy removal and sorting of brooders. Adequate and nourishing food is critical following spawning to ensure healthy brooders for over- wintering. Foods such as or bean cake are supplied at the rate of 1 to 2% of body-weight (BW) per day. Green fodder is also provided until the following spring. When sufficient condition is attained, however, the highly nutritional cake food can be reduced gradually or even stopped. Green fodder should be supplied throughout the entire period, because gonads develop satisfactorily only if sufficient vegetation is eaten. Flushing with new water at a rate to maintain suitable water quality is important. Usually ponds are flushed once or twice a month. The frequency is then increased to three or four times during the month preceding reproduction, with the flow continuing for some four hours on each occasion. One month before spawning, the brooders are separated according to the state of gonad development and condition of the abdomen. In some regions, they are also initially separated by sex. Fully mature males and females are injected with hormones. When the fish are ripe, eggs and milt are taken and mixed by conventional stripping. (Natural spawning may occur among brooders remaining in the ponds). The artificially fertilized eggs are hatched in circular concrete tanks with continuous water flow. Three to four days after hatching, fry are transferred to nursery ponds. 25 For culture of fry, a pond of 0.1-0.2 hectares is recommended. Depths range from 1.2 m for fry to 1.5 m for fingerlings. Monoculture is generally used in raising grass carp fry, which are held in nursery ponds to a size of 2o5 to 3 cmo This size takes another 12 to 20 days to reach e A combination of herbaceous plants, compost, and inorganic and organic manures are added after stocking to increase the production of zooplankton, which is the major food resource for the fry (Table 3-1). This combination of plants and manure is applied at the rate of 2,250-3,000 kg per hectare every few days. Three days after stocking, peanut cake and rice bran (or soy milk and soy cake) are also added daily at the rate of 0.1 to 0.3 kg per 10,000 fry. The types of supplementary foods and feeding rates for fry are similar in many parts of Southeast Asia (Table 3-2). In China, inorganic fertilizer has also been used in nursery ponds. Throughout the culture period, a mixture of ammonium sulfate, urea, and calcium super phosphate (weight ratio of 2:l2l or 2:1:2) is added daily along with 75 to 100 kg/ha of compost every 2 to 4 days*. For the culture of fingerlings, pond conditions and dimensions are similar to those used for raising fry. Fingerlings are raised using either monoculture or polyculture. If the grass carp is the principal species, 20 kg of duckweed or tender grasses are added per day per 10,000 fingerlings. Compost, consisting of plants, inorganic fertilizer, and organic manure, is also added at a rate of 3,750 to 5,250 kg/hectare every 15 days from the second month on, and the amount of duckweed or tender grasses is increased by 30 to 50% during this period. It takes about 40 days with monoculture to raise a fingerling from 3 cm to 6 cm, while fry stocked at 5 cm during September require up to 120 days to double in length. 26 In China, adults are raised in polyculture or integrated culture systems. In polyculture systems, the grass carp is stocked with other selected species of fishes. Ideally, each of the species uses a different habitat or food resource. If the grass carp is to be the major species in the system, it is often stocked with bighead, silver, and mud carp, as well as other species. There may be as many as 10 species in a polyculture system. Supplementary foods such as elephant grass, corn and other vegetables, as well as peanut and soybean cake, are frequently added. The grass carp is an excellent green vegetation consumer and has been used extensively for weed control in some parts of the world. A 1 kg individual can consume half of its body weight in fresh grass each day. Limiting Factors The quality and quantity of water and food are the most important factors that limit production of grass carp. Although the grass carp can tolerate low water quality, growth is reduced. Once water and food requirements are satisfied, fish production may be limited by extrinsic factors such as market potential, "seed" (fry) availability, and disease. The grass carp is par- ticularly vulnerable to certain diseases, and mortalities can be high compared to other carp species. Seed availability In Southeast Asia does not seem to be a major limitation, as there is now an adequate supply of fry in all countries except Malaysia, where seed must still be imported. The Malaysian government does expect to achieve self-sufficiency in fry production in the near future. Nevertheless, it can be cheaper to import seed fish than to produce them in some regions, such as the Philippines. 27 Model Networks The grass carp fry network (Fig. 3-1) is based on fertilization and pro- duction of natural plankton communities. Zooplankton, phytoplankton, and organic fertilizers are all consumed by the fry in varying degrees. With high densities commonly stocked in fry systems, oxygen consumption by the fish may deplete oxygen in the pond; therefore, the network includes an oxygen submodel. The adult network differs considerably (Fig. 3-2 )« While fertilizers may be added, especially in the polyculture systems commonly used for carp, natural plant production is insufficient to feed the high densities of carp stocked. Therefore, supplemental feeding is the main food pathway. Otherwise, this model resembles the fry pond network. There is a wealth of information on the energetics and growth of grass carp which simplifies energetic modeling (Huisman 1979, Huisman and Valentijn 1981, and others). However, data for several key areas are lacking. Particularly, the relationship between dissolved oxygen and mortality rate is tentative. Also, the cropping efficiency for natural and supplemental feed is also uncertain. Another major problem in utilizing this network is variable feeds and feed quality for adults (Table 3-3). This variation, due to differences in assimilation efficiency and energy density of each food type, makes the supplemental food pathway uncertain without very specific information on food quality as well as quantity. 28 \_: X 8 = z o o -J >- III h- o a. OQ 3g UJ z UJ Z 2 o o IZTZT J.. ^ n 4a :^ \\\\\\\w^w -I CO < Q P P x-?rn z o f- ^ z LU < < 3 < > X Q. Z I- uj a. o o a: » :l i^J I ! -! I o I d 2- -^ ■t i < z UJ CO Z O Ui Ui -J }*J Q. U- O. CO §1 a M X UJ ^ '*• X CO O 2 o o BIOCHEMICAL OXYGEN DEMAND CO UJ I U % u 0) C V4 a. u CO o en CO 03 (U SI H I o I— I Pt4 29 CO N o ■^ £•- ■^ =1 o e ^ o »- s s ^ I ^ i 2 O o =4 n nrj 2 g _i z I- ^ Ui CL 5 s i ^ ^\\\\\\\\\<\\\\\\\\\\\-^x^ TT o o Mil ^ en K> ^SSS^^^^SSJ I zl I 81 \, \ 3^ 2 5: -i .i X UJ ^ "" X OT BIOCHEMICAL OXYGEN DEMAND z o X Q. 2 O O a o CD CM I O 30 CHAPTER 4. SILVER CARP (PLA LIN) Introduction The silver carp, Hypophthalmichthys m olitrix , has been cultured in China since some time between 618 and 907 A.D. Naturally produced fry from rivers and streams were long the only source for stocking in culture ponds. The culture of the silver carp was limited, and is still restricted in some regions by both the availability of fry and of suitable means for their trans- port. In the 1960s, the hypophysation of adult Chinese carps eliminated dependence on wild stocks, facilitating their culture in many parts of the world. As the demand for silver carp is relatively limited, however, the species is mainly cultured in Asia (notably China, Japan, Hong Kong, Malaysia, Thailand, and Taiwan), with scattered populations established elsewhere, such as South Africa and the United States. Feeding Habits The silver carp is a planktivore. One- to three-day-old fry of about 7 to 9 mm in length feed mainly on zooplankton, rotifers, and copepod nauplii (Table 3-1). The diet expands considerably after 8 to 12 days to include water fleas, large copepods, and phytoplankton. In fry over 30 mm, an oral sieve membrane is formed. Feeding then shifts primarily to phytoplankton, although some zooplankton continue to be consumed. Although no supplemental food is used for culturing adults, it is provided to the fry. A considerable diversity of feeding schedules and feeds are used in culturing the fry stage in various countries, as summarized in Table 3-2 • 31 Growth Rates The growth rate of silver carp fry is remarkably high in the first 10 days. The average increase in weight is more than 100% every 2 days. The size just after hatching is about 0.7 cm and 0.002 g. Sizes of 1.9 cm and 0.1 g can be attained in about 10 days, 4«7 cm and 1.1 g in 30-40 days, and 17 cm and 55 g in about 80 dayse Weight increases during the initial 10 days range from 0.01- 0.02 g/day (Table 4-1). Growth increases rapidly up to 4.2 g/day during the fingerling stage. Relative growth, however, decreases considerably after the fry stage. Whereas finger lings can double in weight every 10 days, it takes adult fish approximately 100 days to do so. Adult silver carp achieve maximum growth rates in length in the second year, and maximum growth rates in weight in the third year. Growth in both length and weight declines sharply after the third year (Table 4-2). Weight gains during the third year may approach 3 kg. Temperature is an important factor in the growth of the silver carp, as with the grass carp. Chinese carps generally do best at temperatures greater than 20 °C, and temperatures below 15*^0 result in poor appetite. If the temperature is less than 8-10 °C, the fish stop feeding (Lin et al. 1980 )e The silver carp grows best at temperatures near 30 °C, although they grow relatively well near 20 °C (FAO 1983b). In Guangzhou, China, no significant differences in growth rate were noted between fry stocked at densities of 1,500,000 per hectare and 2,000,000 per hectare (Table 4-1). Fry stocked at 1,500,000 per hectare, however, achieved weight increases one to two times higher and length increases 30% higher than did fry stocked at 3,000,000 per hectare. 32 TABLE 4-1. Relationship between growth of fry of bighead and silver carp and stocking density in nursery ponds in China (from Lin et al. 1980). Lty Days of rearing Bighead Carp Silver Carp Stocking densj (fish/ha) Length (mm) Weight (g) Length (mm) Weight (g) 3,000,000 2,000,000 1,500,000 10 10 10 15.2 18.6 19.1 0.050 0.134 0.176 14.5 19.2 21.2 0.086 0.198 0.200 TABLE 4-2. Age and growth of cultivated silver carp. Data are presented for Guangdong Province, China (Lin et al. 1980). Age Size at Age Length Weight (cm) (kg) 15.0 0.07 50.0 1.87 57.6 4.65 60.3 5.34 63.0 6.40 Annual Growth Length Weight (cm) (kg) 15.0 0.07 35.0 1.80 7.6 2.78 2.7 0.69 2.7 1.06 1 2 3 4 5 TABLE 4-3. Age of maturity of silver carp, and water temperatures and growing period at different latitudes in China (Lin et al. 1980). The growing period represents the number of months annually during which the average water temp- erature exceeds 15 °C. Guangsi Districts Guangdong Jiangsu Heilung Jiang Annual growing period (number of months) Average water temp, dur- 12 11 5.5 ing growing period (°C). 27.2 25 24 20.2 Maturity age (years) 2 2-4 3-4 5-6 Accumulated (>15°C) 8,900 6,700- 6,600- 4,400- degree days 13,400 8,800 5,200 33 Reproduction The age at sexual maturity for the silver carp varies greatly with climate and environment, with temperature having a profound direct effect (Table 4-3). Small variations in the age of maturity may be found, however, due to differences in light, nutrition, space, water flow, and food. The silver carp reaches maturity in about 5,000-15,000 total degree-days, as does the grass carp. Accordingly, age at maturity of silver carp differs over the range of latitude in which they are cultured (Tables 4-3 and 4-4). For example, in southern China the silver carp matures in 2 to 3 years, whereas in northern China it takes 5 to 6 years. As low temperature is a limiting factor, artificial elevation in water temperature is sometimes used to accelerate maturation, particularly in northern China. Once maturity is attained, the reproductive life span can extend for 15 years* In general, fecundity of the silver carp ranges from 100,000-150,000 eggs per kg body weight- In Malacca, the equation: F = 156.2 W^'^^, where F is in thousands of eggs and W is body weight in kg, estimated fecundity, with these fecundities being two to three times those observed for cultured stocks in China and India (Zainuddin et al. 1982). A typical 5-kg female might produce about 700,000 eggs, with hatching success depending upon fertilization success. Culture Systems Culture systems for the silver carp can be described according to four stages: (i) brooder, (ii) fry, (iii) fingerling, and (iv) adult. For raising, holding, and maturing silver carp brooders, ponds with water depths of 1.5 to 34 TABLE 4-4. Age at maturation of silver and bighead carp in different geo- graphical regions. Data were taken from Bardach et al. (1972). Silver Carp Region (Age in Years) Males Females Bighead Carp (Age in Years) Males Females China (south) 1-2 2-3 Taiwan 1-2 2-3 China (central) 3-4 4-5 U.S.S.R. (Krasnodor area) 4 5 China (northeast) 4-5 5-6 2-3 3-4 2-3 3-4 3-4 4-5 4 5 5-6 6-7 2.5 m and areas of 0.2 to 0.45 hectares are recommended. The pond bottom should be composed of loam with some humus for regulating pH and enhancing water fertility, and should be flat for easy harvesting. About 1 month before spawning, the brooders are reselected and separated according to gonad development and condition of the abdomen. In some regions, they are also initially separated by sex. For fry and fingerling production of the silver carp, a pond typically greater than 0.2 hectares, with water depth of 1.2 m (fry) to 1.5 m (finger- lings) is used. Monoculture is practiced for silver carp fry. Three or four days after hatching, fry are stocked in ponds for 12 to 20 days until they reach fingerling size (2.5-3 cm). A combination of herbaceous plants, compost, and inorganic and organic manure is added at the rate of 3,000 to 3,250 kg per hectare every 2 or 3 days during this period. If these manures are insufficient, peanut cake is also added at a daily rate of 1 to 3 kg/100,000 fry. The same pond conditions createjd for raising fry are also maintained for fingerlings, which can be reared in either monoculture or polyculture systems. Foods such as peanut cake or rice bran are added at a rate of 1 kg/day/10,000 35 fingerlings along with 3,000 to 4,500 kg of plant and organic compost /ha every 10 days during the first month after stocking. The application of organic manure is changed to 4,500-6,000 kg/ha every 15 days from the second month on* This combination of food is altered as the weather becomes colder, with amounts of compost being reduced, and those of peanut cake or rice bran increased, in order to maintain a high fat content in fingerlings for wintering. A 3-cm carp fingerling takes 20 to 50 days to become 6 to 9^5 cm under these conditions. The silver carp is raised from fingerling to adulthood in either poly- culture systems or integrated systems. In polyculture systems, different species of fish are stocked to most efficiently use the food resources present. However, while silver carp are seldom raised as a major species in polyculture systems, they are often stocked with grass carp where grass carp are the major species. The grass carp consumes large quantities of aquatic macrophytes and produces wastes which act as natural fertilizers that promote phytoplankton growth. As silver carp feed mainly on phytoplankton, a rearing pond with a large phytoplankton biomass is ideal. In integrated culture systems, the silver carp is often raised in conjunction with production of other animals such as pigs, ducks, chickens, and cows. In such instances, the waste from these animals provides the nutrients for phytoplankton growth. Silver carp raised under proper conditions can attain 1 kg in one year, 2 to 3 kg in two years, and 4 to 5 kg in three years. Limiting Factors A sufficient supply of good quality water and a large biomass of phyto- plankton are the keys to success in culturing the silver carp. Mortality due to disease is not particularly high compared with other cultured carp species, 36 and there is little seed limitation, as the demand for silver carp is not high in most parts of Asia. The principal factors limiting production are the maintenance of water quality in ponds and the need to provide incentives to commercial producers through increased market demand and price. Model Networks The energy networks for silver carp fry (Fig. 4-1) and adults (Fig. 4-2) differ considerably. The fry network utilizes natural production of zooplank- ton as well as supplemental feed for consuiaption. This network is based on a system of monoculture at small sizes and low densities. Therefore, the oxygen component is not important. Adult silver carp eat mainly phytoplankton, and their energy network in- cludes no supplemental feeding. As these fish are seldom cultured alone, the network is constructed for polyculture with grass carp as the major species. The oxygen network is not included, as oxygen levels in these polyculture ponds are more dependent on grass carp than on silver carp respiration. Silver carp have received much less attention in physiological research than grass carp. Basic metabolic parameters are largely unknown, although data from grass carp metabolism likely can be substituted for modeling purposes. Conversion and assimilation efficiencies probably differ considerably, however, due to the nature of foods eaten. 37 i2 cr o 2 z g I- ^^^ z o UJ o H UJ UJ u. oc ^ X z -! o? — \ 2 o d it (D Ui s 1 2 o o ^2t o ^ ^ n a -I UJ J 03 < a H O P o t- UU f > i \ 2 ^ 5 • \ ^ fsj i ^ 2 iS *— § !^ y _J » o 5. -J H O < O K >- a. X a. I r J O 1 UJ CO o S o 2 H- ^ UJ _J S <>^ -J Ui O UI a. u. u. CO < o u c u cd o > 0) I s 3 o I— I 38 a: o \ _,_^ i en o h- UI UJ u. ac Ui lO en §=: £ a: "" UJ a. 2 >. 2 d S I- s ^ % S i^ ^ ^ 2 « UJ UJ 2 S O U '^ Si o » ^ -£h n a o o JT z < Q. O O ^ < X 0. o u GO d cd a u a O UJ < d O UJ u. 3 < U > I O M 39 CHAPTER 5. BIGHEAD CARP (PLA SOONG) Introduction As with other native carps, the bighead carp, Aristichthys nobilis , has been cultured in China since some time between 618 and 960 A.D. Prior to the introduction of modern techniques to artificially induce spawning, naturally produced fry were annually exported from the Chinese mainland to other Asian countries • In 1960, techniques for hypophysation eliminated a dependence on natural fry, facilitating the culture of bighead carp in other parts of the world. The present demand for this species is relatively limited. It is mainly cultured in China, Taiwan, Thailand, and Malaysia. Feeding Habits The bighead carp is primarily planktivorous. One- to three-day-old fry feed on zooplankton such as rotifers and copepod nauplii (Table 3-1). Feeding habits of juveniles are essentially the same as those of the silver carp, but water is apparently filtered more rapidly through the gill rakers and the alimentary canal is shorter than in the silver carp. Fingerling bighead feed mainly on zooplankton, but the diet also includes phytoplankton. No supplemental foods are used in rearing and holding adults, but food is added for fry and for other species in polyculture systems. Feeding frequency and choice of feeds vary considerably for cultured fry (Table 3-2). Growth Rates The growth of bighead carp during their first 3 years is remarkably high. Just after hatching, fry are about 0.7 cm and 0.002 g, but they attain a size of 1.3 cm and 0.09 g within 10 days. Growth increments range between 0.01 and 0.02 g/day in the first 10 days, and reach 5 g/day during the fingerling stage. 40 Relative growth, however, decreases continuously. For example, the average increase in weight of fry of more than 50% per day is reduced to 10% per day during the fingerling stage, and to 1% per day for adults. Adult bighead carp grow maximally in length in the second year of life, but maximum growth in weight occurs in the third year. After the third year, growth in both length and weight diminishes sharply (Table 5-1). The maximum growth is about 1 kg per 6 months. At maturity, bighead carp may weigh as much as 10 kg. The same relationships of temperature and stocking density to growth given for silver carp also apply for bighead carp. Reproduction Reproduction of the bighead carp is similar to that of the silver carp, with the following exceptions. The bighead generally requires 1 year longer to mature (as can be seen from Table 4-4). Fecundity is about 100,000-150,000 eggs per kg of body weight, with an average 3-year-old, 10-kg brooder producing a total of about 1.2 million eggs. In Malacca, absolute fecundity of bighead carp was estimated by Zainuddin et al. (1982) by the relationship F = 7.0 W^*^^, where F is in thousands of eggs and W is body weight in kg. Spawning success is roughly 90% and the fertilization rate about 80%. Culture Systems The same culture system used for the silver carp fry (p. 37) is also used for bighead. It takes about 30 days for fingerlings to grow from 3 to 6 cm to reach 6 to 12 cm in length. Bighead differ from silver carp in that they mainly feed on zooplankton whereas silver carp feed primarily on phytoplankton. 41 TABLE 5-1. Age and growth of cultivated bighead carp. Data are presented for Guangdong Province, China (Lin et al. 1980). Size at Age Annual Growth Length Weight Age (cm) (kg) Length Weight (cm) (kg) 17.0 0.12 46.0 3.13 11.6 7.45 0.5 0.20 2.7 0.90 1 17.0 0el2 2 63.0 3.25 3 74.6 10.70 4 75.1 10.90 5 77.8 11.80 Limiting Factors High quality water and a rich supply of zooplankton are the primary factors governing production for this species, as there is usually no seed problem, and mortality due to disease is not particularly high compared to other Chinese carp species . Increase in market demand could certainly stimulate commercial production, and add an important incentive for culture. The availability of ponds with good water quality may be the most important limitation on increased production of this species. Model Network The energy network for fry and adult bighead carp (Fig. 5-1) is similar. Fry are occasionally fed supplementally, but otherwise the food habits remain the same. Culture uses fertilizer which yields zooplankton food but may also be eaten directly. There is no oxygen model included, because (1) fry culture is at densities too low to deplete oxygen, and (2) the adult system is one of polyculture with grass carp as the major species. Thus, grass carp have a larger effect on oxygen levels than do adult silver carp in these ponds. 42 Knowledge of bighead carp physiology is also limited, and modeling can be done only by using data for grass carp which are assumed similar. Once again, as for silver carp, conversion and assimilation efficiencies probably differ considerably. 43 \Z o m r- III « s § en »- Ul UJ u. liJ X z 8 = £ a: z y i UJ 2 12 Ui Ul g 2 8 o I z P ^ UJ CL H 2 2 ^11 ^i ^ o >;■ s o UJ > < o o o a. o q 4: o CO 60 CO < g p > X a. •T3 d CO >> U o s^ < O < O a; c I in M 44 CHAPTER 6. NILE TILAPIA (PLA NIN) Introduction The Nile tilapia, Tilapia nilotica , is native to rivers and lakes in northeastern Africa. Because of its large maximum size and rapid growth, this tilapia has been introduced to natural waters throughout Southeast Asia. It is now more popular for culture in the region than T^. mossambica , and is currently the major freshwater fish marketed in the Philippines. It is also cultured throughout Taiwan and Thailand, but its use is more limited in Malaysia and China. Due to the considerable importance of T_. nilotica in the Philippines, this chapter will emphasize the techniques and systems employed there, and will describe other systems only when they differ substantially from those used in the Philippines . Feeding Habits The Nile tilapia is omnivorous. In nature, it feeds mainly on algae, although detritus, zooplankton, and benthic invertebrates are also eaten. Algae in the diet include not only diatoms and green algae, but also blue- greens such as Microcystis , Spirulina , and Anabaenopsis . The ability of tilapia to digest blue-green algae m^ikes it a favored species for control of algal blooms and maintenance of water quality in many culture systems. The omnivorous food habits of tilapia also make prediction of its con- sumption pattern in culture very difficult « The major natural food pathway in ponds is the direct consumption of phytoplankton and zooplankton, but supplemental feeds such as rice bran or fish meal are also eaten directly. If these foods are not readily available, tilapia will switch to a wider diet, which may include detritus, organic fertilizer, and macrophytes. Often in 45 less intensive culture, duckweed, Lemna minor , is added as a supplemental food. This breadth of diet makes tilapia successful in a variety of culture systems, but trophic relations are very difficult to quantify in a modeling framework. Due to the breadth of diet, few qualitative changes in feeding occur with age. Although slightly different techniques are employed for rearing fry and adults, these are related more to water quality or intensity of management than to differences in diet. Growth Rates The Nile tilapia is a favored culture species because of its rapid growth. Fish from natural waters frequently weigh in excess of 3 kg, and Thai culture systems often produce marketable fish of 0.5 to 0.8 kg after 6 months of culture. Size at marketing and culture duration vary. In the Philippines, fish of 120-150 g are marketed after 4 to 5 months of culture, while in China, fish of 200-300 g are harvested. Differences in size at harvest are mainly related to preferences of local constimers, and the need to avoid natural reproduction. Growth differs during the fry stage and the finger ling to adult stage. Fry take 30 to 45 days to reach 3 cm (at 29% BW per day), while adults take roughly 4-6 months to reach 200-500 g (overall 3.5% BW per day). Male tilapia generally grow more rapidly than females, particularly after the onset of sexual maturation. This has led to an interest in monosex male culture, which takes advantage of greater growth and limited natural repro- duction. 46 Reproduction Tilapia are mouth-brooders in which brood size is limited by the capacity of the oral cavity for incubation and guarding. Thus, numbers of eggs spawned are relatively low (ranging from 200 to 3,000 per female, depending on female size). Fish mature at 3 to 7 months, with size at maturity depending largely on environmental conditions. Generally, fish that attain large maximum sizes mature at later ages • The largest tilapia are usually found in large open lakes, whereas smaller adults are usually found in ponds. This means that cultured fish will often begin to breed at small sizes — often at 30 to 50 g (3 months old). Such early breeding causes problems in tilapia culture, because shifts in energy allocations from body to gonad reduce growth. Additionally, reproduction may create an overabundance of fish, leading to competition among fish for food and reduced growth. This problem will be further described later in this chapter. In most areas of Southeast Asia, tilapia can breed year-round. Mouth brooding takes 10 to 12 days, after which fry are released and begin schooling. Optimum temperatures for reproduction are between 22 and 35*^0. Experience in intensive culture systems in the Philippines indicates that two thirds of the females are able to spax^i recurrently at 2-week intervals. Culture Systems Since tilapia culture varies depending on the brooding system used, culture and brooding will be described together. Techniques will be described in detail for Philippine systems first, then differing methods in the other countries will be considered. 47 Breeders and Fry Prevailing systems of handling breeders for fry production can be divided into three categories: extensive, semi-=intensive, and intensive « (a) Extensive method The most common and least complex seed production systems for tilapia fall into this category. Ponds of approximately 0.02 ha and 1.5 m deep are stocked at a rate of 1 brooder per 2 m^ (sex ratio 1 males 5-7 females). Initial stocking size of breeders is roughly 40 g for females, and reaches 60 g for males. After 3 to 4 months of culture, fry are harvested by seining or total draining. About 200 to 300 fry are produced by each female per month, yielding 50 to 100 fry/m^ /month. Natural food is usually sufficient to sustain production for 4 months. Fry ponds are stocked at a rate of 200 per m^. Chicken manure is applied at a rate of 3,000 kg/ha/month, and rice bran is used as supplemental feed at a rate of 5% BW per day, with twice daily feeding. Survival in the first month is usually from 60 to 70%, yielding 120 to 140 f ingerlings/m^. These fingerlings, which weigh 2 to 5 g each, are then stocked into grow-out ponds. (b) Semi-intensive method This method aims at a higher scale of fry production than does extensive culture. Breeders are stocked in ponds at a higher density (4 per m^) at a ratio of three (50 g) females per each (75 g) male. These breeders are supple- mentally fed with rice bran and fish meal at 3% BW per day. Between 200 and 300 fry are produced per female per spawn, yielding 250 fry/m^. Dipnets are used to harvest fry daily, beginning 12 days after initial stocking. The ponds are maintained for 4 months before draining and restocking. 48 Fry are transferred to cages (hapas) after removal from the breeding ponds. Stocking density in the hapas is about 300 to 350/m^. During this period,, the young are fed supplemental rations (75% rice bran and 25% fish meal) at rates of 5% BW/day fed twice a day. After about 2 weeks, the fish are transferred to nursery ponds (100 to 200/m2) and cultured an additional 2 weeks under a system similar to that described for extensive ponds. Overall fry survival for 30 days is 70%, producing a fingerling yield of 70 to 140 /m^ of pond. (c) Intensive method Intensive fry culture in the Philippines, which is based on large-scale research initiative, is probably most nearly representative of the state of the art for tilapia seed production. The breeders are kept in hapas submerged in fertilized ponds, or in some cases, in lakes or rivers. One male of about 75 g and three 50-g females are put in a hapa (dimensions approximately 1.5mxlmxlm). Supplementary food (75% rice bran, 25% fish meal) is added at 3% BW per day. Under this system, about 500 fry are produced every 2 weeks; they must then be removed or cannibalism will occur. The breeders are changed each month, but may be reused. Hox^ever, most are used only once and are discarded when they reach a weight of 250 g at about 6 to 8 weeks. Fry are transferred to fine mesh cages at densities of 600-1, 200/m2. They are fed four times daily on supplementary diets (60% rice bran, 40% fish meal) at 8 to 10% BW per day. Survival is 70% during 2 weeks of culture. The fish reach a size of 2.5 cm in these 2 weeks and are then transferred into large mesh hapas. These new hapas are stocked at 250/m2. Supplemental feeding follows the same regime. During the next 2 weeks, survival is 80% and the fish reach 1 to 2 g in weight. Because of good survival, rapid growth, and 49 dense stocking, yields of 2,000 fry/m^/yr can be obtained. However, when cage systems with adults and fry in the same pond suffer from declines in dissolved oxygen, aeration may be required at night. Grow-out Ponds Yields of tilapia in grow-out ponds are directly dependent on controlling spawning, as the high rate of natural reproduction produces overpopulation and stunted growth. Control of reproduction is achieved in the Philippines with either short-term culture, or all male culture. For short-term culture in ponds, 2 to 5 g fingerlings are typically stocked at rates of about 10,000- 20,000 per ha. The ponds are fertilized with inorganic fertilizer (16:20:0 at 50 kg/ha once every 2 weeks) or organic manure (500 to 1,000 kg/ha once every 2 weeks). After culture for 4 to 5 months, survival is 85% and fish attain 120 to 150 g in weight. Yields range from about 1,275 to 2,040 kg/ha. Improvements on this basic scheme are "achieved with use of only male fry. In this case, stocking density is from 20,000 to 40, 000 /ha. For the first month of culture, only fertilization is used. During subsequent months, artificial feed is added at 5% BW per day. Survival for 4 to 5 months is 85%, and yields are 2,040 to 3,000 kg/ha/crop, which is about 50% higher than yields from mixed sex culture due to the 50 to 100% higher stocking rate. One additional integrated system is used in the Philippines. About 80 adult pigs are used to supply one hectare of pond with 25,000 kg of manure every 90 days. ^. nilotica are stocked at 20,000/ha, to produce a yield after 90 days of 1,700 kg/ha. This system is reportedly still in experimental stages, although current results are promising. 50 Other Countries As may be expected, variations of the tilapia culture used in the Philip- pines have evolved in other countries. Thai culture is an example of this which vrlll be described. In Thailand, the fry are left to grow in the breeding ponds for 45 days. Two male and three female breeders are stocked per 5 m^. After 45 days, about 100 3 cm fry/m^ are produced. These fry are then stocked directly into grow- out ponds . Grow-out ponds are diverse in character. Integrated systems with 60 pigs/ha or 2,000 chickens/ha have 70% survival of stocked fry after 6 months. The fish are stocked at 10, 000 /ha, and grow to 300 g in culture. Overall yield is commonly 3,125 to 3,750 kg/ha. Monosex culture of manually sorted Z* nilotica is also practiced. In this culture, ponds are stocked at 30,000/ha and the fish are supplementally fed with rice or kitchen waste. Survival of 60% and weights similar to those in integrated systems are attained after 6 months. Overall yield is 5,000 to 6,000 kg/ha. Limiting Factors The major factors limiting production of Nile tilapia in Southeast Asia appear mainly to be biological, and include: (1) the need for control of reproduction, (2) inadequate availability of seed, (3) disease outbreaks, and possibly, (4) poor genetic quality. The problem of natural reproduction and its control has already been described herein. Techniques to avoid this problem have been evaluated by Guerrero (1982). Monosex culture appears to be the most successful, although it may be limited by insufficient fry for purely male culture if the sexes are manually sorted. Techniques being evaluated in the Philippines include hormonal 51 treatments (methyltestosterone) of fry to achieve all male cohorts through sex reversal. Consideration has also been given to the use of x-ray radiation or heat shock to achieve sterility. These and other attempts have succeeded on an experimental basis, but need additional refinement for commercial application. Other controls for overpopulation, such as predator introduction and cage culture, have potential for overcoming the problem but remain largely experimental. Seed availability of Tilapia for commercial producers continues as a problem in the region. In the Philippines, annual fry production is about 200 million, whereas needs there may ultimately be around 1 billion. Similarly, Thailand's annual fry production (40 million) is well below an estimated annual need of 60 million. The intensive culture systems developed in the Philippines could overcome this shortfall in the future if implemented on a large scale. Currently, however, some of this shortfall is met by collection of fry from natural waters. Low genetic quality of fry could result in poor growth in culture sys- tems. This problem could be attributable to inbreeding. Introduction of vigorous brood stock replacements from selective breeding programs may over- come this problem. This topic is presently being studied in the Philippines and elsewhere in the region. Whereas disease problems are not prevalent in Thailand, the Philippines, or mainland China, production in Taiwanese cage culture systems may largely be limited by disease. Major outbreaks of bacterial infections, likely at least in part the result of poor water quality, have caused increased mortality in recent years . 52 Economic incentives do not appear to limit production of tilapia. Both adult and fry culture are highl^r profitable, and fish farmers should be willing to invest additional money to improve their operations, once the materials and knowledge are extended to them. Model Network The Nile tilapia energy network (Fig, 6-1) is similar for fry and adults. It is a very complex energy pathway, as virtually all of the various trophic components are present. The major feeding pathway is through plankton or supplemental feeds but if these become limiting, then almost anything can be eaten. Obviously, feed conversion and growth will differ accordingly. The pathway also includes an oxygen submodel, as high densities and fertilization rates can drive oxygen to low levels. The fish growth submodel is also very complex. Growth is simple with culture of all males, but becomes extremely variable and unpredictable when both sexes are grown together. Differential allocation of energy to reproduction, as well as overpopulation and competition for food, make the ultimate size of fish and the yield under bisexual culture extremely variable. There is a distinct lack of energetics data on the Nile tilapia, which also makes results of energetics modeling speculative. The only major experiments are ones on metabolic rate (Farmer and Beamish 1969), while little is knovjTi on food assimilation or utilization. This, coupled with the variable input and outcome pathways of production, makes culture of Nile tilapia virtually impossible to quantitatively model at present. 53 ^l\\\\\\\\\^\\\\\\\\\\\^X^A\\\\\\\\\\\\\- jrfn < o ^ £ — )^ 3 ? o 2 |5 S -i O H Z GC — Ui u. 1 ^ O UJ < d o yj I zi \^\ i o i ^?^::^ il M Z g X UJ ^ '^ X CO o z < 9 z Q « UJ z •C ^< 2 "^> o g 3 S a: 2 03 u o u o 4-) 0) CI c I I— I 54 CHAPTER 7. GIANT FRESHWATER PRAWN (KUNG KAM-KRAM) Introduction The giant freshwater prax^n, Macrobrachium rosenbergii , is a native of the Indo-Pacif ic region and has been extensively introduced elsewhere. The adults are found in virtually all types of fresh and brackish waters. In Thailand, it is widely distributed and traditionally spawns in the estuarine areas (Sidthimunka and Bhukaswan 1982) « lyUicrobrachium spends its larval stages in brackish water where salinity is 8 to 22 parts per thousand, and then moves upstream (often for long distances) to spend the juvenile and adult stages in freshwater rivers, swamps J, and ditches. Dam construction on rivers has blocked access to required upstream habitats in many areas • Commercial pond culture of this shrimp has begun in Thailand, many other parts of Asia, and America. Macrobrachium rosenbergii is one of the largest and most desirable of the freshwater shrimps. It has rapidly gained a place in commercial markets and has become a very important cash crop in Third World nations . Feeding Habits Under natural conditions, mature freshwater prawns feed mainly on various types of benthic animals. As larvae, zooplankton is the major food. In the absence of an adequate supply of live animals, however, prawn fry will take dead organic material including plant detritus. They begin to feed on benthic animals or organic detritus after reaching 2 to 3 cm and adopting a benthic mode of life. Adults will eat almost any living or dead organic matter of suitable size. When deprived of adequate rations, they may even resort to cannibalism. 55 When cultured in tanks or ponds at elevated densities, artificial feed is provided. For larvae, food such as brine shrimp larvae, fresh fish eggs, fish flesh with chicken-egg custard, or powdered dried chicken blood is added at a rate of approximately 30% BW per day* For adults, supplementary food usually consists of 50% animal material, such as fish, mollusks, earthworms, offal, live insects, and silkworms, plus 50% plant material, such as grains and spoiled fruit. This combination is supplied at a rate of 5% BW per day« Other prepared foods such as fresh mussels and chicken eggs are also used* Growth Rates The growth of the giant freshwater prawn differs considerably among the life stages • Post-larvae stocked at 6.25/m^ in nursery ponds as 0.01 g, 1-cm individuals double in length each month while gaining roughly 7 g over a 90- day period. Survival during these 3 months averages 70%. During the second 3 months of life, the growth in length is about 2 cm monthly, and weight increases average 10 g per month. A typical 7-month-old prawn has reached 45 g at a length of 15 cm. Survival to this stage is about 50%. Similarly, average monthly length increases of 1.3-1.8 cm were observed during a 6-month comparison of cage, ditch, and open pond rearing of 3 to 5-cm juveniles stocked at 5/m2 (Menasveta and Piyatiratitivokul 1982). Ling (1969) found that 5. 5-cm prawns could attain a size of 22.5 cm in 6 months with good growth conditions . ^' rosenbergii grows best at relatively high water temperatures (23° to 32 °C). Optimum temperatures for larval culture are in the vicinity of 28-31 °C (Aniello and Singh 1982). Such temperatures can promote mortality, however, by increasing oxygen consumption of the prawn and by decreasing the amount of dissolved oxygen. The species has little resistance to low water tempera- 56 tares. In water below 5°C, they lie on their sides on the bottom and there is increased mortality. Stocking density is not a major factor limiting growth of the freshwater prawn, provided the density is not more than 10 per m^. Differential growth within the same age group can be eleven fold during a 6-month culture period. Following such culture, an average prawn will weigh 50 g, with the largest being about 110 g and the smallest 10 g. The cause of this variation is attributed to local population density, territoriality, and uneven dis- tribution of food. The food conversion ratio for this species ranges from 1.8 to 3.8:: 1. Terminal size is also important in giant freshwater prawn culture as about 2 to 10% of adults stop growing after attaining a weight of 50 g. This is usually attributed to stock genetics with supplemental effects of poor water quality or unusual sex ratios; the problem is accelerated when the ratio of females to males is greater than 6:1. Reproduction The life span of this prawn in culture commonly exceeds 1 year. Market- able size (40-50 g) is attained in 6-7 months under favorable conditions, while sexual maturity is attained following 4-9 months of rearing. In nature, females may spawn 3 to 4 times annually, while producing up to 120,000 eggs at each spawning. In captivity, however, the^r produce about 10,000 eggs per 10- cm female, with a hatching success of about 90% or more. First broods produced during the first year frequently number less than 20,000 (New and Singholka 1982). 57 Culture Systems The culture systems for the giant freshwater prawn vary according to three life stages: breeders, fry, and adults • For breeding, two to four males are placed together with eight to twenty females in a tank 2 to 3 m long, 1 to 1% m wide, and 40 cm deep. Females brood eggs in a brood pouch for about 19 days at 26 to 28°C« After about the twelfth day the initially bright orange eggs begin to fade to a pale gray. When egg color darkens to a slate gray, hatching is imminent. After hatching, 20,000 larvae are transferred to each 1 m*^ capacity tank. First feeding is with Artemia nauplii. Later a high protein diet (50% eggs, 50% mussel or fish flesh, etc.) is provided. Mortality ranges from 25 to 50% during this early feeding stage. Ten days after metamorphosis, the early fry are transferred to a nursery tank or pond at densities of 15 to 100/m^. This culture stage extends for some 80 days, with a survival rate of about 60% to 70%. Average weights are about 5 g. The young are then transferred to a grow-out pond. When the young reach about 4 cm in length, they are suitable for stocking in 1/2 ha production ponds at about 3 to 5/m^, either alone or in combination with fish. Fish species successfully used in culture with the giant fresh- water prawn include the bighead, grass and silver carp (Malecha et al. 1981, Tunsutapanich et al. 1982), and the snakeskin gourami. Suitable stocking rates depend not only on the numbers and kinds of fish used, but also on the quality of soil and water. There also is some culture of the freshwater prawn in rice paddies. Prawns are stocked after the rice seedlings are fairly well rooted. Due to a short growing season (four months) in paddies , prawns of advanced size are preferred for stocking at about 1/15 m^. Supplementary feeds are not used in 58 paddy culture. When prawns are stocked alone, the survival rate for this stage is 60 to 70%. In 4 months 80 to 950 kg/ha are obtained by conraiercial producers in Thailand, and yields up to 3,000 kg/ha occur on Oahu, Hawaii. Other culture systems such as cages and ditches are also used for raising adult prawns, but yields are not as high as in pond rearing. Limiting Factors The most serious problem in prawn culture is oxygen depletion. The giant freshwater prawn is more sensitive than most fishes as greater than 3-4 ppm D.O. is required for good growth. Wlien the oxygen level is less than 2 ppm, stress is evident; large mortalities can occur if the level dips below 1.5 ppm at any time. Prawns are primarily benthic, thus wind conditions and pond stratification may affect them through impacts on local D.O. conditions. Although there is sufficient seed available for production in areas where culture has been established, the quality of seed needs to be greatly im- proved. It is possible that a decline in seed quality resulting from genera- tions of inbreeding is now a primary cause of slow and terminal growth in prawns. Factors such as territoriality, cannibalism, high mortality during transportation, and uneven distribution of available food can also be important in limiting production levels of the giant freshwater praxm. Model Network The freshwater prawn energy network (Fig. 7-1) is based mainly on sup- plemental feeding. While plankton and benthos are also consumed, problems with dissolved oxygen generally limit fertilization rates and therefore limit natural food production. The prawn growth submodel is very straightforward. The network also includes an oxygen submodel, which is particularly important 59 in Macrobrachium culture, because of their sensitivity to low DO and their benthic habits. There is a wealth of information on Macrobrachium growth and energetics, which simplifies modeling. Bioenergetic studies (Clifford and Brick 1979, Nelson et al . 197 7a) are sufficient for prawn energetics models. Similarly, food utilization and growth are also known (Nelson et al« 1977b). Quantitative models can be made to predict average oxygen levels in ponds. The physics of winds and circulation are more difficult to model, however, making the prediction of oxygen levels near the pond substrate more complex. 60 [\ \\\\\\\\\\\\\\\\\\\\^^X^x\\\\vX\\\\\\V\\\\\V\\\ | p O "T" ro X z z Ui o ^1 I I I I 9. ^ Q .•"; u <« UjI *" z 3: O O u 4^ _. J -r^ k : >n' nWww^S'^'^^^'^'^'^^^'^^^^^^^ ^ ^^^^^^^^^^'-^ n o o o ^ < < g uj a. o z I q: » I z i I SI L^J s < LJ P ^ ^ O UJ < d O UJ 1 J < K Z UI CO ? o UJ UI _J UI u. Q, CO -i ,i z o « X CO Q- O z O o BIOCHEMICAL OXYGEN DEMAND 3 < < CO u Q) 4J CO 0) u o o C V-i 0) c w I 1^ o 61 CHAPTER 8. SNAKESKIN GOURAMI (PLA SALID) Introduction The snakeskin gourami, Trichogaster pectoralis , is native to most of Southeast Asia. It is commonly found in stagnant swamps, rivers, lakes, and rice fields • The species is a traditional Thai food fish which is generally marketed salted and sun-dried. Annual production in Thailand is around 13- 17,000 tons, and production of Trichogaster in Indonesia rivals that of Thailand (Table 1-1). In neighboring countries, this gourami is not commonly cultured, although wild fish are harvested for food. The Thai culture system will be described in this chapter. Most of the data for this description come from Boonsom (1983). Feeding Habits Adult snakeskin gourami feed primarily on zooplankton and benthos, whereas fry eat phy to plankton, small zooplankton, protozoans, and rotifers. Small fry of 5 to 15 mm in length consume mainly protozoans and rotifers, while fry of 15-37 mm are more herbivorous and use mainly phytoplankton and protozoans (Boonsom 1983). Artificial foods are not supplied in most of the culture systems, although experimental brood stocks are fed supplemental diets of 60% rice bran and 40% fish meal at 2% BW per day. Growth Rates Growth rates of the snakeskin gourami under culture are variable, largely due to differing culture densities and food availabilities. Marketing size in Thailand is 100 to 150 g, which is commonly achieved in 8 months of culture. Recent experimental techniques, however, have produced similar sizes in only 6 months. Growth stanzas and maximum growth rates are poorly known. 62 Reproduction Breeding of the snakeskin gourami is mainly during the rainy season (May to September in Thailand). Until recently, spawning was known to occur during every month except December, but supplemental feeding of brooders has extended it into December. Fecundity ranges from 20,000 to 40,000 eggs per female (Boonsom 1983). An individual lays between 3,000 and 8,000 eggs at each spawning in nests in weedy areas. Hatching success is near 90%. The early feeding and growth of larvae are very important determinants of success in gourami farming . Culture Systems Culture of the snakeskin gourami was mainly begun as a fish and rice sys- tem. In the main areas of gourami culture in Thailand, however, rice farming was not profitable due to acid sulphate soils, so the fish became the main crop. Nevertheless, much potential remains for gourami farming in rice fields. Traditional System Traditional culture systems for the snakeskin gourami are summarized as follows. A large field (often 3 to 20 ha) is excavated to form a platform area with a peripheral channel around it, inside the pond dyke. The channel is dug 75 cm or more below the platform level, and is usually 3 to 4 m wide. The field is partly filled with water so only the marginal channels are submerged. Adult fish are then stocked into the channels at densities from 40 to 220 kg/ha. Stocking density is related to availability of breeders, but in general, relatively high densities are stocked. Boonsom (1983) held that overstocking is common in traditionally practiced culture. Sex ratio at stocking is not controlled. The traditional method is to keep the breeders in 63 the small ponds, unfed, for 1 to 2 months in order to make them lean and hungry. The field is then flooded and spawning occurs on the platform. Early survival and growth of fry of the snakeskin gourami are highly dependent on their density, which in turn is controlled by predator abundance, food availability, and number of adult spawners. No supplemental feeding occurs. Organic fertilizer (green manure) is added to the platform area* The most common manure is grass and weeds that have been previously cut from the platform. Other than this, no management is practiced. After 8 months or more, depending on the growth rate, the fields are drained and all fish are forced into the peripheral canal. The canals are dragged to concentrate fish into the lowest point, then mud and debris are pumped up and fish are strained out. Yields obtained from this system have varied from 450 to 1,700 kg/ha per cycle. After harvesting, the pond is left dry for a short time, during which the grass is mowed or burned. Lime may be added to the soil if acidity problems occurred during the past season. After this, flooding of the peripheral channels and stocking of brooders is begun again. New Methods The major limitation in the traditional system was poor survival of fry. Two experimental methods have been studied to find ways to alleviate this problem. One involved fertilization with chicken manure during the fry stage (Boonsom 1983), the other separation of brooding and grow-out ponds. In former method, chicken manure was added to the ponds at 15.6 kg/ha/d for 60 days during fry rearing. The manure was applied once every 10 days at all four corners of the platform. The previous production cycle for the experimental field had yielded 1,000 kg/ha, and long-term averages indicated 64 that 850-1,000 kg/ha were good yields for that farm. The harvest under experimental management was 2,000 kg/ha, which was the most ever recorded. With fertilization, the growth and survival of fry increased in a system otherwise similar to the traditional one. The second experiment involved breeding the fish in a smaller pond (1 ha) for several weeks , and then removing the fry to stock the rearing ponds at a controlled density. Size of fry in the brood ponds reached an average of 70 mm in 45 days. The brood ponds were screened to keep out predators, particularly the snakehead, Ophicephalus s triatus . Additionally, breeders were fed supplements of rice bran and fish meal for a month prior to flooding of the brood field. This brood system resulted in excellent production of young and rapid growth. Controlled stocking of young fish at a larger size also reduced mortality. Growth to marketable size (120 g) occurred over a total culture period of 6 months, compared to 8 to 10 months with the traditional method. Limiting Factors The major factors limiting expansion of snakeskin gourami farming in Thailand appear related to fry growth and survival. Methods for predator control and fertilization of nursery ponds with animal manures need wider application and development. Physico-chemical factors, such as low DO and pH, do not limit production, as the fish is an air breather by virtue of its suprabranchial organ, and is tolerant of acidic conditions. However, low DO could limit reproductive success in brood ponds, and other water quality parameters may reduce adult growth. Also, limited market potential pre- vents widespread use of gourami culture techniques in other countries. 65 Within Thailand, demand is fairly high and market value (12-16 baht per kilo, approx, $.60 U.S.) is good. Widespread adoption of the new gourami culture technologies in Thailand could result in high returns to the farmer. Many of the current problems are related to traditional beliefs such as: (1) the more fish stocked in a field, the better the yield, (2) lean brooders produce fitter young, and (3) no fertilization is needed other than cutting and spreading of weeds. These problems can easily and cheaply be remedied by increasing availability of information from current research on snakeskin gourami culture. Model Network The snakeskin gourami energy network (Fig. 8-1) is a very simple one, due to the largely extensive nature of gourami culture. Either soil inputs or fertilization yield food through zooplankton and benthos production pathways. There is no oxygen submodel, as the fish are air breathers cultured at low densities. The fish energetics submodel is particularly poorly known. Little basic research has been done on this fish, and quantitative pro- duction models are not feasible. The simple energy pathways make modeling possible, but much additional work on metabolism, growth, and food utilization is necessary to enable a fish production model to be constructed. 66 o o O § o h- UjI Ui u. ^ Ui X z **• UJ (!) H- Z <« 55 O (o O m o Ul UJ o ^ ij z .J o III Q. CO ? UJ Z UJ 3 CO z :e 8 5 o £t- nr .\\\\\\\\\\\v^i o a o o M n Is >- cr O UJ 8 z !^ < d 2 >- -1 O UJ Q U. tf) s CO o 60 CO cd C CO o u § (U c w 1 00 o Ft, 67 CHAPTER 9. TAWES (PLA TAPIEN^-KHAO ) Introduction The tawes, Puntius gonionotus , is native to Java and Sumatra « The species also occurs throughout Thailand, where it is most common in rivers of the central region. There, tawes are caught from natural waters in large quantities each year for domestic consumption. The tawes is commonly cultured in Thailand, Indonesia, and Malaysia using the same techniques. Culture in Thailand can therefore serve as our prototype, but differences in Malaysia will also be described. In Thailand in 1980, about 3,000 ha were under tawes culture, with an annual yield of 2,932 tons. This yield was valued at 51,051,040 baht (over $2.5 million U.S.). Feeding Habits In nature, tawes fry commonly feed on plankton, especially zooplankton. Fingerlings and adults are omnivorous, and feed on both plants and animals, but they prefer vegetation. Under culture, the fish are fed a wide range of supplemental feeds, such as chopped and unchopped green vegetation, boiled egg yolk, rice bran, and pelleted diets. Growth Rates The largest tawes from rivers in the central region of Thailand are females, measuring up to 33 cm total length. In nursery culture, the tawes rapidly grows to reach a fingerling size of 0.24 g (3 cm long) in about 6 weeks. When fingerlings are reared in production ponds, a size of 200-250 g is attained after about 8 months. Poor water quality retards growth of the tawes. The fish begin losing appetite when dissolved oxygen drops to 4 mg/L, stop feeding below 3 mg/L, and 68 . die at less than 1.5 mg/L. The pH range for maximum growth is between 6.5 to 8.5. Growth occurs between 24 and 34 °C, but the optimum temperature for growth appears to be between 28 and 32 °C. Reproduction Tawes females attain sexual maturity when about 7-10 months old and 200 to 250 g, while males reach maturity in 6 months at greater than 50 g. In Thailand, the spawning season extends from March to September with a peak in May or June. With good culture techniques, the tawes can reproduce more than once a year. In general, breeders, especially females, are discarded once they weigh more than 1 kg. Fecundity averages about 1,400 eggs per gram in some Thai stocks, with 75-80% of the eggs being laid at one time. Similarly, in Malaysia (Zainuddin et al. 1982), absolute fecundity is estimated by the equation: F = 11.73 W^*^^, where F equals the number of eggs and W is body weight in grams. Boonbrahm (1968) indicated, however, that in some cases females may only average roughly 400 eggs per g of body weight. The newly-laid, semibuoyant eggs measure about 1.0-1.2 mm in greatest diameter. After water absorption, eggs are about 3 mm in diameter. Hatching occurs within 8-12 hours at water temperatures between 25 *" and 32 ""C. Hatching success is over 70%. The newly- hatched larvae are 3 mm in length and begin feeding on minute plankton after 36 hours. 69 Culture Systems Breeding Brood ponds for tawes culture are best designed to create suitable conditions of temperature, light, and water exchangee The ponds commonly used in Thailand are 400 to 800 m^ in area and 0.8 to leO m deep. To such ponds, 150 to 200 kg of organic manure are added when the water is only 0.2 to Oo3 m deep. After 6 or 7 days, the pond is filled and the breeders are introduced « Fifty to 75 kg of manure are subsequently added at 2-week intervals. Stocked densities of brood fish, each weighing 200-400 g, are 3-5/10 m^. The fish are fed daily with 1 to 2% BW of a pelleted diet containing 16% fish meal, 10% soybean meal, 24% peanut meal, 15% broken rice, 30% rice bran, 4% ipil-ipil ( Levcaena levcoceplala ) meal, and 1% vitamin and mineral premix. Nursery Dried earthen nursery ponds are filled with water to 0.6 m deep, and then stocked with 1,000-1,500 2-day-old fry per m^. Beginning 1 day after stocking, the fry are fed for 4 consecutive days with hard-boiled egg yolk. They are then fed rice bran, supplemented with a small amount of crushed pelleted ration containing 16% protein. Within 6 weeks, the fish reach 3 cm and about 0.24 g. Survival rate averages 16%. Grow-out Ponds Various monoculture, polyculture, and integrated rearing systems are m use. (^) Monoculture After being in nursery ponds for about 1 month, tawes fry are moved into earthen production ponds which are usually 400 m^ in area and 1 m deep. 70 These fry are stocked at the rates of 3 to 4 flsh/vor. They are fed twice a day at daily rates of 3 to 5% BW with a 16% protein pelleted diet containing mostly peanut meal and rice bran. The duration of this culture is about 8 months. At harvest, the fish weigh 200 to 250 g. Survival averages 72%. Yield is 5,670 kg/ha/8 months. (^t)) Polyculture Due to its feeding habits, which are similar to those of the grass carp, the tawes is stocked in ponds together with silver, bighead, and common carps. Such local extensive culture is common in Thailand, but only meager data on stocking densities, feeding practices , and production are available for these polyculture sys tems . (c) Integrated culture In Thailand, the tawes is also cultured using only pig droppings as a food source. A typical 1-ha pond is stocked with 25,000 2-3 cm fry, and fertilized with manure from 40 pigs. Fish can grow to marketable size in 8 months, with yields of about 1,500 kg per ha. Limiting Factors The major ecological factors restricting production of the tawes appear to be predators of fry, unsuitable food for fry, and reduced water quality. Tawes culture in Thailand is also restricted from expansion by the high cost of feed. Pelleted diets commonly used for production ponds cost $0.35 to 0.40 per kg. With a food conversion ratio of 1.75, about $0.60 to 0.70 is required for feed to produce 1 kg of fish flesh. This gives little profit margin to farmers who sell to wholesalers or middlemen at $0.75-0.85 per kg. However, as the market retail price of this fish is $0. 90-1. 00 /kg, producers are in a position to gain by direct sales to consiomers or retailers. 71 Predators can cause great mortality of tawes under nursery conditions, especially in the first 7 days after introduction of the larvae to ponds* At that stage, tawes fry are smaller than many kinds of zooplankton. Besides fishes like the snakehead, common predators include aquatic insects and tadpoles. Low survival under nursery conditions may also be due to short- age of food, at least partly resulting from competition with zooplankton. Research on availability of natural food and suitability of prepared feed for fry is necessary • Water quality control is important in breeding, nursery, and grow-out ponds. Low dissolved oxygen can indirectly affect sperm and egg viability, and can be directly lethal to eggs and fry. Food intake and growth also decline when dissolved oxygen falls below desirable ranges. Model Networks The tawes fry energy network (Fig. 9-1) is based on consumption of zooplankton and phytoplankton as well as supplemental feeds. The network includes an oxygen submodel, due to the sensitivity of tawes to low DO. The adult network (Fig. 9-2) differs mainly due to inclusion of macrophytes in the diet. The structures of tawes' energetics models strongly resemble those for grass carp. Relatively little research has been done on tawes, except for work relating to common culture practices. The lack of research on metabolism, food util- ization, and growth limits the ability to produce quantitative models and simulations. Since the energy pathways are somewhat complex, this species may prove difficult to model. 72 K \\\\\\\\\\\\\\\\\\\\\\\\v\\\\\\\\\\\\\\\\\\\\^^ k o 2 UJ o X z s^ £« H 2 OT 55 8 2 o §9 (T. CL UJ ^ g z a . Q ?^ » ^ o (£ a. S ^ i 2 z 3 h lu « Ul '*' 2 ^ -r> 3:^ \v\\\\\\\^^ CO O o p < UJ a. is » f-N g5 I O Ul < =i o £ I zl I gi I ii I oi L._-J i X ^^^^ Ul CO Q. CO o X Ul ^ o o *^ ^ BIOCHEMICAL OXYGEN DEMAND 3 S a: CO < 1- o ►- u. i , ^ 1 i O :a£ X z t- < z _J UJ a. OQ 8 rsi i I ^ 2 O H- :^ ^ 2 ^ < r5 n-' ^ 2^ > X a. _J < o h- o u -J is o a! I*" CO 3 3 O CO T5 cd CO fH CO T3 CO CO CO o o :5 c 4:0 $^ (U c I o _u M 79 CHAPTER 11. SAND GOBY (PLA BU-SAI) Introduction The sand goby, Oxyeleotris marmoratus , is potentially an important culture species because of its high market value. It is commonly found in rivers throughout Southeast Asia. It comprises only a minor part of the total freshwater fish production in Thailand and Indonesia (about 700 tons per year), however, and is only cultured experimentally in Malaysia* In other areas, the goby is primarily collected from the wild. Once again, the Thai culture system will be used as our example. Feeding Habits Food habits of the sand goby change greatly throughout the life history and development. In the wild the young are planktivores which feed mainly on rotifers. Once they reach 3 to 5 cm, insects and larval fish predominate in the diet. As they grow, their diet gradually shifts to include mainly crusta- ceans and insects by the time they reach 25 cm. Fish occasionally are eaten by adults. In culture systems, adults are mainly fed ground fish (90%) and rice bran (10%), or solely ground fish. Supplemental feeding of larvae is with rotifers (for fish up to 5 mm) and water fleas or chironomids (for fish 5 to 18 mm). Growth Rates The growth of gobies both in nature and under culture is generally slow. In pond culture, maximum adult growth rate is only 0.3 to 0.5% BW per day, and in cages growth is even slower (0.2 to 0.3% BW per day). Relative growth rates follow the typical decline with age. Average growth over the first 20 80 days of fry culture is 24% BW per day, but this declines to 5% BW per day during the next 30 days . Reproduction In the wild, the sand goby is a nest building fish. The spawning season in Thailand is from May to October. Fecundity varies from 5,000 to 40,000 eggs per female, depending on size and age. Maturity is attained at a minimum size of 12.5 cm (34 g) , and an age of about 1 yr. Eggs are adhesive, and hatch 1.5 to 5 days after fertilization. Currently, all seed for sand goby culture is obtained from wild fish which have been selected for large size (near 100 g). However, spawning is being experimentally induced at Kasetsart University. Brooders are injected with HCG or common carp pituitary extract, and placed in 38-liter aquaria in pairs. Aquaria are each supplied with a nest constructed of asbestos roofing material. Usually spawning occurs within 3 to 7 days after injection. Data from four hypophysized females (Table 11-1) indicate that an average of 64% of ovarian eggs per female were laid (range 39 to 75%). Overall hatching success was 70%. Survival over 50 days averaged 42% (range 35 to 55%). These fry were cultured with a supplemental feed of rotifers (2,000 to 3,000/L) for 20 days, followed by water fleas (1,000 to 1,500/L) for 30 days. The experiiuient was discontinued when the fry were 50 days old at an average size of 1.8 cm. Recently, eggs for experimental culture have also been collected by placing slates of roofing material in ponds stocked with brooders. The tiles are checked regularly for eggs, which are attached by the female to the slates. Tiles with eggs are removed and placed in aquaria, where hatching occurs. 81 TABLE 11-1. Fecundity, hatching rate, and fry survival for four female sand gobies which were experimentally induced to spawn by injection of HCG or carp pituitary extract. The study was conducted at Kasetsart University, Bangkok, Thailand, during 1982 (Boonbrahm, unpublished data). Survival rate Fish weight Fecundity Eggs laid Eggs hatched over 50 days (g) No. No. (%) No. (%) No. (%) 480 59,100 42,700 (72) 34,600 (81) 12,700 (36) 450 54,200 20,900 (39) 6,700 (32) 3,700 (55) 435 53,700 40,200 (75) 30,200 (75) 9,800 (32) 460 56,300 40,700 (72) 34,500 (88) 14,800 (42) Culture Systems The most common method of culture in Thailand is cage culture. Cages, ranging in dimensions from lto3mx3to8mxlc5m, are floated in rivers. Stocking rate varies with fingerling supply, which is dependent on the success of collecting suitable stocking sizes from the wild. The pre- ferred stocking density is about 100 flsh/xsr or 10 to 20 kg/m^. Fish are given supplemental feeds of ground fish or 90% fish and 10% rice bran at about 5% BW per day. The food mixture is placed in a basket and lowered into the cage* Culture duration varies between 6 and 12 months, depending on growth and the size at stocking. Mortality during culture is dependent on the size of fish stocked. Survival rates after 6 months for fish stocked at less than 100 g are about 60%. Fish of 100 to 200 g, 200 to 300 g, and larger than 300 g, survive at rates of 80%, 90%, and greater than 95%, respectively. Harvested size ranged from 400 to 700 g for one experimental cage culture system (Table 11-2). Average production in cages was about 20 to 35 kg/m^/year . 82 n3 H 0) O T3 a > o cd o CO PQ to CO a u o CO ;2; •H 4J •H CO CO 0) M •H OJ ^ > O -H to C 1=) a u CO M CO CO CO CU CO O 4J iH CO to ^ CO CO a •H iH ^ CO 3 4J CU (U :3 •H a CO X u (U ^ c u o o o »+-< PQ CO >^ CO U O CO ZJ B ^ B C =3 O CO o • u «— I CO 4J • rH ^ C CO CsO^^ CO C "H 00 s F14 :2 0) O CO 4J CO c rH V4 O 3 :3 CO *J a -H ^ ^N CO 4J 60 00 u . 2 c! 5 t G» tt» Q. ai £ 3 ^> 2 =5 ti S S2 2: O ^ z g Z I- ui a. ^ W^^WVW^V^WWWW WWW X^ .\\\\\V\\\^^ o o ^ ii u < 1- z uj en 2 O UJ UJ -J UJ flu u- 0. 3 M Q < CO )i Jl z o Z 8 BIOCHEMICAL OXYGEN DEMAND 3 < O < CO CO o c CTJ u o o 0) d GO I CM 2 3 a fa 95 CHAPTER 13. SUTCH'S CATFISH (PLA SAWAI) Introduction Sutch's catfish, Pangasius sutchi , is native to India and Burma, but was presumably introduced to Indonesia, Thailand, and Malaysia. It primarily inhabits rivers, and in Thailand is most abundant in the central region, particularly in the upper Chao Phya River basin* However, some populations live in still waters including man-made lakes and ponds. There are few reports on culture of this catfish in India, Burma, or Indonesia. In Thailand, its culture is widely practiced, but is most devel- oped in the central part of the country. Total production obtained from culture of this fish in 1980 was 4,200 tons, valued at $2.9 million. Feeding Habits Fry of Sutch's catfish typically feed on zooplankton, whereas fingerllng and adult catfish are more omnivorous but prefer animal matter. Adults con- sume both dead and live animals. In culture, this catfish is fed with supple- ments of trash fish, rice bran, and cooked broken rice. Growth Rates The fragile, newly-hatched larvae measure 3.0 to 3.5 mm. By 7 days after hatching, they are fry 1.0 to 1.5 cm long. Under natural conditions, catfish can attain 80-90 cm in total length and 9-10 kg in body weight within 5 years. When cultured, they can reach 1. 5 to 2.0 kg in a single year. Maximum growth of Sutch's catfish is obtained at water temperatures of 30 to 32 ''C; temperatures less than 25 °C retard grox^h. The species can survive, with no effects on growth, in water with DO as low as 2 mg/L. 96 Reproductio n Sexual maturity is attained at about 1 year of age and 1.0-1.5 kg. Spawning occurs in flowing water during the flood season, which in Thailand is usually between May and September. In the spawning season, fish from backwaters and swamps migrate to areas of current to reproduce. A female weighing 5 kg holds approximately 830,000 eggs which average about 1.1 mm in diameter when newly laid. The eggs are adhesive and those fertilized hatch in 24 to 48 hours in water ranging from 28 to 31 ""C. The sex ratio at mating is one female to two males.. Culture Systems Brooding Brood production of Sutch's catfish is carried out in cages suspended in rivers, and is one of the most successful methods of aquaculture employed in Thailand. Cage culture in flowing water benefits from excellent removal of wastes, which creates suitable conditions for production with controlled water quality. In one method, fifty unsexed catfish averaging 1 kg are stocked in a nylon mesh cage (dimensions 6 m x 2.5 m x 1.5 m) . The fish are fed a sup- plemental diet composed of 60% ground trash fish, 20% cooked broken rice, and 20% duck starter feed at the rate of 1..5% BW per day. After 1 year, the fish average 2.5 kg and can be used as spawners. There is virtually no mortality. In hatcheries, spawners are usually replaced annually. Nursery A satisfactory nursery pond for catfish fry should be earthen, rectangu- 97 lar, and over 0.2 ha. New ponds are generally fertilized with organic manure* < Grow-out ponds do not require fertilization- Nursery ponds are filled with river water to a depth of 0.8 m, and stocked with live Daphnia o Twelve-hour-old catfish larvae are then released into the ponds. The larvae are fed with pulverized hard-boiled egg yolk twice a day (20% BW/day). Ten days later, the larvae have become fry measuring about le5 cm, and are then fed on finely ground trash fish at 20% BW/dayc After 10 days, the fry have reached about 2.5 cm with 60% survival. When between 2 and 3 cm, the young are removed and stocked in other ponds at the rate of 80,000 to 95,000 per ha. They continue to be fed daily with ground trash fish in suitable amounts. Within 30 days, the fry grow to become fingerlings approximately 8 cm long. Survival is 90% during this period. In Thailand, Pangasius cannibalism during the first week of experimental culture is controlled by use of water circulation in circular tanks (through rheotactic effects on orientation) and delivery of zooplankton via this circulation. Upstream orientation of early stages reduces antagonistic encounters and predation. Reduction of water quality problems also reduces levels of antagonistic interactions. Grow-out systems Either earthen ponds or cages are used for rearing fingerlings to market- able sizes. (a) Earthen ponds Earthen grow-out ponds for Sutch's catfish should be larger than 0.1 ha with a minimal water level of 1.5 meters. The ponds are stocked with finger- lings at a rate of 2 to 3/m^. The fish are fed cooked mixtures of 60% water 98 hyacinth, 30% rice bran, and 10% broken rice, and reach 1.5 to 2 kg In one year. Production rates average 30,000 kg/ha. Alternatively, the flngerllngs can be grown to marketable size using pig manure directly as their diet. Small ponds of 0.1 ha are stocked with 4,500 flngerllngs and supplied with manure from 45 pigs. Fish of 2.3 kg are produced In 14 months with a 20% survival rate. The total yield Is 18,000 kg/ha/year. (b) Cages The cages commonly used for culturlng catfish In Thailand are wooden and measure 6 m long, 2.5 m wide, and 1.5 m deep. They are floated In streams, notably the Chao Phya River. Stocking density depends upon the size of the fish. For advanced flngerllngs weighing 300 to 400 g each, the optimum rate Is 80 to 100 flsh/m^ of cage. The fish are fed supplementally with a mixture of 50% broken rice, 40% rice bran, and 10% ground trash fish. After 1 year, the fish are marketable size with an average weight of 1.5 kg. Annual yield Is 100 kg/m^. Limiting Factors Poor water quality, such as low DO, does not grossly limit catfish produc- tion, especially In flowing-water cage culture. The species also Is very tolerant of diseases and parasites. The main factor limiting the expansion of catfish farming In Thailand appears to be market value. Pangaslus sutchl cost less per pound than Clarlas species In Thailand. Although demand is fairly high, the retail market price is only $0.70 per kg. This value is low when compared to the cost of feed ($0.43) needed to produce 1 kg of fish. The low market value of Sutch's catfish is due mainly to the poor quality flesh of fish. This is acute in fish fed nutritionally Incomplete diets, especially those containing very high levels of aquatic weeds. This problem can be remedied by current experiments on catfish nutrition. 99 Model Network The network J for Sutch*s catfish is relatively simple, with supplemental feed and an oxygen submodel (Fig. 13-1 )o There are some differences in fry and adult culture, but not enough to warrant separate models* Adult culture can include fertilization with pig manure, but it is uncertain what portions of supplemental inputs (manures) are directly consumed by the fishe Under cage culture, the oxygen submodel becomes irrelevant due to flowing water renewal c Relatively little has been published on Sutch's catfish, especially in relation to bioenergetics and growth. There is insufficient information to quantitatively model populations of this species at present. 100 p s L t i /~ \_: ^ I- CO UJ o fO ^ 8 2 « (E a. UJ p T r^zzi" ^'-J «? OS UJ LU ~\ =^ J g _j z ^ rf UJ CL o o _J M a o £ o (E C9 I ?=> I z i I ^ I L-r-l < UJ 2 Q UJ UJ -J UJ a. u. a. CO 2 0. M a X Ui ^ »^ X CO ^ _ BIOCHEMICAL OXYGEN DEMAND < s CO M-l O 3 o u o G >^ u c w I M 101 CHAPTER 14 « SUMMARY Introduction The purpose of the workshop was to evaluate the major factors limiting fish production in Southeast Asian freshwater systems, while using a computer modeling framework to assist in future experimentation and planning. With this in mind, an international meeting was conducted in Bangkok during April 1983« The previous chapters have outlined the production systems for the 11 species most commonly propagated in the area. Given as an overview here are the fac~ tors implicated as limiting aquacultural production, plus future directions for research and extension. Major Limiting Factors At the conclusion of the workshop, each participant was asked to list the two factors most important in limiting pond fish production. A total of seven general factors were listed (Table 14-1), with the most important by consensus being water quality and feed availability. In addition, we reviewed each chapter to tabulate limiting factors for these species. Summed over all species, the most important factors were control of diseases, parasites and predators, and market potential (Table 14-1). It is not surprising that the two approaches gave somewhat different results, as the methods of compilation differed. Further, most of these limiting factors do not operate inde- pendently to affect production in culture. For instance, reduced water quality and high stocking densities are directly related to cannibalism, antagonistic interactions, and bacterial diseases in Clarias culture. It was, however, somewhat surprising to see market potential rated so highly in both instances. This factor is one which we can do little to improve, since it 102 •H U 0) CD 4J 4J •P 4J ^ rC O >> O o 0) M C W o cd (0 o ICO o CO P4 o CO i CO a II o ••» Id a O CO u u o •H *H CO • C rH O CO Q- :3 CO C (U CO ^ iH CO CO •H J-I 4-» > C O 4-» CO u il fO * (U u •• M-^ CO CO a T3 tH CU 4J N CO ?-l o CO V4 U-l 0) CD CO CO O O CO I Px3 "U 3 CO u x: CO o o CO u CO 3 CO cr 3 O cr >% o 60 CO •H S ^ Z3 C CO CO O • T3 CO 0) CU CO •H CO CJ ^ a. CO CO 'H u iH CO •H O CO CO Q^ CO U U CO O O U o u o CO 1 + 1*1 + + + I -K I I I I + -K + I 4- + I I I I I + + -K I + + * I I I I + + -JC I I I I + + I I -K I I >i CO u u •H O rH 4J CO CO d CO Cr rH CU CO CO CO ^ O •H O rH m CO O u cr o CO CU u •H CO CO u 4J 4-» M CO CU •H (U CO CO CO TS 0) (U CO (U "H 4J U c u CU . ^ CO C o 60 -H CJ CJ CO CO . I + -K I I I I + +11*11 I I I 4C + I >> CO 4J ^ •H o >. r-^ u 4J CO CO •H Id 'Td U cr (U C u CO Td a Id >. C >^ a*rH CO #« u CU CO •H TJ CU >. (U iH a CJ rM 4J u •H CO CO CO •H •H u ■P •H r^ CO CJ U ^.Td 4J •H CO (U Q) ^ o c JQ >-J M-l •H O (U CO CO rH a 60 #« CO o •H CO 0) 3 >-i CU CO •» CJ N cr o > (U CO •H u CO CO B CO U CO CU CO 0) ^3 ^ nd CU nd 'TS 4J (U M 0) CO C! jd CO (U CO