THESIS LIBRARY Michigan State ‘ University r I Vfi #— This is to certify that the dissertation entitled A MODEL TO FACILITATE OPTIMAL AQUACULTURE PRODUCTION BY QUANTITATIVELY RELATING FISH GROWTH TO FEED AND OTHER ENVIRONMENTAL RESOURCES presented by Charles S. Annett has been accepted towards fulfillment of the requirements for DOCTOR OF PHILOSOPHY degree“, Fisheries and Wildlife K Major professor, > .__ Darrell King Date 3// 5/9: MSU is an Affirmative Action /Equal Opportunity Institution 0-12771 MSU LlBRARlES » ,r__ -_.> RETURNING MATERIAL§z Place in book drop to remove this checkout from your record. flfl§§_will be charged if book is returned after the date ..- i....le stamped below. mew» Wit—1%, ' I I ' ”5510812095 Lififiu‘lmgcéw' A MODEL TO FACILITATE OPTIMAL AQUACULTURE PRODUCTION BY QUANTITATIVELY RELATING FISH GROWTH TO FEED AND OTHER ENVIRONMENTAL RESOURCES By Charles S. Annett A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Fisheries and Wildlife 1985 ABSTRACT A MODEL TO FACILITATE OPTIMAL AQUACULTURE PRODUCTION BY QUANTITATIVELY RELATING FISH GROWTH TO FEED AND OTHER ENVIRONMENTAL RESOURCES BY Charles S. Annett Aquaculture, the pr0pagation and rearing of an aquatic organism, entails the provision of all requirements for growth. The long-term process of natural selection has resulted in each type of organism having specific environ- mental requirements. As a result, optimal production requires the ability to relate the limits of growth of the organism to the availability of particular environmental variables in a quantitative manner. The intent of this research is to determine the usefulness of a model which simplifies the complex biochemical, biOphysical elements of growth, yet also provides the ability to quantitatively relate growth to environmental variables. A review of existing models of growth revealed that the "logistic" curve was important for initiating a considera- tion of rate of growth, and the "enzyme-kinetic" model important for relating growth rate to substrate concentra- tion. These concepts, combined with a consideration of the metabolic expenditures of the organism, resulted in the formulation of a threshold-corrected hyperbolic equation to model growth. The equation provides a means to predict the specific growth rate and conversion efficiency of an organ- ism corresponding to specific feeding levels and the feeding level required to maintain the organism at zero growth. Furthermore, the quantitative nature of the model permits comparisons of growth responses obtained under varying environmental conditions. The threshold-corrected hyperbolic equation was demonstrated to be useful in relating the specific growth rate of Tilapia zillii to three different feeds. Differ- ences in growth were attributed to the energy and protein content of the feeds. Additional experiments were performed which revealed the usefulness of the model to relate growth to fish size and water temperature. Comparison of specific growth rates of I. zillii and Sarotherodon niloticus on identical feeds revealed that the equation provided data for the comparison of different species and their abilities to harvest, digest and assimilate feeds. The threshold-corrected hyperbolic equation has poten- tial application for the following: 1) identify the dietary requirements of organisms, 2) quantify the effect of envi- ronmental variables on growth, 3) quantify the ability of the organism to harvest-digest-assimilate feed, 4) provide data to determine stocking levels and harvest date, and 5) provide a basis for cost benefit considerations. ACKNOWLEDGMENTS A prOper acknowledgment of the support I received in completing this program must begin by expressing my depen- dence on Jesus Christ for guidance and strength. Within that context, I am indebted to many peOple without whom it would not have been possible. I appreciate the support of my family and friends, especially my wife, Dani, who provided encouragement, understanding, patience, and income. Dr. Darrell King, my major professor and mentor, con- tributed untold hours of insightful discussion concerning both this research and life in general. His encouragement and demonstration of excellence in and commitment to the teaching profession is much appreciated. Drs. Don Carling, John Gerrish and Don Hall also contributed significantly with constructive suggestions and encouragement. Many thanks to Dr. Frank D'Itri for his concern, friendship, and the Opportunity to maintain an assistant- ship under his supervision. Craig Spencer and Jane Repko provided examples of excellence in research, and Craig offered much needed Opportunities for recreation. Lastly, the expertise and character of Lynn Kinzel contributed greatly to the completion of this document. LIST OF TABLES . LIST OF FIGURES. . . INTRODUCTION . . . . . MATERIALS AND METHODS. TABLE OF Apparatus and Procedures. Analytical Methods. Temperature. Alkalinity . pH . . . . . . . Ammonia-Nitrogen Dissolved Oxygen Caloric Content. Total Nitrogen-Pr t i Calculations. . . . Growth Rate. . . Determination of 9 Linear Regressian . O S e D u max Non-Linear Regression . Efficiency RESULTS 0 O O O 0 Temperature - Growth. Size - Growth Feed Type - Growth. Species Comparison. . iii and K CONTENTS coca) Constants Page viii 41 41 51 58 63 DISCUSSION 0 O O O O O O O O O O O O O O O O O O O O O 68 variability O O O O O O O O O O O O O O O O O O O O 68 Growth- Feed Relationship. . . . . . . . . . . . . 82 Feed Quality (Protein- -Energy). . . . . . . . . . 82 Harvest and Digestibility. . . . . . . . . . . . 91 Species Comparison . . . . . . . . . . . . . . . 102 Growth-Size Relationship. . . . . . . . . . . . . . lll Growth-Temperature Relationship . . . . . . . . . . 128 SUMMARY 0 O O O O O O O O O O O O O O O O O O O O O O O 155 BIBLIOGRAPHY O O O O O O O O O O O O O O O O O O O O O 168 APPENDICES . . . . . . . . . . . . . . . . . . . . . . 176 ‘ iv Table 1. 2. 10. LIST OF TABLES Summary of experimental conditions . . . . . Weight in grams of Tilapia zillii fed trout Chow at 27.5°C O O O O O O O O C O O O O O O Specific growth rates of Tilapia zillii fed trout chow at 27.5°C . . . . . . . . . . . . Regression data and growth constants calcu- lated for I. zillii fed trout chow at 25.7°C . Growth efficiency data calculated for I. zillii fed trout chow at 27.5°C. . . . . . . Observed specific growth rates and predicted specific growth rate data calculated by both linear and non-linear regression techniques for I. zillii (average weight 3.3 g) fed trout chow at 21.5°C . . . . . . . . . . . . Observed specific growth rates and predicted Specific growth rate data calculated by both linear and non-linear regression techniques for I. zillii (average weight 4.0 g) fed trout chow at 25.0°C . . . . . . . . . . . . Observed specific growth rates and predicted specific growth rate data calculated by both linear and non-linear regression techniques for I. zillii (average weight 10.3 g) fed trout chow at 27.5°C . . . . . . . . . . . . Observed specific growth rates and predicted specific growth rate data calculated by both linear and non-linear regression techniques for T. zillii (average weight 6.8 g) fed trouf chow at 30°C . . . . . . . . . . . . . Observed specific growth rates and predicted specific growth rate data calculated by both linear and non-linear regression techniques for I. zillii (average weight 7.3 g) fed trout chow at 30°C . . . . . . . . . . . . . V Page 21 29-30 31-32 36 39 43 44 45 46 47 Table 11. 12. 13. 14. 15. 16. 17. 18. 19. Observed Specific growth rates and predicted specific growth rate data calculated by both linear and non-linear regression techniques for T. zillii (average weight 7.3 g) fed trout chow at 35°C . . . . . . . . . . . . . Summary of growth constants calculated by both linear and non-linear regression tech- niques for T. zillii (average weight 6.3 g) fed trout chow at five temperatures. . . . . Trout chow particle size corresponding to each size class of T. zillii . . . . . . . . Observed specific growth rates and predicted specific growth rate data calculated by both linear and non-linear regression techniques for 0.0140 g (average weight) I. zillii fed trout chow at 27.5°C . . . . . . . . . . . . Observed specific growth rates and predicted Specific growth rate data calculated by both linear and non-linear regression techniques for 1.76 g (average weight) I. zillii fed trout chow at 27.5°C . . . . . . . . . . . . Observed specific growth rates and predicted specific growth rate data calculated by both linear and non-linear regression techniques for 38.1 g (average weight) I. zillii fed trout chow at 27.5°C . . . . . . . . . . . . Summary of growth constants calculated by both linear and non-linear regression tech- niques for four size classes of I. zillii fed trout chow at 27.5°C . . . . . . . . . . Observed specific growth rates and predicted Specific growth rate data calculated by both linear and non-linear regression techniques for T. zillii (average weight 2.1 g) fed Spirhlina aIgae at 27.5°C. . . . . . . . . . Observed specific growth rates and predicted Specific growth rate data calculated by both linear and non-linear regression techniques for T. zillii (average weight 13.8 g) fed pellEted-EIhaea canadensis at 27.5°C . . . . vi Page 48 49 51 53 54 55 56 59 60 Table 20. 21. 22. 23. 24. 25. 26. 27. Summary of growth constants calculated by both linear and non-linear regression tech- niques for I. zillii fed three feeds at 27.5 C O O O O O O O O O O O C O O O O O O 0 Observed specific growth rates and predicted specific growth rate data calculated by both linear and non-linear regression techniques for S. niloticus (average weight 4.1 g) fed trout chow at 27.5°C . . . . . . . . . . . . Observed specific growth rates and predicted specific growth rate data calculated by both linear and non-linear regression techniques for S. niloticus (average weight 3.9 g) fed pelleted SpiruIina at 27.5°C . . . . . . . . Summary of growth constants calculated by both linear and non-linear regression tech- niques for T. zillii and S. niloticus fed similar feeds at 27.5°C. T . . . . . . . . . Specific growth rates of Sarotherodon nilot- icus fed 3% trout chow per day -- four fish Per tank 0 O O O O O O O O O O O O O C O I 0 Observed specific growth rates, averages, and 95% confidence intervals (two-tailed t test) for variability study. . . . . . . . . . . . Ranges in weight and specific growth rate obtained from three broods of S. niloticus . Concentration of Spirulina Sp. cells in 15.2 1 valume. O O O O O I O I O O O O O O O vii Page 61 64 65 66 71 74 76 107 LIST OF FIGURES Figure Page 1. Dependence of rate on substrate concentration in enzyme catalyzed reaction . . . . . . . . . . 12 2. Threshold-corrected dependence of rate on substrate concentration. . . . . . . . . . . . . 18 3. Observed growth data and the predicted specific growth rate and efficiency curves for I. zillii fed trout chow at 27.5°C . . . . . 40 4. Predicted Specific growth rate responses for I. zillii (average weight 6.3 g) fed trout chow at five temperatures. . . . . . . . . . . . 50 5. Predicted specific growth rate response for four size classes of I. zillii (average weights 0.014 g, 1.76 , 10.0 g and 38.1 g) fed trout chow at 27.5 C . . . . . . . . . . . . 57 6. Predicted specific growth rate response for T. zillii fed trout chow, pelleted Spirulina and peIIeted E. canadensis at 27.5°C . . . . . . . . 62 7. Predicted specific growth rate response for I. zillii and S. niloticus fed trout chow and peIIeted spiruIina at 27.5°C . . . . . . . . . . 67 8. Frequency distribution for the weights of individuals from three broods of S. niloticus thirty days from thh 0 O I O O O O O O O O O O 78 9. Predicted specific growth rate response for four experiments restricting usage of non- linear regresssion analyses. . . . . . . . . . . 80 10. Comparison of the growth response predicted by both linear and non-linear regression analysis for 1.76 g I. zillii fed trout chow at 27.5°C. . . . . . . . . . . . . . . . . . . . 81 viii Figure 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. Predicted specific growth rate response of T. zillii fed trout chow, pelleted Spirulina ahd peIIeted S. canadensis at 27.5°C and compared on a protein-fed Basis . . . . . . . . . . . . Efficiency growth curves for I. zillii fed trout chow, pelleted Spirulina and peIleted S. 'canadensis at 27.5°C . . . . . . . . . . . . . Relationship between the ability of S. nilot- icus to harvest suspended Spirulina celIs and the concentration of suspended celIs . . . . . Efficiency growth curves for four size classes of T. zillii (average weights 0.014 g, 1.76 g, 10.0 g and 38.1 g) fed trout chow at 27.5°C. . Calculated u values for four size classes 0f.I° zilliimféd trout chow at 27.5°C and the exponentiaI curve fit to the data (r - 0.97; n - 4) o o c o o o o o o o c o o 0 Calculated S values for four size classes of T. zillii fed trout chow at 27.5°C and the ‘Iog equation fit to the data (r - 0.94; n - 4) . . . . . . . . . . . . . . Calculated um values for I. zillii (average weighf 6.3 g) fed trout chow at 21.5, 25, 27.5, 30 and 33°C and the quadratic curve fit to the data (r - 0.77; n - 5) . . . . . . Calculated S values for T. zillii (average weight 6.3 g) fed trout chow at 21.5, 25, 27.5, 30 and 33°C and the quadratic curve fit to the data (r - 0.76; n - 5) . . . . . . The three-dimensional relationship between the specific growth rate response and both temper- ature and feeding level of trout chow for I. zillii (average weight 6.3 g). . . . . . . . . Efficiency growth curves for I. zillii (average weight 6.3 g) fed trout chow at 21.5, 25’ 27.5, 30, 35°C 0 O O O O O O O O O O O O C The three-dimensional relationship between feed conversion efficiency and both temperature and feeding level of trout chow for I. zillii (average weight 6.3 g) . . . . . . . . . . . . . ix Page 87 90 109 119 121 125 141 144 147 151 153 INTRODUCTION Aquaculture may be defined as the prOpagation and rearing of an aquatic organism in either an artificial or natural environment. The goal of aquaculture is the cost effective production of an organism. Cost effectiveness may be expressed in either financial terms as monetary reward, or in human terms as the production of an edible product from poorer quality resources. All aquaculture programs share the common purpose of attempting to grow an organism‘ by creating an environment conducive to its most efficient propagation and growth. The long-term process of natural selection has resulted in each type of organism having specific environmental requirements. The unique assemblage of structural, behavioral and biochemical processes charac- teristic of each organism enables it to achieve maximum efficiency under a particular set of environmental con- straints. Alterations in one or more of the environmental constraints results in decreased efficiency and/or yield. The development of an effective aquaculture program requires the ability to relate the limits of growth of the organism in question to the availability of particular environmental resources in a quantitative manner. The first criherion for the growth of an organism is a continued energy supply. In addition, the individual 1 2 elements required in the anabolic process must be provided. Unless a sufficient supply of energy and building components are supplied to the cellular reaction sites, growth cannot occur. The effective promotion of growth requires the identification and quantification of the types of food which efficiently stimulate growth. Unfortunately, the determin-. ation of food requirements for an organism is complicated by the numerous environmental factors which influence the metabolic processes in the organism. Brett (1979) reported on the complex, interactive influence of numerous environmental factors on growth. Some measure of simplification was achieved by classifying biotic and abiotic factors into categories based on how they affected the growth of the organism. The four categories cited by Fry (1971) and Brett (1979) which influenced growth were: 1) controlling factors, such as temperature, which govern reaction rates; 2) limiting factors which restrict reaction rates, such as ration level; 3) masking factors which modify reactions or processes, for example, salinity; and 4) directive factors which cue, or signal particular responses, such as photoperiod. Notwithstanding these simplifications, the growth of an organism is understood to be a complex process. First the organism must be able to successfully harvest, digest and transport all the resources required for growth to the in- ternal reaction sites. Simultaneously, these physiological 3 processes must occur within an environment where all the factors which influence these growth processes are within a functional range. The task of develOping a procedure to quantitatively relate growth to both food and environmental factors appears formidable. For this reason there have been numerous attempts to simplify the complexity by correlating growth with individual factors. These correlations have been investigated from both a limnological and fisheries stand- point and have included both biotic and abiotic parameters. Numerous authors have attempted to relate fish growth or yield to limnological factors. Hayes and Anthony (1964) correlated fish production capacity with the parameters of lake area, depth, and alkalinity. In examining 150 North American lakes, the authors reported that these parameters accounted for 67% of the variability in fish productivity. Ryder (1982) employed the Morphoedaphic Index (MEI) as a means of relating fish growth to environmental parameters. Ryder's original MEI is obtained for any lake by dividing the concentration of total dissolved solids by the mean depth. Thus, the index includes some measure of the chem- ical and physical character of a lake. The MEI was proposed as a "reasonable compromise between unmanageable complexity and ecological oversimplification" (Ryder 1982). While these correlations may provide a first approxi- mation for total fish yield, they are too simplistic to 4 provide species-specific predictions or to relate production to potentially manageable resources in the environment. Because they fail to relate growth to specific dietary or environmental requirements, these correlations provide limited insight for aquaculture management. To a large degree, these attempts to utilize abiotic parameters to predict fish production are actually an indirect means of relating the primary production capacity of the water to fish production. The parameters of area, depth, nutrient concentration, etc. have limited direct impact on fish but greatly influence the levels of primary production which ultimately represent the feed supply level for the fish. Other attempts have been made to relate fish yield to biotic factors, usually primary production. Melack (1976) related the fish yield in tropical lakes to the levels of net phyt0plankton productivity. Fish yields were reported to increase logarithmically in response to arithmetic in- creases in primary production. This correlation was demon- strated to provide predictive ability superior to that of the Morphoedaphic Index. However, the author expressed concern over both the reliability of the primary production data and applicability to more diverse lakes. Liang et al. (1981), using a variety of curve fitting techniques, re- ported that the relationship between fish yield and gross photosynthesis in several Chinese lakes could be described by arithmetic, semilog or log-log regression equations 5 depending on the environmental conditions, diversity of harvested fish and fishing method. The authors suggested that, in general, a logistic curve may describe the full range of the fish yield/gross primary productivity re- lationship. Edwards et a1. (1981) reported that despite constant changes in the phytoplankton community, the growth of Tilapia nilotica (also referred to as Sarotherodon niloticus) was linearly related to the mean phytOplankton concentrations in fish ponds. Similar investigation by Reich (1978) indicated a direct correlation between the harvest of Sarotherodon galilaeus and the total biomass of algae present from the time of fish hatch until harvest. Because the yield of fish has been correlated to primary production, there have been numerous studies investigating the effects of fertilization on fish yield. Seymour (1980) stated that the "production of fish ponds may be increased by fertilization which encourages growth of phytOplankton and in turn the amount of food available to the fish." Swingle and Smith (1939) reported that the addition of inorganic fertilizer greatly increased fish production in ponds. The findings of Dendy et a1. (1968) indicated that production of tilapia increased from 63 to 199% following inorganic fertilization. Schroeder (1974) similarly reported that the harvest of carp and tilapia increased 25 to 100% in ponds fertilized with fluid cowshed manure. The increased fish productivity was attributed to 6 increases in primary production ranging from 10-1000 fold. In cOnsequent research, Schroeder (1978) reported that increases in fish yield occur as a result of increases in both primary production and heterotroPhic production of microorganisms. Following addition to the fish pond, the manure is quickly colonized by a microbial community which digests the organic content. These colonizing bacteria and protozoa are ingested by the fish and become an important stimulant for increased fish production. Burns and Stickney (1980) reported that poultry wastes were effective in increasing the growth rate of Tilapia 52323 and that production paralleled the rates of organic fertilization. Stickney et a1. (1979) recognized that the purpose of organic fertilization was to increase primary productivity which, in turn, would stimulate increased fish production. The authors reported that fertilization with hog manure was effective in increasing the growth rate of Tilapia nilotica. While these correlations have resulted in a measure of increased fish production, they cannot be used to relate the growth of specific organisms to the availability of particular environmental resources in a quantitative manner. Another research approach, utilized by nutritionists, involves the investigation of the ability of formulated diets to promote growth. Smith (1971) prOposed the use of metabolic chambers to determine assimilation efficiencies of 7 food components. By force-feeding the fish and quantifying the separate excrements, the digestibility and nutrient balance of the feeds may be determined. Unfortunately, this method also entails stressing the fish due to force-feeding, restraint in the chamber, abrasion, handling and ammonia concentration in the chamber. While the methodology affords a measure of control to determine the digestion-assimilation efficiencies, it simultaneously results in an alteration of normal function. Application of the results to natural environments requires the questionable assumption of un- altered food processing rates in a stressful environment. An alternative approach involving less stress was utilized by Cho et a1. (1974). Digestibility of feeds are determined by adding 1% chromic oxide (Cr203) to the diet as a tracer and measuring the subsequent concentration in the feces. The relative concentrations of the chromic oxide in feed and feces are used to calculate digestibility and assimilation rates. This method provides the opportunity of studying a population of fish in an unstressed environment. However, the method is limited by the difficulties involved with collection of representative samples of fecal material and a restriction to digestible rather than metabolizable fractions. While these approaches do provide a limited means of evaluating the digestibility and assimilation potential of feeds, they cannot be used to relate the growth of the unstressed organism directly to environmental 8 resources in a quantitative manner. Because of the limitations involved with using correla- tions or empirical models to relate growth to environmental parameters, the practicality of more theoretical models will be reviewed. One well-accepted model proposed to charac- terize the growth of both individuals and populations is the logistic curve. The following summary of the model is based on a review by Kingsland (1982). In 1845 Pierre-Francois Verhulst proposed that during the early stages of growth, a population would increase exponentially until such time that critical resources became limiting. Upon reaching such limits, the rate of growth was retarded resulting in a symmetrical, sigmoidal curve of growth. The sigmoidal curve labeled the "logistic" can be written in differential form as: dN = rN K-N as (T) where K - maximum carrying capacity or upper asymptote (number) N - number of individuals (number) r - maximum rate of growth ( l ) time t - time Raymond Pearl and Lowell Reed independently introduced the same growth curve in 1920 to explain the growth rate of individual organisms as well as populations of bacteria and protozoans. While some researchers sought to define the logistic as a universal law of growth, others recognized it as a logical explanation or model for growth provided 9 certain criteria were met. These conditions were: 1) both r and K are constant, 2) no time lags exist, and 3) all individuals are equal in their effect on the reduction of the growth rate at each density. When the logistic formula is rearranged to express the rate of increase per unit, or individual, the following results: dN - r 1 - N Nd? ( ‘R) The (l -‘g) term represents the multitude of factors which contribute to the decreasing rate of increase in growth as the size of the individual or population in- creases. It was suggested by Lotka in 1925 that the reasons both individuals and pOpulations follow the same curve is that both share the same basic underlying mechanisms of growth Since populations consist basically of "populations" of cells. The fundamental principals underlying the growth process were poorly understood at the time of introduction of the model. Therefore, it is fairly judged to contain both theoretical and empirical elements. Modifications of the model have since been employed to explain r and K selection (MacArthur and Wilson 1967), competition (Hixon 1980), and predator-prey relationships (Chesson 1978). The logistic curve serves the important function of incorporating a rate dependency concept into growth modeling. Verhulst proposed that growth rates will be influenced by the availability of critical, limiting 10 resources. However, the logistic fails to quantitatively relate changing growth rates directly to particular en- vironmental resources. Instead, the model is simply descriptive of typical pOpulation growth rates occurring with time. Another approach to model or characterize the growth of organisms is based on the function of enzymes. In both uni- cellular and multicellular animals, the biochemical reac- tions contributing to growth are regulated by enzymes. Even the simplest manifestations of life involve many complex enzymatically controlled reactions (Hoar 1979). The en- zymes, while not contributing energy to the reactions, are necessary for the activation of the process. Specific proteinaceous enzymes associate themselves with a reactant (substrate) to form an intermediate complex. The formation of this complex allows the reaction to occur at a lower activation energy than would otherwise be possible. As the intermediate complex combines with another reactant to form a product, the enzyme is released and may be reused in the reaction process. The reaction is summarized as follows: E + S ——>I ES -—. P + E (Enzyme) (Substrate) (Enzyme-substrate) (Product) (Enzyme) The relative amount of both enzyme (E) and enzyme specific substrate (S) is critical to the reaction rate. At low substrate concentrations, the reaction rate should be ll proportional to the substrate concentration (S). At these substrate levels, the enzyme (E) is able to completely react with all the substrate (S) forming the enzyme-substrate complex (ES) facilitating the reaction. However, as the concentration of substrate is increased, a higher percentage of the enzyme (E) is bound in the intermediate complex (ES) and less is available to react with substrate. Eventually, if the substrate concentration is increased sufficiently, virtually all the enzyme would be bound in the intermediate complex (ES) with none in reserve to combine with substrate. As fast as the complex breaks down, the released enzyme combines with substrate to form more intermediate complex. Under these circumstances the enzyme is referred to as "saturated," and additional units of substrate yield no increase in reaction rate. The dependence of rate on sub- strate concentration is shown in Figure 1. This enzyme- substrate reaction rate response is described by the Michaelis-Menten equation as follows (Engel 1977): _ S V Vmax (R8 + S) where V - reaction rate ( 1 ) 1me - c o 1 Vmax max1mum react1on rate ( 1me) S - substrate concentration K - substrate concentration at which the reaction rate is one-half the maximum > '.'.‘ g I/2 m Vnox 2! 9. p. C) ‘1 hi on 0 Figure l. 12 Vmox S UBSTRATE CONCENTRATION ($1 Dependence of rate on substrate concentration in enzyme catalyzed reaction. - 13 In the equation, when S becomes very large compared to K the reaction rate V approaches Vm . Conversely, when s’ ax S is very low relative to K8, the reaction rate is prOpor- tional to S. Monod (1949), working with bacterial cultures, stated that "despite the immense complexity of the phenomena to which it testifies, growth generally obeys relatively simple laws, which make it possible to define certain quantitative characteristics of the growth cycle." These growth cycles were recognized by Monod to consist of different phases with the extent of each phase being determined by the composition of the media in which the bacteria grew. Growth was demon- strated to be restricted by limiting factors such as nutrients or adverse environmental factors. When experi- mental conditions were established to provide an excess of all but one limiting nutrient, and the concentration of that limiting nutrient was gradually increased, the growth of the bacteria demonstrated a rapid initial increase, then grad- ually lessened the rate of increase until a plateau was reached at high substrate concentrations. Monod established that the form of the relationship is similar to that deter- mined for the enzyme-catalyzed reactions and prOposed the following equation to relate growth to substrate: 14 _ S u umax (KS+S) where u = growth rate( 1me) u - max1mum growth rate( 1me) max 8 - substrate concentration Ks - substrate concentration at which the growth rate is one-half the maximum. Clearly, this equation is the same form as the previous Michaelis-Menten equation. The constants "max and Ks were used to compare the abilities of different nutrients to stimulate growth of the bacteria. The levels of the constants are an expression of the ability cf the enzyme systems to utilize the particular limiting substrate for growth. The relation of growth to substrate is a simplification in that it does not examine the complex processes which comprise growth. Instead, the organism functions to integrate those complexities into one parameter--growth. Thus the enzyme-kinetic model of growth serves two important functions. First, it incorporates a consideration of rates of growth much like the logistic model. Secondly, the model allows the establishment of quantitative relationships between growth and specific environmental parameters. It would appear that the enzyme-kinetic growth model fulfills the initial objective of providing the ability to relate the limits of growth of an organism to the 15 availability of particular environmental resources in a quantitative manner. However, an examination of Figure 1 reveals that the growth curve passes through the origin allowing no provision for energy expenditures and nutrient utilization for basal metabolic activities. A review of some principals underlying these metabolic expenses will clarify their importance in relating growth to resources. The Second Law of Thermodynamics states that no process involving an energy transformation will occur spontaneously unless there is a degradation of the energy from a concen- trated form into a dispersed form. Therefore, as feed components are processed through the myriad biochemical reactions, a processing cost must be extracted from the feed to drive these transformations. It is expected that since some amount of feed must be allocated to meet these energy expenditures, the growth curve cannot pass through the origin when growth is plotted versus feeding level. Instead, at a zero feeding level, it is expected that negative growth should occur. This is necessitated because the continued functioning of the organism during starvation demonstrates that energy-consuming, life-sustaining reactions continue to occur. These processing costs, when combined with the enzyme-kinetic dynamics, would result in the hyperbolic growth curve being shifted downward to reflect the cost of energy transformations. l6 Organisms with extracellular digestion require that feed components diffuse through membranes prior to utili- zation. The diffusion processes are important in deter- mining the amount and rate at which nutrients are trans- ported to sites for anabolism, and waste materials carried away from catabolic sites. Transport from ingestion to reaction Site often involves diffusion across several membrane barriers. The rates of passive diffusion in an aqueous media are defined by Fick's Law of Diffusion as shown below: Diffusion rate - D S (gg) where S - cross sectional area D = diffusion constant c - concentration x - unit length perpendicular to the plane of S (g;) - concentration gradient At each membrane barrier, some portion of the feed, deter- mined by the gradient, area, etc., will remain outside the membrane. Therefore, the initial level of feed is neces- sarily greater than the level of feed available at the internal reaction Site. For this reason, the growth curve will again experience a downward shift representing a decreased growth rate due to incomplete diffusion and transport of feed components. 17 The growth curve proposed to represent the influence of enzyme kinetics on growth in response to limiting substrate concentration must be modified to incorporate the effects of diffusion processes and the Second Law of Thermodynamics. Both processes similarly shift the curve downward causing the appearance of a threshold feeding level (Figure 2). This feeding level, labeled Sq’ represents the amount of feed required to meet the basal metabolic expenses of the organism at a zero growth rate. The equation used to describe the curve must also be modified to incorporate this shift. By subtracting the maintenance feeding level, Sq, the equation now becomes: S-Sq u ‘ n max - + - EKS Sq) (8 Sq) The equation may be further simplified to: v - um .2_'_~°’_q_ KS+S-ZSq The threshold-corrected hyperbolic equation includes the maintenance feeding level (Sq) and the maximum rate of growth (u ). These values, along with KS, allow the max prediction of growth response to any feeding level of the resource utilized. The equation appears to quantitatively describe the relationship of the Specific growth rate to single limiting environmental parameters in a manner 18 [/2 JJmax . . . REACTION RATE (1J1 SUBSTRATE CONCENTRATION (3) Figure 2. Threshold-corrected dependence of rate on substrate concentration. l9 applicable to aquaculture. The intent of this research is to determine the usefulness of this model in quantitatively relating the growth of fish to the environmental variables temperature, feed type and quality, and the type and size of fish. MATERIALS AND METHODS This research involves numerous laboratory experiments, each from one to five weeks in duration, designed to examine the ability of the model to quantitatively relate the growth of Tilapia zillii and Sarotherodon niloticus (also referred to as Oreochromis niloticus) to particular feeds under con- trolled conditions. The growth of I. zillii was evaluated as a function of the following feeds: commercial trout chow; a pelleted macrOphyte (Elodea canadensis); and a pelleted blue-green alga (Spirulina sp.). The growth response of I. zillii to trout chow was examined at five different tempera- tures. Growth rates were determined for four size classes of ‘1. zillii as a function of trout chow feeding levels at near optimal temperature. The growth of S. niloticus was examined in reSponse to both commercial trout chow and pelleted Spiru- $322. The experimental conditions are summarized in Table l. The Tilapia zillii for the experiments were provided by Dr. Donald Carling at Michigan State University. The Sarotherodon niloticus were obtained from Dr. Ronald Phelps at Auburn University. Their assistance is greatly appre- ciated. Each of the experimental feeds was measured on a dry- weight basis and provided in distinct size categories. The Elodea canadensis was harvested from a 10 acre lake located 20 21 x o.oa macesoaae .m. e x o.o~ mmmmmm .m .em eeaaseamm x o.o~ Meagan .m senescence .m x o.o~ encased“: .m : x H.mm : = x x x x x o.oa : : x en.~ = : x e~o.o aaaaau .M :oeo “sous .o.mm .o.om .m.- .o.m~ .m.H~ Us ouauouomama va mafia owmuo>< mowoomm :mwm was» comm .mcofiuflocoo Hmucoewuomxo mo xuOEESm .H oHan 22 at the Michigan State University, Inland Lakes Research Study Center. The plants were blended, extruded into pellets, dried at 90°C for 24 hours, ground with a mortar and pestle, and sieved to a 1 to 2 mm size range using a U.S. Standard Sieve Series. The pelleted Spirulina Sp. was obtained as a powder from Microalgae International Sales Corp. The powdered algae was mixed with 1% carboxy methyl cellulose binder and sufficient water to form a paste. The paste was then dried at 90°C for 48 hours, ground with a mortar and pestle, and sieved to a 1 to 2 mm size range. The Purina Trout Chow was utilized in several size cate- gories depending on the Size of the fish. In each instance, the feed was ground with a mortar and pestle, then screened to the following size classes: 0.297 to 0.420 mm; 0.420 to 1.0 mm; 1.0 to 2.0 mm; and 2.0 to 4.0 mm. All feeds were stored in sealed containers to prevent spoilage. APPARATUS AND PROCEDURE A series of 30 gallon aquaria each equipped with air stone aerators, heaters, and external tank filters made up the basic system for most experiments. The aquaria were located side by side under a four bulb, 34 watt, fluorescent light fixture which was regulated to a light period from 730 hrs to 1930 hrs each day. All experiments utilized cold tap water from the Michigan State University water supply system which had originated in deep groundwater sources. The 23 aquaria were filled at least three days prior to an experi- ment to allow for aeration, chlorine removal, temperature adjustment and iron precipitation. Each outside tank filter contained synthetic fiber to remove particulate material and clinOptilite resin to reduce ammonia levels. Temperature control was provided by 150 watt Penn Plax aquarium heaters, except for the 35°C experiment which required 200 watt heaters. Plastic netting was placed over the aquaria to limit the renowned jumping ability of the fish. Approximately three weeks prior to each study, all fish were transferred to aquaria adjusted to the forthcoming experimental temperature and fed the apprOpriate feed at approximate maintenance levels. Initially, for each exper- iment, individual fish were captured in a nylon net, meas- ured for length and weight, and from zero to three of the dorsal fin spines clipped. This allowed the individual fish to be identified in succeeding weighing operations through the one to five week experimental period. The only excep- tion to this rule was the weighing process for the fry utilized in the size experiments. Because these fish were so susceptible to injury, an alternative method was devel- Oped. All of the fry utilized in the 0.0140 g experiment were from one brood of about 150 fish. Ten of these were randomly netted for the weighing process. The weights of these ten fish were found to average 0.0140 g and ranged from 0.0139 - 0.0141 g. Therefore, twenty fry were captured 24 with a nylon net and transferred directly to each aquarium, with the initial, individual weights in each case assumed to be 0.0140 g. During the weekly measuring process, the fish were handled using surgical gloves and kept on moist paper toweling. Weight measurements for all the experiments were conducted prior to the morning feeding at 800 hrs. Weights were determined to tenths of a gram using a tap-loading Mettler balance, with the pan covered by a moist paper towel. Due to the hardiness of the fish and the short handling time (less than one minute), the fish showed no adverse effects and demonstrated an affinity for feed within 30 minutes after their return to the aquaria. With the exception of the 0.014 g I. zillii, each ex- periment was conducted with one to four fish per aquarium, and each fish Spine clipped for identificaton. Each group of fish then received different amounts of dry feed based on a percentage of the total wet-weight of the fish per aquar- ium. The percentages of feed usually ranged from zero to five percent body weight per day (dry weight feed/wet weight fish), although in one experiment levels up to 10 percent were included. The feed was measured on a dry-weight basis using a Mettler Balance Model HlOT and was divided into two equal portions which were fed at 800 hrs and again at 1500 hrs. 25 The fish were fed twice per day to increase feed utilization efficiency as reported by Greenland and Gill (1979). Waste material was siphoned daily from each aquarium prior to the first feeding and the water replaced with conditioned water stored in an additional aquarium. Weights and lengths of the individual fish were measured weekly, and the required levels of feed recalculated using these data. Water quality was monitored by periodic determinations of temperature, pH, dissolved oxygen, alkalinity and ammonia levels. After analysis at the beginning of the experiment, the determinations were performed weekly, except for temper- ature which was measured daily. Protein and caloric content determinations were performed on all the feeds utilized. ANALYTICAL METHODS Temperature A standard laboratory mercury bulb thermometer was used to determine the temperature of each aquarium. Values were recorded at the time of the morning feeding. Alkalinity Titratable alkalinity was measured using methyl orange indicator and 0.02 N H2804 as described in Standard Methods (1980). 26 pg Measurements of pH were made using either an Orion model 901 digital microprocessor ionanalyzer with an Orion Ross combination pH electrode or a Beckman Expand-Mate pH meter with a combination electrode. Both units were calibrated with standard buffer solutions. Ammonia-Nitroggg Ammonia-nitrogen (NH4-N) was measured using the phenate method described by Solorzano (1969). The ammonia, phenol and hypochlorite react to form a blue-colored complex . which absorbs at a wavelength of 680 m . Color develOpment of samples and standards was measured with a Bausch and Lamb Spectronic 21 spectrOphotometer. Dissolved Ogyggg Dissolved oxygen levels were determined using a Y.S.I. model 54A dissolved oxygen meter. The instrument was cali- brated weekly using a Winkler-Azide Modification procedure as outlined in Standard Methods (1980). Caloric Content The caloric content of the feed was determined using a Parr Oxygen Bomb Calorimeter Model 1341. Samples were dried at 90°C for 48 hours, then ground in a Wiley Mill with a 20 micron mesh screen. The ground samples were pelletized and the caloric content measured using benzoic acid as a standard. 27 Total Nitrogen-Protein The total nitrogen content of feeds was determined using a Perkin-Elmer Model 240 Elemental Analyzer. Samples were dried at 90°C for 48 hours, then ground in a Wiley Mill with a 20 micron mesh screen. The Elemental Analyzer pro- vides a determination of the total nitrogen content by com- busting a known amount of sample at 1000°C, reducing the total nitrogen content to nitrogen gas, than measuring the thermal conductivity generated relative to that generated from standards of acetanilide. The total nitrogen concen- trations were then multiplied by 6.25 to convert to protein levels. The Elemental Analyzer provides accompanying readouts on hydrogen and carbon content of the samples. These data are included with the nitrogen-protein levels. CALCULATIONS Growth Rate In analyzing the growth data from each experiment, it is necessary to perform a series of calculations. To clarify these calculations, the data from one experiment will be carried through each step of the process with sample calculations performed. The data obtained from a feeding trial using Tilapia zillii fed trout chow at 27.5°C are used for this purpose. The initial data consists of the beginning weight of each fish and the subsequent weight at 28 the end of each week. The data in Table 2 provide the weight for each fish at each feeding level as well as the total weight of fish per aquarium, or feeding level. The specific growth rate of each fish, as well as the average for each feeding level, is calculated for each week's data using the formula: u - ln Wt - ln Wo t where u = s ecific rowth rate ( 1 ) p 3 1me W - weight at time t Wo - initial weight t - time (days) For example, in calculating the specific growth rate for fish #0 at a 0% feeding level during the first week, the equation would appear: u a 1n 12.6 gma- ln 13.5 g2 or u ' -0.00986/day ays Similarly, the average specific growth rate for fish at the J 0% feeding level during the first week would be as follows: u - ln 60.3 m - 1n 64.7 g2 or u = -0.0101/day '7days The specific growth rates calculated for all the weight data in Table 2 are contained in Table 3. 29 “.ma ~.ma o.e~ ~.eH o.ea fleece e.e e.e m.e m.e m.e a H.m e.w ~.w m.~ ~.~ o e sees m see: A see: H see: Heaueee e ease ease N~.o A.ma A.ma m.ma m.ma H.ma Hence 3. 3 3.. ma 3. s H.w m.w e.e ~.m o.m o e see: m sea: N see: A see: assuage a ease emmm an.o m.eH o.e~ ~.mH m.m~ e.m~ fleece qu H.k H.~ a.k H.~ a m.~ m.~ H.m N.» m.m o e game m awe: N see: A some Haasaee a ease seem NN.c m.mn m.em H.mn m.oe N.ee sauce «.5 e.~ H.m e.w m.e e e.HH e.aa e.aa o.~a ~.NH m m.m~ e.m~ A.m~ e.m~ a.ea N m.aa a.~e m.- ~.ma m.ea a e.HH o.~a ~.~H e.- m.ma o e see: m sees N see: A some HeeeaeH a ease some No .uom.k~ as gene scone emu eases» changes we mean» ea sewed: .N edema 30 o.mw N.Hm n.Hm n.0N e some 0.0m ~.HH ¢.NH ¢.< NNoo.o- oooo.o meoo.o- NNoo.o- NNoo.o N emoo.o NNoo.o emoo.o meNo.o oNoo.o- o eweem>< e News N News N News N News a emNN emoa NN.o eNoo.o oooo.o NNoo.o eNoo.o aNoo.o eweuue< eNoo.o mmoo.o anoo.o eoNo.o NNNo.o N mmoo.o- mmoo.o- NNoo.o- omoo.o- Neoo.e- o oweum>< e News N News N News N News N emNN ease Nm.o eNoo.o- emoo.o- NNoo.o- aooo.o- eNoo.o- «Neuese mooo.o- ONoo.o- oooo.o oooo.o oooo.o N meoo.o- emoo.o- emoo.o- NNoo.o- Nmoo.o- o oweum>< e New: m News N News N News N emNN same NN.o Nmoo.o- Nmoo.o- emoo.o- emoo.o- NoNo.o- oweeese NNoo.o- Nmoo.o- Nooo.o- eNoo.o- NNNo.o- e amoo.o- mNoo.o- oooo.o Neoo.o- Nmoo.o- m omoo.o- NNoo.o- oNoo.o- NNoo.o- owoo.o- N NNoo.o- emoo.o- eeoo.o- NNoo.o- SNNo.o- N oeoo.o- mNoo.o- eNoo.o- eeoo.o- oaoo.o- o mweuo>< e News N News N News N News a emNe some No .Uom.m~ an 3020 unouu vow “wages mwmoawfi mo magma nusouw ofimwooaw .m canoe .mmwmum>o Naouo>o SN cocsNocN No: one dude whose .cmumo on cu No: vo>uomoo mos comm Nam cons mxmos wcwuso pounoooo scan: amumu nuzouw acommumou mononucouom :N mouowwm 32 oNNo.o NNNNo.oV NoeNo.ov NNNNo.OV oNNo.o eweeeee - - - - - N NNNo.o NNNo.o NONo.e NNNo.o NNNo.e N NeNo.o NNNo.o oNNo.o NNNo.o NeNo.o N NNNo.o NNNo.o NNoo.o Nooe.o NNNo.o o eweee>< e gees N Nee: N Nee: N xeez e eeNe eeee No.e NeNo.o ANNNo.ov NNNo.o eNNo.o NNNo.o eweeese - - - - - N NONo.o eNNo.o oNNo.o eONowo NNNo.o N eNNo.e NONo.o NNNo.o NNNo.o NNNo.o N NeNo.o eNNo.o NeNo.o NeNo.o NNNo.o o eNeeese e Nee: N Nee: N Nee: N Nee: e eeNe eeee No.N NNNo.o oNNo.o NNNe.o NNNo.o NeNe.o eNeee>< NNNo.o NNNo.o NoNo.o meNo.o NNoo.o N oeNo.o NeNo.o Neoo.o NNNo.o NNNo.o N NNNo.o eNoo.o- NNNo.e NNNo.o NNNo.o N NNoo.o oNNo.o oooo.o oooo.o NNoo.o- o eweeese e sees N Nee: N Nee: N Nee: e eeNe eeee No.N .N.e.eeoev N eNeeN 33 The use of the specific growth rate formula provides advantages over both the incremental and relative growth rate formulas. The incremental growth formula 'Incremental Growth 8 Weightfinal - Weightinitial time has the deficiency of not relating weight gain to initial weight. For example, using the incremental growth formula, a 100 g weight gain by a 1.0 g fish is equal to the growth rate of a 1000 g fish which gains 100 g. The relative growth formula, shown below, is a linear approximation to the logistic formula. Growth Relative Weightfinal - Weightinitial 1 Weight t1me initial This method has the advantage of providing a measure of growth per unit weight of the organism. Using this formula to calculate the specific growth rate of the fish #0 during the first week at the 0% feeding level results in a figure of -0.01020/day, slightly lower than the -0.00986/day ob- tained with the specific growth rate equation. While the relative growth rate formula provides a close approximation of the logistic growth rate, especially over short time periods, the more accurate specific growth rate equation will be employed in this study. Utilization of either the specific growth rate, or its approximation, has two 34 additional advantages. First, the specific growth rate formula is easily rearranged to the exponential equation Mt Wf - Wie as follows: (1n W n = f - 1n wi) t ut (1n Wf - ln Wi) e“t -.Ef Wi - at Wf Wie The rearranged formula is useful in determining the weight of the organism or population at any point in time. A second advantage is the opportunity to utilize the data in relating growth to feed concentrations under defined con- ditions of limiting resources with the base equation 8-8 H g u 0 max (KS+S-25q) and Ks Constants Determ1nat10n of 89, “max Linear Regression. The next stage in the definition of the growth curve is a determination of the maintenance feeding level (Sq). This constant is essential in utilizing the threshold-corrected hyperbolic equation and must be determined prior to its usage. Values for 8Q are obtained by performing a regression of specific growth rate values 35 against the natural log (1n) of the correSponding percent feeding levels. In order to include the values at the 0% feeding level in the regression, a value of one was added to each level of percent feed. The equation for the regression then becomes: u - a + b ln (1 + percent feed). The con- stants a and b were determined by regression. The Specific growth rate (u) was then set at zero and the equation solved for percent feed at a specifc growth rate of zero. The example data in Table 3 yield an Sq value of 0.37 percent feed. Following calculation of the Sq value, the data were used to determine the other constants, Ks and "max’ in the threshold-corrected hyperbolic curve formula. Values of these constants are calculated using a linear transformation of the original formula of Dowd and Riggs (1965) which for the threshold-corrected model becomes as follows: - -K-s+1 -s 3 $9 8 q (S q) ll 11 ll max max When values for S - S are linearl re ressed a ainst ( g) y s s n values for (S - Sq), the reciprocal of the lepe equals "max’ and this value times the intercept of the line equals Ks - Sq. Using the results of the sample experiment, sum- marized in Table 4, "max was calculated to be 0.0641 and K8 was calculated to be 5.04. 36 Table 4. Regression data and growth constants calculated for I. zillii fed trout chow at 27.5°C. q q ——T—q 0 0.37 -0.37 -0.0057 64.91 002 -0017 -000026 65.38 0.5 0.13 0.0018 72.22 0.7 0.33 0.0041 80.49 2 1.63 0.0142 114.79 3 2.63 0.0242 108.68 4 3.63 0.0290 125.17 "max ' 0.0641 Ks ' 5.04 Sq ' 0.37 The data fit the regression line of the transformed equation. 2 of 0.91. with an r This series of calculations has resulted in the deter- mination of values for the three constants of the threshold- corrected hyperbolic curve: K S and "ma . These values 3’ q x are then incorporated into the original equation. Using the data from the example, the equation would appear: - S - 0037 :1 0.0641 (————s + 4.30) The equation now is useful for obtaining the rate of growth resulting from any given ingestion level of the trout chow under similar environmental conditions. The calcula- tion of the three constants has resulted in a quantification of the complete growth range by identifying the maximum rate of growth (u the half saturation value (K8), and the max)’ intercept at zero growth, or maintenance ration (Sq). 37 Two important assumptions are made when utilizing the data to define the growth curves. First, when more than one fish are in one aquarium, each fish cannot be expected to consume equal feed, or the same percentage of its body weight per day. Physical and behavioral differences should cause some individuals to consume a greater and some a lesser share of the total feed. As a result, only the growth in the total weight of the fish per tank is utilized in defining the growth curves. This procedure insures that the weight and specific growth rate of the fish corresponds to the Specified rate of feeding. At least this is certain if all the feed was ingested, which leads to the second assumption. In order for growth data to be incorporated in the defining of the curves, all feed must have been con- sumed. Whenever feed was observed to remain overnight in an aquarium, that week's data were excluded from further calcu- lations since the observed growth of the fish cannot be related to the intended percentage body weight ingested. Occasions of incomplete feeding are indicated on the tables of growth rates for each experiment by enclosing the figures in parentheses. These data are not included in the averages listed in the far right column of the table. Non-linear regression. It has been demonstrated that a linear transformation of the threshold-corrected hyperbolic curve is useful in determining values of "ma and Ks. How- x ever, when the equation is transformed, the possibility 38 exists that some areas of the curve may obtain more power in determining the lepe of the line than others, thus dispro- portionately affecting the calculation of u and Ks, and max the definition of the curve. Therefore, another procedure, non-linear regression was utilized in analyzing the data. Non-linear regression analysis was performed using a BMDP-PAR computer program (Dixon, 1983). In this analysis, the threshold-corrected hyperbolic formula is specified by FORTRAN in a function statement with u Ks and S as q unknown constants. The experimental data from an individual max’ trial are then entered with the average n values being the dependent variable. The values of u Ks and Sq calcu- max’ lated by the linear regression technique are then entered as initial estimates for these parameters, along with upper and lower constraints to contain the search if something should fail to converge. The program then examines the data by an iterative "examine and modify" series of steps. Through a series of iterative-least squares analysis, the prOgram provides an estimate for each of the constants. In addi- tion, standard deviations of the estimates, correlations between the parameters, predicted values, and observed values of the variables are listed. Efficiency An additional calculation provides a means of examining growth relative to the individual units of feed supplied. This is accomplished by dividing the calculated values of u 39 by the corresponding substrate level (S) and plotting this value (u/s) versus S. These plots establish "efficiency" curves which demonstrate the ability of each unit of feed to stimulate growth. The data from the sample experiment are shown below: Table 5. Growth efficiency data calculated for I. zillii fed trout chow at 27.5°C. Calculated u 3 u/5 -0.0055 0 - -0.0024 0.2 -0.012 0.0017 0.5 0.0034 0.0042 0.7 0.0060 0.0076 1 0.0076 0.0166 2 0.0083 0.0231 3 0.0077 0.0280 4 0.0070 0.0319 5 0.0064 0.0350 6 0.0058 0.0376 7 0.0054 The calculated growth curve and "efficiency" curve, as well as the observed data points are plotted in Figure 3. This calculation of efficiency is similar to the more traditional calculation of feed conversion efficiency, where the change in the fishes weight is divided by the amount of feed in- gested. Direct comparisons could be made if the growth rate calculation used in the efficiency determination utilized the relative rather than the specific growth rate equation. 4O oou-(snf) AONBIOIAJE Nd- . ‘IN. 000 ‘Q o 6.0 .oem.um um 30:0 Noouu vow NNHHHN .H.uou mo>uso Nocowonmo can ouch suzouw onNooam uwuoNvoua on» one «New zusouw oo>uomno .m unawwm >40 mum am... .5533 >000 hzmomma n v m N _ o .06. o t . .od 0 15.323 o_m.owam . iuod N .mod N >ozm_o..._mm 30.0 . . 00.0 0.. NT) 3.1th HlMOHO OHIOSdS RESULTS The basic data obtained during the various experiments are presented in appendices. Appendix A contains the weekly length and weight measurements for individual fish and the average fish weight in each aquarium. Appendix B consists of the specific growth rates calculated from the weight data and also indicates which data are excluded from consideration due to incomplete feeding. Appendix C con- tains the results of water quality determinations performed during the experiments. Appendix D contains the results of protein and energy analyses performed on all the feeds utilized in the experiments as well as manufacturers' anal- yses of the feeds. Because these data are provided in appendices, this chapter will be a presentation of the calculated constants and equations which describe the growth curves for each experiment. These data are grouped into four sections: temperature, size, feed type, and Species comparison. Temperature - Growth The first experiment involved the determination of the growth of Tilapia zillii fed trout chow at 27.5°C. Using the procedures described in the Materials and Methods chapter, the average specific growth rate at each feeding 41 42 level was calculated. These data were then used to calcu- late values for the constants u max’ Ks and Sq’ using both the linear and non-linear regression techniques. Table 8 contains the original observed growth data, the constants calculated by both techniques, the predicted growth rates using each data analysis technique, and the efficiencies calculated for each feeding level using the linear regres- sion constants. In addition, the table contains the co- efficient of determination (r2) for the linear regression and the standard deviations for the predicted data points calculated by the non-linear regression analysis. Following this experiment, identical feeding trials were performed with Tilapia zillii fed trout chow, except that water temperatures were maintained as follows: 21.5°C, 25°C, 27.5°C, 30°C, 35°C. Data for each of these experi- ments were similarly calculated and are presented in Tables 6-11. A summary of the constants calculated by both regression techniques, for each temperature is contained in Table 12. Inspection of Table 12 reveals that the maximum growth rate occurred at 27.5°C. Therefore, this temper- ature was used in subsequent experiments. The constants, u max’ Ks and Sq, calculated by linear regression, were utilized to construct growth curves representing the predicted growth throughout the feeding range. The growth curves for each of the five temperatures are included in Figure 4. 43 mnm.o omooc.o meo.o Homo.o ammo.o a nmm.o omooo.o mNNo.o w-o.o mNNo.o m non.o amooo.o wmao.o mmao.o mmao.o N oom.o mmooo.o wmoo.o mmoo.o mmoo.o N 00H.Am\av cowmmouwou mucoumeoo mucoumcoo muou nusouw use can xocofioNuwo umoawauaoc scammmuwou :onmouwou oNuNoomm an N03 wouoNvoua coNNmN>oo .mum moonwauco: umoSNN oo>uomno >oon N susouw vauowvoum u m~.~ ow.o o~¢o.o cowmmouwou umocaauco: mo.o mm.m mm.o anao.o :owmmouwou NoocNH NNN Ne ex em Nee: :ONmmouwmu Noocfia .oom.N~ No 3020 unouu mom Aw m.m uchOS owouo>ov NNNNNN .M you mOSUNSLOON cowmmouwou Noocfiaioo: van ucmcwa nuon >2 oouoaooHoo dump oumu :NSONM camwooam voNoNvoum mam mauou susouw ONMNooam om>uomno .o magma 44 mNm.o mo~oo.o mwNo.o HwNo.o HwNo.o m mao.c ooooo.o woNo.o NONo.o CNNo.o m ooN.o mnooo.o omNo.o Nmao.o omNo.o N o-.o m~coc.o Nooo.c NNoo.o mooo.o N n mNNoo.o oooo.o: omoo.0u oooo.On o ooN.Am\:v conmouwoN mucmumcoo museumcoo much zuzouw Noe vow Nocowonwo Noocwanco: scammouwou :onmouwmu ONMNoomm a: No: ovuofivoua cowumw>ov .vum umocfiaucoc NoocNN vo>umwno Noon N nuaomw emuoNvoum u um.¢ N¢.o Namo.o conmmuwou unocfiaacoc No.0 oN.e NN.o NNNo.o eoNeeeewee NeeeNN Nam Nu m2 cm xoa: scammwuwou Hammad .ooo.mN no 3030 uoouu pom Aw o.¢ unwfloz omnuo>ov NNNHNN .H you moavwcsoou :onmouwou Noosfianco: one NooaNN nuon >5 ovuoasoNoo Sumo ouou zusouw oNuNomam emuoNvon van «much nuzouw onNomam oo>uomno .N mange ooN.o wmfioo.o NmNo.o owNo.o omNo.o q oNN.o mmooo.o NNNo.o NMNo.o NqNo.o m omw.o HNoo.o oqao.o ooHc.o Ncao.o N ooo.o mmooo.o mNoo.o Naoo.o deco.o N.o ocm.o whooo.o mooo.o NHoo.o mHoo.o m.o 0N.Nu maooo.o mNoo.ou «Noo.0u oNoo.0u N.o u oNHoo.o omoo.on mmoo.ou smoo.ou o ooH.Am\:v cowmmouwou museumcoo muscumcoo mama nusouw Nov mom Noaofioammo Noocwauooa conmouwmu :onmouwmu oNuNooam us No3 vauofivmum cowuow>oo .vum Nmmcfiancoc NMOSNN mm>uwmno Noon N nusouw emuofiowum n N.MN m¢.o moa~.o :onmouwou Nooofiauco: NN.o eo.N NN.o Neeo.o eoNeeeeeee NeeeNN Nam Nu mg um xna: :onmouwou Noocwa .oom.NN No 3050 uaouu vow Aw m.oH unwfios owouo>mv NNNHNN .H you moSVNcsooN :onmouwou umocwauco: moo NSOSNH cuon an oeuoasoHoo Sumo mama nusouw owmwoomm vmuofiooua vow moumu nuzouw onwoomm oo>nomno .m mqomfi 46 MNm.o oNNoo.o NeNo.o aMNo.o MMNo.o m oNn.o NoNoo.o mNNo.o Nuao.o oaao.o m mco.o NNNoo.o a-o.o NNHo.o mNHo.o N omm.o NNNoo.o mmoo.o mmoo.o 0400.0 H u mNNoo.o mnoo.ou Naoo.9u mmoo.ou o ooH.Am\=v :onmouwou mucoumcoo muomumcoo much nuaouw Nov vow NOSONonmo Nomcwauco: conmouwou conmouwou mewooam us N03 emuowmoum :oNuoN>wv .vum moonwaucoc umchN vo>uomno Noon N Luzoum vouowvoum I oN.¢ o¢.o Hueo.o :onmouwou Homewauco: mm.o 0N.m Nc.o mwoo.o :onmmuwou uoocfia NNN Ne ex em Nee: conmouwou Noocfia .ooom No 3050 anon» pom Aw w.o unwwmz owmuo>ov NHHHNN .H you mosvwcnoou :onmmuwou Newsflauco: moo Noocwa soon >9 voNoNSONSO ouoo ouou nusouw onwooam vouoNvoua moo mmuou suaouw onNooam vo>ummno .m ONQSH 47 a NaNoo.o NmNo.o mmNo.o NaNo.o m a oNNoo.o owao.o omao.o ¢~N0.o m u NmNoo.o mmNo.o NmNo.o oNHo.o N I meaoo.o omoo.o omoo.o acoo.o N u mONoo.o mNoo.0u meoo.0u o~oo.o: o OOH.Am\av scammouwou mucmumcoo muscumcoo wumu Luzouw New mow Nocowonmo umocwauao: cowmmouwou newmmouwou ofimfioomm on N03 oeuowmoua :oNNmN>ov .Uum uwmcwauco: NoocNH oo>uomno Noon N susoum emuowooum a NN.¢ om.o moao.o :onmouwou Noocflauco: NN.o NN.e Ne.o NNeo.o eoNeeeeNee eeeeNN Nww N m! on x65: N coammouwou umocwa .0 on an 3050 anon» vow Aw m.N unwfioa owmuo>mv «NHHNN .9 you mosmwcsoou :onmouwou moonwauao: van umocfia suon an vouoaooNoo dado mama nuaouw onwoomm vauoNvoua moo mmumu nusouw onNooam vo>uomno .oN manna 48 wmm.c woNooo.o mmNo.o w-o.o mwao.o m omq.o mNaooo.o wNNo.o mNHo.o wNNo.o m mm¢.o wwqooo.o mmoo.o Nooo.o Hwoo.o N on.o wmcooo.o Nmoo.o smoo.o oaoo.o N u NmNooo.o oaoo.01 wqoo.0n Hmoo.Ou o ooH.Am\=v conmoumou mucmumcoo mucmumcoo muou nuaouw Nov com NOSONonmo NoocNNucoc conmoume scammouwou onwomam a: flea oouowvoua coNunN>om .vum unoowanco: NochN Um>uomno Noon N nuaouw vmuowvoum u cq.m mm.o mwmo.o :onmouwmu Noocfianao: Ne.o Nm.e om.o NeNo.o eoNeeeeNee eeeeNN ufim Nu my um xma: :onmouwmu NMOSNH . mm No 30:0 Noouu mom Aw m.n unwflos mwoum>mv NNHHNN .H now moswwcnoou conmmuwmu moonwanco: can Nooawa zuon an vouoasono camp wuou nusouw ONMNooam vauoNomua moo mean» suzouw owwwoomm oo>uomno .NN oNan 49 ¢¢.m mm.¢ mm.o om.o mwmo.o Nemo.o o.mm NN.¢ Nm.¢ om.o Na.o wwao.o quo.o o.om om.a 0N.m m¢.o N¢.o quo.o omao.o 0.0m 0N.mN eo.m o¢.o mm.o moea.o Nqoo.o m.NN mm.q om.a N¢.o .wm.o quo.o mNmo.o o.mN NN.N MN.m ow.o mm.o o~eo.o mnao.o m.NN NoocNNnco: NoocNN NoocNHnooc NoocNH umocfianco: Nomcwa 0o mM am xoa: .moNSNoNomeoNlo>Nu No 3050 Noouu pom mw m.o unwwos owouo>ov NNNHNN .8 you moavwcnoou :onmou mu Noocwaucoc poo unocwa noon an vouoasoaoo mucoumcoo nusouw mo Nuoaaam .NN oNan 50 .moufiWouomeou o>Nw no 3000 usouu vow Aw m.o uswfioa owmum>ov NNHHNN .9 you momco0mou mama nuaouw onwoomm vauoNomum m ><0 mun. 0m... b.1053 >000 Pzwomma n .v d 1 on ._N .a onzwwm m N . o \ ca . u 2N \\ m o H. \ . o , . nu i no 0 . M .86 m . H on 4 W. . 3 RN . ) 0N .No.o.n 51 Size - Growth The next series of experiments was designed to investi- gate the effects of fish size (weight) on growth. Four feeding trials were conducted, each utilizing different average weight Tilapia zillii fed trout chow at 27.5°C. The previous feeding trial using Tilapia zillii at 27.5°C util- ized fish averaging 10.0 g, and this was included as one of the four experiments. The smallest size group averaged 0.0140 g. Special weighing procedures for this size class were noted previously in the Materials and Methods chapter. The two remaining Size classes had average weights of 1.76 and 38.1 g. Trout chow was provided with the particle size ranges adjusted to the fish's weight as listed in Table 13. Table 13. Trout chow particle size corresponding to each size class of I. zillii. Average fish weight Trout Chow Size Range (grams) (mm) 0.0140 0.297 - 0.420 1.76 0.420 - 1.0 10.0 1.0 - 2.0 38.1 2.0 - 4.0 The three constants defining the growth curves ("max’ Sq, and KS) were again calculated by the two regression techniques. These data, observed and predicted growths and efficiencies, are contained in Tables 14-16. The values for the constants used to determine predicted Specific growth rates are shown in Table 17. The constants determined by 52 linear regression were utilized to construct the predicted growth curves for each size class shown in Figure 5. 53 Nw.0 Hm00.0 0000.0 nN00.0 Nm00.0 0H H¢.H Nm00.0 HHN0.0 m0~0.0 NON0.0 m H0.H 0000.0 H000.0 H000.0 0000.0 0 00.H 0000.0 NMN0.0 NHN0.0 MMN0.0 N m0.mu wm00.0 0000.0: m0m0.0u H000.0u H 00H.Am\:v :onmmuwou mucoumcoo muamumcoo muou nuzouw Nov vow NocmHonmm NoocHHuooc conmouwou conmmuwwu OHMHooam as Ho: unuoHvoum :oHumH>ov .vum NomcHHucoc NoosHH vo>uomno >000 N nuzoum vauoHvoum u 0N.m 0¢.H mmoH.0 :onmmuwmu NoocHHuco: Ne.o NN.N NN.N NNoN.o eeNeeeeNee eeeeNN uHm Nu mx cm xma: :onmouwwu NeocHH .uom.NN 06 30:0 Noon» mow HHHHHN .H.Auanos owouo>mv w 0¢H0.0 you mOSUHcsoou :onmouwou NoocHHuco: 0cm NoocHH £000 >0 voumHooHoo ouov moon carouw OHMHooam vouovaum moo ocuou nusouw OHHHooam vo>umm00 .eH OHQQH 54 00N.0 «0000.0 0000.0 «000.0 NN00.0 N NN0.0 0NN00.0 H000.0 H000.0 0000.0 0 N00.H 0NN00.0 HNN0.0 0N00.0 00N0.0 0 0N0.H NON00.0 0NH0.0 0HN0.0 0NH0.0 N 00N.0 00000.0 0No0.0 0N00.0 0000.0 H 00H.Am\=v :onmouwmu museumcoo museumcoo oumu nuaouw >w0\0om NocoHonmo uwmcHHuoo: conmmuwou SOHmmouwou onHowmm ustms mauovaua :oHuwH>oU .0um NomaHHucoc unocHH vo>uomno 0000 N cuzouw wouoHvoum n N0.0N 00.0 000N.0 :onmwuwou unocHHucoc HN.0 00.0 00.0 NON0.0 :onmouwou unocHH uHm N ax 00 xMBNN N :onmoume ummcHH .0e0.NN an 3000 Noon» vow HHHHHN .H.Au00Ho3 o0muo>mv 0 0N.H you moscHasooN conmouwoN umocHHuco: vow NomaHH 0000 an voumHsoHoo wumv oumu nusouw onHooam vauoHvoua 0cm wagon nusouw owuHooam 0o>uom00 .0H oHan 55 000.0 N0000.0 00H0.0 00H0.0 00H0.0 0 000.0 00000.0 N000.0 0000.0 N000.0 N 0N0.0 00000.0 0000.0 0000.0 N000.0 H : H0000.0 N000.0: 0000.0: 0000.0: 0 00H.Am\:v conmmuwmu 000000000 000000000 000p 0030mm >00 000 >ocwHonmm N000HH::oc 00H000N000 conmouwou oHuHommm 00 H03 0000H00u0 coHu0H>00 .000 N000HH:000 N000HH 00>uom0o >000 N 003000 0000H00Nm : 00.0 H0.0 0000.0 :OHmmmuwmu N000HH:coa N0.0 00.0 00.0 00N0.0 conmmuwou N000HH 0H0 NM 00 00 x05: SOHmmmuwou N000HH .oo0.NN 00 3000 usouu 000 HHHHHN .H.A000H03 00000>0v 0 H.00 now mmschnomu conmmuwou N00=HH::o: 0:0 N000HH 00o0 >0 0000HaoH0o 0000 0009 003000 OHMHooam 0000H00u0 000 0000» 0030mm onHooam 00>uom00 .0H 0H00H 56 oN.e 0N.N NN.o Ne.o NNNo.o NNNo.o N N.NN N.NN eo.N Ne.o NN.o NoeN.o Neeo.o N o.oN NN.NN NN.N eN.o NN.o NNNN.o NNNo.o N NN.N NN.N NN.N Ne.N NN.N NNoN.o NNoN.e N eNo.o N000HH:000 N000HH N000HH::00 N000HH N000HH::0: N000HH 0 0 x08 0 N e .0e0.NN 00 3000 0:000 000 HHHHHN .9 mo 00000H0 0000 0000 you 0000000000 00H000000u N000HH:=0: 0:0 N000HH 0000 >0 0000H00H0o 000000000 003000 mo >N0aesm .NH 0H00H 57 .Uo0.NN 00 3000 00000 000 AN N.NN use N o.oN .N NN.N .N eNo.o eeeNNes eeeeeeev NNNNNN .e No 00000H0 0NH0 0000 000 00000000 0000 003000 0H0H0000 000000000 >40 mum 0mm 0.10m; >000 hzmommm e e e N N _ o N \\\m 14.00 \ ed. .2... 0.0.0 . 0.0: .00 No.0 0 0.0 00.0 00.0 0 0.0 .0 0000Hm HT) 31:13 11174089 OIJIOBdS 58 Feed Type - Growth In order to study the effect of feed type on growth, two additional feeding trials were performed with Tilapia zillii at 27.5°C. Pelleted Elodea canadensis and pelleted Spirulina were utilized for comparison against the trout chow study already performed. As previously described, specific growth rates were calculated, regression analyses performed and constants determined. The data from these trials are listed in Tables 18-19 and summarized in Table 20. The constants calculated by linear regression were utilized to construct the predicted growth curves for each feed shown in Figure 6. All the feeds were analyzed for energy content (calor- ies/g), total nitrogen, total carbon, and total hydrogen. These data are contained in Appendix D. The total nitr0gen content was multiplied by 6.25 to convert to protein (Good- hard and Shils, 1978). In addition to the three feeds utilized in this feed comparison, analyses also were per- formed on live, suspended Spirulina sp. algae. These data are also in Appendix D. 59 000.0 00000.0 0000.0 0000.0 0000.0 0 000.0 00000.0 0000.0 0000.0 0000.0 0 000.0 00000.0 0000.0 0000.0 0000.0 N 000.0 00000.0 0000.0 0000.0 0000.0 0 a 00000.0 0000.0: 0000.0: 0000.0: 0 000.A0\:v 0000000000 000000000 000000000 0000 003000 000 000 0000000000 0000001000 0000000000 0000000000 00000000 00 003 000000000 000000>00 .000 000000u000 000000 00>0000o 0000 N 003000 000000000 u 0.00 00.0 0000.0 0000000000 000000u000 00.0 00.0 00.0 0000.0 0000000000 000000 000 0 0M 00 000: N 0000000000 000000 .oom.n~ 00 00000 000000000 000 A0 0.0 000003 00000>0v 000000 .M 000 0000000000 0000000000 000000u0o0 000 000000 0000 00 0000000000 0000 0000 003000 00000000 000000000 000 00000 003000 00000000 00300000 .00 00009 60 000.0 00000.0 0000.0 0000.0 0000.0 0 000.0 00000.0 0000.0 0000.0 0000.0 0 000.0 00000.0 0000.0 0000.0 0000.0 0 000.0 00000.0 0000.0 0000.0 0000.0 0 : 00000.0 0000.0: N000.0: 0000.0: 0 000.A0\:v 0000000000 000000000 000000000 0000 003000 >00\000 0000000000 000000:000 0000000000 0000000000 00000000 000003 000000000 000000>00 .000 000000:0o0 000000 00>00000 0000 N 003000 000000000 : 00.0 00.0 0000.0 0000000000 000000:000 00.o o~.m 00.o 0000.o :o0mmmuwmu 000:00 000 0 00 00 00a: 0 0000000000 000000 .oomunm 00 0000000000 000000 00000000 000 A0 0.00 000003 00000300 000000 .9 000 0000000000 0000000000 000000:000 000 000000 0000 00 0000000000 0000 0000 003000 00000000 000000000 000 00000 003000 00000000 00>00000 .00 00009 61 o.om 00.0 me.o 00.0 0m~0.o Mqu.o .00 000000000 00.0 00.0 00.0 00.0 0000.0 0000.0 00000m0000 000000 0.00 00.0 00.0 00.0 0000.0 0000.0 3000 00009 000000-000 000000 000000-000 000000 000000-000 000000 0M 00 008: I. .000.0~ 00 00000 00000 000 000000 .9 000 0000000000 0000000000 000000-000 000 000000 0000 00 0000000000 000000000 003000 00 0000800 .00 00000 62 .I. a .3ono usouu . .5“ gm mamcmvmcmo m wouwaawa van «awasuwmm kumaam . Umewwaafiu .M.uom mucoammu much nuaouw awwwomam vmuofivmum o wuawwm >000 hzuommm n v m N _ o 1 1 I c I th . _o.o j . ¢Moo._m _ 0.0 (2.4:mEm .1 N 0.0 3010 knock n 0.0 w 0.0 n 0.0 (d‘) any HLMOHS OIleBdS 63 §pecies Comparison After examining the ability of the model to quantify the effects of temperature, size and feed type on the growth of Tilapia zillii, experiments were designed to identify differences in growth experienced by different Species under identical environmental conditions. Therefore, feeding trials were performed using Sarotherodon niloticus fed trout chow and pelleted Spirulina at 27.5°C. The average size of the fish was 4.1 g for the trout chow trial and 3.9 g for the pelleted Spirulina trial. As previously described, the specific growth rates were calculated, regression analyses and other calculations performed. The findings from the trials are provided in Tables 21-22. Comparisons of the data calculated for both species when fed both trout chow and Spirulina are shown in Table 23. Growth curves con- structed from constants calculated by linear regression are shown in Figure 7. 64 me.0 ~0H00.0 0m¢0.0 0¢¢0.0 H¢¢0.0 m m00.0 0~H00.0 mn~0.0 HNN0.0 00N0.0 m wa.0 NHH00.0 0m~0.0 om~0.0 moH0.0 N Hm~.0 00000.0 «500.0 mm00.0 0000.0 a u HOH00.0 ~q00.0n 0000.0: ##00.0n 0 00H.Am\:v cowmmouwou muamumcoo mucoumooo moon nusouw xmv vow zocowowwmm ummcfiauoo: cofimmouwou coflmmouwmu awmwooam on was wouowomua sowumw>mv .vum pomoflaucoa nooowa vo>ummoo xvon N guzouw vmuoflvoum u 0.0H nm.0 NmHN.0 :owmmmuwou moonwauco: Hm.o ~.- sm.o Nmm~.o acummmummu Havana saw me am gas: u N cowmmouwmu nomcwa .oom.mN am 3020 uaouu vow Aw H.¢ unwaos mwmum>mv maowuoaw: .m now mmsvficnomu cowmmouwmu moonwaucoc tam nomcfia coon >0 cocoaooqoo ouov moon nuzouw oflmwommm vmuowvoum vow moody nu3ouw owwwommm vm>uomno .HN magma 65 ~0N.0 N>000.0 H000.0 «000.0 0000.0 0 000.0 00000.0 00N0.0 HON0.0 00N0.0 m an.0 00000.0 00H0.0 00H0.0 00H0.0 N 000.0 ~0000.0 0000.0 0000.0 0000.0 a : 00000.0 0000.0: 0000.0: 0000.0: 0 00H.Am\:v :oflmmmuwou mucmumooo mucoumcoo muou nuzouw >ov vow >ocowo«wwo moonwa::oc coflmmmuwmu cowmmmuwou ofiwfiommm 00 “m3 vmuowvmum :o00Mw>mv .vum Homewa:coc poocaa vm>uom0o >000 N susoum Umuowvmum : 0.0m 00.0 ¢¢~N.0 aowmmouwou umocfia::oc 05.0 0.0a mq.0 ~00H.0 oowmmouwou pooawa any u me am gas: N cowmmmuwmu pomcwa .oo0.n~ 0o mafiasuwmm cmuoaaom vow Aw 0.0 unwfloz owouo>ov maowuoafic 4w you moacwa0oou scammouwou moonwa::oc woo noocfia 00o0 >0 vmumasoaoo ouov ouou 0030mm oMMfiooam voooavmua vow moumu 003ouw ofiufioomw vo>umm00 .NN oH0oH 66 0.00 0.00 00.0 00.0 0050.0 0000.0 .00 000000000 0.00 5.00 00.0 00.0 0000.0 0000.0 3000 00008 000000000 000000000000 0.00 00.5 00.0 00.0 0050.0 0000.0 .00 000000000 5.00 00.0 00.0 50.0 0000.0 0000.0 3000 00008 0.00000 0000009 000000:0oc 000000 u00000:0o: 000000 000000:000 000000 02 00 x08: E0 W0 000000 0000500 00.0 05000000: .0 000 000000 .8 now 0000000000 0000000000 u00:00:0oc 000 000000 0000 >0 0000000000 000000000 003000 00 >00eaam .00 00009 67 .E um 00000003 00...”.000...a 0:0 3000 009.5 000 2600000: .m 000 000000 .0 000 00000000 0000 003000 00000000 000000000 >40 awn. 0w... Pro-u; >000 .__.zmommm o v n N 0 I I d I I T‘D 05.3.00 :3: .N F :23 523...: .0. 05.3.0 m In 3.00:: .0 32.0 38. I 2.0.3:: .0 l J l J .5 003000 00 3.0. u m J o m 3.0 w 0 No.0 1. un no.0 w I. 36 3 00.0 .n 00.0 ( DISCUSSION Discussion concerning the usefulness of the proposed model to quantitatively relate fish growth to environmental resources will be divided into the following four areas: 1) variability, 2) growth-feed relationships, 3) growth-size relationships and 4) growth-temperature relationships. VARIABILITY An examination of the data indicates that great vari- ability is present in the growth rates among fish at the same feeding level. Variability is present both among fish in the same aquarium and for the individual fish from week to week. While the variability is demonstrated in all the feeding trials, the experiment with Tilapia zillii fed trout chow at 27.5°C was chosen as an example. More specifically, an examination of the specific growth rates at the 2% feeding level (Appendix 8-3) reveals a range from -0.0088/day to 0.0315/day during the first week for the individual fish. Variability also was present for each individual fish from week to week. Fish #2 experienced a range in specific growth rate from 0.03ll/day the first week to -0.0034/day the fourth week. Therefore, growth vari- ability, both for individuals with time and between individ- uals, is an important consideration. 68 69 This variability is the result of several complex factors which interact to determine the ingestion rate, the processing efficiency, assimilation rate and eventually the growth rate. Included in these factors are the genetic makeup of the fish, environmental factors, behavior patterns and sex of the individual. All of these factors interact to determine the rates and efficiencies of the processes which determine the growth rate. Because of the complexities of these interactions, a detailed study of the variables has not been attempted. Rather, the variability itself has been examined to determine how to process and interpret the data. Some general observations on the sources of variability are included. One factor with a large potential of affecting vari- ability is the ingestion rate. This also provides a good opportunity to demonstrate the attempt to examine the vari- ability itself without attributing a causative mechanism to it. The best example of the variability of ingestion rate occurred during the feeding trial of Sarotherodon niloticus on Spirulina. Over the course of the five-week study, fish #0 at 1% feed and fish #2 at 3% feed each were observed to ingest little, if any, feed. At the end of the five weeks, the fish had lost 22.6 and 10.7% of their initial body weight respectively, yet refused to ingest the pelleted Spirulina algae which promoted excellent growth in the other fish sharing the aquarium. After the study was completed, 70 these fish were isolated and fed trout chow. Each fish immediately packed its mouth full to the point where its jaw could not close, and even then attempted to pick up addi- tional pieces which could not fit. The explanation of such behavior is left to other researchers. A less bizzare example is encountered in the afore- mentioned feeding trial of Tilapia zillii fed 2% trout chow at 27.5°C. Judging only from the resultant specific growth rates, it appears that the individual fish experienced dif- ferential ingestion rates from week to week. If fish #2 had ingested the same percentage feed from week to week, its growth rate would not have so dramatically decreased each week as was the case. Fish #1 experienced the opposite growth pattern, having the lowest specific growth rate initially (-0.0088/day) and the highest the fourth week (0.0210/day). While it is recognized that other variables have the potential of effecting differential growth rates from week to week, I assume that changing ingestion rates is a major factor. In order to clarify the growth variability associated with differential ingestion, an experiment was designed having three aquaria all containing four Sarotherodon niloticus under similar conditions of feed (3% trout chow), temperature (27.5°C) and size (8.8-14.0 g). The results of the two-week study are shown in Table 24. 71 Table 24. Specific growth rates of Sarotherodon niloticus fed 3% trout chow per day -- four fish per tank. Tank 1 Tank 2 Tank 3 Week 1 Fish 1 0.0296 0.0079 0.0186 Fish 2 0.0164 0.0297 0.0351 Fish 3 0.0307 0.0236 0.0184 Fish 4 0.0319 0.0331 0.0208 Average 0.0273 0.0243 0.0241 all 12 data points 0.024710.0043 average of each tank 0.025210.0030 Week 2 Fish 1 0.0159 0.0132 0.0135 Fish 2 0.0225 0.0237 0.0201 Fish 3 0.0165 0.0203 0.0194 Fish 4 0.0121 0.0118 0.0209 Average 0.0165 0.0181 0.0186 all 12 data points 0.0175:0.0022 average of each tank 0.0177i0.0018 Two-Week Averages Fish 1 0.0223 0.0105 0.0161 Fish 2 0.0195 0.0267 0.0276 Fish 3 0.0236 0.0219 0.0189 Fish 4 0.0220 0.0225 0.0209 Average 0.0219 0.0212 0.0214 all 24 data points 0.021510.0006 average of each week 0.0211t0.0038 two-week average 0.021110.0009 72 Over the two-week period, there were several reversals in the relative magnitude of growth within each aquarium. In the first two aquaria, the fish with the highest specific growth rate the first week experienced the lowest specific growth rate the second week. In aquarium one, the fish having the lowest first week specific growth rate had the highest second week specific growth rate. Overall, the specific growth rate ranged from 0.0079/day to 0.0351/day. The data in Table 24 were used in calculating means and 95% confidence intervals for the means. During the first week, if all 12 data points are used in calculating the mean, an average specific growth rate of 0.024710.0043/day is obtained. If only the averages for each aquarium are used, the mean specific growth rate is 0.025210.0030/day. For the second week, if calculations are based on all 12 data points or just the aquaria averages, the mean specific growth rates are 0.017510.0022/day and 0.017710.0018/day respectively. Over the two-week period, if all data points are utilized, the mean specific growth rate is 0.0215i 0.0006/day; if the six weekly aquaria averages are used, the mean specific growth rate is calculated as 0.02111 0.0038/day; and if only the three two-week averages for each aquarium are utilized, the calculated mean is 0.0211: 0.0009/day. It appears that essentially identical means are calculated if either the individual data are utilized or if aquaria averages are used in the calculations. The 73 variability, expressed as standard deviation or confidence interval, is affected by the calculation technique, but the mean is not. Therefore, throughout this study, the mean was chosen as the most representative expression of the data. The reproducibility between the three aquaria also was monitored in the study. Throughout the two-week study, aquaria two and three contained fish experiencing specific growth rates within 3% of each other. While aquarium one had a slightly higher average specific growth rate the first week, it was slightly lower the second week. The two-week averages demonstrate excellent reproducibility, being within 3.2% of each other. Another study was conducted to investigate growth variability by placing one, two, three, and four fish in separate aquaria, each fed 3% trout chow under identical conditions. The results of the two-week study are shown in Table 25. Once again in this experiment, variability in individual specific growth rates was significant, ranging from 0.0149/day to 0.0339/day. However, whether the overall growth means were calculated from individual or from aquaria average values, the same two-week values were obtained. One trend in the mean specific growth rate data, although not different at a 75% significance level (Stu- dent's t test), might deserve further attention. As the number of fish per aquarium increased, the growth rate steadily decreased. These differences might reflect an Table 25. 74 Observed specific growth rates, avera es, and 95% confidence intervals (two-tailed t test for variability study. Tank 1 Tank 2 Tank 3 Tank 4 Week 1 Fish 1 0.0271 0.0339 0.0257 0.0261 Fish 2 - 0.0215 0.0206 0.0272 Fish 3 - - 0.0310 0.0223 Fish 4 - - - 0.0195 Average 0.0271 0.0277 0.0258 0.0238 $0.0391 $0.0088 10.0042 Week 2 Fish 1 0.0228 0.0171 0.0149 0.0151 Fish 2 - 0.0161 0.0222 0.0151 Fish 3 - - 0.0149 0.0163 Fish 4 - - - 0.0237 Average 0.0228 0.0166 0.0173 0.0175 10.00316 10.0071 t0.00487 Two-Week Averages Fish 1 0.0249 0.0255 0.0203 0.0206 Fish 2 - 0.0188 0.0214 0.0211 Fish 3 - - 0.0230 0.0193 Fish 4 - - - 0.0216 Average 0.0249 0.0222 0.0215 0.0207 10.0063 10.0087 10.0050 10.0032 Mean calculated from individuals 0.0249 0.0222 0.0215 0.0207 Mean calculated from weekly average 0.0249 0.0222 0.0215 0.0207 i0.0017 10.0045 10.0035 10.0026 75 increased energy expenditure caused by increased activity and aggression as the fish become more crowded. A trend which was clearly expressed in the two vari- ability feeding trials (Tables 24-25) is that the growth rate tends to decrease during the second week of the ex- periment. I hypothesize that the growth rate may decrease slightly when fish are experiencing rapid growth. However, the magnitude of decrease seen in all seven variability study aquaria and in 70% of the other feeding trials is too large to attribute to such a mechanism. A possible explanation of the trend may involve the conditioning of the fish prior to experimentation. Before each feeding trial, the fish were held at the experimental temperature they were to experience, and they were given the same feed for a period of at least three weeks. In each case, the amount of feed was approximated to be between maintenance ration and 1% body weight/day. Bilton and Robins (1973) conducted feeding trials investigating the effects of starvation and subsequent feeding on sockeye salmon fry. They reported that following a one to four week starvation period, fish fed ad libitum experienced acceler- ated growth. The authors suggested the fish were able to utilize feed more efficiently following starvation. Al- though the fish used in these studies were not starved prior to the feeding trials, a majority (and in the case of the variability studies, all) of the fish were exposed to higher 76 feeding levels during the experiment. Perhaps the tem- porarily increased efficiency may be attributed to body stores of important growth components which are lacking other depleted, essential reactants. When the missing reactants are supplied during increased feeding levels, the growth is accelerated until the stores are depleted. Because of this important consideration, the maximum possible data were involved in the calculation of growth constants. Only data associated with incomplete feed utilization were excluded. The final study investigating variability in growth involved the rearing of three broods of Sarotherodon niloticus. Each brood was fed an excess of trout chow from the time the fry were first observed free-swimming until 30 days from hatch. (The first brood was weighed 27 days from hatch.) At this time, each fish was individually weighed and the specific growth rate calculated assuming a uniform initial weight at hatch of 0.002 g. The following table lists the growth data for each brood. Table 26. Ranges in weight and specific growth rate obtained from three broods of §. niloticus. Average # weight specific growth specific growth Brood fish range (g) rate range rate (S.D) I 98 0.0617-0.2731 0.1270-0.1821 O.1532t0.0107 II 60 0.0606-0.4805 0.1137-0.1827 0.1513:0.0112 III 105 0.0404-0.7333 0.1002-0.l968 0.1505t0.0260 77 Once again, these data reflect the same pattern of tremendous variability amongst individuals, yet consistency of the mean. When the broods are compared on the basis of frequency for different size ranges (Figure 8), broods I and II have many similarities. Despite the distinctly different frequency distribution of brood III, the mean specific growth rate is still essentially unchanged from that re- corded for broods I and II. The linear and non-linear regression techniques for determining the constants of 0 KS and Sq provided max’ similar results on some feeding trials and very divergent results on others. For the temperature study, the two sets of constants were very similar except for the 27.5°C trial (Table 12). For this experiment, all the non-linear re- gression constants were extremely high and lacked biological reality. An examination of the data calculated for the size experiments (Table 17) reveals good agreement between the regression techniques for the 0.014 g fish and 38.1 g fish, but not the 1.76 g size. Once again, the non-linear tech- nique produced unrealistically high values for u and Ks. max Concerning both the feed type and species comparison exper- iments (Tables 20 and 23), very poor agreement was obtained for all the pelleted Spirulina studies. In both cases, the non-linear regression values for "m and Ks were compar- ax atively high. PERCENT OF BROOD‘ 78 24F ' ' o g, I“ . ' '1 O O. I ZIJr . ' I ' O '0. I o "o. L O ' 0‘ I I IS» I I 0' 0' I ,. I ‘ 0' 0' l2 ° ‘ b I I\ 0 0| | I IN . ' '. .I O I ’I o I 8’ I 'I I I I b ’0 .‘ I ’I. .‘ I . ‘ 4'“ I . ‘ Io . o I 0 o D I. O I . .‘s Io c’. \\ I o.no o.|2 Jo. ‘ (Sr—W . SPECIFIC GROWTH RATE (U) Figure 8. Frequency distribution for the weights of individuals from three broods of §, niloticus thirty days from hatch. 79 An examination of the observed and predicted specific growth rate data reveals that both the linear and non-linear regression techniques provided accurate predictions of growth rates within the experimental feeding range. Despite this similar predictive capacity, some growth constants obtained by non-linear regression were biologically unreal- istic. This paradox may be explained by examining the growth curves representing these problematic experiments (Figure 9). Each instance of divergent non-linear regres- sion constants corresponded to a growth curve exhibiting minimal curvature and relatively high specific growth rates. Apparently, when the computer program used to calculate the nOn-linear regression constants attempted to fit these data to the threshold-corrected hyperbolic equation, the lack of curvature caused the determination of unrealistically high values for "ma and KS. When closer constraints were placed x on the initial estimates for the constants, the program would simply run up to the imposed limits. While these comparatively high constant values are able to predict specific growth rates within the limited experi- mental feeding range, they are unable to continue realistic predictions corresponding to higher feeding levels. This is demonstrated in Figure 10 where the predicted growth curves from both linear and non-linear regression constants are plotted for 1.76 g I. zillii fed trout chow at 27.5°C. 80 .00000000 :o000000w00 000:00u0o: mo 0w00: wc000000000 00003000000 00ou 0cm 00=o0000 0000 :03o0w 00000000 000000000 .0 000w00 >40 mum 0mm 0.1053 >000 hzmommm . 0 ¢ 0 N _ o \ .o.o. \\ . o o......:_._0-.__._~ .0 .06 1(0 N 0.0 mNN .__EN F 00.0 6:330. 0.322.: .0 00.00 .mh; .. SEN F 00.0 (rf) alvu Hlmoas onsloads 00.0 81 00.0.00 00 30:0 00o00 000 000000 .H_w 00.0 0cm 00000000 0o00000w00 000000u0o0 000 000000 £0o0 >0 000000000 000o0000 n03o0w 000 mo 0o00000eo0 .00 000000 >40 000 00.... 0.20.03 >000 0.200000 0m 0.0 0.0 m. m. .m 00. 290000000 0402.... 290000000 0402...- 202 j 00.0 00.0 N..0 0.0 (If) 31W! HlMOIIG OIJIOBdS 82 Because the predicted growth rates are extended over a broad feeding range, the divergence between the regression tech- niques becomes apparent. Since the non-linear regression analysis technique was effective in determining constants when sufficient curvature was present, it is hypothesized that if higher feeding levels were utilized, more curvature in the observed data would be evident and the program successfully used. Due to these difficulties, only the constants determined by linear regression are utilized in the discussion of the data. GROWTH-FEED RELATIONSHIP The capacity of a feed to stimulate growth of an or- ganism is dependent on its ability to provide all the com- ponents necessary for growth to the apprOpriate reaction sites within the organism. In judging the quality of a feed, there are three basic considerations: 1) the feed must contain the necessary components in sufficient con- centration; 2) the organism must be capable of capturing and digesting the feed and transporting the components to the internal reaction sites; and 3) toxins which would hinder growth must not be present. Feed Quality gProtein-Energyl While numerous vitamins, minerals and nutrients have been recognized as essential to the growth of fish, this discussion will be limited to examining the relationship 83 between protein, energy and growth. Proteins are large, complex molecules composed of up to 20 amino acids, 10 of which have been shown to be essential to fish (NRC 1983). Ingested amino acids may be assimilated into new protein or used in repair throughout the body. Being a major component of structural units, protein may represent 65-75% of the total body weight on a dry-weight basis. Both quantity and quality are important considerations for protein supply. If insufficient dietary protein is provided, organisms are cap- able of withdrawing protein from existing tissue to maintain vital functions. Sufficient protein levels must be supplied to provide for maintenance and activity, and to provide an excess for growth. However, if more protein is supplied than can be incorporated into new tissue, the excess will be used for energy, representing an expensive energy supply. In addi- tion, the prOper balance, or quality of amino acids, must be provided. If any essential amino acids are lacking in the diet, tissue production and growth will be limited. The minimum amount of protein (percentage of feed) needed to supply adequate amino acids and produce maximum growth has been determined for several Species of fish (NRC 1983). These levels range from 31-38% for the common carp to 56% for Tilapia aurea fry. Several authors have investigated the growth promoting ability of different protein levels in iso-caloric diets. 84 Davis and Stickney (1978) reported that the growth of Tilapia aurea increased consistently as the percentage of protein was increased in series from 15 to 36%. Mazid et a1. (1979) reported similar findings using Tilapia zillii fingerlings fed variable protein levels. As protein levels were increased, up to 35% corresponding increases of growth occurred. However, when protein levels exceeded 35%, grad- ual growth retardation occurred. Dupree and Sneed (1966) also concluded that increases in protein level up to 40% promote increased growth, but that above 40%, gradual de- creases occur. The authors attributed these decreases to the increased metabolic cost of processing the excess pro- tein. Boyd and Goodyear (1971) summarized the importance of protein content by defining it as the most important aspect of the nutritional value of food. Energy requirements for protein synthesis in fish are reported to be lower than those required by warm-blooded animals due to: l) avoidance of energy expenditures to maintain a constant body temperature, 2) requiring less energy to maintain their position in an aqueous media, and 3) excretion of nitrogen waste more efficiently as ammonia rather than urea or uric acid (NRC 1983). Nevertheless, energy is still required for maintenance, activity and growth. Although warmwater fish can utilize proteins, carbo- hydrates and lipids as energy sources (Carling and Wilson 85 1976), they are not comparable in either cost or effective- ness. Lipids are an important source for energy because they are less expensive than protein, have a high energy content, and are essential in hormone function, membrane structure and vitamin absorption (NRC 1983). Sufficient non-protein energy must be supplied to "spare" the expensive protein for growth, yet an energy excess must be avoided to prevent the organism from de- creasing protein intake as a result of having satisfied all energy requirements. Because of these considerations, dietary energy requirements are often expressed as a func- tion of dietary protein level. Optimal ratios of kcal digestible energy to grams of protein were reported to range from 7.8 to 9.7 for various fish species (NRC 1983). The three pelleted feeds fed to Tilapia zillii in this study had a wide range in both protein and energy levels (Appendix D-4). The commercial trout chow was considered a high quality feed providing 40% protein, an 11.4 energy: protein ratio, and all the vitamins and minerals necessary for growth. (Energy units for all pelleted feeds were expressed in units of kcal/g total energy.) The pelleted Spirulina contained an abundance of protein at 60.5%, but a lower energyzprotein ratio of 7.3. The protein content of the trout chow supplies approximately 47% of the total feed energy, while the protein content of the Spirulina repre- sents over 73% of the total feed energy. According to the 86 content analysis of Spirulina provided by Microalgae Inter- national Sales Corp., lipids supply 13% of the total energy and carbohydrate 14%. The pelleted Elodea canadensis con- tains a much lower protein content at 21% and a lower energy: protein ratio at 3.8. The data obtained from the various feeding trials are listed in Tables 18-20. The data fit the linear transfor- mation of the threshold-corrected hyperbolic equation with a coefficient of determination (r2) of 0.91 (n - 7), 0.96 (n - 5) and 0.53 (n - 5) for trout chow, Elodea, and Spiru- lina respectively. The range in values for the constants, especially n max and Sq, indicate that growth stimulation has differed greatly for the three feeds as shown in Figure 6. To facilitate comparison of the effects of the divergent protein concentrations on the growth of the fish, the same growth values are plotted versus protein in Figure 11. Both sets of growth curves clearly demonstrate that the trout chow with an Optimal blend of protein, energy and essential components, stimulates superior growth. The pelleted Elodea is only capable of producing a maximal growth rate 18% that of trout chow. In addition, the Sq values indicate that twice as much Elodea is needed to maintain constant body weight. These data from the growth curves clearly reflect the protein and energy deficiencies present in the Elodea. The 20% protein content, the reduced total energy content, and the 3.8 energy:protein ratio of the Elodea would be 87 .00000 N00 0000000 0 00 00000000 000 0a m. mm 00 0000000000 .0 00000000 000 0000:0000 00000000 .3000 00000 000 000000 .9 00 00000000 0000 £03000 00000000 000000000 .00 000000 02.0005: >3 5.. amp. 020.03 300 020000.. 0.0 0.. ~._ 0.0 to 0 . .00-»... .9: l\ . 0 w «0800/ x o . 8.0 9 4.0233030 3 .000 m I. 30:0 505. .000 H H. 60.0 w. a. 76.0 n 88 expected to cause energy and especially protein shortages, which would limit growth. Further examination of the two figures reveals that the pelleted Spirulina was capable of producing a "max only 64% of that for trout chow. This occurred despite a higher pro- tein content at 60.5%. A significantly higher feeding level was also required for maintenance feeding with Spirulina. One possible explanation for this reduced growth stimulation is that the high protein represents an excess. The work of Dupree and Sneed (1966), referred to earlier, indicated that a protein excess may incur additional processing costs and lower growth as the excess nitrogen is eliminated. If this were a major factor, it would not be expected to become evident until relatively high feeding levels are reached. Instead, the Spirulina is shown to be a dramatically inferior feed at even 1% body weight per day (bw/d) where protein would not be in excess. Another possible explanation lies in the 7.3 energy: protein ratio of Spirulina. This low ratio, in conjunction with protein providing 73% of the total feed energy, indi- cates that protein will be required to provide a portion of the energy costs of the fish and be unavailable for growth. Therefore, an unknown percentage of protein, probably well below the 60.5%, will actually be available to stimulate growth. This limitation of the protein quality introduces additional considerations. The amino acid content of 89 Spirulina is listed in Appendix D-2. In comparing these values with the requirements listed by the National Research Council (1983), it is possible that lysine, methionine and phenylalanine all could be growth limiting in Spirulina. The combined action of limiting amino acids and protein utilization for energy supply could represent factors which would diminish the ability of the Spirulina protein to stim- ulate growth, despite the high quantity present. A comparison of the growth curves in Figures 6 and 11 also reveals that while the trout chow consistently stimu- lated superior growth, the relative positions of the Elodea and Spirulina growth curves are reversed in the two figures. For equal quantities of protein, the Elodea canadensis stim- ulated superior growth up to a feeding level of approxi- mately 1.5% bw/d-protein. Above this feeding level, the Spirulina stimulated superior growth. However, the 1.5% bw/d-protein feeding level is equivalent to about 7.5% bw/d-Elodea and only 2.3% bw/d-Spirulina. At these feeding levels, there would be dramatic differences in the retention time of feed in the gut, with the Spirulina having increased time for digestion and assimilation. The ability of the pelleted Elodea to promote better growth than the pelleted Spirulina on an equal protein basis again suggests that the quality of the Spirulina protein is lacking. The efficiency curves for each feed, shown in Figure 12, reinforce such interpretation. A comparison of the feeding levels at which 90 .oom.0N 00 00000000M0 .m.00000000 000 000000000 00000000 .3000 00000 000 000000 .9 000 00>000 003000 0000000000 .00 000000 >40 00.. 00... #090? >000 0.200000 0 n 0 n N . 0 \\\ fl . mwavu . 0 4000.00 .\ . «.0 42300.00 . v.0 / 4 00° 3000 0.0000. . 0.0 . 0.. AONSIOIAJB -oon4sxd) 91 maximum efficiency occurs reveals levels of 1.8%, 2.1% and 2.5% (bw/d) for trout chow, Elodea and Spirulina respec- tively. The comparatively high value for Spirulina suggests a deficiency in feed quality which necessitates higher feeding levels for high promotion of growth. Although the consideration to this point has centered on energy and protein, it must be recognized that the fish are integrating the total feed quality into one parameter-- growth. The discussion of feed quality was limited to these parameters because they are perhaps the two major criteria when considering feed quality. The enzyme-kinetic growth model was demonstrated to be capable of producing growth curves and constants which could be used to compare and identify the relative potential of the different feeds to stimulate growth based on protein and energy content. Additional experiments, using other feed components as variables, could similarly quantify the relative importance of these parameters in promoting growth. Eventually, such experimentation could identify Optimal feed compositions for specific species in unstressed environments. Harvest and Digestibility Two other considerations were judged critical in de- termining the ability of feed to stimulate growth: 1) the presence of toxins, and 2) the capability of the organism to capture and digest the feed and transport the components to 92 the internal reaction sites. While toxins are potentially an important factor causing decreased growth, I do not believe them to have been a factor in these trials. The trout chow is a commercial feed formulation, the powdered Spirulina blue-green algae have been approved for human consumption, and the pelleted Elodea canadensis have been successfully utilized in feeding trials involving ruminants. However, the other factors of capture, digestion and assimi- lation efficiency are believed to have considerable import. Feed capture and processing are necessarily related and have been simultaneously influenced by selection pressures. Because of these considerations, the morphological struc- tures and mechanisms associated with feed capture and pro- cessing for various cichlids are discussed. The cichlids, in general, are characterized by unique jaw structures which distinguish them from other families in the order Perciformes. These characteristic structures were described in detail by Liem (1974). The distinctive fea- tures of the jaws are as follows: 1) the joint between the skull base and the upper pharyngeal jaw is synovial, offer- ing controlled mobility; 2) the lower pharyngeal jaws are united; and 3) muscles which in other percoids are attached to the dorsal gill arch and upper pharyngeal are instead attached to the lower pharyngeal. These features accomplish three important functions. First, the united, lower pharyn- geal jaw provides for increased control and degree of jaw 93 movement. Second, the different muscle attachment allows the lower pharyngeal to exert strong biting action against the upper pharyngeal. Third, the free floating, synovial, upper pharyngeal, with its predominant forward-backward movement, working with the other features provides for efficient food mastication (Liem 1974). This unique function of the pharyngeal jaws as an efficient food processor removes the responsibility from the premaxillary and mandibular jaw, which usually function as both collectors and processors of food. As a consequence, the cichlid family is characterized by a high degree of specialization of both the pharyngeal jaws and the mandible and premaxillaries. For example, cichlid pharyngeal denti- tion was shown in scanning electron micrographs to vary from large, blunt teeth in mollusc-eating fish, to pointed, ser- rated, blade-like teeth in piscivors (Liem 1974). Bowen (1982) described the pharyngeal teeth as ranging from fine, thin hooked structures on plankton eaters to coarse, robust structures on macrOphyte consumers. This specialization and versatility has probably been an impor- tant factor which enabled the cichlids to successfully pOp- ulate numerous, diverse niches (Greenwood 1964; Greenwood 1973; Liem 1974). Several authors (Greenwood 1973; Liem 1974; Brooks 1950; Fryer and Iles 1972) have described the high level of endemism or "explosive speciation" exhibited by cichlids in the Rift Valley Lakes. This evidence of 94 selection has been attributed to the continuous spawning of the cichlids, the long history of tectonic activity in the area, and the specialized mouth structures. In summary, the cichlids as a family are characterized by specialized struc- tures which have enabled different species to compete effi- ciently for a variety of feed resources. As a result, it is reasonable to expect that the I. zillii and §. niloticus used in this study possess spec- ialized structures which would enable them to efficiently utilize particular resources. This is especially true since both species originated in a wide, common range across central Africa, including the Rift Lakes where speciation and specialization is so prevalent (Philippart and Ruwet 1982). In concert with the feeding patterns evident during the early develOpment stages of most fishes, young Tilapia (up to two inches) feed on 200p1ankton (LeRoux 1956; Drenner et a1. 1982). Upon maturing, the various species tend to specialize on particular feed resources, although many maintain a varied diet. When confronted with scarcity of their preferential feed, many Species readily utilize other resources (LeRoux 1956). T. zillii are predominantly macrophyte and filamentous algae feeders (Philippart and Ruwet 1982; Bowen 1982; Buddington 1979; Fryer and Iles 1972). MicroscOpic examination of the oral dentition reveals that both jaws contain three rows of teeth with the anterior 95 being bicuspidate, chisel-shaped, and the posterior being tricuspidate. When a leaf is seized, it is gripped firmly by the inner rows and severed by the outer row of blade-like teeth (Fryer and Iles 1972). Because the front teeth are located at the tip of the snout, they have also been observed to function as scrapers of periphyton. Personal observation showed the gill rakers to be short, rigid, sharply pointed and slightly curved inward. The Open Spacing and short length suggest a relatively inefficient filtration mechanism. The upper pharyngeal consists of duplicate pads of dense dentition. Individual teeth are stocky, yet narrow with serrated edges facing posteriorly. The lower pharyngeal is a triangularly-shaped, dense pad of dentition. The anterior portion consists of thin, sharp rods, curved and pointed posteriorly. From anterior to posterior, the teeth gradually become broader and more upright until the last few rows closely resemble the upper pharyngeal teeth, with the notable exception that the serration faces anteriorly. The combined assemblage of jaws, gill rakers and pharyngeals appears to provide an efficient mechanism for biting, tearing and shredding material before passage to the stomach. Adult §. niloticus are reported to effectively feed on phytOplankton (Edwards et a1. 1981; Harbott 1975; Moriarty et a1. 1973). Moriarty and Moriarty (1973) demonstrated that g. niloticus can assimilate 70-80% of the carbon 96 ingested as blue-green algae. Assuming that this species is able to efficiently assimilate phytoplankton, it is expected that mouth structures, as a result of selection pressures, would be shaped and arranged such that phytOplankton could be separated from the water and transported to the eSOph- agus. Greenwood (1953), in an examination of the phytophagus ‘1. esculenta, described an efficient system for harvesting algae through the combined action of mucus secretion, gill rakers and pharyngeal teeth. When suspended phytOplankton is drawn into the buccal cavity, it becomes mixed with the copious mucus which is secreted from the buccal and pharyn- geal epithelium and gill arches. The mucus-algae aggregates are carried posteriorly to the gill rakers which, while not fine enough to filter individual cells, are capable of retaining the aggregates and transporting them to the pharynx. Here the mucus-algae aggregate comes in contact with the anterior pharyngeal teeth. Both the paired upper and fused lower pharyngeal are covered with numerous fine teeth. Those on the anterior, lower pharyngeal are recurved while the posterior teeth are cuspidate and erect. The upper pharyngeals possess a hook-like tip directed toward the eSOphagus. As the aggregates contact this dentition, the upper pharyngeal is moved backward with the hooked tips passing between the lower pharyngeal teeth, raking the aggregates toward the esophagus. 97 Moriarty (1973) suggested that the size and shape of the phytOplankton particles affects the effectiveness of filtration of the mucus-algae aggregates with the larger and filamentous forms more efficiently retained. Furthermore, Fryer and Iles (1972) described the filtering action of micro-branchial Spines located on the gill arches of Tilapia. These numerous, fine projections, each with lateral spines, may act as an efficient sieve trapping very small particles. The particles would be removed from this sieve if the final closure of the Opercula were delayed until after the mouth began to Open, causing a momentary reverse in current which would blow the entrapped particles into the mucus-covered buccal cavity. Drenner et a1. (1984) suggested that the size selective grazing of Tilapia aurea on phytOplankton may be due to the efficient retention of particles by microbranchial spines. This species was stated to size-selectively feed on phyto- plankton larger than 25 um, and escape-selectively feed on ZOOplankton. No apparent relationship of fish size to the size of particle selectively ingested was seen over a range of 4.3 - 18.7 cm standard length fish. A microscOpic examination Of the mouth parts Of §. niloticus used in this study revealed distinctive contrasts to the I. zillii and confirmed the Observations of the previously mentioned authors. The g. niloticus possessed four irregularly aligned rows of tricuspidate teeth on the 98 upper and lower jaw. The teeth resembled the smaller denti- tion found in the posterior rows of the T. zillii jaw. The gill rakers were fleshy, closely packed projections covered with cOpious mucous. The upper pharyngeal consisted of paired pads of densely packed teeth. Individual teeth were tall and thin, aligned on a posterior-anterior axis. Each tooth possessed a hook-shaped tip which faced posteriorly. The lower pharyngeal was a triangular pad with the anterior portion bearing thin, sharp rods curved and pointed poster- iorly. From anterior to posterior, the teeth become broader and more upright until the last few rows closely resemble the upper pharyngeal teeth, with the notable exception that the hook-shaped tip faces anteriorly. The most distinctive difference between the two species is the presence Of the microbranchial spines on the gill arch of the g. niloticus. A single row of projections were observed to exist on the anterior side of the second and third gill arches. Each projection was observed under 400 x magnification to extend lateral spines to each side with Spacing approximated at two microns. These spines appeared fully developed when sampled from a 50 g, 15 cm fish with eight to ten lateral spines present on each side of the projection. However, when viewed on a 0.6 g, 3.2 cm fish, the projections were barely visible and only two lateral Spines, at most, were present. A 2.2 g, 4.7 cm fish had a maximum Of three lateral spines on each side. Once again, 99 when examining the combined assemblage of jaws, gill rakers, microbranchial Spines, and pharyngeals, it appears that g. niloticus is capable of processing a variety of feeds, but is especially suited as an adult for processing algae. Until the microbranchial spines develOp to provide an effi- cient filtering mechanism to harvest suspended algae, it would be dependent on larger food particles. The physiological features involved in the collection, processing and transport of the feed to the eSOphagus are easily discerned and indicate that each species possesses distinctive features which allow it to efficiently harvest a particular feed type. However, the efficient utilization of the feed involves the capability of digestion and assimila- tion into body tissues. Just as each species evidences particular oral struc- tures which enable it to harvest and process particular feeds, so the digestive physiology should evidence features which would facilitate utilization of the feeds. For g. niloticus to successfully utilize phytOplankton, a mechanism must be present to lyse the cell walls, thereby permitting assimilation in the intestine. A biochemical breakdown of the phytoplankton cell walls would require cellulose enzymes, and the mucilagenous sheath of blue-green algae would require hemi-cellulase or RNAase enzymes (Payne 1978). Stickney and Shumway (1974) reported that, based on their investigations into the occurrence of cellulase activity in 100 62 species, fish lack the ability to form cellulase and must obtain such activity either from their food or through the establishment of a cellulase producing bacterial flora in the alimentary tract. Prys and Blaszczyk (1977) further reported that fish are unable to maintain a specialized cellulolytic gut flora. In the absence of a biochemical capability, a physical or chemical mechanism must be relied upon. Payne (1978) determined that several species of mullet which feed on phytOplankton possess a muscular gizzard containing sand particles. Through the grinding action Of this gizzard, the fish is able to physically lyse the cells and render them susceptible to further digestion. However, é. niloticus lacks a gizzard, and the mouth parts described are incapable of lysing small algal cells, so physical disruption of the cell walls seems of little import. Moriarty and Moriarty (1973) and Moriarty et a1. (1973) provided an excellent description of the chemical mechanism §. niloticus employs to lyse phytOplankton cell walls. They reported that acid secretion in the sac-like stomach of the fish follows a diurnal cycle associated with feeding activity. At dawn, before feeding has commenced, the stomach is empty and the pH is 7.0. AS feeding begins, the stomach expands and acid secretion is triggered, dropping the pH to 2.0-3.0. As feeding continues, a large portion of the algae is reportedly not subject to sufficiently low pH, 101 . and the digestion and assimilation are inefficient. Higher pH values are generally apparent in the upper stomach while only in the lower stomach, where the pH drOps to 1.4, is cell lysis effective. After feeding ceases, the acid secretion continues for some time and the accumulated algae are all subjected to pH values well below 2.0. When this occurs, the algal cells were Observed to be completely lysed and the subsequent digestion and assimilation almost com- plete. Harbott (1975) reported similar observations on stomach pH values for §. niloticus and also determined that cells of the blue-green algae Microcystis were lysed in vitro by acid conditions below pH 1.65. The efficiency of digestion and assimilation of phytOplankton was thus demon- strated to be dependent on sufficient exposure to acid con- ditions below pH 2.0 which would lyse the cell walls. While studies are lacking identifying the particular digestive mechanisms in I. zillii, they have been shown capable of efficient digestion of macrOphyteS. Buddington (1979)_reported that when I. zillii were fed Eajas guada- lupensis, they were capable of assimilating 75.1% of the protein, 75.9% of the fat, and 45.4% of the caloric content of the plant. It was previously expressed that an evaluation of feed quality must involve joint considerations of harvestability, composition and digestibility. This seems especially appro- priate in light of the evidence that selective pressures 102 have resulted in Simultaneous physiological refinements in the ability of particular species to harvest and digest particular feeds. If this understanding is correct, then a comparison of the growth of different Species on different feed types should reflect not only the quality of feed composition, but also the relative pre-digestion processing abilities and the digestion-assimilation efficiencies. It was for this reason that I designed an experiment to deter- mine if the proPosed threshold-corrected hyperbolic equation was effective in quantitatively evaluating the digestive- assimilation efficiencies Of different Species utilizing different feeds. Species Comparison In this study, both S. niloticus and I. zillii were each fed both pelleted trout chow and the blue-green algae Spirulina. The results are summarized in Table 23 and Figure 7. In addition, the efficiencies for each of the feeding trials are shown in Figure 12. It is apparent from the constants and the growth curves that each species con- sistently demonstrated superior growth when fed trout chow. The trout chow is formulated to promote excellent growth and may have stimulated growth close to each species' genetic capability. This represents another important capacity of the growth model. Based on this assumption, the comparison of the growths of the two species on Spirulina leads to sig- nificant conclusions. 103 When the efficiencies of growth (from 1-7% bw/day) for each species on Spirulina are calculated as a percentage of their efficiency of growth on trout chow, the I. zillii ranged from 36.3% to 51.2%, while the S. niloticus ranged from 52.8% to 91.5%. While each species was unable to fully utilize the high protein content of the Spirulina to promote growth, the S. niloticus was clearly more efficient than I. zillii. This relative difference in ability to utilize pelleted Spirulina to stimulate growth may be attributed to either different abilities to digest the feed, or different abili- ties to assimilate the components, or both. The ability of fish to digest Spirulina, or any alga, has been noted to be dependent on an ability to break down the cell walls. The digestive system of S. niloticus has been documented as being capable of lysing algal cell walls, and this study confirms its ability to grow well with Spirulina as the sole food source. The inferior growth response exhibited by both Species when fed Spirulina, as compared with trout chow, may again be attributed to either limiting amino acids or protein utilization for energy rather than growth. The literature has documented no physiological adapta- tions which would enable 1. zillii to lyse algal cell walls. Instead, this species has been consistently described as a macrOphyte feeder. The digestive conditions necessary to process macrOphytes would probably not provide the 104 biochemical, chemical or physical mechanisms necessary to lyse algal cell walls. Yet the I. zillii was able to achieve a measure of growth on the pelleted Spirulina. This could be, at least partially, attributed to previous lysing of algal cell walls during the harvesting-drying- pelleting process. MicroscoPic examination of the powdered algae revealed that many of the cell walls were ruptured. If only a portion of the algal cell walls were lysed prior to ingestion by the fish, and only the S. niloticus were capable Of lysing the remaining percentage, then the rela- tive difference in ability to utilize the Spirulina could occur. It is also possible that the relative efficiencies of utilization of pelleted Spirulina by the two Species is due to differential assimilations of equally digested materials. If selective pressures have resulted in one species posses- sing physical mechanisms which enable it to effectively utilize a particular feed, those same pressures may have resulted in changes in assimilation processes which facil- itate the effective incorporation of feed into tissue by the same species. Quite possibly, an interaction of both dif- ferential digestion and assimilation produce the differen- tial efficiencies demonstrated by the growth curves. The threshold-corrected hyperbolic equation is thus shown capable of quantitatively identifying the ability of different species to utilize different feeds. Some Of the 105 general sources of the differential efficiencies may be identified by comparisons Of the growth curves. However, it should be understood that when growth is related to sub- strate, the growth parameter represents the integration of all the complicated mechanisms, processes and reactions associated with feed harvest, digestion and assimilation. Thus far, the model has been effective in relating the quality of different feeds to one Species and also comparing Species on different feeds. These studies, because of their structure, have been limited to examining the growth re- Sponse due to composition of feed and the ability to digest and assimilate. The remaining variable in the consideration of feed and growth was harvestability. Although both I. zillii and S. niloticus were shown capable of growing on pelleted Spirulina, this has little relevance to real-world situations. The mechanical har- vesting of algae is usually inefficient and energy inten- sive. To attain practical relevance as a feed source, the fish must demonstrate an ability to harvest the algae directly from the water and effectively assimilate it. Numerous authors (Melack 1976; Almazon and Boyd 1978; Edwards et al. 1981) have reported correlations between Tilapia growth and levels Of primary production or plankton abundance. While these studies are useful in determining general relationships, they have not related fish growth to particular feeds or identified the source or means of 106 harvest, whether scraping of periphyton, size selective capture Of zOOplankton, indiscriminate filtration, etc. A brief eXperiment was conducted to determine the ability of both fish species to harvest live suSpended Spirulina from the water. Two 20 l aquaria, each equipped with 150 watt heaters and 15 cm air bars, received 15 l of conditioned water which was maintained at 27.5°C. Approxi- mately 200 ml of a dense, live Spirulina culture were added to each aquarium. After extensive aeration and mixing, a 10 ml aliquot was removed, diluted, and the algal cell con- centration determined by microscopic examination. These concentrations were then utilized to calculate the total number of algal cells per aquaria. Following this initial cell count, a 22 g‘S. niloticus was added to one aquarium while the other remained a control without fish. Subsequent sampling for cell count analysis followed the schedule listed in Table 27. Following completion of this experi- ment, an identical trial was performed using the same S. niloticus in one aquarium and a 47 g I. zillii in the other. The results Of cell count analysis in all four trials are contained in Table 27. Over the period of 32 hours, the I. zillii was able to 9 harvest 2.71°10 cells. These cells were estimated to 7mg-dry weight) weigh 0.271 g (an. cell weight was 1.0'10- on a dry-weight basis or 4.44 g on a wet-weight basis (93.9% moisture). Therefore, this fish was able to harvest 9.4% of 107 50.0 no.0 no.m 0.m0 NM 1 mm.0 om.w u mu m¢.0 mm.~ 00.0 n «N 0m.0 0m.~ 0N.o n m.~0 0~.~ 0w.q 00.0 u 00 mm.m 00.0 00.0 n m 00.0 00.0 0N.00 u 0 no.0 w~.00 ww.00 u 0 05.x 00.00 om.o u N 00.N0 00.00 0m.a m.M0 0 00000 000. 00000000. 00000000. 00000000. Am0sonv 00w00 0O 00 00009 0 00008 000000 0000008 0o00so0 0x000 EO00 0809 00000O00: =ovo0050o00m .0az0o> 0 N.m0 00 00000 .00 0:00:00mm 0o GO0000000OSOU .NN 00009 108 its body weight over the study period or 14% of the total algal cells present. The small decrease in cell number in the control aquarium indicates that some adsorption and settling of algal cells did occur. These aggregations of cells could provide Opportunities for capture apart from filtration of suspended algal cells. The S. niloticus, during the first trial, was able 9 to harvest 8.44'10 cells which represented 84.3% of the total cells present. The filtration efficiency of the fish, which had previously received only pelleted feeds, improved during the second trial when it filtered out 11.3‘109 cells or 94.4% of the total algal cells present. Therefore, on the average, the fish was able to harvest 73.6% of its body weight over the study period. Visual observations clearly indicated that, based on either the improved clarity of the water, or the abundant green feces present, the S. niloticus could effectively filter suspended Spirulina cells from the water. MicroscOpic examination of the greenish feces re- vealed that many filaments Of Spirulina were still intact and that limited cell disruption had occurred. Nitrogen analysis of these feces indicated that almost 79% of the original concentration remained. However, the transforma- tion of suspended algae to fecal material does not represent conversion to an unavailable form for S. niloticus. I have Observed that S. niloticus in holding aquaria receiving limited feed are effective in ingesting and recycling fecal material. 109 r o 2.0. m“ o b a: LG- 0 0 I ' O 5 o a. [.2- 0 m P D a) o In > 008' o o: < I .. ° 8 m :3 0.4. o m o 0 o o o Om .0 - r . . I . . 1 a O 2 4 6 8 IO SPIRULINA CELLS PER AQUARIUM (I09) Figure 13. Relationship between the ability of S. niloticus to harvest suspended Spirulina cells and the concentration of suspended cells. 110 Following a Six hour acclimation period, the removal rate of Spirulina cells by S. niloticus decreased dramat- ically with time in each trial. The rate of filtering of S. niloticus was examined by plotting the cells harvested/ hour against the total cell concentration. These data are plotted in Figure 13 (with the exception of the initial six hours of the first trial). The data indicate that harvest rate is directly related to cell concentration up to an 10 cells or average aquarium concentration of almost 1-10 6.58°1O5 cells/ml. It was expected that the microbran- chial spine-mucous harvest mechanism would become saturated at a high cell density and the harvest rate level off. This saturation was not evident at a cell concentration repre- senting an extremely dense algal bloom. Although there was no question concerning the ability of S. niloticus to harvest suspended Spirulina cells from the water, the tremendous passage rate through the gut left some doubt concerning its ability to effectively grow under these conditions. Therefore, the same fish was used on a longer term, four-day experiment, during which the aquarium received daily additions of the Spirulina culture to main- 10 cells/aquarium. 10 tain initial concentrations of 1.62-10 Over the four-day period, the fish harvested 5.36-10 suspended Spirulina cells, or 5.38 g dry-weight (87.9 g wet-weight). Obviously, this represents a tremendous lll ingestion rate--almost 100% bw/day-wet weight. Over the four-day period, the fish grew from 22.0 g to 23.6 g, a specific growth rate of 0.0176/day. Despite the rapid gut passage rate, the fish was able to withdraw sufficient nutrients to sustain growth. The actual growth of the S. niloticus may be Of lesser practical significance than the removal of the suSpended algae from the water column. The conversion of the suspended algae into fecal material accomplishes two additional important functions. First, it serves to greatly clarify the water which would have pro- found consequences for the water chemistry and the biota. Second, it could serve to transform the previously un- available suspended algae into a practical, high quality food source. The fecal material is easily ingested and potentially of high protein content. These factors are of obvious practical usefulness for aquaculture and especially polyculture. GROWTH-SIZE RELATIONSHIP Body size has long been recognized as a factor affecting the growth of organisms. Brett and Shelbourn (1975) stated that the fundamental laws of growth for all animals attest to a continuous decrease in growth rate with increasing Size subject to some variation in the rate Of decrease according to developmental stages. Bertalanffy (1957) identified the following individual parameters as 112 being affected by body size: total metabolism, heart rate, respiratory rate, excretion, chemical composition, and enzyme content. Despite the understanding Of these rela- tionships, the exact causative mechanism relating size and growth are poorly understood. Bertalanffy (1957), in his study of metabolism, in- directly related body size to growth. Initially, a rela- tionship between metabolic rate and body surface area was examined. The surface area Of a body was shown to be directly related to the metabolic rate as shown in Rubner's famous surface rule. Unfortunately, there are considerable difficulties involved in measuring surface area. The sur- face area was, in turn, shown to be directly related to 2/3, expresses that weight. The formula of Meeh, S - bw surface area (S) is obtained by multiplying the 2/3 power Of the weight (W) by an apprOpriate constant (b). By combining these considerations, the surface rule of metabolism accord- ingly states that the metabolic rate (M) is prOportional tO the fractional power of the weight, or M - bWa. Therefore, as the body weight increases, the relative metabolic rate decreases. Bertalanffy then continued by defining three basic types of metabolism. The first type, characteristic of fish, has a metabolic rate which was related to surface area by the 2/3 power of weight. Other types of metabolism were either directly related to weight or by a power intermediate 113 between 2/3 and one. Several purported bases for these relationships were then reviewed and found unsatisfactory. These included: thermoregulation, anatomy and physiology of the circulatory system, types Of respiratory apparatus, ana- tomical or chemical changes related to body size, respir- ation changes at the cellular level, and changes in enzy- matic systems. While each factor may be involved in estab- lishing the size-growth relationship, none were singularly capable of complete explanation. Bertalanffy (1957) postu- lated the presence of unidentified factors in the intact organism which regulated the respiration of tissues and, therefore, metabolism. Following his relating of metabolism to size, Berta- lanffy develOped a direct relationship between growth and metabolism, and also thereby, a relationship between size and growth. The basic consideration was that growth was the result Of a counteraction of synthesis and destruction. This interaction was explained by the general formula: fi%-nWm-kWn Here the change in body weight with time (%%) is equal to the difference between anabolism (nWm) and catabolism (kWn). The exponents m and n indicate that both processes are prOportional to some power of the body weight (W). This important assumption was justified by the consideration that all physiological processes hitherto investigated can be 114 expressed in such a power formula. Based on various physio- logical facts, the catabolism exponent is approximated to be one. The exponent of the anabolism (metabolism) was previ- ously shown to be 2/3 for fish. Therefore, the general formula becomes: 9%- nW2/3 - kW This theoretical formula for growth, based on a weight- metabolic rate relationship, represents a sigmoid growth curve. Paloheimo and Dickie (1965), while appreciating the value of growth curves for comparison purposes, stressed the importance of the biological interpretation of the param- eters describing the statistical fitting. AS a result, the authors questioned the general physiological interpretation Offered by Bertalanffy and suggested an alternative consid- eration by relating growth to energy. Rather than the ab- stract, nonmeasurable concepts of anabolism and catabolism, growth was related to quantifiable energy parameters as shown in the basic energy formula: d_W- - t pR T Here the growth is equal to the metabolizable energy content of the ration (pR) minus the total metabolic energy expendi- ture (T). The metabolic energy cost is again related to weight as T - ‘WG, where ¢ is a constant and 6 defines the 115 rate of change in metabolism with body weight. Therefore, the general formula becomes: -d—"t"-pR-"‘W6 The authors then continued with a lengthy examination of ration energy and feeding efficiency, concluding that growth efficiency decreases for each unit increase in ration. Therefore, as either weight or ration increases, the growth efficiency is expected to decrease. Elliott (1976), in his consideration of the energetics of feeding brown trout, developed formulas which related both the energy of metabolism and food intake to the weight of the fish. Both formulas were identical as shown below: Energy intake (or metabolism) - aWb‘esz Here W represented weight, a, b1 and b2 were constants determined empirically, and T was temperature. Because b1 and b2 are fractional constants, both energy intake and/or energy expenditure per unit would decrease with increased weight. The change in rate of growth would be a function of the relative rate of decrease in both energy intake and metabolism expressed by the empirical constants a, b1 and b2. A comparison of these theoretical considerations re- veals important common factors. Although based on different assumptions and underlying principles, each analysis con- cluded that the metabolic rate (catabolism) would decrease 116 prOportionately to a fractional power of weight of the organism. Bertalanffy (1957) supported this hypothesis even though he expressed catabolism in the form kWn with n - l. The setting of n to l was a means of mathematical simpli- fication justified by the belief that the deviation from unity was small and the basic equation rather insensitive to the change. A second common principle was that the energy intake rate or anabolism would also decrease prOportionately to a fractional power of weight of the organism. Berta- lanffy (1957) expressed anabolism as prOportional to the 2/3 power Of weight for fish, while Elliott (1976) and Paloheimo and Dickie (1965) expressed the energy intake as prOpor- tional to empirically determined fractional powers of the weight. All three authors Similarly conclude that growth rate is related to weight as a function of these interacting processes, both of which decrease with increasing Size. Empirical research tends to confirm many of these theo- retical eXpectationS. Niimi and Beamish (1974) reported that the amount of ration representing satiation decreases with increasing size in largemouth bass. Furthermore, the maintenance costs for zero growth (Sq) was also shown to decrease with increased Size. Brett (1979) reported that increased size had a greater restricting effect on growth than on metabolic rate, though both were affected; and also suggested that age, independent of size, may have some effect on growth rates. In reviewing growth curves reported 117 for three Species of fish, Brett reported that increasing size resulted in a decreasing maximal growth rate and de- creased maintenance levels per unit weight for all species. Gerking (1971), studying the effect of body weight on pro- tein metabolism Of bluegill sunfish, determined that the efficiency Of protein utilization steadily decreased as the weight of the fish increased. The author suggested that the decreased efficiency may, in part, be attributed to a de- creased rate Of feeding with increased size but offered no further explanation. Both theoretical expectations and empirical studies agree that growth rate and maintenance costs decrease with increased Size. This phenomenon has been generally related to numerous physiological processes and feeding energetics. Unfortunately, no definitive explanation has been found which precisely explains the controlling mechanisms. The relationship may be the cumulative expression of several of the aforementioned variables interacting with age and genetic composition. I do not intend to propose that the enzyme-kinetic growth model defines a Single controlling factor producing the weight-growth relationship. Instead, I prOpose the model as a tool to quantitatively monitor the weight-growth relationship by taking a holistic look at the system. Whether the actual mechanisms producing the re- Sponse are controlled at the cellular or organismic level, whether it's more genetically or environmentally influenced, 118 or whether it's any of a number of interacting, complex physiological processes, the model proposes to utilize the fish as an integrator of all the components. Because each of the size-related biochemical processes is governed by enzyme-catalyzed reactions, it is proposed that the integrated response will follow the same pattern which typifies the singular enzyme-substrate reaction. Although the rate Of reaction may change with size, the basic nature of the reaction remains unchanged. Therefore, whether at the singular reaction level or the complex, inte- grated organismic level, the response pattern is similarly characterized by the threshold-corrected hyperbolic growth model. If it were true that the size Of an organism affected, by whatever means, the rate of delivery of substrates or the balance between anabolism and catabolism, these changes should be reflected in the integrated growth reSponse to substrate level. Thus, the model is develOped with con- sideration of the organism's interaction with the environ- ment and how that interaction changes with the increasing Size of the organism. The results of the experiments dealing with size and growth are contained in Tables 14-17 and Figure 5. In addition, the efficiency curve for each size class fish is plotted in Figure 14. It is apparent from Figure 5 that the growth curve for each size class of I. zillii closely 119 .Uom.mm 00 SOLO 0SO00 00w law 0.0m 000 w 0.00 .w 00.0 .w 000.0 m0nw003 0w000>0v 000000 .9 mo 0000000 0000 0aou 0ou m0>030 :03o0w >Oc0000wwm .00 00=w0m >40 mum out .535? >000 bzwommm 01 o .0 m .0 _ o _ p o flu . J: H II. «J éod 0... .0» l N «J / j A 0.2 \ 60.0 .m w 2... m .03 0 Zoo .20 120 follows the threshold-corrected hyperbolic curve model. A regression of the observed growth data, utilizing the trans- formed linear equation, results in coefficients of determin- ation (r2) of 0.99 (n - S), 0.71 (n - S), 0.91 (n - 7) and 0.97 (n - 4) for the 0.014, 1.76, 10.0, and 38.1 g size classes respectively. The data demonstrate that the model is capable of quantitatively eipressing the integrated growth response as a function of varying substrate concen- trations for the various sizes of fish. Although the gen- eral shape of each curve adheres to the model, the individ- ual constants of each curve reflect distinct differences which, by comparison, provides insight into the growth dynamics of the fish. Overall, the most apparent contrast between the curves is that at high feeding levels (above 3%), there is a drama- tic decrease in growth and efficiency with increasing size. These findings correspond with both the theoretical expec- tations and previous research reports on fish growth. This decrease in growth and efficiency is especially significant since the maximum growth rate (umax) has been reduced by 2/3 when the fish has attained only a 38 g size. If the fish were to reach their potential of about 1 kg, the great majority of growth would occur at a specific growth rate below 0.03. The calculated values of u for the different size max classes of fish are plotted in Figure 15, along with the 121 .00 a a .00.0 n 00 «now 0:0 o» 000 10>uso 0000:0comx0 0:0 vc0moom.- wme3o£0 uaouu 000 000000 .9 no 0000000 0000 usow now 00:00> = 0000000000 .m0 0naw0m 324mg .5353 19“.. mm on ¢4~ 0.. 0.. q ‘ . No.0 30°MO.OIO NNQO.OI* .. V0.0 00.0 o 0.0 (mum 3mm Hlmoas ounoads unmwi 122 curve described by the exponential equation "max - 0.0922e-0'0305w. A regression of the calculated "max data, utilizing the exponential equation, results in a coefficient of determination (r2) of 0.97 (n - 4). Whatever the controlling mechanism for the growth process, tremendous emphasis is placed on early, rapid growth. As the fish quickly accumulate size, perhaps enough to escape predation or to facilitate utilization of differ- ent feed resources, the growth rate rapidly diminishes. And yet, these considerations have only involved growth at high feeding levels above 3%. Below this level, quite different growth dynamics are evidenced. At low percentage feeding levels, numerous crossover points occur between the growth curves for the different sized fish (Figure 5). In examining the growth curve of the 0.0140 g fish, it is noted that the line intersects the other growth curves at feeding levels between 1.5 and 2% bw/d. At feeding levels of 1.5%, the relative positions have entirely reversed, and the smallest size fish exPerience the lowest growth rate. These contrasts and intersections are the result of widely diver- gent maintenance requirements for the various size classes. While the smallest fry demonstrate superior growth at high feeding levels, they simultaneously expend high levels of energy to maintain that level of performance. The 0.014 g fish are shown to require a ration level of 1.55% bw/d for maintenance. At this same percentage of 123 feed, the 1.76 g fish would be growing at a specific growth rate of 0.015. These dynamics are also demonstrated in the efficiency curves. Above a 2% bw/d feeding level, the 0.014 g fish are superior to larger fish in their ability to process feed and convert it to fish tissue. However, this relative advantage in efficiency changes drastically when feeding levels drOp below 2% bw/d. This size class is most competitive at high percentage feed levels with their feed processing geared for fast growth at high cost. When the high maintenance cost has been paid, the potential for growth is tremendous. The feeding level for maximum effi- ciency of this size fish is approximately 3.2% bw/d. Above this level, a gradual reduction occurs due to decreased diffusion-assimilation efficiencies and the saturation of enzyme reaction rates. At ration levels above 3.2% bw/d, each additional unit of feed added contributes to an already nearly saturated system, and the efficiency of processing each unit of feed decreases. Below the maximum efficiency feeding level, the growth per unit feed decreases rapidly. This may be attributed to the high maintenance costs re- quiring an increasing percentage of the total feed, leaving little for growth. The 1.76 g fish demonstrates different priorities of Operation. At higher percentage levels of feed, this size is clearly less efficient than the smaller fish. However, at approximately 2% bw/d feed for each, the two experience 124 equal growth rates. At lower, equal percentage body weight feeding levels, the larger fish, with a 43% reduction in maintenance costs, becomes the more efficient processor. This trend continues with the 10 g fish whose mainten- ance costs are 76% lower than for the smallest fish. The maintenance feeding levels (Sq) are plotted versus size in Figure 16. Also shown in the figure is the curve repre- sented by the equation: Sq - 0.906 - (0.149an). A regression of the maintenance feeding level data, utilizing the log equation, results in a coefficient of determination (r2) of 0.94 (n - 4). Reduced maintenance costs with increased size also are reflected in the steadily decreasing feeding levels for max- imum growth efficiency. Feeding levels for maximum effic- iency were 3.2%, 2.5%, and 1.7% bw/d for the 0.014 g, 1.76 g and 10 g fish respectively. At high percentage feeding levels, the 38.1 g fish exhibited relatively low specific growth rates and levels of efficiency indicating a limited ability to process high feed levels. Because of its lower maintenance requirements, the 38.1 g fish are able to experience specific growth rates higher than the 0.014 g fish at equal percentage feed levels below 1.7% bw/d. Similarly, the 38.1 g fish are able to grow faster than the 1.76 g fish when feed levels are below 1.12 bw/d. However, the 38.1 g fish is never able to pro- cess equal percentages of feed more efficiently than the 10 g 125 .A0 I a 000.0 I Nuv 000m 0m» O0 000 cO000=v0 wO0 0:0 0:0 oommmw 00 3O£0 00Op0 000 000000 .9 no 0000000 0000 uaom you 00:00> m 0000000000 .00 0uaw0m 825.2 20.0; :0... 00 on .0 o. 0. 0 o a} q q a J q a q 1 4 q u q o L d O 1 ¢.o .3... 03.8 -0006 1.0 .06 O L 0.. 0.. (”3) 13A3'I 9010335 aouvnalmvw 126 fish, due to higher maintenance costs. While the small difference in maintenance costs (0.11% bw/d) may not be statistically significant, it may represent an important trend in the anabolism-catabolism balance indicating the increased costs necessary to maintain an aging fish. It would be interesting to define the growth curve of a rela- tively large fish (200-250 g) to compare maintenance costs on this size class. It is important to realize that these comparisons of growth have been made on the basis of feeding levels ex- pressed in percentages of body weight. Therefore, these comparisons reflect only the abilities of each size class fish to utilize similar percentages of feed -- not equal amounts of feed. When a 10 g fish and a 0.014 g fish are both fed 1.5% bw/d trout chow, the 10 g fish will experience a higher specific growth rate and demonstrate a higher effi- ciency of feed utilization. However, the 1.5% bw/d for the 10 g fish equals 0.15 g of trout chow per day, which corres- ponds tO over 1,000% bw/d for the 0.014 g fish. Therefore, the smaller size classes are more efficient processors of equal weights of feed. In summary, the enzyme-kinetic growth model was demon- strated to be useful for quantifying the differential growth response of various sizes of Tilapia zillii. The growth curves indicated that at high feeding levels, both the growth rate and efficiency decrease steadily with increasing 127 size. However, at lower feeding levels, the relative growth rates and efficiencies are greatly affected by maintenance requirements. The data also suggest that although mainten- ance costs decrease rapidly during the fry-fingerling tran- sition, eventually, as adults, the maintenance costs stabil- ize or may even increase slightly. These changes in "m and Sq with size may be incor- ax porated into the original growth formula and utilized to predict the growth of fish at any size for the given condi- tions. The maximum growth rate (umax) was effectively related to weight with the exponential formula: u - 0.0922e'0'0305w max This value for “max may be substituted into the original threshold-corrected hyperbolic curve which then would appear: II ' (0.0922e'0.0305W)( S " Sq ) KS + S - ZSq The maintenance feeding level was related to weight by the 10g formula Sq - 0.906 - (0.149an). This value also may be incorporated into the original formula as follows: u _ (0.0922e'0'0305W) s - (0.906 - 0.149an) Ks + s - 2(O.906 - 0.1491nW) Using this formula, the growth of any size I. zillii at any 128 given feeding level of trout chow for the given conditions could be calculated. The formula could be further refined by incorporating a mathematical expression which related Ks to weight. The data obtained for Ks may be related to weight by a quadratic equation with a coefficient of determination (r2) of 0.98 (n - 4). However, this quadratic equation projects an un- realistic cessation of growth at a 55 g size level. Obvi- ously, this is unrealistic and again points out the desir- ability of feeding trials involving larger sized fish. GROWTH-TEMPERATURE RELATIONSHIP Temperature is recognized as an important variable affecting the well-being of organisms (Hill 1976). Because fish are poikilothermic, temperature has been reported as the primary environmental factor influencing their physi- ology (Caulton 1983). However, the well-being or growth of an organism may also be viewed as the integration of the efficiencies at which each of the physiological processes are occurring within the organism. It is at this level, the biochemical reactions and biOphysical mechanisms, that tem- perature has its direct effect. An important principal underlying all reactions is that as temperature increases, the movement of particles in- crease. The increased movement of particles causes both an increase in the kinetic energy and an increase in the 129 frequency of collision between molecules. More importantly, the number or percentage of collisions having sufficient energy to participate in a reaction are greatly increased (Solomons 1976). This principle and a consideration of activation energies of reactions led Arrhenius to relate temperature to the metabolic rate of organisms as expressed in the equation: metabolic rate - a10bT , where a and b are constants and T is temperature (Hill 1976). In summary, biochemical reactions have been shown to be temperature dependent. BiOphysical mechanisms such as diffusion and adsorption are also important factors in physiological processes. For organisms with extracellular digestion, the diffusion pro- cesses are important in determining the rate at which nutri- ents are transported to sites for anabolism, and waste materials carried away from catabolism sites. Transport from ingestion to reaction site involves diffusion across several membrane barriers. Passive diffusion usually occurs in an aqueous medium, wherein rates are predicted according to Fick's Law of Diffusion. The Fick's Law diffusion con- stant is determined by the amount of random motion of the particles in the aqueous solution. Therefore, as the tem- perature increases, the kinetic energy of the particles increase and the diffusion rate increases (Florey 1969). Because both the rate of supply of reactants and the kinetic energy of the molecules increase with higher 130 temperature, it would seem that the rates of the physiologi- cal processes should also increase. However, this relation- ship is clearly limited in biological systems where organ- isms experience both upper and lower temperature limits for their efficient Operation and existence. Other factors must be involved affecting the basic relationship between temper- ature and the physiological processes. In both unicellular and multicellular animals, the biochemical reactions are regulated by enzymes. Even the simplest manifestations of life involve many complex enzy- matically controlled reactions (Hoar 1979). The enzymes, while not contributing energy to the reactions, are neces- sary for the activation of the process. Enzyme reactions, like other reactions, are dependent on the amount of molecular motion and kinetic energy. As the temperature increases, the reaction velocity of the enzyme with the substrate increases (Hill 1976). However, because enzymes are proteins, they are susceptible to structural disruption and denaturation at higher tempera- tures (Atlas and Bartha 1981). Without the prOper func- tioning enzyme, the biochemical reactions are stifled at high temperatures. A consideration Of enzymes has added new dimension to the influence of temperature on organisms. Rather than a system of reactions whose rates increase pro- gressively with increasing temperature, instead, an inte- grated complex of enzymatically controlled reactions are 131 occurring within specific temperature bounds. Indeed, the system is further complicated by a con- sideration of isozymes, different chemical forms of the same enzymes which function effectively at two different tempera- ture levels. Organisms have been shown capable of intercon- verting from the production of one isozyme to another, thereby maintaining reaction rates over a broader tempera- ture range (Hill 1976). Thus, rather than a dramatic change in reaction rates with changing temperature, a gradual change might be experienced as a series of isozymes Operate over a temperature range. Temperature boundaries are also affected by the solidification of lipids at low temperatures. Lipids are important components of membranes, and their solidification can seriously disrupt diffusional processes. The extent of solidification is affected by the degree of saturation of the lipids. The higher the degree of saturation, the higher the solidification temperature. Organisms adapted for low temperatures are usually characterized by lipids of low saturation (Hill 1976). As temperatures decline, the degree of solidification will affect the diffusion rates across membranes. This, in turn, will affect the rates of the enzymatically controlled reactions. These effects of low temperature on rates of reactions contributes to causing a decreased protein synthesis rate in bacteria (Atlas and Bartha 1981). 132 Temperature is thus shown to have a general effect on the physiological processes occurring in organisms. The factors of activation energy levels, enzyme function and denaturation, diffusion and lipid solidification all inter- act to delineate a temperature range within which the organ- ism can survive. Rather than abrupt boundaries to these ranges, it is expected that isozymes would function to dampen the negative repercussion of high or low temperature extremes. In addition to the general influence of temperature on reactions, it is possible to identify specific effects of temperature on individual physiological processes. For example, Smit (1967) reported that the digestive rate in fish is determined by three processes whose intensities are temperature dependent: l) enzymatic action, 2) gastric juice secretion, and 3) stomachal motility. It was deter- mined that both the quantity and composition of gastric secretion in brown bullheads is affected by temperature. Gastric juice output was shown to be maximal at 25°C with decreasing volume at both higher and lower temperatures. Food intake rate is another variable affecting the digestive efficiency of fish. The rate of food intake has been shown to be dependent on temperature for several fish species. Peters and Boyd (1972) reported that the maximum feeding rate of the hogchoker depended primarily on temper- ature, and that the relationship was characterized by a 133 quadratic equation. Kinne (1960), working with desert pupfish, similarly found that ingestion increased as the temperature was raised from 15°C to 30°C, but decreased when the temperature was further elevated. Caulton (1977) re- ported that the ingestion rate of Tilapia rendalli was Opti- mal at about 27°C and decreased significantly above or below that temperature. The efficiency of digestion is also affected by tem- perature. Jobling and Davies (1983) reported that increased temperature decreased the gut detention time and also increased the rates of digestion and evacuation. Caulton (1983) similarly reported that increased temperature caused an increase in assimilation efficiency due to more complete feed trituration and improved preassimilatory processing of feed. Because the changing temperature affected various aspects of feed processing differentially, the author sug- gested that the overall effect of temperature on the ingestion-digestion-assimilation process must be considered. Brett (1967) reported on the effects of temperature on the swimming performances of young sockeye salmon. Those studies demonstrated that the capacity of the fish to per- form was adversely affected by temperatures above and below an Optimal temperature of 15°C. Brett theorized that an internal control was Operating as a limiting factor, pos- sibly in the capacity to deliver oxygen, remove metabolic products, provide adequate substrate or activate enzymatic 134 processes. Kutty and Sukumaron (1975) reported similar findings that swimming performance varied in response to temperatures with high and low extremes causing deterior- ation and failure. The authors suggested that temperature acts through the central nervous system to control the rates of processes. Many studies have examined the relationship between temperature and metabolic rate. Denzer (1968) reported that the respiratory rate for both rainbow trout and Sarotherodon niloticus varied with temperature. Caulton (1978) reported data obtained using a continuous flow reSpirometer which indicated that increases in temperature from 16°C to 37°C resulted in increases in oxygen uptake and routine metabolic energy expenditures by Sarotherodon mossambicus. Elliott (1976) demonstrated that respiration of brown trout in- creased with increasing temperature until a maximum was reached at l7.8°C, followed by decreases at two higher tem- peratures. The same pattern of a maximum energy flow occur- ring at an intermediate temperature with decreased energy at lower or higher temperatures was reported for the processes of food intake, waste excretion, specific dynamic action and reSpiration attributed to physical activity. Only the standard metabolism was found to continually increase with increased temperature. Caulton (1983), in summarizing the effect of temperature on feeding and metabolism, reported that an increasing temperature leads to increased 135 assimilation but also to increasing costs due to elevated specific dynamic action. Therefore, in order to prOperly gauge the effect of temperature on an organism, the overall effect must be considered. Fish species are commonly reported to exist within discrete temperature ranges. Broad ranges are reported in reference to the ability to survive and much narrower ranges for Optimal growth and reproduction. A variance in temper- ature ranges for different fish Species has resulted in the usage of terms such as thermOphilic (loving high tempera- tures), eurythermal (having a broad temperature range), and stenothermal (having a narrow temperature range) (Philippart and Ruwet 1982). For example, numerous authors have docu- mented the relatively high temperature range of Tilapia. Denzer (1968) reported that Tilapia nilotica could survive within a range of 11°C to 42°C. Chervinski and Lahav (1976) confirmed the lower limit of Sarotherodon niloticus by noting that most fingerling died upon a two-day exposure to 11°C water. Philippart and Ruwet (1982), experimenting with S. niloticus, determined a lower temperature limit of 10- l3°C and an upper limit of 34-36°C. The same authors, work- ing with I. zillii, reported a lower range of 6-10°C and an upper limit of 36-37°C. Brett et a1. (1969) determined that sockeye salmon evidenced a dramatically lower range extending from 1-25°C. 136 It thus appears that the physiological processes of organisms are bounded by temperature limits and that within these limits the reaction rates vary with temperature. The temperature dependent rates of some physiological reactions and processes may be quantitatively defined. However, when these processes are examined in the context of the complex, interdependent matrix of systems within the organism, the complexity of the network defies complete quantitative definition. One means of simplifying the diverse effects of temperature on the various components of the system is to quantify growth and utilize this parameter as an integrator of the complex interactions. It is expected that the growth-temperature relationship, since it is a cumulative expression of the processes described earlier, will evidence two trends similar to those experienced by the contributing factors. First, because all the reactions involved in the growth process are enzymatically controlled, it is antici- pated that at any given temperature the growth reSponse to substrate will also follow the threshold-corrected hyper- bolic curve characteristic of enzyme-substrate reactions. Secondly, if factors such as diffusion, ingestion, diges- tion, respiration, etc. exhibit distinct similar reSponses over a range of temperatures, it is anticipated that growth will exhibit a similar response over the same temperature range. This anticipated response is again based on enzyme 137‘ kinetics. As temperature changes, the rates of the enzyme- controlled reactions will change. These rate changes should cause a shift in the growth response curve but should not change the basic characteristic of the reSponse. With the exception of the standard metabolic rate, the processes contributing to growth have been reported to ex- hibit an Optimal reaction rate at an intermediate tempera- ture and decreasing rates with either decreasing or increas- ing temperature. Therefore, it is hypothesized that growth, the integrator of these components, will express similar trends. Kinne (1960) determined that desert pupfish experienced a maximal growth rate at 30°C with decreasing growth at both increased and decreased temperature. Growth was reported to decrease in the following order: 30°; 25°, 35°, 20°, and 15°C. Dwyer et a1. (1983) reported similar responses in studying the effect of temperature on brook trout growth. The highest rate of growth occurred at 16°C and decreased in the following series of temperatures: 13°, 19°, 10°, 7°, 4°C. Platt and Hauser (1978), studying the growth rate of I. zillii at various temperatures, determined an Optimal maximum growth rate occurred between 28.8°C and 31.4°C. When the temperature was elevated or lowered, growth was reported to decrease. When the growth data of Platt and Hauser (1978) are plotted against temperature, the data are fit by a quadratic curve with an r2 of 0.79. This 138 quadratic curve had roots at 21.2°C and 37.6°C and Optimal growth at 29.4°C. Brett et al. (1969) reported a similar. type response for sockeye salmon in a much lower temperature range. Optimal growth was reported at 15°C with progressive decreases at 10°, 20°, 5°, 1° and 24°C. The observed and calculated data resulting from my feeding studies where I. zillii were fed trout chow at five different temperatures are listed in Tables 6-11 and summar- ized in Table 12. The growth curves constructed from the calculated constants u Ks and Sq are shown in Figure 4. max’ The regression of the specific growth rate data against the linear transformation of the threshold-corrected hyperbolic equation results in coefficient of determination values (r2) of 0.99 (n - 4), 0.97 (n - 5), 0.91 (n - 7), 0.88 (n - 5), 0.91 (n - 5) and 0.95 (n - 5) for temperatures of 21.5°, 25.0°, 27.5°, 30.0°, 30.0° and 35.0°C respectively. The high degree of correlation of the data to the threshold-corrected hyperbolic curve indicates that the enzyme-kinetic model is useful in quantitatively monitoring growth-substrate responses over a range of temperatures. The close comparison between the observed and predicted data throughout the 0-5% bw/d feeding range also emphasizes the accuracy of the model. This is especially significant since the feeding levels encountered in natural environments would usually fall within this range. These data from the model confirm the first of the two expected trends: growth at 139 each temperature is quantitatively related to substrate by the threshold-corrected hyperbolic curve. Despite the similarity in shape of the growth curves attributed to a common mechanism, there are still signifi- cant differences between actual specific growth rates found at different temperatures. Although both feed and tempera- ture were variables in these experiments, a similar sequence of feeding levels was employed for each series of tempera- ture experiments. Therefore, when growth response patterns are compared between each series of temperature experiments, the differences may be attributed to temperature. The in- troductory discussion of temperature leads to the expecta- tion that changes in temperature should affect both the supply of reactants and rates of the enzyme-catalyzed reactions. The enzyme-kinetic base of the model, in turn, leads to the expectation that changes in temperature should also affect the growth constants. The model has the advan- tage of facilitating comparison between growth curves by providing the three constants which define the curve and the ability to calculate growth corresponding to any feeding level. A comparison of the constants for the feeding trials reveals significant trends. The calculated values for "max indicate that at high feeding levels, the greatest growth occurred at 27.5°C and tapered off sharply at either higher or lower temperatures. Calculated "max values are plotted 140 as a function of feeding level in Figure 17, along with a curve described by the quadratic equation umax--0.l863+ 0.0181T-3.37-10'°T2. The um 2 ax data fit the quadratic equa- tion with an r of 0.77 (n - 5). From the equation, Optimal "max is predicted to occur at 26.8°C and zero growth at both 14 and 40°C. These data follow the same general trend for I. zillii reported by Platt and Hauser (1978). These calculations also were performed using the data of Brett et a1. (1969) on sockeye salmon. The authors hand fit growth curves to data obtained at 1°, 5°, 10°, 15° and 20°C. Maximum specific growth rates were then determined from the asymptotes to each curve. These data were fit with 3 1 a quadratic curve (umax-6.87-1o' +1.96-1o' T-6.73°10-3T2) with an r2 of 0.99 (n - 5). The Optimal "max for sockeye salmon determined from the equation was predicted at about 15°C with zero growth occurring at both 0°C and 29°C. Despite the excellent fit to the data, there is reason for caution in interpretation. Both the lower threshold and Optimal temperatures correspond well with previous research findings. However, the predicted upper temperature for growth threshold appears very high for sockeye salmon. Indeed, the authors had conducted an additional experiment at 24°C in which half the fish died and negative growth was reported. Therefore, caution should be taken in extending apparent trends beyond the experimental temperature range. Apparently the sockeye salmon are well adapted for cold MAXIMUM SPECIFIC GROWTH RATE (pmax) 0.06 - 0.04 .. 0.03 - 141 umax . -O.I863 +o.|s'r- 3.3746472 0.02 I6 ‘ 26 24 2s 32 36 .l A L J j 4 J TEMPERATURE (°C) Figure 17. Calculated “m values for T. zillii (average weight 6.3 g) fed trout Show at 21.5, 25, 27.5, 30 and 35°C and thezquadratic curve fit to the data (r - 0.77; n - 5). 142 water temperatures with their enzyme-catalyzed reactions capable of performing down to 0°C. However, at higher temperatures, somewhere between 20°C and 24°C, they exhibit a drastic loss of ability to process feed. Therefore, the literature data and the data from this study confirm the second expected trend, that growth acting as an integrator of the complex physiological processes would exhibit optimal rates at intermediate temperatures with decreased rates at both extremes. As temperature affects the supply rates of reactants, molecular activity, and enzyme efficiencies, both the rates of individual reactions and the integrated growth response (specific growth rate) shift similarly. The model provides not only the ability to quantitatively compare the "max at each ex- perimental temperature level but also to use the quadratic fit to predict "max for any temperature. Furthermore, the quadratic equation may be substituted for "ma in the orig- x inal growth equation as shown below: -4 2 5.8 u - (-0.186+0.18lT-3.37°10 T ) __‘1_ S+K -ZS 3 q Thus the model permits the determination of I. zillii growth at any given feeding level of trout chow, under the controlled conditions, partially corrected for any temper- ature within the experimental range, for that size class of fish. 143 Despite the importance associated with maximum growth rates, there are other factors involved in evaluating feeds and growth. The natural environment rarely provides suffi- cient feed levels to support maximum growth rates of fish. Therefore, such factors as maintenance costs and growth efficiency at low feeding levels are important criteria for assessing feed quality and fish growth. The threshold- corrected hyperbolic curve provides quantitative mechanisms for comparison at these low feeding levels. The maintenance feeding level of an organism, Sq, is useful because it identifies the amount of feed required to maintain the organism with zero growth. The Sq levels calculated for each temperature are summarized in Table 12 and plotted against temperature in Figure 18. The data points in Figure 18 are fit by the quadratic equation Sq- 4.9oe-o.3oer+5.18-1o'3 2 2 T with an r of 0.76 (n - 5). Mini- mum maintenance costs are predicted to occur at a tempera- ture of 29°C. As the temperature decreases from this point, it lowers the level of molecular activity, enzyme efficiency and diffusion rate. Therefore, the amount of substrate required to provide the maintenance level of reactants would increase. As temperature increased from the Optimal, it would be expected to increase molecular activity but de- crease enzyme efficiency and possibly cause structural disruption and denaturation. As a result, the amount of substrate required to supply a given unit of reactant would increase. Lab "61 L4 » l.2 - I.0 - 0.8 ~ 0.6' 0.4 - 0.2. MAINTENANCE FEEDING LEVEL (Sq) l6 Figure 6.3 g) 144 Sq- 4.906- O.306'l’+ 5.I8-I63 12 42‘0‘54 {8:312:56 TEMPERATURE ('c) _L 18. Calculated S values for T. zillii (average weight . fed trout chow a? 21.5, 25, 27.5,230 and 35°C and the quadratic curve fit to the data (r - 0.76; n - 5). 145 The importance of the Sq values is demonstrated in examining the growth curves in Figure 4. At a 3% feeding level, the fish at 21.5°C have the second highest growth rate. However, at a 0.9% feeding level, the same tempera- ture stimulates the lowest growth rate. The relatively rapid decrease in growth for fish at 21.5°C at low feed levels is caused by the high Sq of 0.77% bw/d. The elevated Sq at low temperatures suggest that T. zillii is poorly adapted to a cold water environment, eSpecially at low feeding levels. Because the Sq data may be fitted by a quadratic func-’ tion, the data may again be incorporated into the growth equation already temperature-corrected for "m as shown ax below: u I (-0.186+0.0181T-3.337°10-4T2) -32 S-(4.906-O.306 0+5. 18' 10'31‘2) )) +K8-2(4.906-0.3O6 T+5.18°10 T The model now permits the determination of I, zillii growth at any given level of trout chow, under the controlled con- ditions more fully corrected for any temperature. Brett et al. (1969) also reported on the maintenance feeding requirements of sockeye salmon over a temperature range of 1°-20°C. The S data of Brett et a1. (1969) can q be related to temperature by the exponential equation: (0.105'10T) 2 Sq - 0.321010e quadratic nature of the I. zillii Sq data, these cold water with an r of 1.00. Unlike the fish demonstrate maintenance costs which continue to 146 increase with rising temperature. At 20°C the fish have an Sq of approximately 2.6% body weight/day. The enzyme- controlled reactions involved with maintenance and growth are well-adapted to cold water situations and rapidly lose efficiency with rising temperature. The remaining constant used to define the growth curve is Ks. This constant is important in defining the slope of the curve, and the data collected in this study are summar- ized in Table 12. When the Ks values are plotted against temperature, the data fit the quadratic equation Ks - -15.066+1.318T-0.0216T2 2 , with an r of 0.87 (n - 5). This equation may now be fitted into the threshold-corrected hyperbolic equation to obtain the rather imposing equation shown below: -4 2 u a (-O.l86+0.018lT-3.37'10 T ) S-(4.906-0.306T+5.18-10-3T2) S+(-15.066+l.318T-0.0216T2)-2(4.906-0.306T+5.18'10-3T2) Thus, the model now is useful for predicting the specific growth rate opr. zillii for any level of trout chow, and at any temperature under the conditions of this study in the manner shown in Figure 19. To insure against adding to the complexity of the equa- tion, the discussion will quickly shift to another important consideration which justifiably receives much attention-- feed conversion efficiency. Feed costs are often a major , // 'il’ o...E////”“‘ \ o 02 7 \ 0.0: 7 I..‘\ \ i 4 M ' °‘Wo’ "°4§3\ ‘ /’.. " . °‘*‘ SPECIFIC GROWTH RATE (JJ) ‘5 e 0 . e C) 0 Ole Figure 19. The three-dimensional relationship between the specific growth rate response and both temperature and feeding level of trout chow for I. zillii (average weight 6.3 g). 148 financial consideration in aquaculture. Therefore, it is important to understand the efficiency with which feed is being converted into edible tissue. Frequently this process is quantified by a simple conversion factor calculated by dividing the final weight of the organism by the total weight of feed utilized. This conversion factor may be transformed to percent conversion efficiency by multiplying by 100. Conversion efficiency is determined both by the quality of the feed and the ability of the organism to assimilate the feed. Hopefully the considerations to this point have made it clear that measures of feed conversion are actually measures of the efficiency at which the enzyme-catalyzed reactions involved with growth process the feed. Therefore, it is anticipated that the threshold-corrected hyperbolic equation should be useful for providing a quantitative basis for determining feed conversion efficiency at any feeding level. Furthermore, it should be understood that before growth can occur, the maintenance costs of the organism (Sq) must be met. Thereafter, growth or feed conversion efficiency is a measure of the relative amount of feed re- quired for maintenance and that which is left as an excess available for growth. As a result, the feed conversion efficiency should fluctuate dependent on the feeding level, as a function of u S , and Ks which characterize the max’ q shape of the curve. Any variable factor which causes a 149 change in either u S or Ks will also cause a shift in max’ q conversion efficiency. Despite the emphasis placed on rapid growth, the high feeding rates necessary to produce such growth may not represent Optimal efficiency rates. It was previously reported that Kinne (1960) discovered that desert pupfish experienced superior growth at 30°C. However, when these same data are expressed in terms of weight gain per weight of feed, the fish reared at 20°C experienced the greatest conversion efficiency. The higher temperature stimulated higher ingestion and growth rates, but the digestion- assimilation process was reduced in efficiency. The conversion efficiencies were related to temperature in a pattern similar to growth since the efficiencies were reported to decrease with the following order of temper- atures: 20°, 15°, 25°, 30°, 36°C. Intermediate temper- atures stimulated the maximum feed conversion efficiency and decreased with both lower and higher temperatures. The author suggested that conversion efficiency is affected by age, heredity, environmental history of the fish and meta- bolic rate. Brett et a1. (1969) also determined the maximum effi- ciency feeding levels for sockeye salmon reared at different temperatures. These values were obtained by hand fitting a curve through growth data plotted against feeding level and drawing a tangent from the origin to the curve. The feeding 150 level corresponding to the tangent junction with the curve represented the maximum efficiency feeding rate. When these feeding levels of Optimal efficiency were plotted against temperature, the data points were fitted by the straight line equation: efficiency a 0.542+0.308T with an r2 of 0.99. The method Of calculation of efficiencies for this study was explained in the Materials and Methods section. The determination simply involves dividing the specific growth rate (u) by the feeding level (%S). If u is expres- sed as 1 and %S is expressed as Feed, then the Time Fish 18 efficiency units become F3335 thus giving a measure of fish' Time growth relative to both feed and time. Efficiency curves from each of the temperature experiments are plotted in Figure 20. Each of the curves has a similar shape with an abruptly ascending limb tapering to a maximum point followed by a gradual descending slope. The enzyme-kinetic consider- ation suggests that the initial sharp increase is due to the fact that at low feeding levels all the feed components will be efficiently diffused and processed through the enzyme- catalyzed reactions. Efficiencies will be negative until sufficient feed is provided at Sq to meet maintenance costs. As feeding levels continue to increase, the additional units are effi- ciently processed until competition begins at diffusion sites and the recycling enzymes cannot process the increas- ing supply as effectively. At yet higher feed levels, 151 0.00 .omunmkh .00 .040 um 32.0 2.95 0.00 00 To 0nw003 0w0u0>0v 000000 .H you 00>000 :03O0w 000000000m .ou 0usw0m >oom bzmomwm m 0 ¢ .0 N _ .¢0.0- 00.0. 0 .mod .I on r! .cod l/on .mod nNN 60.0 .0... 00| “(S/IT) AON3IOI333 152 growth is still increasing at this point, but because some feed components escape diffusion through the membranes, and the enzyme processing rate cannot increase as fast as the supply, the growth per unit feed increases more slowly. Eventually the trade-off between increased supply and less complete diffusion and diminished increase in reaction rate will result in a peak in efficiency. Thereafter, the enzyme-catalyzed reactions approach saturation and cannot respond as well to further increased supply, causing the reaction rate per unit feed to decrease. A comparison of the curves in Figure 20 and the three-‘ dimensional Figure 21 reveals that temperature has a pro- found effect on efficiency. The efficiencies demonstrated at 27.5°C are superior to all other feeding trials through- out the feeding range, and at times are almost twice that at 35°C. The changing Sq levels associated with different temperatures also have a dramatic effect. This relationship is most evident when examining the 21.5°C efficiency curve (Figure 20). At feeding levels above 2% bw/d, this feeding trial demonstrated the second highest efficiency levels. Yet because of the high maintenance costs of 0.77% bw/d at feeding levels below 1% bw/d, this trial experiences the lowest efficiency. The feeding levels stimulating the maximum conversion efficiency for all five temperatures lie within 1.5-2.1% bw/d. As feeding levels increase beyond this range, they 153 \\ / 1 \ x 0.4 / J C12 ’ / / EFFICIENCY (p/sHoo C) .0 \ \ Figure 21. The three-dimensional relationship between feed conversion efficiency and both temperature and feeding level of trout chow for I. zillii (average weight 6.3 g). 154 continue to stimulate faster growth but at lessened effi- ciency due to diminished diffusion-enzyme processing rates. The 27.5°C temperature for Optimal processing seems to represent a compromise between the Optimal temperatures for the three growth constants which are 26.8°C for "max’ 29.0°C O for Sq and 30 C for KS. SUMMARY The effective rearing of an organism requires the pro- vision of all feed components and environmental conditions necessary for growth. Whether the rearing involves an ex- tensive, natural feed situation or an intensive, artificial feeding program, the basic consideration is the same--to produce an organism by providing all the components neces- sary to stimulate growth in a cost-effective manner. Effec- tive management thus requires the following information: 1) the environmental requirements of the organism, 2) di- etary requirements of the organism, and 3) identification of the ability of the organism to harvest, digest and assimi- late the form of feed. In addition, there is usually a con- sideration of whether these conditions can be provided in a cost-effective manner. In seeking to effectively rear an organism, the over- whelming complexity necessitates some form of simplification to avoid confusion concerning the interactions and feedbacks in the system. And yet, to manage effectively requires some level of understanding concerning these processes and the relationship of the organism to the environment. Historically, a rather pragmatic approach has been taken in seeking to increase aquaculture productivity. The complexity of the natural system has resulted in heavy 155 156 reliance on empiricism wherein fish productivity is related to a single environmental parameter such as fertilization. These studies align closely with the "empirical ecology approach" prOposed by Rigler (1982). Rigler prOposed the empirical approach as a preferable, common ground for both theoretically-based limnologists and more practically-based fisheries managers. Rigler stated that when patterns in the behavior of the natural system are described by correla- tions, they are useful in establishing management practices. The literature supported the usefulness of correlations to increase fish productivity. For instance, correlations between fertilization and fish productivity have resulted in increased fish production. Unfortunately, such correlations have not led to the identification of the mechanisms and rates which stimulated these increases. It is ironic that Rigler, in attempting to unite the sciences of limnology and fisheries management under the banner of empiricism, found it necessary to discredit both. All too often different investigative techniques are ap- proached with an either/or attitude. Seldom are different methodologies discussed in an attempt to explore, combine and utilize the applicable elements of each. Rigler (1982), in summarizing the usefulness of the empirical approach, states that "a really useful and interesting correlation inevitably stimulates some of us to move on to the next phase of science--that of replacing the empirical theory 157 with an explanatory theory." Thus, while degrading the attempts of the theoretical ecolOgists to develoP models to explain and predict the workings of the natural system, Rigler ignores that the ultimate goal of each is the same--a desire to understand the natural system. The distinction between the approaches is not the goal but rather the means to a common goal. Admittedly, correlations by themselves are inadequate to quantitatively relate growth to particular environmental resources or to explain the basis of the Ob- served patterns. Following consideration of the strengths and weaknesses of the empirical approach, it is advantageous to question the necessity, or possibility, of identifying the mechanisms and rates by which growth is related to environmental param- eters. If these relationships could be quantified, it would unquestionably enhance the understanding of the natural sys- tem, which was expressed as a common goal of all approaches. Therefore, the issue becomes a singular question of whether or not it is possible to quantitatively define the relation- ships. The practicality of relating growth to environmental resources was explored by reviewing previous theoretical approaches to modeling growth. This review revealed that the "logistic" model had initiated a consideration of rate Of growth during different phases and that the "enzyme- kinetic" model had incorporated the capacity to 158 quantitatively relate specific growth rate to individual environmental parameters. The enzyme-kinetic model was subsequently altered to incorporate an allowance for energy-feed component expenditures associated with basal metabolism. The threshold-corrected hyperbolic equation is intended as a research tool which includes aspects of both theoret- ical models and correlations to achieve the goal of under- standing and prediction of the natural system. In addition to providing insight into the rates and mechanism of growth, it is anticipated that this research methodology will pro- vide Opportunity to recognize the commonalities inherent in the disciplines of limnology and fisheries management. Kitaka (1972) expressed the belief that the fields of geology, chemistry, physics and biOlOgy (including limnology and fisheries management) are all interdependent. The author stated that "limnological factors have also been shown to influence feeding, growth rates and survival of fish." As the growth rate of fish is shown to be dependent on environmental resources, the interdependence of limnology and fisheries management is made evident. Because the model has both empirical and theoretical elements, the appearance of a mutual exclusiveness between these modeling approaches is also avoided. The prOposed growth model, provides a research method- ology which, on the one hand simplifies the system, yet also 159 facilitates the quantitative relating of growth to feed, en- vironmental variables and digestive abilities of the organ- ism. Simplification is achieved by allowing the organism to become an integrator of its internal biochemical-biOphysical complexity. While the individual components are not quanti- tatively examined by the model, the integrated response to feed and environmental variables is monitored and related directly to these factors. The threshold-corrected hyperbolic equation was first examined to determine its usefulness in quantifying the effects of the amount and type of different feeds on growth. Controlled experiments were conducted to quantitatively relate the growth rate of I. zillii to three feeds: commer- cial trout chow, pelleted Elodea canadensis, and pelleted Spirulina. It was demonstrated that the threshold-corrected hyperbolic equation and the constants which define it (”max’ Ks and Sq) could be used to effectively predict the fish growth response to each feed. These quantitative relation- ships were also shown effective in relating fish growth to the protein and energy content of each feed. The growth response of E. niloticus to both trout chow and pelleted Spirulina also was quantitatively defined by the model. The determination of growth constants for each organism on similar feeds permits comparisons between the species and the relating of differences to the unique mor- phology Of each species. 160 The effectiveness of the model was next investigated concerning its ability to quantitatively define the rela- tionship between growth-substrate and size. Controlled experiments were conducted using various sizes of I. zillii. It was determined that for each size class of fish, the threshold-corrected hyperbolic curve accurately described the growth-substrate relationship. As the size of the fish changed, the functions of the growth equation (u Ks and max’ Sq) also changed, and these functions were related to size by either logarithmic, quadratic or exponential equations. When these relationships (equations) were incorporated into' the growth equation, it was useful for predicting the growth response of fish at any size and at any feeding level under the controlled conditions. The model also was examined relative to its usefulness in quantifying the effects of temperature on growth. Based on controlled experiments conducted on Tilapia zillii with temperature as a variable, it was demonstrated that growth could be quantitatively related to substrate at each temper- ature. Furthermore, each component of the growth curve (u temperature. These relationships could be described by K Sq) was shown to change as a variable relative to max’ 8’ either a quadratic or a power function formula. When these temperature dependent factors were incorporated together in the growth formula, the equation became useful for predict- ing the effect of temperature on the growth of I. zillii at 161 any feeding level throughout a broad temperature range. The practicality of the model was thus demonstrated in accur- ately quantifying a growth-substrate-temperature relation- ship in agreement with theoretical expectations and previous research findings. These findings demonstrate that use of the threshold- corrected hyperbolic equation is effective in quantifying the effect of environmental and feed variables on growth. The experiments involving both feed types and species relate to the ability to identify dietary requirements Of an organ- ism and the ability to utilize the feed. Growth curves con- structed using the model were shown to accurately reflect the growth response of 1. zillii to different feeds. The constants defining the curves could then be compared and related to protein and energy content of the feeds. Other components (protein, essential amino acids, vitamins, etc.) could similarly be used as the variable in such experiments and the optimal concentration of each determined from the growth response. Optimal species-specific feed formulations could thus be developed based on the integrated ability of the fish to digest and assimilate the feed under non- stressed conditions. Furthermore, similar studies could be conducted on various forms of "natural" feeds. By comparing the growth curves produced when feed is presented to the organism in different forms, conclusions may be drawn con- cerning the organism's abilty to harvest, digest and assimi- late the feed. 162 The temperature experiments demonstrated the model to be useful in quantitatively defining the effect of an envi- ronmental variable on growth. Similar controlled studies using other variable factors such as ammonia concentrations, dissolved oxygen levels, pH, concentrations of Specific toxins, etc. could be conducted to similarly identify the environmental range for individual species concerning these factors. Through a series of such experiments, the Optimal environmental conditions for any species could be identi- fied. The quantitative nature of the model also facilitates the determination of an additional critical dimension of growth--conversion efficiency. Because the model permitted the calculation of growth rates corresponding to any feeding level, the growth per unit of feed also could be calculated for the entire feeding range. These data demonstrate the efficiency with which each unit of feed is converted into fish tissue. Because the profit motive underlies most aqua- culture Operations, these conversion efficiencies are of critical importance. Profit margins of any animal husbandry Operation are basically a function of sale price and cost of production with feed being a major production expense. Initially, when examining the growth curves, it might appear that since growth continues to increase with higher feeding levels, the fish should always receive a high 163 feeding level which would stimulate near maximal growth. However, when examining any of the efficiency curves, it is evident that at elevated feeding levels, the ability of each unit of feed to stimulate growth is decreased. Optimal efficiency feeding levels for both species under all condi- tions examined ranged from 1.6 to 3.2% bw/d. At these feeding rates the maintenance costs of the organism are provided and additional feed remains to stimulate growth. This additional feed is efficiently processed by the organ- ism and converted to tissue. When feed levels are increased beyond this maximal efficiency point, additional feed is available for growth but each unit of feed is less effi- ciently diffused and processed through the system of enzyme- catalyzed reactions. As a result, the economically Optimal feeding rate becomes a complex consideration of balancing the market price Of the organism against the feeding costs at different efficiencies. The threshold-corrected hyperbolic equation also has useful application to other considerations important in aquaculture. The model was demonstrated to facilitate the prediction of Specific growth rates of organisms of various sizes, at different feeding rates of specific feeds. These data permit the prediction of weight of the organism with time. Therefore, the time to reach marketable Size at vari- ous feeding levels, and different efficiencies could be cal- culated. A decrease in feeding level might increase the 164 efficiency of feed utilization but would increase the time to harvest due to a decreased growth rate. These calcula- tions, along with considerations of feed cost and market value, would permit the determination of cost-effective feeding levels. Another factor, indirectly involved in all considera- tions to this point, is the stocking level Of fish. The level of stocking, along with the amount of feed added, determines the percent body weight of feed available to the fish. The prOposed model would permit the calculation of stocking levels which would result in Optimal, cost-effec- tive, feeding levels and allow for changes in the size of the fish. Extensive empirical research has been useful in estab- lishing effective feeding tables for hatching Operations. For example, Leitritz and Lewis (1980) compiled feeding tables for rainbow trout based on the recognition that the feeding ration for optimal growth is affected by both size and temperature. The threshold-corrected hyperbolic equa- tion has a two-fold advantage over the empirical process. First, the model provides insight into the underlying mechanisms of growth and how it is influenced by numerous variables. Secondly, through inexpensive, short-term ex- perimental procedures which avoid stress to the fish, the model facilitates the relating of specific growth rate and conversion efficiency to a broad range of feeding levels and 165 provides the maintenance ration level. The potential usefulness of the model also applies to natural feeding situations. Here the growth of the fish cannot be related to known increments of external feed. Instead, growth must somehow be correlated with the amount of natural feeds produced in the environment. The produc- tivity of natural waters has frequently been estimated through correlations to biotic and abiotic variables. Un— fortunately, these correlations only provide rough esti- mates of total productivity and are not capable of quanti- fying the production of Specific feed types. The growth model has demonstrated the ability to accurately predict fish growth relative to particular levels of various feed types. If the availability of Specific natural feeds could be determined, then the growth of the fish could not only be predicted but also predetermined by controlling the number and size of fish in the environment. This predictive capacity for fish growth requires the identification of both standing crap and growth rate of the natural feed. This information could be obtained by utilizing the growth model to relate the growth of natural feeds to environmental parameters. The quantification of growth of natural feeds becomes the basis for predicting the growth of the fish. In each case, growth is based on the dietary requirements of the organism, the existing environ- mental conditions, and the ability of the organism to util- ize the particular substrate for growth. 166 In summary, the threshold-corrected hyperbolic equation has the following potential application: 1) identify the dietary requirements of organisms, 2) quantify the effect of environmental variables on growth, 3) quantify the ability of the organism to harvest-digest-assimilate feeds, 4) pro- vide a basis for cost-benefit considerations by identifying feed conversion efficiencies, and 5) provide growth data which relates to stocking rate and harvest time. These applications are also possible in diverse situa- tions due to the inexpensive, uncomplicated nature of the analytical technique. The basic data acquisition requires only a series of aquaria, an accurate balance to weigh the fish and feed, and routine laboratory equipment. More SOphisticated analyses of protein content, vitamins, amino acids, etc. are desirable to enhance data interpretation. However, the basic procedure of quantitatively relating the Specific growth response of fish to substrate concentrations under controlled conditions provides important data identi- fying Optimal growth conditions. As a result, this research technique has practical application for enhancing aqua- culture in remote areas and lesser develOped countries. While this research has been limited to aquaculture applications, the potential scope is much broader. The growth model is based on principals of diffusion mechanisms, the first and second laws of thermodynamics and enzyme- catalyzed reactions. These underlying principles are 167 believed to function similarly in most organisms. There- fore, this model may provide a basis for quantitatively defining growth processes in other natural systems. BIBLIOGRAPHY Almazan, G. and C.E. Boyd. 1978. Plankton production and Tilapia yield in ponds. Aquaculture 15:75-77. Atlas, R.M. and R. Bartha. 1981. Microbial Ecology. Addison-Wesley Publishing CO., Reading, MA. 560 pp. American Public Health Association. 1980. Standard Methods for the Examination of water and wastewater. Amer. PubIT Health Assoc., Inc. 15th ed., New York. Bertalanffy, L. von. 1957. Quantitative laws in metabolism and growth. Quart. Rev. Biol. 32(3):217-231. Bilton, H.T. and G.L. Robins. 1973. The effects of starvation and subsequent feeding on survival and growth of Fulton Channel sockeye salmon fry. J. Fish. Res. Bd. Canada 30(1):1-5. Bowen, S.H. 1982. Feeding, digestion and growth-qualitative considerations. Pp 141-156. In Pullin, R.S.V. and R.H. Lowe-McConnell (eds). The biology and culture of tilapias. ICLARM Conference Proceedings.'International Center for Living Aquatic Resources Management. Manila, Philippines. Boyd, C.E. and C.P. Goodyear. 1971. Nutritive quality of food in ecological systems. Arch. Hydrobiol. 69(2):256-270. Brett, J.R. 1967. Swimming performance of sockeye salmon (Oncorhynchus nerka) in relation to fati ue time and temperature. J. Fish. Res. Bd. Canada 24%8):173l-1741. . 1974. Tank experiments on the culture of pan- ?ize sockeye §Oncorhynchus nerka) and pink salmon Q. gorbuscha using envifonmental control. Aquaculture 4:34 " o . 1979. Environmental factors and growth. In Fish Physiology, vol. 8, ch. 10. 599-675. Academic Press Inc. 168 169 Brett, J.R. and J.E. Shelbourn. 1975. Growth rate of young sockeye salmon, Oncorhynchus nerka, in relation to fish size and ration level. J. Fish. Res. Bd. Canada 32(11): 2103-2110. Brett, J.R., J.E. Shelbourn and C.T. Shoop. 1969. Growth rate and body composition of fingerling sockeye salmon, Oncorhynchus nerka, in relation to temperature and ration Size. 3. Fish. Res. Bd. Canada 26(9):2363-2394. Brooks, J.L. 1950. Speciation in ancient lakes. Quarterly Review of Biology 25:30-44; 148-176. Buddington, R.R. 1979. Digestion of an aquatic macrOphyte by Tilapia zillii (Gervais). J. Fish. Biol. 15:449-455. Burns, R.P. and R.R. Stickney. 1980. Growth of Tila ia aurea in ponds receiving poultry wastes. Aquaculture 20: 117-121. Caulton, M.S. 1977. The effect of temperature on routine metabolism in Tila ia rendalli boulenger. J. Fish. Biol. 11:549-5 . . 1978. The effect of temperature and mass on routine metabolism in Sarotherodon (Tila ia) mossambicus (Peters). J. Fish Biol. 13:195-281. . 1982. Feeding, metabolism and growth of Tilapias: some quantitative considerations. In R.S.V. Pullin and R.H. Lowe-McConnell (eds). The biology and culture of tilapias. ICLARM ConferenceiProceedings 7, 432 pp. International Center for Living Aquatic Resource Management. Manila, Philippines. ' Chervinski, J. and M. Lahav. 1976. The effect of exposure to low temperature on fingerlings of local Tilapia (Tila ia aurea) (Steindachner) and imported Tilapia (Tila ia vuIcani) (Trewavas) and Tila ia nilotica Einne) :25-39. ( in Israel. Bamidgeh 28(1) Chesson, P. 1978. Predator-prey theory and variability. Ann. Rev. of Ecol. and Systematics. 9:323-348. Chimits, P. 1957. The tilapias and their culture. FAO Fisheries Bulletin lO(l):l-24. Cho, C.Y., H.S. Bayley and S.J. Slinger. 1974. Partial replacement of herring meal with soybean meal and other changes in a diet for rainbow trout (Salmo gairdneri). J. Fish. Res. Bd. Can. 31(9):1523-1528. 170 Davis, A.T. and R.R. Stickney. 1978. Growth responses of Tilapia aurea to dietary protein quality and quantity. Trans. Am. Fish Soc. 107(3):479-483. Dendy, J.S., V. Varikul, K. Sumawidjaja and M. Potaros. 1968. Production of Tilapia mossambica Peters, plankton and benthos as parameters for evaluating nitrogen in pond fertilizers. F.A.O. Fish. Rep. 44(1):226-240. Denzer, H.W. 1968. Studies on the ph siology of young Tilapia. F.A.O. Fish. Rep. 44(4 :357-366. Dixon, W.J. 1983. BMDP Statistical Software. University of California Press, Berkeley, CA. 733 pp. Dowd, J.E. and D.S. Riggs. 1965. A comparison of estimates of Michaelis-Menton kinetic constants from various linear transformations. J. Biol. Chem. 240:863-869. Drenner, R.W., G.L. Vinyard, M. Gophen and S.R. McComas. 1982. Feeding behavior of the cichlid, Sarotherodon alilaeum: selective predation on Lake Kinneret zooplanEton. Hydrobiologig 87:17-20. Dupree, H.K. and K.E. Sneed. 1966. Response of channel catfish fingerlin s to different levels of major nutrientslin puri ied diets. Tech. Paper 9. Bureau of Sport Fisheries and Wildlife, U.S. Gov. Printing Office. Washington, D.C. 21 pp. Dwyer, W.P., R.G. Piper and C.E. Smith. 1983. Brook trout growth efficiency as affected by temperature. Prog. Fish Cult. 45(3):161-163. Edwards, P., Sinchumpasak, O. and M. Tabucanon. 1981. The harvest of microalgae from the effluent of a sewage fed high rate stabilization pond by Tilapia nilotica. Part 2: Studies of the fish ponds. quaculture 23:107-147. Elliott, J.M. 1976. The energetics of feeding, metabolism and growth of brown trout (Salmo trutta L.) in relation to body weight, water temperature and ration size. J. Animal Ecol. 45:923-948. Engel, P.C. 1977. Enzyme Kinetics: The Steady State Approach. John Wiley and Sons, New York. Florey, E. 1969. An introduction to_genera1 and comparative animal physiology. W.B. Saunders Co. Philadelphia, PA. 171 Fry, F.E.J. 1971. The effect of environmental factors on the physiology of fish. In Fish Physiology; W.S. Hoar and D.J. Randall, eds. Academic Press, N. . Fryer, G. and T.D. Iles. 1972. The cichlid fishes of the Great Lakes of Africa. Oliver & Boyd, Edinburgh. 641 PP. Garling, D.L., Jr. and R.P. Wilson. 1976. Optimum dietary protein to energy ratio for channel catfish finger- lings, Ictalurus punctatus. J. Nutr. 106:1368-1375. Gerking, S.H. 1971. Influence of rate of feeding and body weight on protein metabolism of bluegill sunfish. PhiSiOlo 20°10 44(1):9-190 Goodhart, R.S. and M.E. Shils. 1980. Modern nutrition in health and disease. Lea and Febiger, Philadelphia, PA. 1370 pp. Greenland, D.C. and R.L. Gill. 1979. Multiple daily feedings with automatic feeders improve growth and feed conver- sion rates of channel catfish. Prog. Fish Cult. 41:3, 151-153. Greenwood, P.H. 1953. Feeding mechanism of the cichlid fish, Tilapia esculenta. Nature 172:207-208. . 1964. Explosive Speciation in African lakes. Proc. R. Inst. Gt. Br. 40:256-269. . 1973. Morphology, endemism and speciation in African cichlid fishes. Verh. Deutsch. Zool. Gesellsch. 115-124. Harbott, B.J. 1975. Preliminary observations on the feeding of Tilapia nilotica Linn. in Lake Rudolf. Afr. J. Trop. HydPobiol. & Fish 4(1):27-37. Hayes, F.R. and E.H. Anthony. 1964. Productive capacity of North American lakes as related to the quantity and the trOphic level of fish, the lake dimensions, and the water chemistry. Trans. Amer. Fish. Soc. 93:53-57. Hill, R.W. 1976. Comparative physiology of animals, an environmental approach. Harper & Row Pdhlishers. New York, N.Y. Hixon, M. 1980. Competitive interactions between California reef fishes of the genus Embiotoca. Ecology 61(4): 918-931. 172 Hoar, W.S. 1979. Regulatory Mechanisms, Chapter 2, in General and Comparative Physiology. Huisman, E.A. and P. Valentijn. 1981. Conversion efficien- cies in grass carp (Ctenopharyn eal idella, Val.) using a feed for commercial production. quaculture 22: 279-2900 Jobling, M. and P.S. Davies. 1980. Effects of feeding on metabolic rate, and the Specific dynamic action in plaice, Pleuronectes platessa L. J. Fish Biol. 16:629-638. Kingsland, S. 1982. The refractory model: the logistic curve and the history of pOpulation ecology. anrt. Rev. 8101. 57:29-52. Kinne, O. 1960. Growth, food intake, and food conversion in a euryplastic fish exposed to different temperatures and salinities. Physiol. Zool. 33:288-317. Kitaka, G.E.B. 1972. The relevance of limnological informa- tion in the develOpment and management of inland fisheries. Afr. J. TrOp. Hydrobiol. Fish. 2:77-85. Kutty, M.N. and N. Sukumaran. 1975. Influence of upper and lower temperature extremes on the swimming performance of Tilapia mossambica. Trans. Am. Fish. Soc. 4:7 - o Leitritz, E. and R.G. Lewis. 1980. Trout and Salmon Culture (Hatchery Methods). Calif. Dept. Fish and Game, Fish Bull. 164. 197 pp. LeRoux, P.J. 1956. Feeding habits of the young of four species of tilapia. South Afr. J. Science Sept. 33-37. Liang, Y., J.M. Melack and J. Wang. 1981. Primary production and fish yields in Chinese ponds and lakes. Trans. Am. Fish Soc. 110:346-350. Liem, K.F. 1974. Evolutionary strategies and morphological innovations: cichlid pharyngeal jaws. Systematic Zoology 22:425-441. MacArthur, R.H. and E.O.Wilson. 1967. The_theory of island biogeography. Princeton Univ. Press. Princeton N.J. 173 Mazid, M.A., Y. Tanaka, T. Katayama, M.A. Rahman, K.L. Simpson and C.O. Chichester. 1979. Growth response of Tila ia zillii fingerlings fed iso-caloric diets with variaEle protein levels. Aquaculture 18:115-122. Melack, J.M. 1976. Primary productivity and fish yields in trOpical lakes. Trans. Am. Fish. Soc. 105(5):575-580. Monod, J. 1949. The growth of bacterial cultures. A. Rev. Microbiol. 3, 371-394. Moriarty, D.J.W. 1973. The physiology of digestion of blue- green algae in the cichlid fish, Tilapia nilotica. Journal of Zoology 171:25-39. . 1976. Quantitative studies on bacteria and algae in the food of the mullet Mu i1 cephalus L. and the prawn Meta enaeus bennettae (Race - . EXp. Mar. Biol. Ecol. 22:131:143. Moriarty, D.J.W. and C.M. Moriarty. 1973. The assimilation of carbon from phytoplankton by two herbivorous fishes: Tila ia nilotica and Haplochromis nigripinnes. Journal of Zoology 17l:4l-55. National Research Council. 1983. Nutrient requirements Of warmwater fishes and shellfishes. National Academy Press, Washington, D.C. 102 pp. Niimi, A.J. and F.W.H. Beamish. 1974. Bioenergetics and growth of largemouth bass (Micropterus salmoides) in relation to body weight and temperature. Can. J. Zool. 52:447-456. Paloheimo, J.E. and L.M. Dickie. 1965. Food and growth of fishes. I. A growth curve derived from experimental data. J. Fish. Res. Bd. Canada 22(2):521-542. Payne, A.I. 1978. Gut pH and digestive strategies in estuarine rey mullet (Mugilidae) and tilapia (Cichlidael. J. Fish Biol. 13:627-629. Peters, D.S. and M.T. Boyd. 1972. The effect of temperature, salinity, and availability of food on the feeding and growth of the hogchoker, Trinectes maculatus (Bloch & Schneider). J. Exp. Mar. Bidl. Ecol. 7:201-207. Philippart, J-Cl. and J-Cl. Ruwet. 1982. Ecology and distri- bution of tilapias. Pp. 15-59. In R.S.V. Pullin and R.H. Lowe-McConnell (eds). The biology and culture of tilapias. ICLARM Conference Proceedings. International Center for Living Aquatic Resources Management. Manila, Philippines. 174 Platt, S. and W.J. Hauser. 1978. Optimum temperature for feeding and growth of Tilapia zillii. Prog. Fish Cult. 40(3):105-107. Prejs, A. and M. Blaszczyk. 1977. Relationship between food and cellulose activity in freshwater fishes. J. Fish Biol. (11):447-452. Reich, K. 1978. A possible correlation between algal number and S. galilaeus yield in the Kinneret. Bamidgeh 30(47:1 - o Rigler, E.H. 1982. The relation between fisheries management and limnology. Trans. Am. Fish. Soc. 111(2):121-130. Ryder, R.A. 1982. The morphoedaphic index--use, abuse and fundamental concepts. Trans. Am. Fish. Soc. 111:154-164. Schroeder, G.L. 1974. Use of fluid cowshed manure in fish ponds. Bamidgeh 26:84-96. . 1978. AutotrOphic and heterotrOphic production of microorganisms in intensely-manured fish ponds, and related fish yields. Aquaculture 14:303-325. Seymour, E.A. 1980. The effects and control of algal blooms in fish ponds. Aquaculture 19:55-74. Smit, H. 1967. Influence of temperature on the rate of gastric juice secretion in the brown bullhead, Ictalurus nebulosus. Comp. Biochem. Physiol. IITI253I32. Smith, R.R. 1971. A method for measuring digestibility and metabolizable energy of fish feeds. Prvg. Fish-Cult. 33(3):132-134. Solomons, T.W.G. 1976. Organic Chemistry. John Wiley & Sons, Inc. New York, N.Y. 1056 pp. Solorzano, L. 1969. Determination of ammonia in natural waters by the phenolhypochlorite method. Limnol. Oceanogr. 14:799-801. Stickney, R.R., J.H. Hesby, R.B. McGeachin and W.A. Isbell. 1979. Growth of Tilapia nilotica in ponds with differing histories of organic fertilization. Aquaculture 17:189-194. 175 Stickney, R.R. and S.E. Shumway. 1974. Occurrences of cellulose activity in the stomachs of fishes. J. Fish Biol. 6:779-790. Swingle, H.S. and E.V. Smith. 1939. Fertilizers for increasing the natural food for fish in ponds. Trans. Am. FiSho SOC. 68:126-135. APPENDIX A 176 Weights and lengths of Tila ia zillii fed trout chow at 21. Table A-1. 1% Feed Week 2 Week 1 Initial CID gm cm gm C111 gm Fish # 2Aw45 3233 7509 5565 0123 12.7 Total Weight 2% Feed Week 2 cm Week 1 Initial C111 cm gm gm gm Fish # 0123 14.4 Total Weight 3% Feed Week 2 Week 1 Initial cm gm C111 gm CID gm Fish # 50Rw9 0 L433 1 3090. 665& 0.525 2 4.333 4 1887 6555 05.089 3322 12.2 0666 6555 0123 Total Weight 4% Feed Week 2 Week 1 Initial CID gm Discontinued CID gm CID gm Fish # 4.874 3]..”43 6449 5665 7838 234.9“ 3226 5665 0123 3 6 13.6 Total Weight Weights and lengths of Tila ia zillii fed trout chow at 25°C. Table A-2. 0% Feed Week 5 cm Week 4 cm Week 3 cm Week 2 cm Week 1 cm Initial cm gm gm gm gm gm gm Fish # ”war-4 o o o o HNNQ 0005 O 0 W000 OOON JMMQ 000': 0000 amuse-4 O Nmmm 0900 O O 0000 OHNM TOtal Weight In 11.6 14.2 1% Feed Week 5 cm Week 4 cm Week 3 cm Week 2 cm Week 1 cm Initial gm gm gm gm gm gm Fish # oorxow O C O mmqm QHMN ouch NOO‘O mmmo H®MH . O O coon mmom O 0 O O mqmm ”\ONH o U300“ NMQO qum €¢m0 O 0 W000 OHNM 16.0 Total Weight 2% Feed Week 5 cm Week 4 cm Week 3 cm Week 2 Week 1 cm n1t1a gm gm cm gm gm gm gm cm Fish # N—cmnn 0000‘ Cho-CM NON® BN0"? 0 l-fid’0O‘ 0006! N0ng MN¢0 O l-fifl'lfiw QNO‘H O O 0 000m Ov-INM ._q B N 19.9 21.8 23.5 . 17.4 Total Weight Table A-2 (cont'd.). 3% Feed Week 5 cm Week 4 cm Week 3 cm Week 2 cm Week 1 Initial cm cm gm gm gm gm gm gm Fish # O‘WHDF\ QWD< 0000.0 0000.0 0000.0 0000.0 0000.0 owmuo>< 0000.0: N000.0: 0000.0: 0000.0: 0000.0: owmum>< Aokoo.ov 0000.0 0500.0 5HH0.0 0N~0.0 m xmmz 0500.0 moao.o o~oo.o meoo.o oo.o m xmoz Naoo.o- oo.o omoo.o- oo.o oo.o m xmma 0~H0.0 NHH0.0 H000.0 HOH0.0 ¢0~0.0 q xmmz 0000.0 0000.0: 0000.0 50H0.0 0000.0 4 xmmz 0000.0: 0000.0: 0000.0: 0000.0: 5500.0: 4 gum: 50H0.0 0000.0 N000.0 HOH0.0 NHH0.0 m xmm: 0000.0 < m N H 0 4 3.50 60mm NN wwwuo>< m N a 0 s swam vwmm NH wwwum>< m N H 0 4 swam comm N0 .oomu um 30:0 uaouu vow wflaafln camoawh mo mmumu nuaouw 00000000 .~-m manna 201 .mmmwum>w HHmum>o :H vmvsHocH no: muo wuwv muons .qumm 05 OH HO...— U0>HOQQO mw3 UQQW HHw cos: mxwma wcHusv vmuuzooo :oHs3 moan“ nuzouw uammwuamu mummnucwuwa :H moustm H000.0 000.0 0H0.0 500.0 000.0 wwmum>< 0~N0.0 00N0.0 00H0.0 m5~0.0 00N0.0 owmum>< 50~H0.0v 0000.0 50N0.0 0000.0 05H0.0 n 3mm: AoHHo.OV 00H0.0 50H0.0 0000.0 NNH0.0 m xwmz HmoHo.OV oo.o maoo.o 00N0.0 0N00.0 0 x0»: HomHo.00 HOH0.0 00H0.0 HON0.0 0000.0 0 xmms HmNHo.ov oo.c HwHo.o 05~0.0 0000.0 m 3mm: HNHHo.OV 0HH0.0 HOH0.0 0000.0 00H0.0 m xmm: HemHo.ov NHo.o mHo.o- a~o.o H~o.o N mam: 00H0.0 5Ho.0 5H0.0 NN0.0 0N0.0 N gum: Hw~o.o mNo.o «Ho.o “mo.o mmo.o H 3mm: m0~0.0 000.0 000.0 000.0 000.0 H gum: wwwum>< 0 N H 0 0 :mHm vwmm N0 wwwum>< .A.v.u:oov ~-m «Hang 202 Heoo.o oqoo.o aNoo.o mooo.o HNoo.o mwmum>< HHoo.o- oooo.o mqoo.o- HNoo.o: HNoo.o H 0woo.o Hmoo.o omoo.o quo.o ONoo.o: o omnum>< 0 xmos N gm»: N xmmz H gum: 0 :mHm cmmm NN.o «Hoo.o oooo.o NHoo.o mHoo.o aHoo.o mwmum>< aNoo.o Nmoo.o mmoo.o «oHo.o NHHo.o H mmoo.o- mmoo.o- NHoo.o- enoo.o- Nqoo.o: o mwaum>< 0 gums m 30m: N mum: H xmmz 0 :mHm cmom Nm.o oNoo.o- omoo.o- mNoo.o- aooo.o- oNoo.o- mwmum>< mooo.o- 0Noo.o- oooo.o oooo.o oooo.o H meoo.o- omoo.o- «moo.o- NHoo.o- Hmoo.o- o mm¢pm>< 0 xmmz m xmma N xmmz H 3mm: 0 smHm comm NN.o Nmoc.o: Nmoo.o- 0Noo.o- omoo.o- HoHo.o- wwwum>< Nwoo.o: Nmoo.o- Haoo.o- omoo.o- NHHo.o- 0 mmoo.o- mNoo.o- oooo.o Nqoo.o- Hmoo.o- N omoo.o- HHoo.o- oHoo.o- HNoo.o- omoo.o- N MNoo.o- omoo.o- oqoo.o- ”Noo.o- quo.o- H coco.o- mNoo.o- «Noo.o- oqoo.o- maoo.o- o mwaum>< 0 3mm: m mam: N xmo: H 3603 0 :mHm ummm No .oom.5~ um 30:0 uaouu 0mm HHHHHN mHmmHHH mo moumu :u3ou0 UHMHomam .0:0 mHan .mowmum>m HHmum>o :H wovaHocH uo: mum wumv mmone .cwumm on cu no: vm>uomno was vmmw HHm can: mxoo3 wcHusv vmuusooo noH£3 moumu nuzouw ucommumwu mmmmnuomumm :H moustm 203 oNNo.o ANNNo.OV AooHo.ov ANoHo.o0 oNNo.o owauu>< : : - : : N NNNo.o NNNo.o NoNo.o omHo.o NmNo.o N oeNo.o NHNo.o oNHo.o oNNo.o eqNo.o H NNNo.o oNNo.o NHoo.o Nooo.o NNNo.o o awmum>< e xams N gum: N xmoz H xmmz 0 smHm cmmm No.0 NqNo.o AHNHo.ov HNNo.o 0NNo.o HNNo.o owmum>< : : : - - N NONo.o «NHo.o oNNo.o «ONo.o oNHo.o N «NNo.o NONo.o HNNo.o NHNo.o NNNo.o H quo.o «HNo.o NoNo.o NqNo.o HNHo.o o mwmum>< 0 xmmz N xwms N gum: H 3mm: 0 anm ummm No.N NNHo.o ONHo.o NNHo.o oNHo.o NeHo.o owe~o>< NoHo.o NNHo.o NoNo.o quo.o oHoo.o N ooHo.o NoHo.o mqoo.o NHHo.o NHNo.o N HNHo.o «Noo.o: HNHo.o NoNo.o HHNo.o H HNoo.o oHNo.o oooo.o oooo.o NNoo.o- o ammuo>< 0 xow: N wi2 N gum: H sum; 0 :mHm nmam No.N .A.v.u:oov 0:0 oHnt 204 0HH0.0 HNH0.0 05H0.0 0500.0 0000.0: 00000>< 0000.0 5000.0: 0H00.0: HOH0.0 H000.0 00000>< 0000.0: 0000.0: 0000.0: 0500.0: 0000.0: 00000>< 0NH0.0 omNo.o ooHo.o o.o NNoo.o- N x003 0000.0 0.0 00H0.0: 00H0.0 NOH0.0 N Nam: 0000.0: 0000.0: 0000.0: 5N00.0: 0N00.0: N 0003 HHH0.0 HNH0.0 HOH0.0 0H00.0 0000.0: 0 x00: 0000.0 NHoo.o «Noo.o o.o ooHo.o 0 0003 N000.0: 0H00.0: 0000.0: HOH0.0: 0000.0: 0 0003 0000.0 NNoo.o NNHo.o NNoo.o o.o N Nam: 0000.0 0H00.0 H500.0: 5000.0 00H0.0 N xmm: 0000.0: N000.0: 0.0 5000.0: 5N00.0: N 0002 00H0.0 0000.0 55H0.0 0NH0.0 0000.0: N 3003 N000.0 5000.0 0N00.0 00H0.0 0N00.0: N 3003 0000.0: H000.0: 0000.0: 0000.0: N000.0: N x003 00H0.0 0000.0 00N0.0 00H0.0 5000.0: H 0003 0000.0 00H0.0: 0000.0 0N00.0 NOH0.0: H gum: 0000.0: H000.0: N000.0: 50H0.0: 0000.0: H No»: 00000>< 0 N H 0 N NNHN U000 NN 00000>< 0 N H 0 0 anm 0000 NH 00000><. 0 N H 0 0 :NHm 0000 N0 .Uo00 00 koso 0:000 000 HHHHHN 0Hm0HHH no 00000 nuzouw onHo000 .0-N «Hams 205 .000000>0 HH000>o c0 0000H000 no: 000 0000 00008 .00000 00 O0 00: 00>000no 003 0000 HHO 0003 03003 000000 0000:00o 00003 00000 £03o0w 000000000 00002000000 :0 0000000 00N0.0 5NN0.0 00H0.0 50H0.0 0H00.0 00000>< 00H0.0 HOH0.0 05H0.0 00H0.0 HON0.0 00000>< AOHH0.0v 0000.0 05H0.0 0000.0: 00H0.0 N 000: 0NH0.0 00N0.0 0000.0 05H0.0 0H00.0 N 000: HOH0.0 NHH0.0 00H0.0 N000.0 00H0.0 0 0003 00H0.0 NON0.0 0N00.0 0000.0 N000.0 0 0003 00H0.0 00H0.0 05N0.0 00N0.0 0500.0 N 0003 00H0.0 50H0.0 00N0.0 0NH0.0 HOH0.0 N 0003 NON0.0 00N0.0 0500.0 HON0.0 5N00.0 N 0003 0NN0.0 NHN0.0 00N0.0 0HH0.0 0000.0 N 0003 0500.0 5500.0 0HN0.0 0.0 0000.0 H M003 00N0.0 00H0.0 00H0.0 0000.0 0000.0 H 0003 00000>< 0 N. H 0 0 0000 0000 N0 00000>< .H.v.0=000 0-N mHnae 206 H000.0 0NH0.0 H000.0 0000.0 0000.0 00000>< 0000.0 0000.0 0000.0 0N00.0 00000>< H000.0: N500.0: 5000.0: 0000.0: 0000.0: 00000>< 0000.0 «NHo.o NNHo.o o.o ONoo.o N 000: N000.0 NNHo.o NHoo.o- o.o N 0003 H000.0: 5000.0: 0000.0: 5000.0: NN00.0: N 000: 0500.0 NHN0.0 0HH0.0: 0000.0 0000.0 0 gum: 0000.0 5000.0 0500.0 0000 0N00.0 0 0003 NNoo.o- 0oHo.o- H000.0- oNoo.o: HNoo.o- 0 0003 5000.0 NOH0.0 0HH0.0 00H0.0 H000.0: N 0003 AHH00.0:V 0.0 5HH0.0 0NN0.0: 0N00.0 0 M003 N0oo.o- oHHo.o- ONoo.o- o.o o.o N 0003 0Noo.o o0Ho.o oNoo.o- NNoo.o o.o N 0002 5N00.0 0.0 0NH0.0 0000.0: 0N00.0 N 0003 0N00.0: 0N00.0: 0.0 0000.0: N000.0: N 0003 0000.0 0000.0: 00H0.0 5HH0.0 00H0.0 H 0003 5000.0 0H00.0 NN00.0: 50N0.0 0000.0 H 3003 5000.0: 0500.0: 0000.0: 0HH0.0: 5HH0.0: H 0003 00000>< 0 N H 0 0 0000 0000 NN 00000>< 0 N H 0 0 0000 Nmom NH 00000>< 0 N H 0 0 0000 0000 No .oo00 00 3o00 00000 000 00HH0N 0000H0H mo 00000 :03o0w 00000000 .0:m 0H00H 207 .000000>0 00000>o :0 00000000 000 000 0000 000:8 ocmuflm GD 0..— uOG U0>h0m£0 was» 600% H00 0003 00003 000000 00000000 00003 00000 003000 000000000 00000000000 :0 0000000 00H0.0 NNH0.0 N000.0 00N0.0 00000>< 0NH0.0 0500.0 00N0.0 0N00.0 00000>< “0000.0v 00N0.0 N000.0 0000.0 N 0003 A0000.ov 0000.0 0000.0 N 0003 AO0H0.ov 0000.0 00N0.0 0NH0.0 0 0003 AqNoo.ov 0000.0 0000.0: 0 0003 00000.00 0N00.0 0000.0 ~n00.0 N 0003 AO0H0.ov 0000 0N00.0 0N00.0 N 0003 00000.00 N000.0 NN00.0: 0N00.0: N 0003 ANN00.0:V 0000.0: NHN0.0 0000.0 N 0003 0000.0 0000- muH0.0 N000.0 00N0.0 0 0003 0N00.0 0000.0 00N0.0 0N00.0 0000 0 0003 00000>< 0 N H 0 0 0000 0000 N0 00000>< 0 N H 0 0 0000 0000 N0 .A.0.0coov N:m 00000 208 Table B-6. Specific growth rates of Tila ia zillii fry (0.0140 g average weight) fed trout chw at 27.5°C. Average Specific Growth 1% Feed -0.0401 2% Feed 0.0233 3% Feed 0.0459 5% Feed 0.0779 10% Feed 0.0854 Table 8-7 0 209 Specific growth rates of Tilapia zillii fry (1.76 g average weight) fed trout c ow at 27.5°C. 1% Feed 2% Feed 3% Feed 5% Feed 7% Feed Fish # wNo—IO Fish # wNI-‘O Fish # ri-oo Fish # lev-‘O Fish # wNp—IO Specific Growth 0.0319 0.0058 0.0191 2.92.7.1 Average 0.0085 Specific Growth 0.0151 0.0141 0.0143 0.0297 Average 0.0170 Specific Growth 0.0261 0.0061 0.0465 0.0261 Average 0.0239 Specific Growth 0.0638 0.0397 0.0338 0.0556 Average 0.0483 Specific Growth 0.0554 0.0607 0.0436 0.0671 Average 0.0572 210 Table B-8. Specific growth rates of 38.1 g average weight Tilapia zillii fed trout chow at 27.5°C. 0% Feed Week 1 Week 2 Average -0.0089 -0.0016 -0.0053 1% Feed Week 1 Week 2 Average 0.00425 0.00413 0.0042 2% Feed Week 1 Week 2 Average 0.00986 0.00778 0.0088 3% Feed Week 1 Week 2 Average 0.01385 (0.00629) 0.0138 Figures in parentheses represent growth rates which occurred during weeks when all feed was observed not to be eaten. These data are not included in overall averages. 211 N000.0 0000.0 0N00.0 N000.0 0000.0 0000.0 ommum>< 0N00.0 0000.0 00.0 0000.0 0000.0 0000.0 0 N000.0 N000.0 0N00.0 0N00.0- 0N00.0 NN00.0 N NN00.0 0000.0- 00.0 00.0 00.0 NNN0.0 N 0000.0 NN00.0 NN00.0 0000.0 00N0.0 NNN0.0 0 mwmuw>< 0 x00: 0 x00: 0 x00: N 000: N x00: * nmNe 000m NN 0N00.0 NN00.0 0000.0- NN00.0- NN00.0- 0N00.0 mwaum>< 0000.0 0N00.0 0N00.0- NN00.0- 00.0- 0N00.0 m NN00.0 0000.0 0N00.0 NN00.0- 0000.0- «000.0 N 0N00.0 0N00.0 NN00.0- NN00.0- 00.0 0000.0 N - - - - - - 0 mwmuo>< 0 x00: 0 emu: 0 x00; N 0003 N 3003 N nmNm 0000 NH NN00.0- NN00.0- 0000.0- 0N00.0- 0N00.0- 0N00.0- mwmum>< 0N00.0- NN00.0- N000.0- 0000.0- NN00.0- 00.0 0 0000.0- NN00.0- 00.0 NN00.0- 00.0 00.0 N N000.0- NN00.0- N000.0- N000.0- 0000.0- 0N00.0- N 0N00.0- NN00.0- NN00.0- N000.0- 00.0 N000.0- 0 mmmpo>< m x00: 0 x00: N 3002 N 3003 N x00: 0 nmNm 0000 N0 .Upm.hm no mwmcopocoo movoam pouoaaom 0mm NNNNNN «NmaNNe mo 0000“ nozouw onNoQO .0-0 oNnme 212 0000.0 0000.0 N000.0 ~000.0 5000.0 owwuo>< 0000.0 0000.0 0000.0 «000.0 0000.0 wwwuo>< 0000.0 0N00.0 0000.0 0000.0 0000.0 0 0003 0N00.0 0000.0 0000.0 5000.0 0000.0 0 x00: 0000.0 0000.0 0000.0 0000.0 0000.0 0 x00: 0000.0 0000.0 0000.0 00.0 0000.0 0 x00: 0000.0 0000.0 0000.0 0m00.0 0n00.0 m 0003 0000.0 0000.0 0000.0 0000.0 N000.0 m 0003 0000.0 0000.0 0000.0 0000.0 0000.0 N 000: 0000.0 0000.0 0000.0 0000.0 0000.0 N 000: 0000.0 0000.0 0000.0 0000.0 0000.0 0 0003 0N00.0 N000.0 0000.0 0000.0 0000.0 0 x00: owmuo>< 0 N 0 0 00 :0000 0000 N0 owwum>< .A.0.00000 0-0 00000 213 Table B-10. Specific growth rates of Tila ia zillii fed Spirulina sp at 27.5°C. 1% Feed Fish # Week 1 Week 2 Week 3 Week 4 Average 0 0.0056 -0.0114 -0.0061 0.0 -0.0030 1 0.0 0.0 0.0 0.0 0.0 2 0.0 -0.0066 0.0066 0.0 0.0033 3 -0.0070 -0.0073 0.0143 0.0066 0.0079 Average (0.0) (-0.0061) 0.0031 0.0015 (0.0021) 2% Feed Fish # Week 1 Week 2 Week 3 Week 4 Average 0 0.0114 0.0 (0.0106 -0.0052) 0.0114 1 0.0 0.0050 (-0.0102 -0.0054) 0.0 2 0.0056 -0.0114 (-0.0130) -0.0070) 0.0056 3 0.0130 0.0061 (-0.0061) (0.0) 0.0130 Average 0.0071 (0.0) (-0.0014) (-0.0044) 0.0071 3% Feed Fish # Week 1 Week 2 Week 3 Week 4 Average 0 0.0050 -0.0156 -0.0056 0.0109 0.0050 1 0.0267 0.0101 0.0154 0.0165 0.0267 2 0.0087 0.0267 0.0255 0.0085 0.0087 3 0.0 -0.0087 0.0 0.0 0.0 Average 0.0123 (0.0069) (0.0096) (0.0109) 0.0123 5% Feed Fish # Week 1 Week 2 Week 3 Week 4 Average 0 0.0128 -0.0128 0.0087 0.0232 0.0128 1 0.0411 -0.0191 0.0145 0.0045 0.0411 2 0.0168 -0.0168 0.0087 0.0267 0.0168 3 0.0220 0.0220 0.0313 0.0256 0.0220 Average 0.0224 (-0.0040) (0.0181) (0.0215) 0.0224 Figures in parentheses represent growth rates which occurred during weeks when all feed was observed not to be eaten. averages. These data are not included in overall 214 0000.0 N000.0 0000.0 0000.0 0000.0 0000.0 wwmum>< 0000.0 0000.0 0000.0 0000.0 N000.0 0000.0 0 0000.0 0000.0 0000.0 0000.0 0000.0 0000.0 N 0000.0 0000.0 0000.0 0000.0 0000.0 0000.0 0 0000.0 0000.0 0000.0 0000.0 0000.0 0000.0 0 mwmum>< 0 x003 0 3003 0 3003 0 x00: 0 0003 0 £000 0000 Na 0000.0 0000.0 0000.0 0000.0 0000.0 0000.0 wwwum>< 0000.0 0000.0 N000.0 0000.0 0000.0 0000.0 0 0000.0 0000.0 0000.0 0.0 0000.0 0000.0 N 0000.0 0000.0 0000.0 0000.0 0000.0 0000.0 0 0000.0 0000.0 0.0 0.0 0000.0 0.0 0 0w0u0>< 0 x00: 0 0003 0 0003 N 000: 0 x00: 0 0000 0000 N0 0000.0: 0000.0: 0000.0: 0.0 0000.0: 0000.0: wwwuw>< 0000.0: 0000.0: 0000.0: 0.0 0.0 0000.0: 0 0000.0: 0.0 0000.0: 0.0 0000.0: 0000.0: N 0000.0: 0000.0: 0000.0: 0.0 0000.0: N000.0: 0 0000.0: 0000.0: 0000.0: 0.0 0.0 0000.0: 0 ammuo>< 0 x003 0 £003 0 0003 N 1003 0 x003 0 £000 @000 No .000.0N um 3ono uaouu 000 mao0uo00: novoumnuoumm mo mmuwu £u3ou0 00000000 .00:0 @0008 215 0000.0 0000.0 0000.0 0000.0 0000.0 0wmu0>< 0000.0 0000.0 0000.0 0000.0 0000.0 00000>< 0000.0 0000.0 0000.0 0000.0 0000.0 m 000: 0000.0 0000.0 0000.0 0000.0 0000.0 0 0003 0000.0 0000.0 0000.0 0000.0 0000.0 0 3003 0000.0 0000.0 0000.0 0000.0 0000.0 0 0003 0000.0 0000.0 0000.0 0000.0 0000.0 m x003 0000.0 0000.0 0000.0 0000.0 0000.0 m 0003 0000.0 0000.0 0000.0 0000.0 0000.0 N 0003 0000.0 0000.0 0000.0 0000.0 0000.0 N 0003 0000.0 0000.0 0000.0 0000.0 0000.0 0 3003 0000.0 0000.0 0000.0 0000.0 0000.0 0 0003 00000>< 0 0 0 0 0 500.0 0000 00 0wmu0>< .A.v.ucoov 00-m «Hams 216 Heao.o MHHO.o mNdo.o 0000.o mmfio.o mm0o.o mwauo>< 000o.o Haao.o 0N00.0 ano.o mHNo.o onNo.o m N000.0 ONHO.o 000o.o 000o.o mm0o.o 000°.o N oNoo.o aooo.o N0oo.o NoHo.o Nmoo.o 0moo.o 0 N00o.o Noflo.o 0N0o.o 000°.o mmdo.o mmHO.o o owuuo>< m 0003 0 000: m 0003 N 0003 0 000: 0 0000 0000 NN mmoo.o «moo.o o0oo.o omoo.o 00oo.o Aoooo.o-0 mwaum>< mmoo.o mNoo.o 0Noo.o NNHo.o Nmoo.o Nmoo.o- m Nooo.o mwoo.o N000.0 nooo.o 0moo.o o.o N 000o.o Nmoo.o «00¢.o «moo.o «00o.o 0o0o.o 0 0000.0- mmoo.o- omoo.o- oo~o.o- o.o m00o.o- o ammuo>< m 0003 0 0002 m 0003 N 0002 0 000: 0 0000 0000 NH mmoo.o- 00oc.o- mNoo.o- 0moo.o- omoo.o- N000.o- mm~um>< 0noo.o- N000.0- mqoo.o- o.o o.o on~o.o- m N000.0- 0000 009°.o- mqoo.o- o.c Nmoo.o- N 0000.0- o.o o.o nqoo.o- mNfio.o- 000o.o- 0 nooo.o- 0000 o.o oNoo.o- omHo.o- Heoo.o- o umaum>< n 0003 0 0003 m 0003 N 000: 0 0003 0 5000 0000 No .o.m.0~ 00 0m 0:00:0000 00000000 000 03000000: 000000500000 00 00000 503000 00000000 .00:0 00008 217 .00w000>0 00000>0 00 00000000 000 000 0000 00009 .00000 00 00 000 00>000no 003 0000 000 0003 00003 000000 00000000 00003 00000 00300w 000000000 00000000000 00 0000w00 0000.0 0000.0 memo.o «mmo.o mm¢o.o 0w000>< 00N0.0 mqmo.o mmoo.ou Mnmo.c 0.0 0w000>< memo.o oomo.o Nnmo.o 0000.0 mm0o.o 0 000: wmmo.o nmmo.o 0.0 wmmo.o mqoo.o m 000: 00mo.o 00mo.o 0wmo.o ammo.o mm0o.o 0 0003 00N0.0 0000.0 0000.0- 0000.0 0.0 0 0003 moeo.o memo.o mmqo.o 00N0.0 0mmo.o 0 0003 00N0.0 0000.0 0.0 0000.0 0.0 m 0003 0o¢o.o omqo.o No~0.o @0mo.o nmmo.o N 0003 mmmo.o Nmmo.o Nmoo.on 0000.0 omoo.o N 0003 Nmao.o mmmo.o mmqo.o m0mo.o Nmmo.o 0 000: ow0o.o 0~0o.o Nmoo.oa ommo.o Nm0o.ou 0 x00: 0w000>< m N 0 o 0 0000 0000 Nn 0w000>< .0.0.00000 N0-0 00000 218 oNNo.o «NNo.o ooNo.o NNNo.o quo.o owoo.o mwmuo>< Nooo.o Nooo.o NoNo.o oNNo.o oNo.o Noo.o m mmNo.o omfio.o ooNo.o «oNo.o Hmo.o oNo.o- N moNo.o moNo.o «NNo.o emNo.o NNo.o NNo.o N NNNo.o mNNo.o cooo.o oo.o Noo.o Nmo.o o umouo>< m xmma a Mama m xomz N 3mm: N xoma N :mNm vumm NN cho.o NNoo.o mmoo.o mqoo.o amoo.c ANoNo.o-v owaum>< mNNo.o coco.o oNoo.o NmNo.o NNo.o mNo.o m Nqoo.o omoo.o mmoo.o o¢oo.o moo.o oNo.o- N «Noo.o Nmoo.o «moo.o «moo.o- oo.o «No.0- N Nqoo.o- Nwoo.o- Nmoo.o- oo.o oo.o moo.o- o mwauu>< n xmms q gums m Now: N sums N xmmz N nmNm comm NH oNoo.o- mNoo.o- Nooo.o- omoo.o- mqoo.o- mmNo.o- mwmum>< Nooo.o- oo.o mqoo.o- Nqoo.o- moo.o- NNo.o- m «moo.o- oo.o oo.o quo.o- oo.o NNo.o- N Nooo.o- oo.o omNo.o- oo.o ooo.o- NNo.o- N oNoo.o- Nwoo.o- NNoo.o- Nmoo.o- Noo.o- oNo.o- o mwmum>< m 3mm: q gums m xmmz N xmmz N 3mm: N smNN nmom No \hmeuu mumowamsvv .ooom am 3050 usouu vwm «waawu dammAHH mo mmuwu nuzouw camaommm .mHum manMH 219 .mmwmum>w NNmum>o :N vmszocw you mad wumv mmmnH .cmuwm ma 0» No: vm>ummno was vmmw NNw 50:3 mxmm3 wcwusv vmuuaooo sofln3 mmuwu nusouw accumummu mmmwsucmuma cw mmuzwfim memo.o omo.o mmo.o sco.o wNo.o wwwum>< e-o.o wnao.o wmfio.o mNNo.o mamo.o mwmum>< NomNo.ov om~o.o NoHo.o NHN0.0 como.o m Nam: m¢~o.o o-o.o < m N H o N nmNN vmmm Na mwmum>< .N.n.usoov NN-m mNnmN APPENDIX C 220 Table C-1. Results of water chemistry analyses performed while TilaEia zillii were fed trout chow at 21.5°C. Dissolved Oxygen (mg/1 D.C.) 2 Feed Initial Week 1 Week 2 Average 1 9.0 8.9 8.9 8.93 2 8.9 809 9.1 h 8096 3 8.9 8.8 9.0 8.90 4 9.0 8.6 8.9 8.83 Ammonia Nitrogen (mg/l-N) 2 Feed Initial Week 1 Week 2 Average 1 0.01 0.43 0.43 0.29 2 0.01 0.66 0.74 0.47 3 0.01 0.81 0.89 0.57 4 0.01 0.98 0.82 0.60 2g 2 Feed Initial Week 1 Week 2 Average 1 8.30 8.20 8.20 8.23 2 8.30 8.25 8.25 8.27 3 8.35 8.25 8.20 8.27 4 8.30 8.25 8.20 8.25 Temperature (°C -- 7 day averages) 2 Feed Initial Week 1 Week 2 Average 1 21.7 21.6 21.4 21.5 2 21.9 21.6 21.4 21.5 3 21.4 21.3 21.4 21.3 4 121.8 21.7 21.5 21.6 221 mo.w ow.” mm.m - oo.m om.» mm.» m mN.m NN.N NN.N - MN.» oN.N Nw.m N mw.m ow.m Nm.m - mm.» oN.m oo.m N mw.w oN.w om.m - Nm.w mw.m NN.N N mm.m aN.m mm.m . Nw.w om.m Na.m o awaum>< m Now: s Nmmz m Now: N Nam: N gums NmNNNcN comm N mm Nm.o Ns.o oe.o - om.o - NN.o m oN.o am.o mm.o - oN.o - No.0 m NN.o mm.o mm.o - Nm.o - oN.o N NN.o 0N.o NN.o - NN.o - oN.o N 0N.o NN.o mN.o - NN.o - NN.o o ammuo>< n Moms a 3mm: m xmmz N xwma N xmmB NmNuNcH comm N AzuN\wev cowouuwz mwsoaa< N¢.N m.N o.w m.N w.N o.m o.» m mw.N m.N m.N w.N m.N o.» o.w m mw.N w.N o.N N.N o.N o.m m.N N om.N N.N m.N N.N 0.x o.m N.N N om.N m.N m.N m.N o.» o.N m.N o mwaum>< m Now: q Now: m Now: N Nam: N xmma NmNuNcN comm N A.o.o N\wav mmwwxo vo>aommwa .UomN ad 3050 usouu vow mum: NNNNN mNQmNNH mafia: anuowuom momhamcm xuumfiawso uwumz mo muNsmmm .Nuu mNan "-4 222 m .mwumavm Nam pom Como N\we ¢mm wcfiwmum>m :oNumsuost mNuuNN woufinwsxw moaam> xuwcfiNmMN< NNH5NNQMN< o.mN o.mN o.mN o.mN o.mN m.¢N o.mN m N.mN o.mN o.mN o.mN o.mN m.mN m.oN m o.mN o.mN o.mN o.¢N o.mN o.¢N m.mN N m.¢N o.mN o.¢N o.mN o.mN N.NN m.cN N o.mN o.mN o.mN o.mN o.¢N o.¢N o.mN o owauo>< m Nam: N News N 3mm: N Now: N Now: NmNNNcN emmm N Ammwmuo>m xmv N In oov muaumuomaoe .N.v.N=oov N-o manN 223 Table C-3. Results of water chemistry analyses erformed while Tilapia zillii were fed trout chow at 2 .5°C. Dissolved Oxyggn (mg/1 D.C.) 2 Feed Initial Week 1 Week 2 Week 3 Week 4 Average 0 8.0 7.9 - 7.9 7.8 7.70 0.2 8.1 8.1 8.1 8.4 8.17 0.5 8.1 8.2 8.1 8.4 8.20 0.7 8.1 8.1 8.2 8.4 8.20 1 7.9 7.9 - 8.0 8.0 7.95 2 8.1 8.0 - 7.8 7.6 7.88 3 8.0 7.8 - 7.7 7.9 7.85 4 8.0 7.6 - 7.7 7.8 7.78 Ammonia Nitrogen (mg/l-N) 2 Feed Initial Week 1 Week 2 Week 3 Week 4 Average 0 0.04 0.08 - 0.13 0.18 0.11 0.2 0.03 0.28 0.15 0.01 0.12 0.5 0.03 0.18 0.02 0.01 0.06 0.7 0.03 0.15 0.02 0.02 0.06 1 0.03 0.17 - 0.27 0.24 0.18 2 0.03 0.31 - 0.22 0.26 0.21 3 0.01 0.47 - 0.49 0.40 0.34 4 0.01 0.58 - 0.61 0.38 0.40 2E Z Feed Initial Week 1 Week 2 Week 3 Week 4 Average 0 8.93 8.89 - 8.89 8.86 8.89 0.2 8.25 8.50 8.55 8.45 8.44 0.5 8.30 8.50 8.60 8.50 8.47 0.7 8.35 8.50 8.60 8.50 8.49 l 8.93 8.91 - 8.80 8.78 8.86 2 8.89 8.83 - 8.67 8.56 8.74 3 8.89 8.80 - 8.66 8.62 8.74 4 8.94 8.81 - 8.59 8.50 8.71 224 Table C-3 (cont'd.). Temperature (°C -- 7 day averages) Z Feed Initial Week 1 Week 2 Week 3 Week 4 Average 0 26.0 27.2 27.5 27.3 27.2 27.3 0.2 27.5 27.7 27.4 26.6 28.4 27.5 0.5 27.4 27.7 27.5 27.6 27.5 27.5 0.7 27.7 27.7 27.5 27.5 27.5 27.6 1 26.0 27.0 27.6 27.5 27.3 27.4 2 26.0 27.7 27.6 27.4 26.9 27.5 3 26.5 27.0 27.5 27.5 27.3 27.3 4 27.0 27.2 27.7 27.6 27.5 27.5 Alkalinity Alkalinity values exhibited little fluctuation averaging 338 mg/l CaCO3 for all aquaria. 225 »4.» N».» N8.» me.» 08.» »m.» «».» » Nm.» Nm.» «8.» N8.» »q.» »8.» »N.» » »».» N».» »».» »».» m».» »».» »N.» N NN.» 8N.» NN.» oN.» »N.» »N.» »N.» N »N.» N».» »N.» o».» »».» »».» »».» o m»mmm>< » Nmmz N xmm: » xmms N xmm: N xmm: NmNmNcN »mmN N mm NN.o »N.c NN.o NN.o no.» No.» No.» » ON.» no.» 0N.» NN.o no.» NN.o No.» » »o.» «o.» No.» »o.» no.» »N.o No.» N »o.» No.» No.» No.» »o.o »o.» No.» N mo.» No.» No.» No.» 4o.» »o.o No.» o owmum>< m x003 ¢ xmma m xomz N zoos N xomz NmNuNGH comm N AZuN\wev mmmouuwz cwcoaa< »4.» N.» ».» ».» - N.» ».» m »N.» 4.» N.» 4.» - ».» ».» m «m.» ».» ».» e.» - ».» N.» N »».» ».» ».» ».» - ».» ».» N «N.» N.» ».» N.» - ».» ».» o m»mmm>< » 3mm: 4 xmm: » Nmmz N xmms N »mms NmNmNcN mmmm N A.o.o N\wav cowwxo vo>Hommwa .ooom No 3050 uaouu vow mum: NNNHHN m mmN.H mNch vmauomuoa momemcm xuumNEmno noun: mo muNsmwm .qnu mNan 226 .mNumsom NNm Now coco N\wa «mm chwmuo>m coNumauost oNuNNN wouNnNnxo mwst> NuNcNmeN< NuNaNmeN< o.om 0.0m N.om o.mN 0.0m 0.0m 0.0m m m.mN o.om o.om N.oN o.mN m.mN m.om m m.mN o.om a.¢N o.oN m.mN a.mN m.aN N o.om o.om o.mN o.mN o.om m.oN m.mN N w.mN o.om m.aN w.mN m.oN m.¢N o.aN o m»mmm>< » Nmm: N Nmms » xmms N xmmz N xmm: NmNNNcN »mmm N Ammwmuo>m xmv N 11 oov muaumummamm .N.m.maomv »-» mNnmN 226 .mNumaom NNm you oomo N\wa «mm chwmum>m coNumsuost mNuuNN wouNnNnxo mmst> NuNcNmeN< NNNcNNmNN< 0.0m 0.0m N.om ¢.@N o.om 0.0m 0.0m m ¢.oN 0.0m 0.0m N.mN 0.0N w.oN m.om m m.mN 0.0m ¢.¢N m.mN o.a~ a.¢N m.mN N 0.0m 0.0m m.oN m.mN 0.0m o.oN m.mN N m.¢N 0.0m ¢.oN m.oN $.oN m.@N o.aN o m»mum>< m Nmm: » Nmm: » Nmmz N Nmm: N Nmm: NmNNNcN mmmm N Ammwmum>m va N :1 oov ouaumummamw .N.».mcomv »-» mNnmN 227 No.w NN.w wo.w mo.w No.m oo.w om.w om.w No.w co.m ow.w No.w wo.m mo.w oo.w oo.w mm.m qw.m mo.m «m.w ¢o.w oN.w Nw.m oo.w No.w No.w mo.w MN.m Nw.w om.w No.w «0.x om.w Nm.w NN.m mwmmm>< m xmm: N xmm: » xmms N xmmz N xmmz NmNNN qm.c NN.o NN.o No.9 «0.0 m»mum>< » Nmm: N xmmz » xmmz N 3mm: N 3mm: NmNNNcN NN.o mN.o om.o Nm.o «N.o no.0 mo.o co.o NN.o mm.o Nm.o No.0 No.0 No.0 00.0 m¢.o no.0 No.0 No.0 No.o No.0 co.o wN.o No.0 mo.o No.0 No.0 No.0 mN.o No.o :H CHNMW mmmm N mm OHNMW coom N A21N\wav mmwouuNz oNcoaa< NN.» N.N ».N ».» - ».» N.» » Nm.w N.w m.w o.w I w.m w. w m wo.w m.w o.w N.m a w.m w. m N NN.w w.w 0.x N.w 1 o.w m. m N oo.w o.w m.w m.m I N.w N. m o owouo>< m xooz q xooz m xooz N xoo: N goo: NoNuNcN coom N A.o.o N\wev mmmwxo co>NommNQ .08 mm No 30:0 anon» cow ouoz NNNN NNN onoNNH oNan coauouuom momNNoco NuumNeoso pogo; mo mNNSmom .muo oNnoH 228 .oNuozco NNo now ooou N\wa ¢Nm chwouo>o coNuosuost oNuuNN couNnNnxo mosNo> NuNcNmeN< NuNcNNoxN< N.mm o.mm m.¢m o.nm m.mm w.< » xmmz 8 xmm: » xmms N Nmm: N xmmz NmNNNcN cmmm N Amowono>o Noc N :1 oov ouauouomaoH .N.».Ncoov »-» mNnmN 229 Table C-6. Results of water(chemistry analyses performed while Tilapia zillii fry 0.0140 g average weight were fed trout chow at 27.5°C. Dissolved Oxygen (mg/1 D.O.) Z Feed Initial Week 1 Average 1 8.0 7.8 7.9 2 8.0 7.7 7.8 3 7.9 7.7 7.8 5 7.9 7.9 7.9 10 8.0 7.8 7.9 Ammonia Nitrogen (mg/l-N) 2 Feed Initial Week 1 Average 1 0.01 0.04 0.02 2 0.01 0.02 0.01 3 0.01 0.01 0.01 5 0.01 0.01 0.01 10 0.01 0.05 0.03 2n 2 Feed Initial Week 1 Average 1 8.60 8.60 8.60 2 8.65 8.60 8.62 3 8.70 8.65 8.67 5 8.70 8.65 8.67 10 8.70 8.65 8.67 Temperature (°C -- 7 day averages) 2 Feed Initial Week 1 Average 1 27.0 27.1 27.1 2 28.0 28.7 28.7 3 27.5 28.1 28.1 5 27.5 27.4 27.4 10 27.5 27.6 27.6 230 Table C-7. Results of water chemistry analyses perfgrmed while Tilapia zillii fry 1.75 g average weight were fed trout chow at 27.5°C. Dissolved Oxygen (mg/l D.O.) 2 Feed Initial Week 1 Average 1 7.8 7.7 7.7 2 7.9 7.8 7.8 3 7.8 7.8 7.8 5 7.8 7.8 7.8 7 7.9 7.8 7.8 Ammonia Nitrogen (mg/l-N) 2 Feed Initial Week 1 Average 1 0.01 0.07 0.04 2 0.02 0.12 0.07 3 0.01 0.14 0.07 5 0.01 0.26 0.13 7 0.02 0.37 0.19 Re Z Feed Initial Week 1 Average 1 8.65 8.35 8.50 2 8.60 8.15 8.37 3 8.70 8.30 8.50 5 8.70 8.15 8.42 7 8.70 8.25 8.47 Temperature (°C -— 7 day averages) 2 Feed Initial Week 1 Average 1 27.0 27.4 4 27.4 2 28.0 27.6 27.6 3 27.5 27.5 27.5 5 26.0 27.4 27.4 7 27.5 27.5 27.5 231 Table C-8. Results of water chemistry analyses performed while Tilapia zillii 38.1 g average weight were fed trout chow at 27.5°C. Dissolved Oxygen (mg/l D.O.) Z Feed Initial Week 1 Week 2 Average 0 8.0 7.9 8.0 7.97 1 8.0 8.0 7.9 7.97 2 7.9 8.0 7.9 7.93 3 8.1 8.0 8.0 8.03 25 Z Feed Initial Week 1 Week 2 Average 0 8.81 8.79 8.68 8.76 l 8.73 8.70 8.76 8.73 2 8.80 8.62 8.59 8.67 3 8.75 8.63 8.51 8.63 Temperature (°C -- 7 day averages) Z Feed Initial Week 1 Week 2 Average 0 27.5 27.5 27.7 27.6 1 27.5 27.5 27.8 27.6 2 27.5 27.5 27.7 27.6 3 27.5 27.5 27.7 27.6 Alkalinity Alkalinity values exhibited little fluctuation averaging 329 mg/l CaCO3 for all aquaria. 232 mm.n om.n nm.n Nq.n om.m ww.w mn.w m No.0 N0.n nn.n 0m.m Nm.n 00.x nw.n n 0m.n Nq.n Nm.n Nm.w 00.w co.m N0.w N NN.n N0.n No.0 00.x n0.n 00.0 mn.n N mN.w Nn.n NN.n 0n.m 0N.n Nn.n nN.n 0 owouo>< m xoos o Moos n xooz N xoos N zoo: NoNuNcN coom N mm 00.0 no.0 «0.0 no.0 NN.0 0N.0 No.0 m No.0 no.0 no.0 no.0 nN.o NN.0 No.0 n mN.0 no.0 «0.0 00.0 nN.0 no.0 No.0 N mN.0 no.0 no.0 no.0 on.0 on.o No.0 N NN.0 no.0 NN.0 wn.0 nn.0 No.0 No.0 0 mwmum>< m Nmms N xmms » xmmz N xmm: N Nmm: NmNNNcN »mmm N Aan\wev cowouuNz oNcoaa< on.n 0m.w 00.0 0n.o mN.0 00.0 on.N m no.w No.0 on.n 00.0 mN.0 0N.n mN.N n 00.0 N<.n 00.0 no.0 0n.o mN.n nN.N N 00.0 N0.n no.0 00.0 0N.0 mN.n m0.n N 00.0 Nn.m 00.0 00.0 0¢.0 0n.n 0N.N 0 owouo>< m xooB q xooa n xoos N xoos N xoos NoNuNcN coom N A.0.0 N\wav cowwxo co>NommN0 .mNmaococoo oocONm couoNNoa cow ouos NNNNNN o moNNH oNNss coENONNoQ mooNNocm NuumNeoso noun: «0 muNsmom .010 oNnoH 233 .oom.¢N chwouo>o coNuoa NcSNm onu uaoswsounu ouaumuomeou eoou um cocNmaou oNuoaao NNo pom mouzuouomaos ouauouomaofi .oNuoavm NNo now nooou N\wa omn chwouo>o cONuosuoDNw oNuuNN coNNncho mozNo> NNNcNNoxN< NNNcNmeN< .N.».Ncomv »-» mNnmN 234 Table C-lO. Results of water chemistry analyses performed while Tilapia zillii were fed pelleted Spirulina sp. Dissolved Oxygen (mg/1 D.O.) Z Feed Initial Week 1 Week 2 Week 3 Week 4 Average 1 8.1 8.1 7.9 8.1 8.0 8.04 2 8.2 8.1 7.9 8.1 7.9 8.04 3 8.1 8.1 7.8 8.0 8.1 8.02 5 8.1 8.0 7.9 7.9 7.8 7.94 Ammonia Nitrogen (mg/l-N) Z Feed Initial Week 1 Week 2 Week 3 Week 4 Average 1 0.03 0.31 0.28 0.02 0.05 0.14 2 0.02 0.50 0.03 0.05 0.09 0.14 3 0.03 0.70 0.02 0.02 0.05 0.16 5 0.03 1.22 0.03 0.13 0.07 0.30 pg Z Feed Initial Week 1 Week 2 Week 3 Week 4 Average 1 8.35 8.45 8.35 8.35 8.50 8.40 2 8.30 8.45 8.40 8.50 8.55 8.44 3 8.30 8.45 8.35 8.50 8.55 8.43 5 8.30 8.45 8.35 8.40 8.20 8.34 Temperature (°C -- 7 day averages) Z Feed Initial Week 1 Week 2 Week 3 Week 4 Average 1 28.5 27.7 27.6 27.6 27.5 27.6 2 28.5 27.6 27.6 27.6 27.4 27.6 3 28.5 27.7 27.6 27.8 27.4 27.6 5 27.5 27.4 27.6 27.6 27.7 27.5 Alkalinity Alkalinity values exhibited little fluctuation averaging 343 mg/l CaCO3 for all aquaria. 235 NN.N mn.w o¢.n no.0 mn.n 1 om.n 00.n mm.n 00.n 0N.N 1 nm.n 00.n mm.n 00.n 0<.n 1 0m.n m0.n no.0 00.n 0¢.n 1 nm.n no.0 00.0 ON.» no.» 1 mwmum>< m xmmz N Nmmz » xmm: N xmmz N xmmz No.0 No.0 00.0 00.0 00.0 1 no.0 00.0 no.0 «0.0 No.0 1 no.0 00.0 No.0 «0.0 No.0 1 «0.0 00.0 no.0 «0.0 00.0 1 «0.0 «0.0 no.0 No.0 00.0 1 m»mmm>< m Nmmz N xmmz » Nmms N Nmm: N xmmz mm.» mm.» on.» on.» on.» NmNmNcN No.» No.» No.» No.» No.» NmNmNcN CHNMI-n mmmN N OHNMLfi »mmN N mm A21N\wav cowouuNz oNnoaa< NN.N mn.N mn.N 0N.N on.N 1 on.N mN.N on.N mn.N mo.N 1 Nn.N 0N.N on.N oo.N mn.N 1 on.N 0N.N 00.N 00.N mo.N 1 nn.N 0m.N on.N 00.N mn.N 1 mwmum>< m Nmm: 8 xmmz » xmmz N 3mm: N xmmz ON.» oN.» oN.» om.N oN.» NmNmNaN OHNMW »mmm N A.0.0 N\wav cowwxo co>NommN0 .3050 usouu com ouo: msoNNONN: cocouozuouom oNNss coauowuoa momNNocm NuumNeono Noam: mo mNNamom .NN-o mNnmN 236 you n.NN n.NN n.NN N.NN ¢.NN owouo>< n.NN 0.NN n.NN n.NN n.NN » 3mm: noooo N\we omn chwouo>o n.NN n.NN n.NN n.NN «.NN N xmms n.NN N.NN n.NN n.NN N.NN » Nmmz .oNuosco NNo cONNmaumsNN mNNmNN »mmNnNzxm mmst> NmNaNmeN< NuNcNNoxN< n.NN n.NN n.NN m N.NN n.NN n.NN n n.NN N.NN n.NN N n.0N 0.0N 0.0N N N.NN N.NN 0.nN 0 N xoos N zoo: NoNuNcN coom N Amowouo>o Noc N 11 any ouauouomaoa .A.c.ucoov NN10 oNan 237 on.N mN.n 0N.N mN.n on.N 1 0n.n 0¢.n mN.n 0¢.n mn.w 1 nn.n 0N.N 0N.n mn.n mN.n 1 0¢.0 mm.n mN.n n¢.n on.n 1 0n.n 0N.n m0.N 0m.n 0N.N 1 mwmum>< » xmms q xmmz » Nmm: N xmmz N xmmz No.0 no.0 no.0 00.0 nN.0 1 No.0 «0.0 NN.0 «0.0 nN.o 1 «0.0 «0.0 no.0 00.0 No.0 1 «0.0 «0.0 no.0 «0.0 No.0 1 o0.o «0.0 «0.0 «0.0 No.0 1 m»mmm>< n 3mm: 4 Nmm: » Nmms N Nmm: N xmms on.» o».» »4.» no.» me.» NmNNNaN No.» No.» No.» No.» No.» NmNNNcN OHNM‘D coom N OHNMl-fi coon N mm A21N\wav mmwouuNz oNcoaa< NN.N m0.N on.N 0N.N on.N 1 0<.N o¢.N 00.N on.N m¢.N 1 00.N m¢.N on.N m<.N 0N.N 1 0N.N 0N.N on.N 00.N on.N 1 no.N on.N 00.n 00.N 00.0 1 m»mmm>< » Nmms N Nmm: » Nmms N Nmms N Nmm: oN.» oN.» oN.» oN.» oN.» NmNmNcN OHNMW cook N A.0.0 N\wav cowwxo co>NommN0 .am ocNNauNmm couoNNom cow ouo: maoNuoNN: cocouosuouom oNan coauowuom mooNcho NuumNaoso wouo3 mo muNSmom .NN1U oNan 238 .oNuoaoo NNo pom noooo N\wa nNn chwouo>o coNuosuosNu oNNuNN couNnNsxo moaNo> NuNcNNoxN< NuNsNNoxN< n.NN N.NN N.NN n.NN n.NN N.wN 0.NN m n.NN N.NN N.NN N.NN N.NN n.NN n.NN n n.NN n.NN n.NN n.NN n.NN n.NN n.NN N N.NN n.NN N.NN n.NN N.NN N.NN n.NN N N.NN n.NN N.NN n.NN n.NN N.NN n.NN 0 owouo>< m moo: q xooz n xooz N xoos N xooz NoNNNcN coom N Amowouo>o Noc N 11 oov ouauouomaow .A.c.ucoov NN1U oNnme APPENDIX D 239 Table D-l. Chemical analysis of Purina Trout Chow, Grower, Size #3, (provided by manufacturer). Guaranteed Analysis Crude protein not less than.......40.0Z Crude fat not less than........... 9.0% Crude fiber not more than......... 5.0Z Ash not more than.................10.0Z Added minerals not more than...... 1.5% Ingredients Fish meal, soybean meal, wheat middlings, ground yellow corn, dried whey, soybean oil, dried yeast, dehydrated alfalfa meal, corn gluten meal, calcium carbonate, salt, ethoxyquin (a preservative), vitamin A supplement, D activated animal sterol (source of vitamin D-3), menadione sodium bisulfite (source of vitamin K activity), vitamin E supple- ment, DL methionine, vitamin B-12 supplement, ascorbic acid, biotin, choline chloride, folic acid, pyridoxine hydrochloride, thiamin, niacin, calcium pantothenate, riboflavin supplement, copper sulfate, manganous oxide, ferrous carbonate, calcium iodate, cobalt carbonate, zinc sulfate. 240 Table D-2. Typical chemical analysis of Spirulina sp. algae used in pelleted algae feeding trials (analysis provided by manufacturers). Chemical Composition Nonessential Amino Acids Moisture 7.0Z Alanine 5.82% Ash 9.0Z Arginine 5.98Z Proteins 71.0Z Aspartic Acid 6.43Z Crude fiber 0.9Z Cystine 0.67Z Glutamic Acid 8.94Z Glycine 3.46Z Essential Amino Acids Histidine 1.08Z Proline 2.97Z Isoleucine 4.13Z Serine 4.00Z Leucine 5.80Z Tyrosine 4.60Z Lysine 4.00Z Methionine 2.17Z Phenylalanine 3.95Z Threonine 4.17Z Tryptophan 1.13Z Valine 6.00Z Vitamins Biotin (H) average 0.4 mg/kg Cyanocobalamin (B-12) average 2 mg/kg d-Ca-Pantothenate average 11 mg/kg Folic Acid average 0.5 mg/kg Inositol average 350 mg/kg Nicotinic Acid (PP) average 118 mg/kg Pyridoxine (B-6) average 3 mg/kg Riboflavine (B-2) average 40 mg/kg Thiamine (B-l) average 55 mg/kg Toc0pherol (E) average 190 mg/kg Minerals Calcium (Ca) 1,315 mg/kg Phosphorus (P) 8,942 mg/kg Iron (Fe) 580 mg/kg Sodium (Na) 412 mg/kg Chloride (CI) 4,400 mg/kg Magnesium (Mg) 1,915 mg/kg Manganese Mn) 25 mg/kg Zinc (Zn) 39 mg/kg Potassium (K) 15,400 mg/kg Others 57,000 mg/kg Table D-3 o 241 Results of analyses performed on feeds to determine carbon, hydrogen, nitrogen, protein and caloric content. Spirulina--Pelleted Z Z Nitrogen Protein 9.64 60.2 9.71 60.7 Avg 9.67 60.5 Spirulina--Cultured Z Z Nitrogen Protein 10.56 66.0 10.61 66.3 10.65 66.6 Avg 10.61 66.3 Commercial Trout Chow Z Z Nitrogen Protein 6.55 40.9 6.37 39.8 6.41 40.1 Avg 6.44 40.2 Elodea canadensis--Pelleted Z Z Nitrogen Protein 3.40 21.2 3.30 20.6 3.30 20.6 Avg 3.33 20.8 Z Carbon 43.7 43.9 43.8 Z Carbon 49.1 49.2 49.3 49.2 Z Carbon 42.7 41.4 42.4 42.2 Z Carbon 40.7 40.3 40.1 40.4 Z Hydrogen 7.10 7.11 7.10 Z Hydrogen 7.27 7.70 7.64 7.54 Z Hydrogen 7.03 6.68 .9339. 6.84 Z Hydrogen K cal/g 4.65 4.62 4.63 K cal/g K cal/g 4.84 4.74 4.79 K cal/g 3.86 3.84 3.85 APPENDIX E 242 Table E-l. Composition of Spirulina media. Nutrient Concentration NaHCO3 13.61 g/l Na2C03 4.03 g/l KZHPO4 0.50 g/l NaNO3 2.50 g/l K2804 1.00 g/l NaCl 1.00 g/l MgSO4-7H20 0.20 g/l CaC12-2H20 0.04 g/l Metal Solution 6.0 ml/l Micronutrient Solution 1.0 ml/l Thiamin-HCl 0.10 mg/l Biotin 0.5 ug/l 312 0.5 ug/l Composition of Metal Solution Naz-EDTA 750. mg/l Na2M004°2H20 4. mg/l CoC12-6HZO 2. mg/l FeCl3°6H20 97. mg/l MnC12-4H20 41. mg/l Composition of Micronutrient Solution Naz-EDTA 50.0 mg/l Cu804-5H20 19.6 mg/l CoC12-6H20 20.0 mg/l H3BO3 618.4 mg/l MnClZ'4H20 36.0 mg/l ZnS04°7H20 44.0 mg/l m187121117173171]1mmu