DIGESTIBILITY, FEEDING VALUE AND LIMITING AMINOftCIDS IN HIGH-FIBRE AND FIBRE-REDUCED SUNFLOWER CAKES FED TO TILAPIA {OREOCHROMIS NJLOTICUS) B Y JOYCE GICHIKU M A I N A B.Sc. The University of Nairobi, 1981. M . S c , The University of Nairobi, 1992, A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE F A C U L T Y OF GRADUATE STUDIES (FACULTY OF AGRICULTURE) We accept this thesis as conforming to the required standard. THE UNIVERSITY OF BRITISH COLUMBIA FEBRUARY 2001. (g) Joyce Gichiku Maina, 2001 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of ffrCuL-7Y frf AQAldyLTUje/H-. S ^ i B v C e The University of British Columbia Vancouver, Canada Date Fg&%Uff*~V g P p / DE-6 (2/88) \ Abstract Four experiments were conducted at the University of Nairobi, in Kenya, to determine the effect of reducing the amount of fibre in sunflower cake on nutrient digestibility and feed utilization in tilapia (O. niloticus), and to compare this low-fibre cake with a commercially available high-fibre sunflower cake. The extent to which protein from a high-fibre and a fibre-reduced sunflower cake could replace fishmeal protein in tilapia diets, and the effects of supplementing diets made from a low-fibre sunflower cake with amino acids lysine, methionine, and threonine on growth, feed intake, and feed utilization were also investigated. Also of interest was to compare digestibility and feeding value of Kenyan omena fishmeal with that of Low-Temperature (LT) anchovy fishmeal. Tilapia (O. niloticus) fingerlings were used in all the experiments. Water temperatures and dissolved oxygen concentrations were maintained above 26 °C and 5.5 mg/litre respectively. Dehulling of sunflower seeds was done using a manual dehuller. Crude fibre levels in the dehulled cakes were all below 15% (DM basis). Protein from the low-fibre and high-fibre sunflower cakes was well digested by tilapia. The apparent digestibilities of protein in the sunflower cakes and the fishmeals were not significantly different. Reduction of fibre in sunflower cake had no effect on the digestibility of protein. Digestibility of energy in the sunflower cakes was low. Apparent digestibility coefficient for energy (ADC-E) and digestible energy concentration (DE) were higher in the low-fibre sunflower cake than in the high fibre cake, but the differences were only significant for DE. There were no differences in the apparent digestibilities of protein, energy and organic matter between omena and anchovy fishmeals. ii In Experiment 2, the feeding value of a high-fibre and a low-fibre sunflower cake, omena and anchovy fishmeals was evaluated at two dietary protein levels (20% and 30%). There was no significant interaction between protein level and protein source. Fish fed at the 30% protein level gained more weight and had better feed conversion efficiency (FCE) than those fed at the 20% level. There were no significant differences in weight gain between fish fed diets based on anchovy and omena fishmeals and the low-fibre sunflower cake. Fish fed diets based on the high-fibre cake gained significantly (P < 0.05) less weight than those fed diets based on anchovy fishmeal. The low-fibre and high-fibre sunflower cakes were tested over a wide range of dietary inclusion in Experiment 3, each supplying 30%, 60%, and 80% of the dietary protein. The extent to which body fatty acids in tilapia reflect dietary fatty acids was also investigated. The low-fibre and high-fibre sunflower cakes could comprise up to 60% and 30% of the dietary protein respectively without compromising the performance of the fish. The inclusion of higher levels of the cakes in the diets caused a depression in feed intake, which resulted in lower weight gains of the fish fed these diets compared to those fed the control diet. Body fatty acid composition closely reflected dietary fatty acid composition. In Experiment 4, a basal diet in which a fibre-reduced sunflower cake provided 80% of the dietary protein was supplemented with amino acids lysine, methionine and threonine. The levels of these amino acids in the basal diet were 1.17%, 0.75% and 1.05% for lysine, methionine and threonine respectively, while the stipulated requirements (NRC, 1993) are 1.54%, 0.8% and 1.2% respectively. There was a trend to improved growth rate and FCE in fish fed diets supplemented with lysine and threonine, i i i but the improvement did not attain statistical significance. Methionine, added alone or together with threonine did not elicit any response in fish. iv TABLE OF CONTENTS Abstract ii Table of contents v List of Tables viii List of Appendices ix Acknowledgements x Chapter 1: General introduction 1 1.2 References 5 Chapter 2: Review of Literature 6 2.1 Aquaculture in Kenya 6 2.2 Fish species cultured 9 2.3 Dietary protein and amino acids requirement in fish 10 2.3.1 Search for new sources of protein 16 2.4 Sunflower (Helianthus annuus) 26 2.4.1 Taxonomy and origin of the domesticated sunflower 26 2.4.2 Morphology of the sunflower plant 27 2.4.3 Chemical and physical composition of sunflower seeds 27 2.4.4 Sunflower seed oil 28 2.4.5 Sunflower seed proteins 29 2.4.6 Composition of sunflower hulls 29 2.4.7 Anti-nutritive factors in sunflower seeds 31 2.4.8 Processing methods 32 2.4.9 Sunflower meal 32 2.5 Use of sunflower meal in animal feeds 36 2.5.1 Ruminants 36 2.5.2 Non-ruminants 37 2.5.2.1 Sunflower meal in swine diets 37 2.5.2.2 Sunflower meal in poultry diets 39 2.5.2.2.1 Broilers 39 2.5.2.2.2 Sunflower meal in Layer diets 40 2.6 Use of sunflower meal in fish diets 40 2.7 References 45 Chapter 3: Experiment 1: Digestibility of nutrients and energy in wheat bran, high-fibre and fibre-reduced sunflower cakes, anchovy fishmeal and omena fishmeal by Oreochromis niloticus 56 3.0 Abstract 56 3.1 Introduction and objectives 57 3.2 Materials and methods 60 3.2.1 Sunflower cakes 60 3.2.1.1 Fibre-reduced sunflower cake 60 v 3.2.1.2 High-fibre sunflower cake 60 3.2.2 Ingredients other than sunflower cakes 60 3.2.3 Chemical analyses 61 3.2.4 Experimental diets 61 3.2.5 Supply and maintenance of fish 63 3.2.6 Fecal collection 64 3.2.7 Digestibility assessment 64 3.2.8 Data collection and analytical procedures 66 3.2.9 Statistical analyses 66 3.3 Results and discussion 67 3.3.1 Chemical composition of the reference diet, test diets and test ingredients. 67 3.3.2 Fish performance 67 3.3.3 Apparent digestibility of nutrients in test ingredients 70 3.3.4 Apparent digestibility coefficient for protein (ADC-P) 70 3.3.5 Apparent digestibility coefficient for energy (ADC-E) and digestible energy concentration (DE) in test ingredients 75 3.3.6 Apparent digestibility coefficient for organic matter (ADC-OM) 78 3.4 Conclusions 79 3.5 References 80 Chapter 4: Experiment 2: The feeding value and protein quality in high-fibre and fibre-reduced sunflower cakes and Kenya's "omena" fishmeal for tilapia (Oreochromis niloticus) 83 4.0. Abstract 83 4.1 Introduction and objectives 85 4.2 Materials and methods 85 4.2.1 Experimental diets and design 87 4.2.2 Fish sampling 90 4.2.3 Data collection and analytical procedures 91 4.2.4 Chemical analyses 91 4.2.5 Statistical analysis 92 4.3 Results and discussion 93 4.3.1 Chemical composition of the diets 93 4.3.2 Fish performance, PER, and PPV 97 4.3.3 Effect of diets on whole body composition 106 4.4 Conclusions 110 4.5 References 113 Chapter 5: Experiment 3: Partial replacement of fishmeal with high-fibre and low-fibre sunflower cakes in diets for tilapia (O. niloticus): Effect on fish performance and whole body fatty acids. 117 5.0 Abstract 117 5.1 Introduction and objectives 119 vi 5.2 Materials and methods 120 5.2.1 Experimental diets and design 120 5.2.2 Fish sampling 122 5.2.3 Data collection and analytical procedures 123 5.2.4 Chemical analyses 123 5.2.5 Statistical analysis 124 5.3 Results and discussion 125 5.3.1 Chemical composition of the diets 125 5.3.2 Fish performance PER, PPV, body and fatty acid composition 127 5.4 Conclusions 142 5.5 References 144 Chapter 6: Experiment 4. Evaluation of the most limiting amino acids in diets based on sunflower cake fed to tilapia (O. niloticus). 148 6.0 Abstract 148 6.1 Introduction and objectives 149 6.2 Materials and methods 151 6.2.1 Experimental diets and design 151 6.2.2 Fish sampling 153 6.2.3 Data collection and analytical procedures 153 6.2.4 Chemical analyses 153 6.2.5 Statistical analyses 154 6.3 Results and discussion 155 6.3.1 Chemical compositions of the diets 155 6.3.2 Fish performance 156 6.4 Conclusions 166 6.2 References 167 Chapter 7: General discussion, conclusions and recommendations 170 7.1 References 176 vii LIST OF T A B L E S Table # 2.1 Fish production in Kenya 8 2.2 Some fibre components of sunflower seed hulls and other agricultural residues 30 2.3 Proximate compositions of Kenyan sunflower seed cakes 34 2.4 Average amino acid compositions of sunflower meal, soybean meal, cottonseed meal and rapeseed (or canola) meal. 35 3.1 Compositions of the diets used in Experiment 1 62 3.2 Compositions of the ingredients used in Experiment 1 68 3.3 Performance of O. niloticus after 50 days of feeding on the experimental diets 69 3;'4 Apparent digestibility coefficients (ADCs) and apparent digestible energy (ADE) values of the reference and test diets 71 3.5 Apparent digestibility coefficients (ADCs) and digestible energy values for the fibre-reduced and high-fibre sunflower cakes, omena fishmeal, anchovy fishmeal and wheat bran 72 4.1 Chemical compositions of the ingredients 88 4.2 Compositions of the diets used in Experiment 2 89 4.3 Amino acid compositions of the test diets 95 4.4 Effect of protein level on fish performance 98 4.5 Effect of source of protein on fish performance 99 4.6 Performance of O. niloticus fed diets containing high-fibre and fibre-reduced sunflower cakes, and LT. anchovy and omena fishmeals for 78 days. 100 4.7 Effect of feeding diets based on high-fibre and fibre-reduced sunflower seed cakes, LT. anchovy and omena fishmeals on whole body composition of O. niloticus after 78 days 107 5.1 Compositions of the diets used in Experiment 3 121 5.2 Amino acid compositions of diets used in Experiment 3 126 5.3 Percentages of fatty acid in the diets 128 5.4 Fatty acid compositions of corn oil, sunflower oil and herring oil. 129 5.5 Fish performance in relation to diet treatment after 70 days 130 5.6 Effect of protein source and level of sunflower cake on fish performance 131 5.7 Percentages of body proximate constituents viz., moisture, protein fat and ash (Air-dry basis) at 70 days in relation to diet treatments. 132 5.8 Percentages of fatty acid levels in the whole body of fish in relation to diet treatment. 137 5.9 Effect of type of sunflower cake and level in the diet on percentages of whole body fatty acids 138 6.1 Compositions and chemical analyses of diets used in Experiment 4 152 6.2 Determined Amino acid compositions of the diets used in Experiment 4 (DM basis) 157 viii 6.3 Determined amino acid compositions of the diets used in Experiment 4 (% of dietary protein) 6.4 Performance of fish (absolute weight, weight gains, specific growth rates (SGR), feed intake, and feed conversion efficiency) in relation to diet treatment LIST OF APPENDIXES Appendix # 1 Kabete water quality parameters assessed at the start of the study ix Acknowledgments This work was made possible by many people to whom I am greatly indebted. To Dr. Beames, my thesis supervisor, I say a big thank you for going beyond the call of duty to ensure that the work was completed. Though you retired from active teaching more than three years ago, you always found the time to go through my work patiently, meticulously and thoroughly, and were always ready for discussions. Special thanks to members of my committee, Dr. Higgs, Dr. Mbugua, Dr. Iwama and Dr. Kisia for going through all those drafts, and providing very valuable ideas. I am also grateful to Gay Huchelega for proof reading the thesis, valuable comments, and for being a very genuine friend. Thanks also to Mike Pitt for the concern you showed to all of us who were on the CIDA program. My family deserves special mention. To my husband, Julius Maina, I could not have done it without you. Thank you for taking very good care of our daughter during the years I have been away. Nobody would have done a better job. I am very grateful for your love, encouragement and support. I greatly appreciated all those long distance calls; emails, and letters which helped me maintain my sanity in a distant land. To my daughter, Beatrice, I wish to apologize for the years that I could not be physically present with you. I am very grateful that you understood, and survived without me. To my parents, Ester and Lasidslas Mwangi, thank you for sowing the seed, and believing that I could do it. I am also very appreciative to my brothers and sisters, for your concern for me, and for taking care of mum and dad. I owe special gratitude to all my friends. Gracias for making U B C and Vancouver a home away from home, and for welcoming me into your homes and your lives. I am immensely grateful to the Gichane's, Kamabu's, Tidyebwa's, Grace Wangu, Muthoni and Mugo Kimari, The Njenga's, Kangethe's, Wanjau's, Kamande's, Emmah Kishindo, Charles Ochieng, John Agak, Lucy Karanja, Abba Hammond, Rebeccca Biegon, and many others. Thank you for your support. I also wish to acknowledge my friend, Dr. Lucy Kabuage. I am grateful for your fellowship and prayers. To Giles, Sylvia, and Siva, thank you for your help in the lab. To Dr. Thompson, and Joyce Tom, I appreciated the interest you showed in my work, and for spurring me on to the finish line. To Rachel Njoroge and Thomas Njau, who helped with the experiments, thank you for your patience and dedication. I also wish to thank all the people who helped with the data analyses, C. Matere, Dr. Wanjau, and Dr. Charagu -God bless you all. Lastly, I wish to acknowledge some organizations that contributed greatly to the accomplishment of this work. I am grateful to the Canadian International Development Agency (CIDA) and Rockefeller Foundation for funding the work, and the University of Nairobi, for giving me the opportunity. It has been a long journey. I am grateful to each and everyone of you who helped along the way. Chapter 1 General Introduction Nile tilapia (O. niloticus) has become increasingly important as an inexpensive source of dietary protein in many countries. Tilapia culture is widespread in Africa and Asia on account of the fast growth, adaptability to a wide range of culture conditions and high consumer acceptability of this genus of fish. Nile tilapias, like all fish, require energy, protein, lipids, vitamins and minerals in their diets. In the wild these nutrients may be provided by the natural feed in the ponds. However, as fish biomass increases, e.g. in aquaculture, the provision of artificial feeds becomes essential. The characteristic diet of tilapia in the wild is a mixture of plant matter and detritus of plant origin. Blue green algae, diatoms, macrophytes and amorphous detritus are all common natural constituents of an adult tilapia diet (Bowen, 1990). Tilapia possess morphological and physiological adaptation mechanisms for utilization of these dietary components. Pharyngeal teeth break food particles into smaller units for easier peristaltic mixing and increased exposure to digestive enzymes. Gastric acid secreted to an unusually low pH lyses prokaryotic and eucaryotic cell walls to expose the cytoplasm to intestinal enzymes. This ability of tilapia to digest high-fibre materials has not been fully exploited in the development of tilapia diets for intensive culture. The protein component of most commercial fish diets generally includes a large proportion of fishmeal, usually 30-50% of the diet. Fishmeal is expensive and therefore considerable effort has gone into research to evaluate new protein sources to totally or partially replace the fishmeal. Soybean meal has an acceptable amino acid profile for the growth of most fish species, and therefore has been widely used to partially replace 1 fishmeal (Lovell, 1991). Soybeans, however, are not suitable as a replacement for fishmeal in some countries because they would have to be imported. In Kenya, for example, in 1990 and 1991 more than five million tonnes of soybean meal were imported for the Animal Feed Industry (Dept. of Animal Production, 1992). Thus there is a need to evaluate the more readily available locally produced sources of protein. Sunflower seed cake is an inexpensive and common oil-processing byproduct in many countries. In Kenya, sunflower farming was revitalized after a period of decline when the government decided to decontrol consumer prices for edible oils, fats, and animal feeds. Despite the ready availability of sunflower seed cake, its very high crude fibre content limits its use in animal feeds, especially for monogastric species, which account for 70% of the total feed produced. The crude fibre level of sunflower seed cakes in Kenya ranges from 24.1% to 40.2% (Jacob, 1993). Complete industrial dehulling of sunflower seeds has not been achieved (Tranchino et al, 1984; Cargill, 1980). Partial dehulling of the seed (10 - 12% removal) is common in the oil-seed industry. In the present study, one of the primary objectives was to determine the effect of fibre reduction of sunflower seed cake on the utilization of this protein product by fish. A second objective was to evaluate Kenya's "omena" fishmeal, made from Rastrineobola argentea. This is a small cyprinid fish endemic to Lake Victoria, Lake Kyoga and Lake Nabugabo in Uganda. It is locally known as "omena" in Kenya, "dagaa" in Tanzania and "mukene" in Uganda (Manyala et el., 1992; Wandera, 1992). It has a short life span of 1-2 years and its total length rarely exceeds 100 mm (Wanink, 1989). Prior to 1960, R.. argentea was of little economic importance in Kenya, forming an insignificant proportion of fish landed from Lake Victoria (Chitamwemba, 1992; 2 FAO, 1992). Catches of this fish have undergone explosive changes in the last 15 - 20 years (Manyala et al., 1992). It has become very important commercially, especially in the animal feed industry. Prior to 1991, Kenya relied heavily on imported fishmeal, mainly herring meal from Denmark. In 1992, Kenya adopted a Structural Adjustment Program as recommended by the PMF. As a result of this, the value of the Kenya shilling fell sharply against the major currencies of the world and importation of goods became very expensive. Feed manufacturers turned to omena fishmeal to replace imported fishmeal. According to the Kenya Bureau of Statistics, before 1991 the country imported more than 350,000 metric tonnes of fishmeal every year. This figure fell to 800 metric tonnes in 1991. Currently, omena fishmeal is widely used in the animal feeds industry. Despite this widespread use, no studies have been done to assess its quality and feeding value. The Kenya Bureau of Standards (KBS), which sets standards for consumer products, has not set any specifications for omena fishmeal due to lack of scientific research data. Hence there is need to assess the quality and feeding value of omena fishmeal relative to the imported fishmeals. Specific objectives of this study were: a) To determine the effect of reducing fibre content in sunflower cake on the apparent digestibility of protein, energy, and organic matter using tilapia (O. niloticus) as the test animal. b) To compare the nutritional values and protein qualities of diets based on high-fibre and low-fibre sunflower cakes, omena and anchovy fishmeals, when fed to tilapia (O. niloticus) at two levels of dietary protein. 3 c) To establish the highest level at which high-fibre and low-fibre sunflower cakes could replace fishmeal in diets of tilapia (0. niloticus) without affecting growth, and to evaluate the effect of substituting sunflower cake for fishmeal on whole body fatty acid composition. d) To determine the most limiting amino acids in diets based on sunflower cake fed to O. niloticus, and to evaluate the effect of supplementing these diets with the fore-going amino acids on fish performance. 4 1.2 References Bowen, S.H., 1982. Feeding, digestion and growth - Qualitative considerations In: The biology and culture of tilapia. R.S.V. Pullin and R.H. Lowe Mclonell (Eds.) I C L A R M , Manila Philipines. Cargill Inc., 1980. Industry News: America's native oilseed crop rediscovered. Journal of American Oil Chemists Society, 57: 264 - 268 Chitamwemba, D B F . , 1992. The fishery of Rastrineobola argentea in Southern Sector of Lake Victoria. In: The Lake Victoria dagaa 7?. argentea. Report of the first meeting of the working group on the Lake Victoria R. argentea, 9 - 1 1 Dec. 1991, Kisumu, Kenya. UNDP/FAO Regional Project for Inland Fisheries Planning. IFIP FAO., 1992. Report of the sixth session of the CLFA subcommittee for the development & management of the fisheries of Lake Victoria, 10-13 February, 1992. FAO Fish Reports. Jacob, J.P., 1993. The feeding value of Kenyan sorghum, sunflower seed cake, and sesame seed cake for poultry. Ph.D. Thesis, The University of British Columbia. Lovell, T., 1991. Nutrition of aquaculture species. J. Anim. Science, 69: 4193 - 4200 Manyala, J.O., Nyawade, C O . , and Rabour, C O . , 1992. The Dagaa (Rastrineobola argentea Pellegrin) Fishery in the Kenyan Waters of Lake Victoria: A natural review and proposal for future research. In: The Lake Victoria dagaa (R. argentea). Report of the first meeting of the working group on L. Victoria R. argentea. 9 - 1 1 December, 1991, Kisumu, Kenya. UNDP/FAO Regional Project for Inland Fisheries Planning IFIP. Tranchino, L. , Melle, F., and Sodini G., 1984. Almost complete dehulling of high oil sunflower seed. Journal of American Oil Chemists Society., 61: 1261 - 1265. Wandera, S B . , 1992. A study of R. argentea in Ugandan lakes. In: The Lake Victoria Dagaa. Report of the first meeting of the working group on Lake Victoria R. argentea 9 -11 December, 1991, Kisumu, Kenya. UNDP/FAO Regional Project for Inland Fisheries Planning IFIP. Wanink, J. H . , 1989. The Ecology and the Fisheries of dagaa (R. argentea). In: Fish Stock and Fisheries in Lake Victoria. A handbook to the Hest/TAFIRI & FAO/DANTDA regional seminar, Mwanza, January - February 1989. Report of the Ftaplocromis Ecology Survey Team, (HEST) and the Tanzanian Fisheries Research Institute (TAFIRI) no. 53, Leiden, The Netherlands, RUL. 5 Chapter 2 Review of literature 2.1 Aquaculture in Kenya Development of aquaculture in Sub-Saharan Africa is relatively recent. Most of the aquaculture systems were introduced in the last 35 years. Trout farming in high altitude areas was first introduced in South Africa in 1859, and in Kenya in 1910. According to FAO statistics (Coche et al. 1994), total aquaculture production in Africa in 1990 was 14,700 metric tonnes, which was equivalent to 0.5% of world aquaculture production. The estimated value of this production was US$ 25 million. Nigeria, Ivory Coast, Zambia, and Kenya were among the largest producers. More than 30 indigenous and exotic species are cultured in the region. Tilapia, particularly Oreochromis niloticus are the major species cultured, but other species like Clarias gariepinus (catfish) and Cyrpinus carpio (carp) are also important. Vincke (1995) lists three production systems practiced in Sub-Saharan Africa, viz., extensive, semi-intensive and intensive systems. The extensive system is the oldest, where aquaculture is mainly rural and directed to satisfying nutritional needs of the family. Small and large-scale commercial farmers prefer the semi-intensive system where aquaculture is integrated with raising farm animals. This system is becoming increasingly important in the development of aquaculture in the region. Intensive farming has not been fully developed (Coche et al., 1994). There are only a few private commercial farms in Kenya, Malawi, Nigeria, Zambia and Zimbabwe. At the Continental level, there are various constraints to fish farming, such as a lack of a good national data bank, a lack of good statistical production data, scarcity of public funds and a lack of good co-ordination between researchers and 6 producers (Coche et al., 1994). There are also social and technological constraints such as an inaccessibility to credit for small- scale fish farmers, an excessively low technological level of the farmers and a shortage of various key feed ingredients because of competition for food for humans and other animal species such as poultry and swine. Kenya has 10,000 square kilometers of inland lakes and 6500 km of coastline. Eighty percent of the fish landings are from fresh water lakes, and 19% from marine sources . Aquaculture contributes only 0.5% of the total fish production (Table 2.1). About 86% of the fish from inland waters come from Lake Victoria, while 6% comes from Lake Turkana. Other lakes and Rivers contribute 8%. The main species in the wild catch are Lates niloticus (Nile perch), Rastrineobola argentea (Omena), Oreochromis niloticus (Nile tilapia), Cyprinus carpio (Common carp) and Micro salmoides (black bass). The history of fish farming in Kenya dates back to 1910 during the colonial era. European settlers, unfamiliar with Kenyan indigenous fish, imported trout (Onchorhycus mykiss and Salmo trutta), black bass, and common carp. The fish were stocked into various rivers for sport fishing. Black bass was also stocked into Lake Naivasha. Government involvement in fisheries started around 1926, with allocation of funds for the care of trout and trout fishing. From 1926 to 1937, the fisheries program was administered by the Game Department. In 1954, a separate department for fisheries was formed and a trout hatchery established at Kiganjo. After the Second World War, the government of Kenya started showing an interest in raising indigenous fish, particularly tilapia, as a potential food crop for the rural population. Table 2.1: Fish Production in Kenya (metric tonnes) Year Aquaculture Inland Marine Total capture capture Production 1987 310 124,096 6,875 131,281 1989 530 131,000 14,566 146,096 1993 1,014 167,510 14,966 183,490 1994 1,114 173,500 28,249 202,863 1995 1,083 154,164 38,541 193,788 Source: Fisheries Department, Kenya (1995). 8 A program for stocking dams and ponds was started in Western Kenya and in the sixties, a campaign "eat more fish" was launched and quickly spread to various parts of Kenya, including the non-fish-eating communities in Central Province (Ochieng, 1994). Currently, fish farming is mostly practiced as part of other farming activities. At the national level, the contribution of fish farming to fish production is insignificant, but it has an important effect on nutrition and income at the farmer level. Besides, the main sources of fish, which have traditionally been the fresh-water lakes, particularly Lake Victoria, are having problems with water hyacinth, pollution, over-fishing, and the disappearance of some species from the catches. Consequently, the gap between the national fish requirement and production can only be met through aquaculture. 2.2 Fish species cultured According to Balarin (1985), the warm-water species currently cultured in Kenya are tilapias (O. niloticus, O. mossambicus, T. rendali and T. zilli), common and mirror carp (Cyprinus carpio), and black bass (Micropterus salmoides). Rainbow trout (Onchorhynchus mykiss) and to a lesser extent, brown trout (Salmo trutta) are cultured in high-altitude cold-water areas. Marine shrimp (Penaeus indicus and P. monodon) are cultured at the coast. There is a wide range of culture practices. Small family fish farms consist of earthen ponds (130 m2 to 1000 m2), stocked with tilapia (Western Kenya), or tilapia and carp (Central Kenya) (Ochieng, 1994). Water may be stagnant or flowing through. On-farm organic fertilizers may be applied at varying rates. Productivity is in the order of 500 - 2000 kg/ha/year (Ochieng 1994), with the fish being consumed by the house-hold. Tilapia fry are produced at the Department of Fisheries in Sagana, and at the Lake Basin Development Authority production centers. Trout farming is done on 9 commercial farms at the slopes of Mt. Kenya. Trout farming requires clean, clear, cold (10-18 °C) water flowing in large quantities; this restricts its practice in Kenya. In addition, investment and operating costs are high. 2.3 Dietary Protein and Amino Acid Requirements of fish Dietary protein quality and quantity are major factors that influence fish performance. Protein is the most expensive component of fish diets. Most studies done on protein requirement in fish have been designed to maximize growth; hence growth rate has been the main criterion used to determine requirement (Siddiqui et al, 1988; Wang et al, 1985; Santiago et al, 1982; De Silva and Perera 1985). On small farms, such as the family farms in Kenya, where the fixed costs are low and feed costs would be the major components of the variable costs, a more suitable measure of diet quality would be feed conversion ratio. Protein requirement as a percentage of diet is higher for fish than for most terrestrial animals. Some researchers have explained this by relating requirement to the feeding habits of fish, pointing out that most fish are carnivorous and hence the high protein requirement (Cowey, 1975 and Watanabe et al, 1979). However, high protein requirements are also a characteristic of omnivorous and herbivorous fish such as common carp, tilapia and grass carp. Not much difference has been noted between the requirements of these latter fish and the carnivorous fish. A plausible explanation may be that a portion of the ingested protein is catabolized for energy. In fish, both lipids and proteins are readily available energy sources, while the value of carbohydrates as an energy source varies among species. It has been shown that tilapia (O. niloticus) (Popma, 1982) and channel catfish (Wilson and Poe, 1985), which are warm water herbivorous 10 and omnivorous fish respectively, digest over 70% of the gross energy in non-cooked starch, while rainbow trout, which are cold water carnivorous fish, digest less than 50% (Cho and Slinger 1979). Another reason for the higher dietary protein concentration is that fish have lower dietary energy requirement because they exert relatively less energy to maintain position, do not maintain a constant body temperature, and excrete most of their nitrogenous waste as ammonia. Wilson and Halver (1986), stated that fish do not have a requirement for protein per se, but rather require amino acids that are usually obtained from the diet by digestion of protein. All fish require the 10 indispensable amino acids that are required by other animals (Cowey, 1994), i.e. arginine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, threonine, tryptophan, and valine. Tyrosine is a non-essential amino acid that has a sparing effect on phenylalanine, while cystine has been shown to spare methionine in fish diets (Page, 1978). Maintaining an optimum amino acid balance is essential for optimal fish growth. Excess methionine has a depressing effect on growth, which may be due to its inhibitory effect on the absorption of neutral amino acids (Ingham and Arme, 1977). Similarly, a growth reduction has been observed when the ratio between leucine and isoleucine is increased (Nose, 1979). Antagonism between branched chain amino acids has been reported in mammals (May et al, 1991; Hargrove et al, 1988; Calvert et al, 1982), and in fish (Hughes et al, 1984). Tilapia, compared to other species of fish, require relatively low concentrations of dietary crude protein (NRC 1993). There have been several attempts to determine protein requirements, and a wide range of dietary protein levels has been proposed. One of the major problems with the stated requirements (Jauncey, 1982; Cruz and Laudencia, 1977; 11 Davies and Stickney, 1978), is that diets used to determine them were not formulated to meet the requirements of essential amino acids, which were not quantified until recently (Santiago and Lovell, 1988). Scott et al. (1982) stipulated that amino acid requirements might differ depending on the balance (excesses and deficiencies) of amino acids > in the diet. It therefore follows that diets formulated to determine protein requirements must first and foremost meet the requirement for amino acids. Lack of knowledge of amino acid requirements may explain some of the variation in the stated requirements. There are only three studies that have investigated the requirements of some tilapia species for some or all of the essential amino acids. Jackson and Capper (1982) studied the requirements of O. mossambicus for lysine, methionine and arginine, while Jauncey et al. (1983) studied the essential amino acid requirement of the same species based on the amino acid analysis of fish flesh protein. Santiago and Lovell (1988) studied the requirements of O. niloticus fry (15 to 87 mg) for 10 essential amino acids using casein /gelatin diets. The latter study represents the most extensive and complete study on the essential amino acid requirements of tilapia. The values reported were significantly higher than those reported by Jauncey et al. (1983) for O. mossambicus. Santiago and Lovell (1988) postulated that essential amino acid requirements of the two related species differ considerably. Research in this area encounters methodological problems such as lower growth rates when fish are fed diets based on synthetic amino acids (Mazid et al, 1978; Yamada et al, 1982). The reason for this has not been established. A further confounding factor is that the minimum requirement for crude protein varies with the rearing system used. Clark et al. (1990) observed no significant differences in weight gain or FCR among tilapia grown in outdoor seawater pools and fed 12 diets containing 20, 25, and 30% crude protein. The authors postulated that fish on the lowest level of dietary protein maintained a good growth rate by feeding on algae present in the pools. A wide range of estimates of the optimal dietary crude protein concentration for tilapia has been reported. Winfree and Stickney (1981) reported that 56% crude protein promoted maximum weight gain in tilapia (O. niloticus) weighing 2.5 g. Shiau and Huang (1989) reported 24% crude protein as the optimum crude protein concentration for tilapia (O. niloticus x 0. aureus) weighing 2.9 g. Fish in the latter study were maintained at a salinity of 32 -34 ppt. Water salinity has been observed to influence protein requirement, being lower at full salinity than in fresh water (Shiau and Huang, 1989; Clark, 1990). Luquet (1991) reviewed several studies, and recommended 30-35% crude protein as the optimum for tilapia. In making this recommendation, the author relied on studies that utilized good protein sources such as fish meal and casein. The quality of dietary protein affects protein requirements. Both fishmeal and casein have good amino acid profiles and good protein digestibility and are considered to be high quality protein sources. The quality of fishmeal, however, may vary depending on the species of fish, processing method and freshness of the raw material used (Anderson, 1996). In most of the published studies, the type of fishmeal used has not been specified. McCallum and Higgs (1989) reported that low-temperature dried herring meal had a slightly reduced protein quality compared to freeze-dried herring meal, whereas the high-temperature-dried meal had a dramatically reduced protein quality. 13 Estimates of optimal dietary protein levels have typically been in the range of 35 -40% for tilapia weighing less than 5 g (Mazid et al, 1979; Jauncey, 1982; Siddiqui et al., 1988). The crude protein requirement for tilapia is inversely related to their size. Winfree and Stickney (1981), observed that maximum weight gain in tilapia (O. aureus), weighing 2.5 g, was realized when diets contained 56% crude protein, while in fish weighing 7.5 g, 34% crude protein was adequate. Siddiqui et al. (1988) observed that the optimal dietary protein level for Nile tilapia (O. niloticus) fry weighing 0.8 g was 40%, while the corresponding level for fish weighing 40 g was 30%. Twibell and Brown (1998), determined that the crude protein requirement for tilapia with an initial weight of 21 g was 28%). In the latter study, the protein content of the diet was increased by increasing the level of soybean meal. The diet containing the highest level of crude protein (34%) contained 35% soybean meal. Soybean meal contains anti-nutritive factors, the content of which may vary depending on the processing method. It is not clear whether the higher level of soybean meal in the high-protein diets may have caused the depressed growth rates observed at the higher protein levels. Thus, the stated requirements for protein show a wide variation, reflecting the different environmental conditions in which the studies were done. Fish factors such as size, and stocking density also affect the requirements. Similarly, dietary factors such as protein quality, the ingredients used and the way they were processed would all affect requirements. The stated values for optimal dietary protein level have been estimated from growth response curves. There are various problems when the requirements are estimated in this way. Growth is a non-specific response, and it is affected by many 14 factors such as temperature, water quality, biomass density and water flow rates. Furthermore, different growth rates were attained for the "optimal" diet in the various experiments, and, in some of the studies, the growth rates were quite low for the "optimal diet", indicating that the environmental or dietary factors or both were not really optimal. The other major problem in stating protein requirements for tilapia is that there is very little information available on the digestibility of feedstuffs. Digestible energy (DE) concentration is an important factor affecting protein requirement (Scott et al., 1982). If the DE content of the diet is low, most of the protein will be catabolized for provision of energy. Unfortunately, data on the digestible energy content of major dietary components generally used in tilapia diets are inadequate, and this has hindered the expression of protein requirements in relation to digestible energy content of the diet. Protein to energy ratio is important in determining the requirement of protein. A low protein to energy ratio will lead to slow growth, while a ratio that is too high would lead to catabolism of proteins for provision of energy. Cisse (1996) observed that in tilapia S. melanotheron a protein:energy ratio of 70mg protein.kcal"1 was optimal. Values higher or lower than this resulted in poorer growth performance. In determining protein requirements, some authors have treated energy and protein as independent variables, although their effects are not independent. Protein can be catabolized for energy, and energy is used in the synthesis of half of the 21 amino acids used in growth and metabolism (Bowen et al, 1995). Furthermore, feed intake is regulated by the available energy content of the diet. The other reason for the different values reported by various authors would be in the source of protein used. Low-temperature (LT) high-quality fishmeal would contain a favorable amino acid profile for the growth of most fish (Anderson, 1996). In contrast, 15 most plant proteins may be limiting in one of more of the indispensable amino acids. It is therefore unlikely that diets with LT fish meal as the main source of protein would result in the same optimal dietary protein estimate as diets composed entirely of plant feedstuffs. 2.3.1 Search for new sources of protein Tilapia are the third largest group of farmed fish species after carp and salmonids (FAO, 1997). Nile tilapia was the sixth most cultured fish species in the world in 1995, with a total production of 473,641 m.t, and an average annual increase in production of 12% per annum since 1986. Between 1984 and 1997, the global production of farmed tilapia increased more than three-fold i.e., from 186,544 tonnes to 659,000 tonnes, and in 1995, it represented more than 4.48% of the total farmed fish, with a value of US$ 925 million (Tacon, 1997). Feed represents the highest operating cost in intensive fish aquaculture, with protein being the most expensive dietary component. Traditionally, fish diets have been based on fishmeals as the main protein sources due to their high protein content, good amino acid profile, and excellent supply of essential fatty acids, minerals and vitamins of high digestibilities. Fishmeal is also the single most expensive ingredient in aquaculture feeds (Tacon, 1993). A reduction in world production of fishmeal, coupled with increased demand and competition with terrestrial domestic animals for the limited supply, has further increased fishmeal prices. Many developing countries are unable to afford fish meal for inclusion in feed for fish and other domestic animals. Furthermore, high dependence on fishmeal would make fish prices high and less competitive compared to other meats. For that reason, considerable research has been done to evaluate new 16 protein sources. El-Sayed (1999) reviewed the alternative protein sources tested as replacements for fishmeal in tilapia diets. Fishery by-products have shown promising results. One such product is fish silage which is prepared from fish or fish-processing wastes. Feeding experiments have indicated that fish silage can replace fishmeal in tilapia diets. The nutritional value of silage depends on the source of the fish species, and on the care taken during silage preparation (Fagbrenro and Jauncey, 1994). In particular, good knowledge is needed regarding the chemical changes that occur during the digestion and storage of silage. The nutritional quality of fish silage can be improved by limiting the extent to which proteins are hydrolyzed to polypeptides and free amino acids. Termination of the ensiling process after 3 - 7 days was shown to result in improved weight gain, protein efficiency ratio, biological value and net protein utilization when these products were fed to mink (Screde, 1981), calves (Offer and Hussain, 1987) and salmonids (Lall, 1991). Fish silage that has been acid or enzymatically digested, is a viscous liquid that is difficult to transport, store, or feed to animals. It has a low solid and high moisture content which makes it difficult to dry. Carbohydrates, cereals, crop residues and by-products have been used as filler materials, making it possible to dry the silage in conventional driers. In tilapia, fish silage can successfully replace fishmeal in diets. Lapie and Bigueras (1992) fed Nile tilapia fish silage preserved in formic acid, and blended with fishmeal in a 1:1 ratio, and observed that growth rate was similar to that of the fish fed on the fishmeal control diet. When silage to fishmeal ratio was increased to 3:1, growth was significantly reduced, presumably due to the high acidity of the diet, which may have depressed the appetite of the fish. Formaldehyde formed in the ensiling process inhibits protein hydrolysis (Haard 17 et al., 1985; Hussain and Offer, 1987), but may be toxic to some animals at high concentration. Fagbenro and Jauncey (1993) found that fermented fish silage blended with soybean meal, hydrolyzed feather meal, or meat and bone meal in a ratio of 1:1, could replace 75% of the fish meal in diets for Nile tilapia, with no significant differences in weight gain, or hemoglobin and hematocrit levels in the fish. Terrestrial animal by-products have been used successfully in tilapia feeds. Poultry by-product meal, hydrolyzed feather meal, blood meal, and meat and bone meal have high protein contents. Unfortunately, most of these protein sources are deficient in one or more of the essential amino acids, particularly lysine, isoleucine, and methionine (Tacon and Jackson, 1985). When the limiting amino acids are supplemented, the diet quality is improved. Tacon et al. (1983) found that hexane-extracted meat and bone meal alone, or mixed with blood meal, in the ratio of 4:1, and supplemented with methionine, successfully replaced 50% of fish meal protein in diets fed to Nile tilapia fry. When blood meal was used alone, the results were still comparable to those of the control. This was contrary to a later study by El-Sayed (1998), who observed significantly reduced growth rates and feed efficiencies when fish meal was replaced by blood meal. The differences between the two studies could be due to the fact that methionine, which is deficient in blood meal was supplemented in the first but not the second study. Hydrolyzed feather meal has been used as a protein source for tilapia with contradictory results. In studies by Tacon et al. (1983), Viola and Zohar (1984), and Davies et al. (1989) with O. niloticus, O. mossambicus, and all male tilapia hybrids respectively, fish fed on diets based on hydrolyzed feather meal exhibited poor performance, presumably due to poor digestibility and low levels of lysine in the meal. 18 On the contrary, Falaye (1982), Bishop et al. (1995) and Gaber (1996) observed that hydrolyzed feather meal could replace between 40% and 66% of the fishmeal in tilapia diets. Chicken offal silage has also been tested (Belal et al, 1995) in O. niloticus fingerlings weighing 10.8 g. The authors only tested the range of 0-20% inclusion level and found that it could replace fishmeal up to the 20% level. Additional studies should have been conducted to determine the highest level to which chicken offal silage could replace fishmeal. Animal manures have also been used as protein sources for tilapia. Alhadrami and Yousif (1994) reported that camel and cow manures could be successfully incorporated in tilapia diets at 10% and 20% levels, respectively. It is not clear, however, whether the increased growth of the fish was the result of direct consumption of the manures, or whether the manures increased natural food productivity in the ponds. Plant protein sources have also received considerable attention as full or partial replacements for fishmeal. Among them, soybean meal has been the most widely used. It has a high protein content and a good essential amino acid profile, but is limiting in lysine, and the sulfur amino acids, methionine and cystine. Raw and under-heated soybeans contain proteins that inactivate the digestive enzymes trypsin and chymotrypsin, and cause agglutination of red blood cells in-vitro (Scott et al., 1982). Heat treatment inactivates these proteins, making soybean meal a major protein source in diets of many fish species. In tilapia (O. niloticus), studies to evaluate the potential of soybean meal to wholly or partially replace fishmeal have yielded varying results. In most of the studies 19 conducted, soybean meal could replace between 67% and 100% of the fishmeal, depending on fish species and size, dietary protein level, source of the soybean meal, processing methods and the culture system used. Studies done also indicate that there are no added benefits in supplementing diets based on soybean meal with the assumed limiting amino acids. For instance, in studies conducted by Tacon et al. (1983) and Jackson et al. (1982) pre-pressed solvent extracted meal, with or without methionine supplementation, was found to successfully replace 75% of the fishmeal in diets fed to tilapia (0. niloticus) and (0. mossambicus), respectively. Also, Shiau et al. (1989), found that 67% of the fishmeal in diets for hybrid tilapia (O. niloticus x O. aureus) could be replaced by soybean meal at a low dietary protein level (24%). Despite the low protein level and the high content of soybean meal in the diets, addition of methionine did not improve fish performance. Similarly, Viola et al. (1988) did not observe any improvement in the growth of hybrid tilapia (O. niloticus x O. aureus) fed diets in which 50% of the protein originated from soybean meal, supplemented with the amino acids lysine and methionine versus soybean meal alone. The lack of response to amino acid supplementation in the quoted studies is surprising considering that most of the soybean meal used was heat treated to inactivate anti-nutritive factors. The Maillard reaction between reducing sugars and amino acids, particularly lysine, results in linkages that are not hydrolyzed by digestive enzymes. These amino acids become unavailable to the fish, even though they are chemically present. It would therefore be expected that supplementing such diets with amino acids such as lysine would improve fish performance, but this is contrary to what has been observed. Viola et a/.(1988) postulated that tilapia are able to utilize lysine from heat-20 treated soybean meal because of their stronger gastric digestion compared to that of most other fish. Also, it can be postulated that the lysine and methionine requirements of tilapia are lower than those generally assumed. Minerals, rather than amino acids may be the factors that limit the efficient utilization of soybean meal in tilapia. Ogino et al. (1979) found that availability of phosphorus from bone and from plant protein sources was low in carp, which are stomachless, but in contrast, tilapia could effectively utilize mineral phosphorus, but were also unable to utilize phytin phosphorus. Thus, replacement of animal proteins with plant proteins creates a phosphorus deficiency in tilapia, which should be balanced by dietary inclusion of a suitable mineral supplement. The extent to which soybean meal protein can replace fishmeal protein in tilapia diets is affected by dietary protein level. Davis and Stickney (1978) found that at a low dietary protein level (15%), replacement of fishmeal protein with soybean meal protein caused growth depression, while at a high protein level (36%), soybean meal protein could totally replace fish meal protein in the diets without any significant decline in fish performance. This was contrary to the observations of Shiau et al. (1989), who reported that at 24% dietary protein, soybean meal protein could effectively replace 67% of the fishmeal protein in diets of tilapia (O. niloticus x O. aureus). When the diets contained 32% protein, replacing 30% of the fishmeal protein with soybean meal protein significantly decreased fish growth and FCR. The differences in findings between the two studies may have been caused by the type of soybean meal, and the way it had been processed. In the studies by Shiau et al. (1989), the anti-nutritive factors in soybean meal may not have been completely destroyed, thus causing the reduced growth observed 21 when soybean was added at high levels in the diet. Indeed, El Sayed (1999) postulated that the differences in findings between the two studies may have been related to the quality and processing of the soybean meal, fish species, size and culture systems used. The potential of cotton seed cake as a replacement for fishmeal has been studied in many fish species. Cotton seed cake is one of the most available plant protein sources in the world. It is relatively cheap, has a protein content ranging from 26% to 54%, air-dry basis), and a reasonably good amino acid profile for a plant protein. However, it has relatively low levels of lysine, cystine and methionine and contains a phenolic anti-nutritive compound, gossypol which is toxic to many animal species and also binds to lysine, reducing its availability (Jauncey and Ross, 1982). Its effect on fish is species specific. In rainbow trout, Herman (1970) reported that 0.03% free gossypol was toxic, while Dorsa et al. (1982) found that channel catfish could tolerate up to 0.09% free gossypol without any reduction in growth. In tilapia (O. aureus), Robinson et al. (1984) reported that a free dietary gossypol content of 0.2% had no adverse effect on fish performance. Various authors have studied the use of cotton seed cake as a protein supplement for tilapia, with inconsistent results. Ofojekwu and Ejike (1984) found that O. niloticus fed diets with cotton seed cake grew at slower rates than fish fed the fishmeal control diet. Similarly, Abdel-Fattah and El-Sayed. (1990), working with O. niloticus fingerlings, observed that fish fed on diets where 65% to 80% of the protein originated from cotton seed cake had poor weight gains compared to fish fed on the fishmeal control diet. Supplemention of the diets containing cotton seed cake with lysine did not improve fish performance. In a later study, Abdul-Aziz et al. (1999), using Nile tilapia 22 fingerlings, found that fish fed on diets in which 25% of the protein was from cotton seed cake had lower growth rates compared to those fed on the control diet based on fishmeal. In contrast to the above findings, Jackson et al. (1982) successfully used diets in which 50% of the fishmeal protein was replaced with cottonseed meal. There were differences in the way the various studies were done, which may account for the differences in the results. In the study by Ofojekwu and Ejike (1984) the diets contained unconventional feed ingredients, which may have affected the results. Gari, defined as grated cassava partially fermented, dried under the sun, and dried to 13.5% moisture level, was one of the ingredients used. Similarly, the cellulose used in the study was prepared by soaking filter paper in hot water and extracting it with 18% KOH for 24 hours. The growth rates obtained for fish of initial weight 3.71 g ranged from 0.08 % to 0.38%, while growth rates of fish with an initial weight of 45 g were 0.23 to 1.05%. These growth rates were low even for fish fed the control diet, indicating that there were other factors that affected the fish negatively. Abdel-Fattah and El-Sayed. (1990) used a control diet where wheat-bran was incorporated at a level of 60%. Recent studies have shown that the digestibility of protein and energy in wheat bran is low in tilapia (Popma, 1982; Anderson et al., 1991). The use of such high levels of wheat bran in the control diet, therefore, would have negatively influenced the digestibility of the energy and protein. In the studies by Jackson et al. (1982), the fish on the control diets also performed poorly, making it difficult to draw any firm conclusions from the findings. Other oilseed by-products that have been tested in diets for tilapia have included groundnut, sunflower meal, canola meal, rapeseed meal, sesame meal, copra, macadamia nuts, palm kernel and defatted cocoa cake (Jackson et al., 1982; Pereira and Pezzato, 23 1999; Higgs et al, 1990; Davies et al, 1990; Hossain et al, 1992; Guerrero, 1995; Balogun and Fagbenro, 1995; Omoregie and Ogbemudia, 1993; Fagbenro, 1988). All may have good potential as protein sources for tilapia, but further work needs to be done on their utilization. Leguminous and cereal plants and their by-products have been tested as partial replacements for fishmeal in tilapia diets. Leucaena leaf meal (LLM, 30% CP) from the plant Leucaena leucocephala is a potential protein source. The plant is drought resistant and the leaves have a high protein content. It is widely used as an animal feed, particularly for ruminants. However, the presence of the toxic amino acid mimosine limits its use in diets for monogastric animals. It is also low in the essential amino acids, arginine, threonine, isoleucine, histidine and methionine (Lim and Dominy, 1991). Several studies have been undertaken to assess the potential of using leucaena leaf meal as a protein source in tilapia diets, with varying results. Salaro et al. (1995) observed that Leucaena seed meal could comprise only 20% of the dietary protein in O. niloticus fry weighing 0.5 g. In larger fish, Santiago et al. (1988), noted that the fish performed poorly when leucaena leaf meal exceeded 40% of the diet. Mimosine in leucaena can be degraded to a relatively less-toxic form, through various processing methods, thereby increasing its nutritive value. Wee and Wang (1987) found that fish fed diets with leucaena leaf meal that had been soaked in tap-water for 48 hours and sun-dried for 12 hours had better growth rates than those fed the diets in which the leaf protein had only been sun-dried. The soaked leucaena leaf meal could supply 25%) of the total protein in the diet. In studies by Osman et al. (1996), the best performance was found for tilapia that were fed diets containing leaves that had been 24 dried for 48 hours or autoclaved for 15 minutes, compared to the leaves that had been treated with sodium hydroxide or incubated with rumen liquor. In conclusion, leucaena leaf protein can be used successfully as a protein source in tilapia diets. It can supply between 20% and 40% of the protein depending on the fish size and the processing method used to detoxify the mimosine. Despite the fact that mimosine can be completely or partially destroyed, the nutritive value of the leaf protein appears to be limited by other factors. Single cell proteins (SCP) such as unicellular algae, bacteria, cyanobacteria, and yeast have received a lot of attention in tilapia culture. Of particular interest has been the the biosynthesis and utilization of SCP by tilapia within intensive and semi-intensive farming systems. The production of single-cell proteins is simple, inexpensive and an effective way of producing natural fish food. Chamberlain and Hopkins (1994) reported that spraying a source of carbon such as wheat bran or cellulose on the surface of pond water with continuous aeration, at the optimum carbon:nitrogen ratio (15:1), would increase bacterial growth. Bacteria that are produced consume the carbon as an energy source and reduce ammonia concentration through nitrification. Fish may feed on the bacteria so produced directly, or they can be harvested and used commercially as a protein source. Viola and Zohar (1984) found that a commercial single cell protein diet (Pruteen, 70% CP) could replace 50% of the fishmeal protein in diets fed to tilapia hybrid (O. niloticus x O. aureus). 25 2.4 Sunflower (Helianthus annuus) 2.4.1 Taxonomy and origin of the domesticated sunflower Commercial sunflower (Helianthus annuus) belongs to the compositae family. All species of Helianthus are native to the Americas, where archeological evidence reveals that wild sunflower was used by American Indians as a source of food, and also in medicine and ceremonies (Heiser, 1955). Sunflower seed meal was mixed with flour to make bread. Oil from the seeds was used to season food and anoint hair, and it also served as a base for pigments. Further, sunflower was used as a medicine to cure rattlesnake bites and for other remedies (Heiser, 1976). After the discovery of the Americas, sunflower was introduced to Europe, where it spread eastward and northward, eventually reaching Russia in the middle of the eighteenth century (Zukovsky, 1950). By the beginning of the twentieth century, sunflower became a major edible oil crop in Russia. Russian plant breeders devoted much effort to the improvement of sunflower productivity and disease resistance in cultivated plants. They increased the oil content of the seeds from 28% in the 1920's to 43% by 1935, and to 49% in 1955. Presently, some varieties have oil contents that exeed 50%. Cultivated sunflower was re-introduced to North America by the immigrants from Europe around 1875. Currently, Russia is the leading producer of the crop followed by France, USA, China and Spain (Putt, 1997). Eighty percent of the monetary value of sunflower is derived from the oil. The meal is the main by-product after oil extraction, and it contains 30 - 45% protein. European settlers introduced sunflower into Kenya in the nineteenth century. It was cultivated as an export crop for bird feed (Zulberti, 989). When independence was attained in 1963, it was grown as a cash crop in the high-potential areas of the country. 26 Production of the crop declined steadily due to the fact that monetary returns from it were much lower than those from other cash crops such as wheat and corn, which could be grown in the same areas. Production has recently been revitalized, however, following government decontrol of consumer prices of edible oils and fats and animal feeds. Sunflower farming is also moving from high-potential to marginal areas where few alternative crops can be grown. 2.4.2 Morphology of the sunflower plant The most striking feature of the sunflower plant is the head inflorescence which carries the seeds. The floral head consists of individual small flowers which are congested and attached on a single horizontal plane to simulate a large individual flower. The whole intricate arrangement of the head and structure of the flowers is believed to be an adaptation to improve the efficiency of pollination by insects and by other means (Heiser, 1976). Each flower has a single ovary containing one seed that ripens into the fruit or achene. The achene consists of a seed (kernel), and adhering pericarp (hull). 2.4.3 Chemical and physical composition of sunflower seeds Two different types of sunflower cultivars are cultivated, viz., the oilseed varieties that have an oil content of 40 to 51%, and the low oil seed varieties in which the oil content varies between 21 and 32% (Earles et al, 1968). The high-oil seed cultivars are black-seeded and have thin hulls, which adhere to the kernel tightly. Edible oil is the main product of the oil-seed cultivars, with the meal being an important by-product. In addition to high oil content, high-oil seed cultivars generally have low hull content and smaller size compared to the low-oil cultivars. The low-oil seed varieties, referred to as "confectionary" sunflowers, have large stripped seeds, and relatively thick hulls which 27 remain loosely attached to the kernel (Vaughan, 1970; Park et al., 1997) They are mainly used in snack, confectionary, bakery and bird food markets. Within each type of cultivar, the composition of the seeds varies with location, year of planting, type of soil and cultural practices. The oil content of the high-oil cultivars compares favorably with that of other oilseeds (Unilever, 1976). Due to extensive selection for high oil content, sunflower seeds have a higher oil content than found in most other oil seeds except peanuts. 2.4.4 Sunflower seed oil Intensive selection has been done in the sunflower plant for high oil content (Senkoylu and Dale, 1999). The oil content was increased from 28% to over 50% during a period of seven decades (Zatari, 1989). In the seeds of any one cultivar, the oil content may vary depending on geographical location, environmental temperature, planting season and other cultural practices (Zatari, 1989). The fatty acid composition of sunflower oil is characterized by a low level of linolenic acid (NRC, 1993), and for that reason, the oil has excellent storage qualities. It has also has a lower level of palmitic acid than soybean oil, and a higher level of linoleic acid. Sunflower oil is considered to be a highly desirable oil for human consumption because of its light color, bland flavor, high smoke point, and high level of linoleic acid. The fatty acid composition of sunflower oil is affected by the environmental temperatures that occur during growth of the plant (Canvin, 1965). Seeds produced in cool climates contain 70% or more linoleic acid, while those produced in hot climates may contain as little as 25% (Unger and Thompson, 1982). The use of breeding to modify the fatty acid composition of sunflower oil has received little attention and, for 28 that reason, the fatty acid composition of the oil from seeds of different cultivars within similar environments is quite uniform (Kharchenko and Borodulina, 1976). Sunflower oil has high levels of oleic and linoleic acids. 2.4.5 Sunflower seed proteins The protein content of the sunflower kernel ranges from 9% to 24 %, and depends on variety, climate, soil and cultivation conditions (Dorrell and Vick, 1997). Selection of sunflower for a high oil content has resulted in an attendant decrease in protein content because protein and oil contents are negatively correlated with one another. Environmental temperatures also affect protein quality and content. In studies by Canvin (1965), the protein content of the seeds increased from 14% to 20% as the mean environmental temperatures increased from 10°C to 26°C. Under optimal processing and dehulling conditions, the protein quality of sunflower meal is equivalent to that of soybean meal. However, when processing conditions are harsh, or when excess heat is used to desolventize the meal, some decline in the biological value of the protein occurs due to destruction of lysine, arginine and tryptophan (Clandinin, 1958). The amino acid composition of sunflower meal and soybean meal is shown in Table 2.4. Sunflower meal protein is relatively deficient in lysine, but rich in sulfur amino acids when compared with soybean meal protein. 2.4.6 Composition of sunflower hulls Sunflower hulls are the outer covering of the sunflower seed. They make up 22% to 30% of the total weight of the seed. Cellulose, lignin and hemicellulose comprise 74% to 90% of the total components (Table 2.2), and are highly indigestible by animals. Lipids, proteins and minerals make up the rest. Generally, sunflower hulls contain more lignin 29 -v Table 2.2: Some fibre components of sunflower seed hulls and other agricultural residues (Earles etal, 1968) % (Air-dry basis) and range Material Lignin Pentosans Cellulose Oat hulls 17 39 36 Wheat straw 18 (15-21) 30 (27-32) 33 (29-37) Sugarcane bagasse 19 (16-22) 30 (27-32) 33 (30-37) Corncobs 14(8-17) 41 (31-45) 32 (22-39) Sunflower hulls 27 (25-30) 27 (25-31) 30 (29-32) 30 and less pentosans and cellulose compared to other agricultural residue materials. They can be utilized as a roughage source, but they are low in nutrient content, poorly digested and highly unpalatable (Park et al., 1997). Hulls derived from oil-seed processing may be of slightly higher quality and contain more protein and fat than the larger confectionery hulls. 2.4.7 Anti-nutritive factors in sunflower seeds Unlike most other oilseed meals, anti-nutritive factors are not a major problem in sunflower. The seeds, however, contain arginase and trypsin inhibitors which are heat labile and easily inactivated (Roy and Bhat, 1974). The potential use of sunflower protein isolates for human food is limited by the presence of phenolic compounds in the seed. During protein extraction in alkaline medium, chlorogenic acid and other phenolic compounds are oxidized to o-quinones and form covalent linkages with proteins giving dark-green or brown products (Sosulski and McCleary, 1972). While chlorogenic acid is not considered to be toxic, Delic et al. (1975) found that a 2% inclusion in the feed of mice resulted in a depressed feed intake, and a reduced weight gain. Methionine and choline chloride partially offset these effects. Both genotype of the seeds and environmental conditions during seed maturation have a direct effect on the concentration of chlorogenic acids in the seed. Dorrell, (1976) analyzed 38 inbred lines and found that the concentration of chlorogenic acid ranged from 1.4% to 4%. Early seeding and warm temperatures during seed maturation favored higher levels of chlorogenic acid. Eliminating or reducing the amount of chlorogenic acid in the seed through genetic selection is difficult because oil and chlorogenic acid content are positively correlated. 31 2.4.8 Processing methods The method used to process sunflower seeds is one of the most important factors that determines the composition of the meal. The seeds may or may not be dehulled prior to oil extraction, depending primarily on the design of the processing plant. In the older plants, dehulling equipment is not available, and the seeds are crushed whole. Modern processing plants dehull from 40 to 75% of the achenes, but even a very efficient dehulling system can only remove a maximum of 90% of the hulls from the seeds. The hulls are discarded as trash or used as fuel for plant operations (Dorrell and Vick, 1997). Dehulling has many advantages. First, it reduces the movement of unnecessary mass through the system, and second, it reduces wear and tear in the expeller. Third, it reduces the wax content of the oil and lastly, it reduces the fibre content of the meal (Dorrell, 1976). Three basic methods are available for extraction of sunflower oil, namely; mechanical screw press, direct solvent extraction, and a combination of screw-press and solvent extraction. 2.4.9 Sunflower meal Sunflower is grown for the oil, but the meal left behind after oil extraction is a valuable and nutritious by-product. The chemical composition of the meal compares favourably with that of most other vegetable-type meals. Exceptions are the higher fibre and ash contents of sunflower meal, which reduce its metabolizable energy. Approximately 8.3 million tons of sunflower meal were produced world-wide in 1990/1991, making sunflower meal the fourth largest source of oil seed meal following soybean, cottonseed and rapeseed meals (Dorrell and Vick, 1997). Almost all sunflower meal that is marketed comes from the processing of the black-hulled oilseed type sunflower. The chemical 32 composition of sunflower meal depends on the variety of the seed, the processing method, and the degree of dehulling or decortication (Earles et al, 1968; Ravindran and Blair, 1992). The oil content of the meal varies with the type and efficiency of the oil extraction process. If seeds are completely dehulled before oil extraction, a meal with a protein content in excess of 4 0 % can be achieved with a solvent extraction system, and 37%) protein with a mechanical extraction system. In contrast, if no dehulling occurs, the meal contains only 2 8 % protein after oil extraction by either method. Intermediate dehulling results in a sunflower meal with about 34%> protein. Sixty to sixty five percent of the sunflower meal produced in the USA and Canada is this type. The reminder is in the 28%) protein category. In Kenya, where the present project was done, the crude protein content of sunflower meal is typically in the range of 2 5 % to 30 %. The protein and fibre contents of the sunflower meal are inversely related. The fibre content has a negative impact on nutrient availability and the digestible energy level (Villamide and San Juan, 1998). Kondra et al, (1974) found that feeding a high- fibre diet (19 .6% fibre DM basis) to chicken layers and broilers resulted in significant reduction in food consumption and body weight gain, but had no effect on FCR during the growing period. The levels of proximate constituents in some Kenyan sunflower seed cakes are given in Table 2.3. There was considerable variation among the samples, which was mainly caused by differences in processing methods (Jacob, 1993). The levels of crude fibre were high, and ranged from 2 4 . 1 % to 40.2%o. Crude protein ranged from 24 .9% to 37.6%. The amino acid compositions of sunflower meal, soybean meal, cottonseed meal and rapeseed (or canola) meal are presented in table 2.4. 33 Table 2.3: Proximate compositions of Kenyan sunflower seed cakes (adapted from Jacob, 1993). A l l samples were produced by the expeller process. % ( D M basis) M i l l Sample no. % D M 1 C P E E C F Ash N F E *A 1 94.1 30.8 16.6 28.4 7.5 16.7 A 2 97.2 24.9 12.7 33.5 7.0 21.9 A 3 96.9 25.8 13.1 29.0 8.7 23.4 A 4 96.9 25.0 13.7 32.1 7.8 21.4 B 1 93.1 32.4 12.2 24.7 9.2 21.4 B 2 96.0 37.6 8.0 24.8 8.9 20.7 B 3 96.8 33.4 10.3 24.1 10.3 21.9 C 1 91.0 29.9 14.4 35.7 5.5 14.4 C 2 96.4 28.9 12.6 29.7 7.0 21.9 D 1 93.2 33.5 12.4 25.8 6.8 19.7 D 2 93.5 29.8 14.9 33.8 6.6 14.9 D 3 94.4 32.2 12.0 33.1 5.9 16.8 D 4 94.3 32.7 12.0 32.8 5.9 16.6 E 1 95.5 25.3 11.0 40.2 5.1 18.3 E 2 93.9 32.5 13.0 27.4 7.8 19.4 E 4 93.6 34.0 11.6 24.9 7.3 22.2 F 1 95.6 27.0 12.3 33.1 9.4 18.2 Misc . 1 92.6 27.5 14.5 29.6 6.3 22.1 Misc. 2 93.8 26.0 11.4 38.3 4.6 19.7 Mean 94.6 29.8 12.5 30.8 7.2 19.7 SD 1.7 3.7 1.8 4.6 1.5 2.6 M a x 97.2 37.6 16.6 40.2 10.3 23.4 M i n 91.0 24.9 8.0 24.1 4.6 14.4 'DM- Dry matter, CP-Crude protein, EE- Ether extract, CF-Crude fibre, NFE- Nitrogen-free extract * A , B , etc. refer to mills from which the samples were obtained. 34 CD CX 60 cd CD O c cd O 1 3 CD CD c/a CD & s-T3 C cd *cd CD £ -a o o "3 > O c/5 CD CD O 3 o c o O a o o o Cd o cd CD 00 Cd i-CD C N -CD ' (50 O o " N? i & 5P.9 t-. ^ 2 ex J3 +-» 6 .g '53 -*-» 2 PH | Q 2 CJ T3 U r-~ O 0 0 O S t-O C N i — i C N Tt >-< C N C N O O O N Tt 0 0 00_ V~i © C N CO Tt C N 0 0 o o o V) 0 0 Tt V O CN CO CN CN Tt O 0 0 N O 0 0 0 0 o O V) i n co >—i o\ ^-J ^ o O N © ro co vo co Tt oo o C N O Tt o m NO v~) © t— vo ro r- Tt Tt o ©' o ©' O C N C N H m H (S V"> Tt Tt Tt t vo i/1 o o o o Tt V} CN C— vo f v~i vo d d ©' ©' o o o o Tt 0 0 m Tt Tt C N O N O N 0 0 O C N O N O N T3 CJ CJ 1/1 CJ I 0 0 O N O N Q 1/-) CN co vo O V O co N O Tt r~ Tt O O N d O N Tt 0 0 O N 0 0 O N co 0 0 VO t. O V~l CO CN m O N v i r-i r- co «/i cN CN CO Tt CN i i r- M • — i v-> Tt V"> CN CO CN CN N O N O 0 0 C N vp I - ; r- oo_ C N i-H rt oo in r - . 0 0 C N - H < d ~ _• Z ON O CO i—i CO NO CO Tt C N <-i «-i I C ) N VO H CN 0 0 CO CN O O N 0 0 . Tt Tt < d o d Z ON NO C N C N r-4 0 0 r-I - H C N C N ON ON O i — l VI V O t ON d d d d r~ vo r- Tt r- vo vo r-© o d d V) N O O N N O co' vi t-i v-| CO Tt Tt CO 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 •a s -a •a 0 ) a 73 e "3 u a a> o OH a o O C N O Ul Ul Tt oo N O co NO VO i—I CN CO CN CN CO CO 1—I 0 0 Tt 0 0 O C N 1 VI O N ~ Tt Tt CO O N O N Tt C N r-r-i d CO v-> v i vo CO Tt N O Tt oo r-o i-1 O N 0 0 CN 0 0 VI t VO t d d d Tt CO oo v i Tt CO o o o O N O N O N VH o *3 I f a 3 O « CD > > 3 CD a *o I/) CD c o CD H DH OH H CD" 'c« >N 173 >N CD e O U CD '8 o +-» B Xi CD CD O cZ CD c 'H IS CD OH crude fibre). They speculated that the increased feed consumption response was an attempt by the animals to satisfy their nutrient requirements, and that the pigs were unable to consume enough feed for normal growth because of its high level of fibre. It is reasonable to conclude from the data of the above-mentioned studies that, when lysine level is adequate in the diet, adjustment in feed intake to satisfy energy requirement could be impaired, at least partially by the bulkiness of the diets containing sunflower cake. The sunflower meal used by Seerley et al. (1974) contained only 3% crude fibre, a value comparable to that of soybean meal. Hence, energy level and consequently feed consumption were not significantly affected by replacement of soybean meal with sunflower meal. On the other hand, in the studies by Moser et al. (1985), the sunflower meal had a high fibre content (20.5%). 38 2.5.2.2 Sunflower meal in poultry diets 2.5.2.2.1 Broilers High fibre content limits the use of sunflower meal in broiler diets. The dehulled meal contains at least 11% to 12% fibre, which is quite high compared to dehulled soybean meal that contains 3% crude fibre (NRC, 1994). This characteristic of sunflower meal may lead to bulky diets which could be a problem for young chicks due to the limited capacity of their digestive system. The density of the diet is of prime concern in terms of nutrient intake and the resultant growth rates. Pelleting has been shown to improve the utilization of sunflower meal in diets for poultry (Waldroup et al., 1970; Nir et al., 1994). An important factor to be considered when sunflower meal is included at high levels in diets for broilers is the energy status of the diet. Large amounts of dietary fibre reduce the available energy density of the diet. The metabolisable energy value of the meal depends on the effectiveness of the dehulling process. If sunflower meal is incorporated into diets at high levels, the nutrient and energy densities of the resulting diets may be significantly diluted and consequently growth is retarded. The addition of fat has been used to improve energy densities in broiler feeds. Zatari and Sell (1990), for example, found that adding 6% of an animal-vegetable fat blend to a broiler diet containing 20% sunflower meal containing 33%> crude protein and 18% crude fibre improved weight gain and feed conversion ratio. Lysine has been observed to be the first limiting amino acid in SFM diets fed to poultry. Supplementing lysine to these diets however, has given mixed results, depending on the other ingredients in such diets and their total lysine content. 39 2.5.2.2.2 Sunflower meal in layer diets Egg-type chicks are more tolerant to fibrous feeds than broilers because of their slower growth rates and greater capacity to adjust feed intake according to energy needs. McNaughton and Deaton (1981) reported that sunflower meal could be included at up to 30% in layers diets without adversely affecting body weight, egg production or egg weight. Birds responded to decreased energy in the diets by increasing feed intake. Similar observations were made by Deaton et al. (1979) using a high-fibre SFM (36% crude protein, and 24% crude fibre). At high dietary levels of inclusion of sunflower cake (30%>), FCR was significantly decreased, but egg production, egg weight and shell strength were not affected. The gizzards and intestines of the birds were enlarged. These findings suggest that layers tend to consume more feed to maintain the same rate of egg production when the diet contains high levels of SFM. The pelleting of such diets has been shown by several workers to increase the feed intake of layers (Waldroup et al., 1970; Nir era/., 1994; Jensen, 1998). 2.6 Use of sunflower meal in fish diets There is little information on the utilization of sunflower meal by fish. The few studies available have mainly been done with rainbow trout (Salmo gairdneri) and the results are not consistent. This is mainly due to differences in experimental methodology, and differences in the quality of the sunflower meal between studies. The maximum level of sunflower meal that can be included in fish diets without affecting growth rate or feed efficiency depends on a number of factors, These include the composition of the remainder of the diet and the nutritive value of the sunflower meal itself. Some of the 40 reported studies do not make reference to the crude fibre content, which is the main factor limiting the use of sunflower meal in fish diets. Tacon et al. (1984) fed rainbow trout diets containing 0 to 37% of solvent extracted sunflower meal with a fibre content of 24.7% (AD basis) and found no difference in performance between fish fed the diet with the highest level of sunflower meal and those fed the soybean meal control diet. Similarly, Morales et al. (1993) also evaluated sunflower meal as a protein source for rainbow trout. The sunflower meal used in their study was solvent extracted, with a fibre content of 16.5%. They found that sunflower meal could replace all the soybean meal in the diet, and could be included at a level up to 40%). In the study by Morales et al. (1993), rainbow trout fed on the soybean and sunflower meal diets had better feed intake than those fed on the fishmeal control diet, which had a reduced feed intake and performed poorly. These results suggest that the fishmeal used in the latter study was of poor quality. Scott and co-workers (1982) also noted that fish fed diets based on sunflower meal (17.5% fibre) had better growth than those fed on the soybean meal control diet. The soybean meal used in the study was solvent extracted and had undergone the normal degree of heat processing associated with the solvent extraction process, but the level of trypsin inhibitors was still quite high, indicating that the treatment did not completely destroy the inhibitor, which may account for the poor growth of the fish fed these diets. The results of studies on rainbow trout in which diets based on sunflower meal have been supplemented with amino acids are not consistent. Sanz et al. (1994) fed rainbow trout diets containing 36% sunflower meal (16.4% crude fibre), and observed that without amino acid supplementation the fish had lower feed intake, growth and feed 41 efficiency compared to those fed a fishmeal control diet. Addition of the amino acids lysine, leucine and methionine improved the growth rate to a level comparable with those fed the fishmeal control diet. Contrary to the above findings, Tacon et al. (1984), however, did not observe any beneficial effects on growth and feed utilization when rainbow trout were fed diets containing 36% sunflower meal and 0.2%> supplemental methionine. This may be due to the fact that the sunflower meal had high levels of sulfur amino acids. Indeed it may have been more worthwhile to supplement the diets containing sunflower meal with lysine, which was very low in these diets (1.1% of diet), compared to the fishmeal control diet (5.9%> of diet). In conclusion, the studies available so far indicate that rainbow trout can utilize sunflower meal at levels up to 40% in the diet. It is possible that a higher level than this could be used if a good source of digestible energy is also included in the diet. However, at higher levels, it may be necessary to supplement the diets with amino acids. The question of whether or not supplementing diets with synthetic amino acids is of any benefit has not been adequately resolved. Further work needs to be done to establish the limiting amino acids. There are only limited studies on the use of sunflower seed cake in tilapia diets. Abdul-Aziz et al. (1999), working with O. niloticus fingerlings with an initial weight of 19.4 g, replaced 25% to 50%> of soybean meal with sunflower cake. The digestibilities of protein and organic matter were significantly lower in the diets that contained the sunflower cake compared to the soybean meal control diet. Among the two diets based on sunflower cake, the digestibility of protein and organic matter was lower in the diet where sunflower cake replaced 50% of the soybean meal protein than in the diet where the 42 replacement level was 25%. Sintayehu et al. (1996) evaluated the digestibility of high-fibre sunflower cake by tilapia (O. niloticus), and observed that the apparent digestibility of crude protein was higher in soybean than in cotton seed cake and sunflower cake. The digestibility of organic matter and gross energy was lower in sunflower meal than in soybean meal and cotton seed cake. In the growth trial, sunflower cake protein replaced 32% of the fishmeal protein with no adverse effects on fish performance. Supplementation of the diets containing sunflower cake with lysine and methionine did not improve fish performance. Fish used in the study by Sintayehu and co-workers (1996) had an initial weight of 90 g., and the diet containing sunflower cake also had 30% fishmeal, which would have adequately supplied all of the lysine needed by fish of that size. Jackson et al. (1982) fed tilapia (O. mossambicus) (initial weight 13 g) diets where 75% of the protein originated from sunflower cake and observed that there were no differences in fish performance between fish fed the diet with the high level of sunflower cake, and those fed the fishmeal control diet. The results of this experiment, however, should be interpreted with caution, because the growth rates that were obtained for fish in this study were well below optimal (0.9% to 1.25%). Furthermore, fish fed the positive control diet had a specific growth rate of 1.09% per day, indicating that there were other factors that may have adversely affected fish in this trial. Growth rates for tilapia of similar size and fed on a positive control diet have typically been in the range of 1.7% to over 2% (El Sayed, 1998; Abdel-Fattah and El-Sayed, 1990) In summary, the high fibre content of sunflower cake limits its broad use in fish diets. In other animal species e.g., swine and poultry, the low lysine content of sunflower cake has also been identified as a factor limiting its use, whereas in fish, the response to 43 dietary lysine supplementation has not always been consistent. Further work is certainly needed on this topic. The high fibre content of sunflower cake could be a strong hindrance to its extensive use in fish diets because fibres are highly indigestible by most fish species. Dietary fibres also act as diluents, and they may reduce diet palatability, as well as mineral bioavailability. 44 2.7 References Abdel-Fattah, M., and El-Sayed, 1990. Long term evaluation of cotton seed meal as a protein source for Nile Tilapia, Oreochromis niloticus (Linn). Aquaculture, 84: 315-320. Abdul-Aziz, G.M., El-Nady, M.A, Shalaby, AS., and Mahmoud, S.H., 1999. Partial substitution of soybean meal protein by different protein sources in diets for Nile tilapia fingerlings. Bulletin of Faculty of Agriculture, University of Cairo, Vol. 50: Issue 2: Anderson, J., Capper, B.S., and Bromage, N.R., 1991. Measurement and prediction of digestible energy value in feedstuffs for the herbivorous fish tilapia (Oreochromis niloticus Linn). British Journal of Nutrition, 66: 37- 48. Anderson, J.S., 1996. Dietary protein quality and quantity for Atlantic Salmon (Salmo salar) reared in sea water. PhD. thesis, The University of British Columbia. 156 pp. Balarin, J.D., 1985. National reviews for aquaculture development in Africa. 7. Kenya. FAO fish circ, 770: (7) 96pp. Balogun, A.M., and Fagbenro, OA., 1995. Use of macadamia press-cake as a protein feedstuff in practical diets for tilapia, (Oreochromis niloticus) (L). Aquaculture and Research, 26: (6) 371-377. Belal, I.E. H., Al-Owaifeir, A , and Al-Dosari, M., 1995. Replacing fishmeal with chicken offal silage in commercial Oreochromis niloticus (L) feed. Aquaculture Research, 26: 855 - 858. Bishop, CD. , and Angus, R.A., and Watts, A., 1995. The use of feather meal as a replacement for fishmeal in the diet of Oreochromis niloticus fry. Bio-Resource Technology, 54: 291-295. Bowen, S.H., Lutz, E.V., and Ahlgren, M.O., 1995. Dietary protein and energy determinants of food quality. Ecology, Vol 76: (3) 899-907. Calvert, C C , Klasing, K . C , and Austic, R.E., 1982. Involvement of food intake and amino acid catabolism in the branched-chain amino acids. Antagonism in chicks. Journal of Nutrition, 112: (4) 627-635. Canvin, D.T., 1965. The effect of temperature on the oil content and fatty acid composition of the oils from several oilseed crops. Canadian Journal of Botany, 43: 63-69. Chamberlain, G.N., and Hopkins, J.S., 1994. Reducing waste use and feed cost in intensive ponds. World aquaculture, 25: (3) 29 - 32. 45 Cho, C.V., and Slinger, S.J., 1979. Apparent digestibility measurements in feedstuffs for rainbow trout. In: Finfish Nutrition and Fish-feed Technology. J.E. Halver and K. Tiews, (Eds.) Vol. 2 239-247 Cisse, A , 1996. Effects of varying dietary protein:energy ratios on food consumption, growth and body composition of S. melanotheron. The Third International Symposium on Tilapia in Aquaculture. ICLARM Conference Proceedings, Manila No. 41. Clandinin, DR., 1958. Sunflower seed oil meal. In: Processed plant protein foodstuffs. A.M. Attschul (Ed.) Academic Press, New York. 557-575 Clark, A.E., Watanabe, W.O., Olla, B.L., and Wicklund, R.L, 1990. Growth, feed conversion and protein utilization of Florida red tilapia fed isocaloric diets with different protein levels in sea water pools. Aquaculture, 88: 75-85. Coche, A G . , Haight, B.A., and Vincke, M.J., 1994. Aquaculture development and research in Sub-Saharan Africa. - Synthesis of national reviews and indicative action plan for research. FAO, Rome, CEFA Technical Paper 23, 140pp. Cowey, C.B., 1975. Aspects of protein utilization by fish. Proceedings of Nutrition Society, 34: 57-63. Cowey, C.B., 1994. Amino acid requirement of fish - A critical appraisal of present values. Aquaculture, 124: 1-11. Cruz, E.M., and Laudencia, I.L., 1977. Protein requirements of tilapia (mossambica) fingerlings, Kalisasan Phil. J. Biol., 6: 177 - 181. Davies, A T . , and Stickney, R.R., 1978. Growth responses of tilapia (aureus) to dietary protein quality and quantity. Trans. Am. Fish Soc, 107: 479- 483. Davies, S., Williamson, J., Robinson, M., and Bateson, R.L, 1989. Practical inclusion levels of common animal by-products in complete diets for tilapia (O. mossambicus Peters). In: Proceedings of the 3rd International symposium on feeding and nutrition in fish. Toba, Japan, pp 325 - 332.. Davies, S.J., Mclonnel, S., and Bateson, R.L, 1990. Potential use of rapeseed meal as an alternative protein source in complete diets for tilapia (O. mossambicus Peters). Aquaculture, 87 (2) 145-154. De Silva, S.S. and Perera, P.A.B., 1985. Effects of dietary protein level on growth, food conversion and protein use in young tilapia (nilotica) at four salinities, Trans. Am. Fish. Soc, 114: 584-587. Deaton, J.W., McNaughton, J.L. and Burdick D., 1979. High fibre sunflower meal as a replacement for soybean meal in layer diets. British Poultry Science, 20: 159-162. 46 Degussa, A., 1998. The Amino Acid Compositions of Feed Ingredients. Degussa Feed Additives, Frankfurt, Germany. Delic, I., Vucurevic, N . , and Stojanovic, S., 1975. Investigation of inactivation of chlorogenic acid from sunflower meal under invitro conditions in mice. Acta Vet. (Belgrade) 25: 115-119. Dorrell, D.G., 1976. Chlorogenic acid content of meals from cultivated and wild flowers. Crop Science, 16: 422-424 Dorrell, D. G. and Vick, B.A. , 1997. Properties and processing of oilseed sunflower, in Sunflower Technology and Production. A .A. Schneiter (Ed). American Society of Agronomy Inc., Crop Science Society of America, Inc., Soil Science Society of America, Inc., Madison, Wisconsin, USA. pp 834. Dorsa, W.J., Robinette, H.R., Robinson, E.H. , and Poe, N.E. , 1982. Effects of dietary cotton seed meal and gossypol on growth of young channel catfish. Trans. Am. Fish. Soc, 3: 651-655 Earles, F.R., Vanetten, C.H., Clark, T.F., and Wolff J.A., 1968. Composition of data on sunflower seeds. Journal of American Oil Chemists Society, 45: 876-879. El-Sayed, A.F., 1999. Alternative dietary protein sources for farmed tilapia Oreochromis spp. Aquaculture, 179: 149- 168. El-Sayed, A .F .M. , 1990. Long term evaluation of cotton seed meal as a protein source for Nile tilapia, (Oreochromis niloticus, Linn). Aquaculture, 84: 315 - 320. El-Sayed, A . F . M . , 1998. Total replacement of fishmeal with animal protein sources in Nile tilapia (Oreochromis niloticus) feeds. Aquaculture Res., 29 (4) 275 - 280. Erickson, G .M. , Edgerly, C M . Fisher, G.R., Lindberg, G.L., and Park C.S., 1984. Feeding complete mixed diet supplemented with either soybean or sunflower meals according to lactation curve. J.Dairy Sci. 67, (Suppl. 1, Abs): 142 Fagbenro, O , and Jauncey, K. , 1993. Chemical and nutritional quality of raw, cooked and salted fish silages. Food Chemistry, 48: 331-335. Fagbenro, O , and Jauncey, K., 1994. Chemical and nutritional quality of dried fermented fish silages and their nutritive value for tilapia (Oreochromis niloticus). Animal Feed Science and Technology, 45: 167-176. Fagbenro, O A , 1988. Evaluation of defatted cocoa cake as a direct feed in the monosex culture of tilapia guineensis (pisces: cichlidae). Aquaculture, 73: (1-4) 201-206. 47 Falaye, A.E. , 1982. The use of hydrolyzed feather meal alone or in combination with supplemental amino acids as dietary source for tilapia, (Oreochromis niloticus) MSc. Theis, University of Stirling, U.K. FAO (Food and Agriculture Organization of the United Nations) 1997. Aquaculture production statistics 1986-1995. FAO Fish. Circ, No 815, Rev, 9. Rome, Italy. 179 pp. Feedstuffs, (1998). Feedstuff's ingredient analysis table 1998 Edition. Prepared by Nick Dale, University of Georgia, Athens, Georgia, USA. Fisheries Department, 1995. Fisheries Annual Statistical Bulletin, 1995. Prepared by the Fisheries Department, Ministry of Tourism and Wildlife, Kenya. Gaber, M . M . A . , 1996. Partial and complete replacement of fishmeal by poultry by-products and feathermeal in diets for Nile tilapia (Oreochromis niloticus). Annals Agric. Sci., Moshtohor 1, 203 - 214. Green, S. and Kiener, T., 1989. Digestibilities of nitrogen and amino acids in soya-bean, sunflower, meat and rapeseed meals measured with pigs and poultry. Animal Production., 48 :157- 179. Guerrero, R.D., 1995. Evaluation of home-made feeds used for commercial tilapia production in the Philipines. International Co-operation for Fisheries and Aquaculture th Devt: Proceedings of the 7 Biennial Conference of the International Institute of Fisheries Economics and Trade. Keeluna (Taiwan): National Taiwan Ocean University, 1995. 2, 254-257. Haard, N.F., Kariel, N . , Herzberg, G., Feltham, L A W . , and Winter, K. , 1985. Stabilization of protein and oil in fish silage for use as ruminant feed supplement. J. Sci. Food Agric, 36: 229-241. Hargrove, D . M . , Rogers, Q.R., Calvert, C C , and Morris, J.G., 1988. Effect of excesses of the branched chain amino acids on growth, food intake and plasma amino acid concentration of kittens. Journal of Nutrition, 118:3,311-320. Heiser, C.B., 1955. Origin and Development of the cultivated sunflower. Am: Biol. Teach., 17: 161-167. Heiser, C.B., 1976. The sunflower. Univ. Okla. Press, Norman, Okla. Herman, R.L. , 1970. Effects of gossypol on the rainbow trout Salmo gairdneri Richardson. J. Fish Biol., 2: (4) 293-304. Higgs, D.A., Dosanjh, B.S., Little, M . , Roy, R.J.J., and McBride, J.R., 1990. Potential for including canola products (meal and oil) in diets for Oreochromis mossambicus x O. aureus hybrids. The current status of fish nutrition in aquaculture. The Proceedings of the 48 Third Intl. Symp. on Feeding and Nutrition in Fish. August 28-September 1, 1989. Toba, Japan. 301-314. Hossain, M.A . , Nahar, N . , Kamal, M . , and Islam, M . N . , 1992. Nutrient digestibility coefficients of some plant and animal proteins for tilapia (Oreochromis mossamhicus). Journal of Aquaculture in the tropics. 7 (2) 257-266. Hughes, S.G., Rumsey, G.L., and Nesheim, M.C. , 1984. Effects of dietary excesses of branched-chain amino acids on the metabolism and tissue composition of Lake trout (Salvelinus namaycush). Comp. Biochem. Physiol., A: Comparative Physiology, Vol. 78 A: (3)413-418. Hussain, R .A.K. , and Offer, N.W., 1987. Effect of formaldehyde treatment on degradation of acid preserved fish silage protein in-vitro. Anim. Feed. Sci., 16: 297 -304. Ingham, L.D. , and Arme, C , 1977. Intestinal absorption of amino acids by rainbow trout. J. Comparative Physiol., 117: 323 - 324. Jackson, A.J., and Capper, B.S., 1982. Investigation into the requirement of tilapia (Oreochromis mossamhicus) for dietary methionine, lysine and arginine in semi-synthetic diets. Aquaculture, 29: 289 - 297 Jackson, A.J., Capper, B.S., and Matty, A.J., 1982. Evaluation of some plant proteins in complete diets for tilapia (Oreochromis mossamhicus). Aquaculture, 27: 92-109. Jacob, J.P., 1993. The feeding value of Kenyan sorghum, sunflower seed cake and sesame seed cake for poultry. Ph.D thesis, The University of British Columbia. Jauncey, K. , 1982. The effects of varying dietary protein level on growth, feed conversion, protein utilization and body composition of juvenile tilapia (Oreochromis mossamhicus). Aquaculture, 27: 43 - 54. Jauncey, K. , Tacon, A.C.J. , and Jackson, A.J. , 1983. The quantitative essential amino acid requirements of Oreochromis mossamhicus, In: Proc. 1st Int. Symp. Tilapia in Aquaculture. Fish-Elson, J. and Z. Yaron, (Eds). Tel Aviv University, Tel Aviv, Israel, 1983, 328 pp Jensen, L.S., 1998. Pellet quality and performance of broilers. Department of Poultry Science Bulletin, University of Georgia, Athens, Georgia, USA. Kharchenko, L . N . , and Borodulina, A.A. , 1976. Accumulation and metabolism of fatty acids in seeds of high oleic mutant sunflowers. In. Proc. 7 t h Intl. Sunflower Conference. Krasnodar, USSR. 178 - 180. 49 Kondra, P. A., Sell, J.L., and Guenter, W., 1974. Response of meat and egg-type chickens to a high fibre diet. Canadian Journal of Animal Science, 54: 651-658. Lall, S.P., 1991. Nutritional value of fish silage in salmonid diets. In: Bull. Aquaculture Assoc. Canada, no 91-1. R.G. Ackman and J.O Dor (Eds.) Aquaculture Association of Canada, Ontario. 63 - 74. Luquet, P., 1991. Tilapia Oreochromis species. In: Handbook of Nutrient Requirements of finfish. R P Wilson (Ed.) CRC Press. Boca Raton, Florida, U S A 169 - 180. Lapie, L P . , and Bigueras-Benitez C M . , 1992. Feeding studies on tilapia (Oreochromis sp) using fish silage. Paper presented at the 8 t h Session of the Indo-Pacific Fishery Commission Working Party on Fish Technology and Marketing. Yogjakarat, Indonesia, 24 - 27 Sept. 1991. FAO Fish Rep., No. 470 suppl. Lim, C , and Dominy, W.G., 1991. Utilization of plant proteins by warm water fish. In: Proc. Aquaculture Feed Processing and Nutrition workshop. D M . Akiyama and Tan, R .K.H. (Eds). Thailand, Indonesia, 19-25 September, 1991. May, R . C , Piepenbrock, N . , Kelly, R.A., and Mitch, W.E. 1991. Leucine-induced amino acid antagonism in rats: muscle valine metabolism and growth impairment. Journal of Nutrition, 121: (3)293-301. Mazid, M.A . , Tanaka, Y. , Katayama, T., Rahman, M . A , Simpson, K . L , and Chichester, C O . , 1979. Growth response of Tilapia zilli fingerlings fed isocaloric diets with variable protein levels. Aquaculture, 18: 115-122. Mazid, M.A . , Tanaka, Y . , Katayama, T., Simpson, K . L . , and Chichester, C O . , 1978. Metabolism of amino acids in aquatic animals. III. Indispensable amino acids for Tilapia zilli, Bull. Jpn. Soc. Sci. Fish., 44 McCallum, I .M., and Higgs, D.A., 1989. An assessment of processing effects on the nutritive value of marine protein sources for juvenile chinook salmon (Oncorhynchus tshawytshd). Aquaculture, 77: 181-200. McNaughton, J.L., and Deaton, J.W., 1981. Sunflower poultry applications. Feed Mgt. 32: 27 -28 . Morales, A.E. , Cardenete, G., Sanz, A. and De la Higuera M . , 1993. Re-evaluation of crude fibre and acid-insoluble ash as inert markers, alternative to chromic oxide, in digestibility studies with rainbow trout (Oncorhynchus mykiss). Aquaculture, 179: 71 -79. Moser, R.L. , Cornelius, S.G., Pettigrew, J.E., and Hanke, H E . , 1985. Efficacy of 34% crude protein sunflower meal for growing pigs. Nutrition Reports International 31: (3) 583- 591. 50 M r , I., Twina, Y . , Grossman, E., and Nitsan, Z., 1994. Quantative effect of pelleting on performance, gastrointestinal tract and behavior of meat-type chickens. British Poultry Science, 35: 589-602. Nose, T., 1979. Summary report on the requirements of essential amino acids for carp. In: K.Tiews, and J.E.Halver, (Eds), Finfish Nutrition and Fish Technology. Heenemann, GmbH and Co., Berlin, 145 - 156. NRC, 1993. Nutrient requirements for fish. National Research Council, National Academy Press. Washington, D. C., U . S. A. 114pp N R C (1994) Nutrient Requirements of Poultry, 9 t h Revised Edition, National Research Council, National Academy Press, Washington, DC Ochieng, P.A., (1994). Aquaculture development and research in Kenya. In: Aquaculture development and research in sub-Saharan Africa. National reviews. CIFA Technical Paper, 23 suppl., 397 pp Offer, N.W. and Hussain, R.A.K. , 1987. Fish silage as a protein supplement for early weaned calves. Animal Feed Science Technology, 17: 165 - 177. Ofojekwu, P C , and Ejike, C , 1984. Growth response and feed utilization in the tropical cichlid (Oreochromis niloticus (Linn)) fed on cotton seed based diets. Aquaculture, 42: 27-36 Ogino, C , Tekeuchi, L . , Takeda, H. , and Watanabe, T., 1979. Availability of dietary phosphorus in carp and rainbow trout. Bull. Jpn. Soc. Sci. Fish., 45: (12) 1527 - 1532. Omoregie, E., and Ogbemudia, F.I., 1993. Effect of substituting fishmeal with palm kernel meal on growth and food utilization of the Nile tilapia (Oreochromis niloticus). Israeli Journal of Aquaculture, Bamidgeh. 45: (3) 113-119. Osman, M.F. , Omar, A.E. , and Nour, A . M . , 1996. The use of leucaena leaf meal in feeding Nile tilapia. Aquaculture International, Vol 4: No. 1 9 - 18. Page, J.N., 1978. Dietary sulfur requirements of fish. Nutritional, pathological and biochemical criteria. Ph.D Thesis. Cornell University, Ithaka, N Y . Park, C.S., Edgerly, C M . , Erickson, G.M. , and Fisher, G.R., 1981. Response of dairy cows to sunflower meal with varying dietary protein and fibre. J. Dairy Sci., 64: Suppl 1. (Abs) 141. Park, C.S., Moon, Y.S. , Wiesenborn, D., Chang, K.C.S. , and Hoffman, V. , 1997. Alternative uses of sunflower. In: A .A. Schneiter (Ed.) Sunflower technology and production. American Society of Agronomy Inc., Crop Science Society of America Inc., Soil Science Society of America Inc., Madison, Wisconsin, USA. 834 pp. 51 Pereira, da Silva, and Pezzato, E . M . , 1999. Alimentary ingredients and feeding behaviour of Nile tilapia (Oreochromis niloticus). (Portuguese ) Acta Scientiarum. 21: (2) 297-301. Popma, T.J., 1982. Digestibility of selected feedstuffs and naturally occurring algae by tilapias. Ph.D. Dissertation. Auburn University, Alabama. Putt, E.D., 1997. Early History of sunflower. In: Sunflower Technology and Production. A A . Schneiter (Ed). American Society of Agronomy Inc., Crop Science Society of America, Inc., Soil Science Society of America, Inc., Madison, Wisconsin, USA. pp 834. Ravindran, V . and Blair, R., 1992. Feed resources for poultry production in Asia and the Pacific. II. Plant Protein Sources. World's Poultry Science Journal, 48: 205 - 231. Rhone Poulehc, 1993. Feed ingredients formulation in digestible amino acids. In Rhodimet Nutrition Guide. 2 n d Edition, Rhone Poulenc, Animal Nutrition, Antony France. Robinson, E.H. , Rawles, S.D., Oldenbury, P.W., and Stickeney, R.R., 1984. Effects of feeding glandless or glanded cotton seed products and gossypol to tilapia (aureus). Aquaculture, 38: 145 - 154. Roy, N.D. and Bhat, R.V. , 1974. Trypsin inhibitor content in some varieties of soybean (glycine max) and sunflower seed (Helianthus annuus). Journal of the Science of Food and Agriculture, 25: 265 - 269. Salaro, A .L . , Pezzato, L .E . , Luvizotto, M.C.R., Carratore, C.R., and Del Rosa, G.J.M., 1995. Productive performance and anatomopathology changes of Nile tilapia (Oreochromis niloticus) fingerlings fed with diets containing Leucaena (Leucaena leucocephola seed meal). Proceedings of the 6 t h Rio-Grande Meeting of Aquaculture Experts and 3 r d South Brazil Meeting on Aquaculture. 94-102 Santiago, C.B., Aldaba, M.B. , and Lavon, M.A. , 1982. Dietary crude protein requirement of tilapia nilotica fry. Kalikasan, Phil. J. Biol., 11: 61-71. Santiago, C.B., Aldaba, M B . , Lavon, M.A. , and Reyes, OS . , 1988. Reproductive performance and growth of Nile tilapia (Oreochromis niloticus) brood stock fed diets containing Leucaena leucocephola leaf meal. Aquaculture, 70: (1) 53-61. Santiago, C.B., and Lovell, R.T., 1988. Amino acid requirements for growth of Nile Tilapia. Journal of Nutrition, 118: 1540-1546. Sanz, A., Morales, A.E. , De la Higuera, M . , and Cardenete, G., 1994. Sunflower meal compared with soybean meal as partial substitutes for fishmeal in rainbow trout (Oncorhynchus mykiss) diets: protein and energy utilization. Aquaculture, 128: 287-300. Schingoethe, D.J., 1981. Sunflowers: An ingredient with potential: Dairy applications. Feed Management, 32 (6): 18. 52 Scott, J.R., Newton, S.H. and Katayama, R.W., 1982. Evaluation of sunflower meal as a soybean meal replacement in rainbow trout diets. Proceedings of the Thirty-Sixth Annual Conference, South-Eastern Association of Fish and Wildlife Agencies: October 31 to November 3, 1982, Jacksonville, Florida. Scott, M L . , Nesheim, M.C. and Young, R.J., 1982. Nutrition of the chicken. 3 r d Edition, M . L . Scott and Associates, Ithaca, New York. 562 pp. Screde, A., 1981. Utilization of fish and animal by-product in mink nutrition VI. The digestibility of amino acids in fish viscera products. Effect of feeding on growth, fur characteristics and reproduction. Acta Agric Scandinavia, 31: 171-178. Seerley, R.N. , Burdick, D., Russom, W.C., Lowrey, R.S., McCampbell, H.C., and Amos, H.E., 1974. Sunflower meal as a replacement for soybean meal in growing swine and rat diets. Journal of Animal Science, 38: 947 Senkoylu, N . , and Dale N . , 1999. Sunflower in poultry diets: a review. World's Poultry Science Journal, Vol 56. 153-174. Shiau, S.Y., and Huang, S.L., (1989) Optimal dietary protein level for hybrid tilapia (O. niloticus* O. aureus), reared in seawater, Aquaculture, 81: 119-Shiau, S.Y., Kwok, C C , Huang, J.Y., Chen, C M , and Lee, S.L., 1989. Replacement of fishmeal with soybean meal in male tilapia (Oreochromis niloticus x O. aureus) fingerling diets at sub-optimal protein level. J. World Aquacult, Soc. 20 (4) 230 -235. Siddiqui, A.Q., Howlader, M.S., and Adam A. A , 1988. Effects of dietary protein levels on growth, feed conversion and protein utilization in fry and young Nile tilapia (Oreochromis niloticus). Aquaculture, 70: 63-73. Sintayehu, A. , Mathies, E., Meyer-Burgdorff, K. H. , Rosenow, H. , and Gunther, K.D. , 1996. Apparent digestibilities and growth experiments with tilapia (Oreochromis niloticus) fed soybean meal, cottonseed meal and sunflower seed meal. Journal of Applied Icthyology, Vol. 12: (2) 125-130. Sosulski, F.W., and McCleary, C.W., 1972. Diffusion extraction of chlorogenic acid from sunflower kernels. Journal of Food Science, 27: 253-256. Tacon, A.G.J. , 1993. Feed ingredients for warm water fish. Fishmeal and other processed feedstuff's, FAO Fish. Circ , No. 856, FAO, Rome, Italy, 64pp. Tacon, A.G.J. , 1997. Global trends in aquaculture and aqua feed production 1984 - 1995. Intl Aqua feed Directory and Buyers Guide, 1987/1988. 5 - 37. 53 Tacon, A.G.J. , and Jackson, A.J., 1985. Utilization of conventional and unconventional protein sources in practical fish feeds. In Cowey, C.B., Mackie, A . M . , and J.G. Bell (Eds). Nutrition and feeding in fish. Academic Press, London 119-145. Tacon, A.G.J. , Jauncey, K. , Falaye, A , Pantah, M . , MacGowen, I., and Stafford, E., 1983. The use of meat and bone meal, hydrolyzed feather meal, and soybean meal in practical fry and fingerling diets for O. niloticus. In: Proc. of the 1s t International symposium on Tilapia in Aquaculture. J. Fishelson, and X . Yaron, (Eds.) Tel Aviv Univ. Prs, Israel, pp 356-365. Tacon, A.G.J. , Webster, J.L., and Martinez, C.A., 1984. Use of solvent extracted sunflower meal in complete diets for fingerling rainbow trout (Salmo gairdneri Richardson). Aquaculture, 43: 381-389. Twibell, R.G., and Brown, P.B., 1998. Optimal dietary protein concentration for hybrid tilapia (O. niloticus x O. aureus) fed an all plant diet. World Aquaculture Society, Vol 29: 19-16. Unger, P.W., and Thompson, T.E., 1982. Planting date effects on sunflower head and seed development. Agron. J., 74: 389-395. Unilever, 1976. Unilever Educational Booklet. Plant Protein Food. Unilever, London, U K . Vaughan, J.G., 1970. The Structure and Utilization of Oil Seeds. Chapman and Hall (Eds), London. 35-40. Villamide, M.J. and San Juan, L.D. , 1998. Effect of chemical composition of sunflower seed meal on its true metabolisable energy and amino acid digestibility. Poultry Science, 77: 1884-1892. Vincke, M.M.J . , 1995. The present state of development in continental aquaculture in Africa. In: The Management of integrated freshwater Agro-Piscicultural Ecosystems in tropical areas. J.J. Symoens and J.C. Micha (Eds.) 28-61. Viola, S., and Zohar, G., 1984. Nutrition studies with market size hybrids of tilapia (Oreochromis) in intensive culture. Bamidgeh, 36: 3 - 15. Viola, S., Ariel, Y . , and Zohar, G., 1988. Animal protein free feeds for hybrid tilapia (Oreochromis niloticus x O. aureus) in intensive aquaculture. Aquaculture, 75: 115 -125. Waldroup, P.W., Hillard, C M . , and Mitchell, R.J., 1970. Sunflower meal as a protein supplement for broiler diets. Feedstuffs, 24: 42-43 54 Wang, K.W., Tekeuchi, T. and Watanabe, T., 1985. Effect of dietary protein levels on growth, of tilapia O. nilotica. Bull. Japan. Soc. Sci. Fish., 51: 141 - 146. Watanabe, T., Tekeuchi, T., and Ogino, C , 1979. Studies on the sparing effect of lipids on dietary protein in rainbow trout diets. In: Finfish Nutrition and Fish Feed Technology. K. Tiews and J.E., Halver, (Eds.) Heenemann GmbH and CO., Berlin, 113 - 125. Wee, K . L . , and Wang, S.S., 1987. Nutritive value of leucaena leaf meal in pelleted feeds for Nile tilapia. Aquaculture, 62: (2) 9 7 - 108. Wilson, R.P., and Halver, J.E., 1986. Protein and amino acid requirements of fishes. Ann. Rev. Nut., 6: 225-244.. Wilson, R.P., and Poe, W.E., 1985. Apparent digestibility coefficients of protein and energy coefficiens of feed ingredients for channel catfish. Prog. Fish Culturist, 47: 154 — 158 Winfree, R. A. , and Stickney, R.R., 1981. Effects of dietary protein and energy on growth, feed conversion efficiency and body composition of tilapia (aureus). Journal of Nutrition, 111 1001-Yamada, S., Tanaka, Y.„ Katayama, T., Sameshima, ML, and Simpson, K . L . , 1982. Plasma amino acid changes in tilapia nilotica fed a casein and a corresponding free amino acid diet. Bull. Jpn. Soc. Sci. Fish., 48: 1783 -Zatari I. M . , 1989. Evaluation of sunflower meal as a feed ingredient in broiler chicken diets. The effect of using supplemental fat or pelleting. Ph.D. thesis, Iowa State University, Ames, Iowa. 153pp. Zatari, I .M. and Sell, J.L., 1990. Sunflower meal as a component of fat supplemented diets for broiler chickens. Poultry Science, 69: 1503-1507. Zukovsky, F . M . , 1950. Cultivated plants and their wild relatives. Commonwealth Agric. Bull. Furnham, Royal, U . K . Zulberti, C , Lugugo, C.J., 1989. The Vegetable oil-protein system in Kenya. Summary report phase 1. Vegetable-oil protein system program. Working paper no. 10. Egerton University, Njoro, Kenya 55 Chapter 3 Experiment 1: Digestibility of nutrients and energy in wheat bran, high-fibre and fibre-reduced sunflower cakes, anchovy fishmeal and omena fishmeal by Oreochromis niloticus 3.0 Abstract The apparent digestibility of protein, organic matter, and energy of high-fibre and fibre-reduced sunflower cake, wheat bran, omena fishmeal and anchovy fishmeal was investigated in tilapia (O. niloticus) fingerlings. A reference diet in which anchovy fish meal and soybean meal were the main sources of protein was formulated. Test diets were made by combining 70% of the reference diet with 30% of each of the following: wheat bran, high-fibre and fibre-reduced sunflower cakes, anchovy and omena fishmeals. Each diet was provided to 3 tanks offish, each with 12 fish. Water temperature was maintained at 26°C, and dissolved oxygen concentration above 5.5 mg/litre. Digestibility of the diets, nutrients and energy was done by the indirect method using chromic oxide as a marker. Fecal collection was done by stripping. Among the fishmeals and the sunflower cakes, apparent protein digestibility coefficients were similar. Protein digestibility was lower (P < 0.05) in wheat bran than in the other ingredients. Apparent digestibility coefficients for energy and organic matter, and the digestible energy concentrations were significantly higher for the fishmeals (P < 0.05) than the plant protein ingredients. There were no significant differences in the digestibility of energy and organic matter between anchovy fishmeal and omena fishmeal, and this was also true for the digestible energy concentrations (P > 0.05). The dehulling of sunflower cake increased the digestibility of energy by 12% relative to the undehulled meal. 56 3.1 Introduction and objectives In order to formulate practical diets for feeding fish, we need to know the nutrient compositions of the feedstuffs that are potential ingredients for inclusion into such diets. We also need to know the biological availability of the nutrients and energy in each of the ingredients for the species under consideration. The diets for intensive fish culture are formulated where possible on the basis of literature values for digestible energy and protein in the feedstuffs. For some warm water fish such as tilapia, however, comprehensive tables showing the availability of nutrients and energy from various ingredients are not available. Consequently, diets for tilapia are formulated using published data extrapolated from other fish species such as carp (Cyprinus carpio), and catfish (Clarias gariepinus). There are differences in the ability of tilapia, carp and catfish to digest proteins, fats and carbohydrates (Degani and Revach, 1991). Tilapia are herbivorous fish and they digest carbohydrates efficiently (Anderson et al., 1984; Viola and Arieli, 1983). Tilapia have also been shown to digest animal proteins better than carp (Degan et al, 1991). Carp are omnivorous, and under natural conditions, they find their feeds from benthic sources. By contrast, tilapia feed at all trophic levels, with the result that the natural diet of tilapia contains a higher percentage of plant life. On the other hand, tilpia cope less efficiently with fat than carp and catfish (Degani and Revach, 1991). Methods for determining digestibility coefficients involve either direct or indirect measurements of the amounts of nutrients ingested, and subsequently excreted. The direct method involves measuring all the feed consumed by the fish, and all the resulting excreta (Smith,1971; Smith et al, 1980). In studies by Smith et al, (1980), individual 57 1 fish were confined in metabolic chambers and force-fed a measured amount of feed. The excrements were quantitatively collected and analyzed for nutrient content. This method was very stressful to the fish, and likely compromised feed utilization. The indirect method involves the use of an indigestible marker (Maynard et al, 1979). Chromic oxide, crude fibre, polyethylene and acid-insoluble ash have all been used as markers. In using markers, several assumptions are made. For example, it is assumed that the marker is inert, indigestible, does not stream ahead of the intestinal contents, and is not preferentially retained at some level in the gastrointestinal tract. The relative concentration of the marker is also observed to increase in the intestinal contents as the ingesta passes along the tract and the nutrients are absorbed. In digestibility studies in fish by direct or indirect methods, several problems are encountered in the collection of feces from water. One problem is the likelihood of leaching of soluble fecal compounds and the fragmentation of the feces with the dispersion of fine particulate matter into the water. Besides, it is difficult to separate feces from the water and to avoid contamination of the feces with the uneaten feed. In view of the above, several methods of collecting samples of feces where a marker has been used have been devised such as intestinal dissection, (Smith and Lovell, 1971), stripping (Nose, 1960), anal suction (Windell et al, 1978) and the use of special collection tanks (Cho and slinger, 1979; Cho et al. 1982; Talbot, 1985). In the present study, chromic oxide was used as a marker and feces were collected by stripping. Cho and Slinger (1979), Cho et al. (1982) and Talbot (1985) showed that the results of digestibility studies obtained by intestinal dissection, stripping and special collection 58 tanks were not significantly different for digestibility of dry matter, crude protein and lipids. The main objective of this experiment was to determine the effect of reducing fibre content in sunflower cake on the apparent digestibility of protein, energy and organic matter using tilapia (0. niloticus) as the test animal. Also, it was of interest to compare the digestibility of protein, energy and organic matter of Kenyan omena fishmeal with that of prime quality anchovy fishmeal for the reasons provided previously. A growth study was also undertaken to assess whether all of the diets were acceptable to the fish. 59 3.2 Materials and Methods 3.2.1. Fibre-reduced sunflower cake Hybrid sunflower seeds (Kenya Fedha) were purchased from a commercial trader (Rift Valley Products, Nakuru, Kenya). The seeds were partly dehulled using a manually-operated Cecoco dehuller (Ibaraki, Osaka 567 Japan) which incorporated a dehuller and a sorting machine. Throughput was 25 kg/hr to achieve a 75% dehulling efficiency. Efficiency depended on the seed type and moisture content. Seeds with low moisture content were easier to dehull than those with a high moisture content, and for that reason, all seeds were dried to less than 10% moisture before dehulling. The oil content of the partly dehulled seeds was reduced by a commercial screw press oil extractor (Gold Feeds, Nairobi, Kenya). 3.2.1.2 High-fibre sunflower cake The high-fibre sunflower cake used in this experiment and all other experiments reported in this thesis was a commercial cake that is marketed as prime quality sunflower cake. It is processed by Rift Valley Products, Nakuru, Kenya, for Unga Feed Millers in Nairobi, the largest feed manufacturers in the country. The fibre content is considerably lower than in other cakes available in the country and it fetches a premium price. It was processed from the same variety of sunflower as the low-fibre sunflower cake. 3.2.2 Ingredients other than sunflower cakes. Omena fishmeal was purchased from Tamfeeds, (Nairobi, Kenya). It was made from the cyprinid fish, Rastrineobola argentea. The anchovy meal was a high quality Chilean LT. meal purchased from Moore Clark, B.C., Canada. This fishmeal was purchased in Canada because, at the outset of the study, it was not possible to obtain prime quality herring 60 meal elsewhere to act as a positive control. Herring meal became available in Kenya later in the project, and was used in Experiments 3 and 4. The oc-cellulose and cornstarch were bought from ICN, Canada, and chromic oxide from Fisher Scientific (Ontario, Canada). Soybean meal and wheat bran were bought from Sigma Feeds, (Nairobi, Kenya). 3.2.3 Chemical analyses All ingredients and the experimental diets were analyzed in duplicate for dry matter, ash, crude fibre, fat and protein according to standard procedures (AOAC, 1984). Gross energy was determined by adiabatic bomb calorimeter at the Kenya Industrial Research Development Institute, Nairobi, Kenya. Acid detergent fibre (ADF) and neutral detergent fibre (NDF) contents were determined by the method of Waldern (1971), using an Ankom technoloanalyzer (Ankom Technology, 140 Turk Hill Park, Fairport, N Y 14450). The chromic oxide contents in the diets and feces were determined by the acid digestion method of Farukawa and Tsukahara (1966). Water quality parameters (Appendix 1) (pH, colour, turbidity, and the levels of iron, manganese, calcium, magnesium, sodium, potassium, aluminium, chlorides, fluorides, nitrates, nitrites, ammonia, total nitrogen, sulphate, orthophosphates, total hardness, total alkalinity, total suspended solids, free carbon dioxide, total dissolved solids, dissolved oxygen and residual chlorine were analysed at the Ministry of Water Development, Water Quality and Pollution Control laboratory in Nairobi. 3.2.4 Experimental diets Six diets, whose compositions are shown in Table 3.1, were formulated. A reference diet based on prime quality anchovy fishmeal (Moore Clark, BC, Canada), soybean meal, 61 & O H -a t/5 P t4-c O c _o O o O ^ a .o. g PH 1) .-§1 SO SO Csl SO sd o o o c o o o s m o r ~ s o o o o o m o o s > / - > o r -^ o o q o o o o t 1 ; l ; | n o s d o t ^ H r ^ s d ^ o < N ' o o fN --H O so o O sd o o o m o o s m o t ^ q O O O O O « ; * ; t w i o K H t ^ s d ^ o ' r s i o o so SO C S o o o c o o o s s n o r ~ O O O O O O T r r j - T j - u - i o c ^ r ^ r - ^ s d ^ o c N O o 00 r-' m o O T r o o r ^ s n i n - - ; O O u - i o < N O r < - i o O ffi 00 rH K r n vb oo ^ m ^ H h l J i o t s q o « oo ^- S os M O O T f ^ H ^ Q oo \o M c> O 00 ^H O SO Os os - H ^ H OS O0 TT 00 t n Tt so co oo j , o\ r . oo m os os o m oo t-; § SO ~ t~-d oo ^ oo oi K as ^ M . — "s? ^ >s-S « CJ 6 2 £ Q -S o w < O •c C N SO 00 a •a CS B O SO ^ co oo a o oo a wheat bran, and corn-starch was formulated. Soybean meal and anchovy fishmeal were the main sources of protein while wheat bran was used as an inexpensive binder to ensure good water durability of the pellets. Test diets were formulated by combining 70% of the reference diet, with 30% each of the test ingredients (high-fibre and fibre-reduced sunflower cakes, "omena" fishmeal, anchovy fishmeal, and wheat bran). Chromic oxide (0.5%) was used as an indigestible marker, and was added at the expense of cellulose. Al l diets were pelleted at the Naivasha Station of the Kenya Agricultural Research Institute using an Ottevanger pelleting machine (Ottevanger Machine Fabrieken B.V, 2750 A A Moekapelle, Holland). Three tanks, each containing 12 fish, were used for each diet. The fish were fed to satiation 3 times daily at 8.30 a.m., 12.30 p.m. and 4.30 p.m. 3.2.5 Supply and maintenance of fish Male O. niloticus fingerlings weighing 59g ± 5.4 g (+ sd) were purchased from Baobab Fish Farm (Bamburi Nature Trail, Mombasa) and transferred to the University of Nairobi for this experiment. They were acclimated to laboratory conditions for a period of three weeks. During the acclimation period they were fed on commercial feed (Baobab tilapia pellets) for a period of one week, and then the reference diet for another 2 weeks before the onset of the digestibility trial. After the acclimation period, they were weighed in groups of 12 fish selected at random and allocated to the experimental circular tanks. Eighteen tanks with a diameter of one metre and filled with water to a depth of 0.26 metres were used at the start of the experiment. Water level was adjusted as the fish biomass increased to maintain a stocking density below 0.1 kilograms of fish per litre of 63 water. The chemical parameters of the well water, which was used in all experiments are shown in Appendix 1. A Sweetwater T M Regenerative blower was used for aeration. Each tank was fitted with an AS8-1 (3 inches) diffuser. Water temperatures and dissolved oxygen concentration in the tanks was maintained at 26°C ± 2°C and above 5.5 mg/liter, respectively. Water was completely exchanged in each tank every 48 hours. The fish were kept under a natural photoperiod (Nairobi, Kenya, 1° 16' S, 36° 48' E). Duration of the growth trial was 50 days, but feces collection continued for 3 months. All fish that died were replaced by others of similar weight from a reserve tank to ensure that there were enough fish to provide the required amount of the fecal samples. Fish were starved 24 hours before handling. 3.2.6 Fecal collection Feces were collected by stripping, (Nose, 1960), once a fortnight during the growth trial (50 days), and thereafter, once every week for the next 56 days to the end of the trial. Fish were fed at 8.30 a.m and 10.30 a.m and fecal collection was done approximately 5 to 7 hours after the first feeding. Al l fish in a tank were first anaesthetized with MS 222 and blotted dry. Urine was then pressed out by gentle pressure applied along the lateral line towards the vent. Pressure was then applied on the region between the anal fin and the vent and the expressed feces collected into aluminium dishes. Feces from each tank were dried immediately after collection at 60°C for a period of 24 hours. They were then labeled and and frozen (-5°C) until the end of the trial, when all feces from each tank were pooled. Fecal samples were analyzed for moisture, chromic oxide, protein, gross energy and ash. i 64 3.2.7 Digestibility assessment Apparent digestibility coefficients of the test diets and test ingredients were calculated by the indicator method of difference as described by Maynard et al. (1979). Apparent digestibility coefficients (%) for each of the 6 test diets and test ingredients were determined for crude protein, gross energy and organic matter. The apparent digestibility of each nutrient was calculated as follows; 100 [ 1 - marker cone, in diet X nutrient cone, in feces] marker cone, in feces nutrient cone, in diet Digestible energy concentration in the diet was calculated as follows; Gross energy in diet - [fecal energy concentration X marker cone, in diet] marker cone, in feces Apparent digestibility coefficient for a nutrient in the test ingredient was calculated from the digestibility coefficient for the reference and the test diets according to the following equation of Forster (1999). ADCNingr = {(a + b) * ADCNcom - (a)*ADCN ref}/b ADCNingr = apparent digestibility coefficient of a nutrient in test ingredient A D C N ref - apparent digestibility coefficient of a nutrient n the reference diet. a = nutrient contribution of reference diet to nutrient content of combined diet, = (% nutrient in reference diet) * (100 - i) b = nutrient contribution of test ingredient to nutrient content of combined diet, = (% nutrient in test ingredient)*i i = percent of test ingredient in the combined diet. ADCNcom = Apparent digestibility coefficient of a nutrient in the diet consisting of a combination of the reference diet and the test ingredient 65 3.2.8 Data collection and analytical procedures Fish growth and performance was assessed by the following parameters; absolute weight, weight gain, specific growth rate, feed intake and feed conversion ratio. Specific growth rate (% per day was calculated as follows: 100(ln final wt (g) - ln initial wt (g))/number of experimental days. Feed conversion ratio was calculated as ingested feed (g)/wet weight gain. 3.2.9 Statistical analyses Data on final weight of fish, weight gain, specific growth, feed consumption, feed conversion ratio and digestibility of various nutrients and energy in the diets were analyzed using PROC. G L M of the SAS statistical package. Means were compared using Tukeys's test with the level of significance set at P < 0.05. In analyzing data on final weight, weight gain, and feed consumption, an analysis of covariance was done with the initial weight of the 12 fish as the covariate, to control for the differences in the initial weight of the fish. The statistical model used in the analyses of various fish performance parameters and digestibility was; Yij = p + Ai + Bj + eij where, p = overall mean A = effect of diet e = random error 66 3.3 Results and discussion 3.3.1. Chemical composition of the reference diet, test diets and test ingredients The results of proximate analyses of the reference and test diets are presented in Table 3.1, while those of the test ingredients are presented on Table 3.2. Dehulling sunflower seeds considerably reduced the fibre content and increased the percentage of protein in the cake. Dehulled sunflower seeds are difficult to press using the conventional screw press machines and consequently the fibre-reduced cake had a slightly higher percentage of lipid than the high-fibre cakes. Gross energy was also higher in the fibre-reduced cake than noted in the high-fibre cake when the results were calculated on a dry weight basis. The dehulled cake had a dry matter content of 92%. Crude protein, fibre, fat and ash contents (DM basis) were 44.6%, 12%, 11%, and 9.8% respectively. Gross energy content was 5.01 kcal/g., while acid detergent fibre and neutral detergent fibre were 14% and 26%, respectively. The high-fibre sunflower cake had a dry matter content of 92% and contained the following levels of nutrients and energy (DM basis); crude protein, 30%; crude fibre, 29%; crude fat, 9.7%; gross energy, 5.00 kcal/g; ADF, 24.5%; NDF, 48.9%. Gross energy contents in the diets did not vary appreciably, ranging from 4807 kcal/kg to 4954 kcal/kg ( D M basis), while crude protein levels were in the range of 31.1% to 44.6% (DM basis). Crude fibre, ADF and NDF values were highest in the diets containing high-fibre sunflower cake, while ash content was highest in the diets that contained both fishmeals. 3.3.2. Fish performance Final weights, weight gains, specific growth rates, feed intakes, feed efficiencies and percent mortalities are presented in Table 3.3. For the 50 days of the feeding trial, feed intake and feed 67 IH 1 3 3* 2? O 5^ O H S P 3 b 1 m ol so o O to O M H S O S O cn co o i—i o in SO O Os o) o o OI o OS OS co ^ H i n o so co os in so I I os oi oi OI in Os so r- TJ-<—< O T f OO so ^ H oi ^ H co in in so oo os 60 I—I ! > v ' CD P 0 0 CN CN CD -3 CD O c NO 0 0 r-' m 0 0 O in m in m m in O 0 0 .0 r -XI T f VO •8 m r-•§ O N VO •s CN r-CN O 0 0 O 0 0 0 O 0 CN 0 0 O 0 in 0 >/-> 0 0 0 r -ON O CN VO CN T f CN C--' CN vd CN T f CN O ro T f r- r- VO r- O CN 0 0 m 0 0 i> 0 0 t - ' 0 0 O ON 0 0 ro T f VO CN 0 in m O N m ro VO O VO T f VO r--' m 1 T3 CD O CD t-, I CD \~ X U3 CD o 3 t/3 CD s-X X ! 60 X , ti CD o q=l C2 3 cd CD B X cd C CD s o cd B X i o X o 3 c cd cd C/3 / - S V O ro ro Cfl C/5 C s s 1 ~ H CN < I ON NO conversion ratios did not vary significantly between fish fed the diets, indicating that inclusion of high-fibre sunflower cake at the 30% level in the diet did not depress feed intake. Mortality was highest for the fish fed the diet with a high level of anchovy fishmeal. The specific growth rates observed were 0.8%, 0.75%, 0.73%, 0.72%, 0.69% and 0.64% for fish fed the reference diet, and diets based on omena fishmeal, fibre-reduced sunflower cake, wheat bran, anchovy fishmeal, and high-fibre sunflower cake. The growth rate of fish fed on the high-fibre cake was significantly less than that found for the fish fed the reference diet. This could be due to the low digestible energy content of the diet based on the high-fibre cake. The reason that fish performance was assessed was mainly to determine whether all diets were acceptable to the fish, and whether each would support positive nitrogen balance. There were no significant differences between groups in feed intake indicating that all diets were equally acceptable. 3.3.3 Apparent digestibility of nutrients in test ingredients. The apparent digestibilities of energy, protein and organic matter of the reference diet and test diets are presented in Table 3.4. The digestibility coefficients were subsequently used to calculate the digestibility of nutrients and energy in the test ingredients. In regard to the latter, the digestibility of nutrients and energy was determined according to the equation of Forster (1999) considering that each of the test diets consisted of 70 % of the reference diet and 30%> of the test ingredient. 3.3.4 Apparent digestibility coefficient for protein (ADC-P) The apparent digestibility coefficients for protein (ADC-P) in the test ingredients are listed in Table 3.5. Anchovy and omena fishmeals each had an ADC-P of 90%. The 70 Table 3.4: Apparent digestibility coefficients (ADCs) and apparent digestible energy (ADE) values of the reference and test diets (70% reference diet, 30% test material). Protein Energy DE Organic matter ADC ADC (kcal/kg) ADC (%) (%) (DM basis) (%) Reference 92.8 78.5 3840 76.2 Ref-Fibre reduced SFC1 91.3 67.1 3341 63.1 Ref-High fibre SFC 90.9 63.7 3101 61.6 Ref-Omena fishmeal 91.4 78.4 3771 75.8 Ref-Anchovy 91.5 80.8 3889 73.5 Ref -wheat bran 89.8 66.2 3228 60.2 'SFC Sunflower Seed Cake ADC Apparent digestibility coefficient DE Digestible energy 71 Table 3.5: Apparent digestibility coefficients (ADCs) and digestible energy values for the fibre-reduced and high-fibre sunflower cakes, omena fishmeal, anchovy fishmeal, and wheat bran. Protein Energy DE Organic matter ADC ADC (kcal/kg) ADC (%) (%) (DM basis) (%) Test Ingredient Fibre-reduced SFC1 88.60" High-fibre cake 85.60a Omena Fish Meal 89.70a Anchovy Fish meal , 90.00a Wheat bran 75.20b SEM 1.90 42.10b 2203b 37.20b 29.70b 1401° 30.80b 78.203 3624a 74.70" 86.00a 4003a 76.50a 37.10b 1787bc 30.30b 2.65 132.7 2.57 A D C Apparent digestibility coefficient D E Digestible energy 1 SFC Sunflower cake Means (n = 3) within a column with a common superscript are not significantly different (P>0.05) 72 fibre-reduced and high-fibre sunflower cakes had ADC-P values of 89% and 86% respectively, which were not significantly different. Differences in protein digestibility between the fishmeals and the sunflower cakes were not significant (P > 0.05). Wheat bran had an apparent protein digestibility coefficient of 75%, which was significantly lower (P < 0.05) than those found for the other test ingredients. The results of this study are in agreement with the findings of other authors, despite the fact that different methodologies were used in the various studies. For example, with respect to fishmeal, Watanabe et al. (1996), observed an apparent protein digestibility of 92% for white fishmeal and a local fishmeal fed to tilapia (O. niloticus). Further, Hanley (1987) reported an ADC-P value of 87%> for a fishmeal made from different species of fish fed to O. niloticus. Watanabe and co-workers (1996) used digestibility chambers (as described by Cho et al, 1982) for collection of feces, while Hanley (1987) collected feces by intestinal dissection. There is limited information on the digestibility of protein in sunflower cake by tilapia (O. niloticus). Sintayehu et al. (1996) determined the protein digestibility of sunflower cake that had a crude fibre content of 29.4% (AD basis) using HCL-insoluble ash as the indigestible marker. The fish were hand fed, and the feces were collected by the sedimentation technique. The apparent protein digestibility coefficient of sunflower cake in that study was 89.8%, which was not appreciably different from the values of 86% and 89% observed in the current study for the high-fibre and the fibre-reduced cakes, respectively. In studies with carp by Eid and Matty (1989), true protein digestibilities of two sunflower cakes (compositions not defined) were reported as 65%> and 68%. These values were much lower than those obtained for tilapia in the current study. In the studies 73 by Eid and Matty (1989), an in-vitro digestion method was used, where the protein sources were incubated in-vitro with carp intestinal extract, and then the digestibility of the various nutrients was determined. The method used by Eid and Matty (1989) would give true digestibility values, which should be higher than the apparent values. Differences between the two studies may be due to the different methods used to assess digestibility and possibly to differences between the two species in utilization of plant proteins. Differences in digestive capabilities between tilapia and carp have been reported by Degani and Revach (1991), who compared the digestive capabilities of tilapia, carp, and catfish, and by Watanabe et al. (1996), who compared protein digestibility in rainbow trout, carp, tilapia, and ayu (Plecoglossus altivelis). The studies by Degani and Revach (1991) showed that tilapia digest protein from animal sources better than carp, while Watanabe et al. (1996) observed that the digestibility of protein from plant protein sources by carp and tilapia was similar. It therefore follows that the observed differences in protein digestibility for sunflower cake between the current study and that of Eid and Matty (1989) may have been due to the different methodologies used in the two studies. Commercial feed binders are expensive, and wheat and wheat products are often used to increase pellet stability in water. Wheat bran is an inexpensive product that could be used as a binder in tilapia pellets. Information on the digestibility of protein in wheat bran for tilapia species is limited. Popma (1982) determined that the apparent protein digestibility of wheat bran in tilapia was 71%, which was close to the value of 75% observed in the present study. It is not clear why the digestibility of protein in wheat bran was lower than that of the other ingredients assessed in this study, particulary the high-74 fibre and fibre-reduced sunflower cake. Popma (1982) attributed the low digestibility of wheat bran protein to the feeding habits of tilapia, pointing out that tilapia repeatedly pick up and expel pelleted feeds before swallowing. In this process, the pellet disintegrates which may lead to selective feeding on individual components of the diet. The author observed selective feeding in coarsely-textured and less palatable diets (coffee pulp, wheat bran and alfalfa meal). Furthermore, for the above mentioned diets, Popma (1982) found that the concentration of chromic oxide in the diets was more than twice the concentration in the faeces, which would indicate a negative digestibility. The author concluded that chromic oxide was not suitable for determination of digestibility in such feeds. In contrast to the above argument, the digestibility of protein in the high-fibre sunflower cake used in the current study was not significantly different from that of fishmeal, despite being "coarsely" textured. Furthermore, in the study by Popma (1982), raw corn which had the same texture as wheat bran had an apparent protein digestibility of 84%, which was not significantly different from that of the fishmeal used in that study. This may indicate that there was another factor that affected the digestibility of protein in the wheat bran which needs further investigation. 3.3.5 Apparent digestibility coefficient for energy (ADC-E) and digestible energy concentration (DE) in test ingredients The results of the apparent digestibility coefficients for energy (ADC-E) and digestible energy (DE) concentrations of the test ingredients are listed in Table 3.5. Anchovy fishmeal had an ADC-E of 86%, while omena fishmeal had an ADC-E of 78%>, which were not significantly different. 75 The A D C - E values for the two fishmeals in this study are comparable to those reported by other authors for tilapia species. Hanley (1987) reported an A D C - E of 87%, and a digestible energy concentration of 3283 kcal/kg D M for a fishmeal made from different species of fish, while Anderson et al. (1991) noted a digestible energy content of 3876 kcal/kg D M . The type of fishmeal used in the latter study was not defined. In a study with carp, Hossain and Jauncey (1989) reported an apparent energy digestibility coefficient of 80% for fish meal (type not denned), which is closer to the value of 78% observed for omena fish meal in this study, but lower than that of 86% for anchovy fish meal. The A D C - E of fishmeal is affected by many factors, including the source, composition and freshness of the fish, and the processing temperatures involved in the production of the meal (Anderson, 1996). Fish lipids are high in polyunsaturated fatty acids, which are susceptible to oxidation, during processing. Fligh temperatures cause lipid peroxidation, resulting in reduced digestibility (Davadasan et al, 1985). Opstvedt (1974), working with poultry, observed that oxidation of lipids in fishmeal reduced the energy value of the meal. The digestibility of energy in fibre-reduced sunflower cake has not been reported previously for tilapia or carp. A D C - E values in the high-fibre and fibre-reduced sunflower cakes in the present study were 30% and 42%, respectively, while the digestible energy contents were 1401 kcal/kg D M and 2203 kcal/kg D M , respectively. However, these preceding differences in the A D C - E coefficients were not significant (P < 0.05), although the difference in energy value, when expressed as D E concentration was significant (P < 0.05). The lack of a significant difference in the energy digestibility coefficients may have been caused by the large random variation in values observed for 76 the fibre-reduced sunflower, high-fibre cake, and wheat bran. Despite the reduction in fibre content, the fibre-reduced sunflower cake had a low A D C - E compared to the fishmeals. Some authors have reported low digestibility of carbohydrates in sunflower cake. In rainbow trout, Sanz et al. (1994) found a digestibility of 40% for carbohydrates (NFE plus fibre) in sunflower cake compared to 50% for soybean meal carbohydrates. Bendi and Spandorf (1953) also found low (26%) digestibility of sunflower cake carbohydrates in carp. It was not possible to assess the digestibility of all nutrients in the present experiment. It is plausible that the digestibility of sunflower carbohydrates in tilapia is also low, and that this accounted for the observed low values for digestible energy content. The high-fibre sunflower cake was markedly higher in fibre content as measured by the different fibre analysis methods (crude fibre, ADF and NDF). Anderson et al. (1991) and Kirchgessner et al. (1986) stipulated that the digestibility of energy in plant protein sources is determined by the amount of cell wall fractions in the respective feedstuffs. In studies by Anderson et al. (1991), NDF was well correlated (inversely) with the D E content in feedstuffs fed to tilapia (O. niloticus), while Kirchgessner et al. (1986) related energy digestibility to the ADF content of feedstuffs. Sintayehu et al. (1996), reported an apparent digestibility coefficient for energy of 49%, and a D E content of 2298 kcal/kg D M in high-fibre sunflower cake, while Anderson et al. (1991) observed a D E value of 867 kcal/kg D M . The crude fibre contents of the sunflower cake in the studies refered to were 26%-29%. In the study by Anderson and Associates (1991), ADF and NDF values for the sunflower cake were 31% and 40%, respectively, while in the study by Sintayehu et al. (1996) these values were not stated. 77 The apparent digestibility coefficient for energy of 30%, and digestible energy content of 1401 kcal/kg D M observed in the present study for the high-fibre sunflower cake are closer to the values observed by Anderson et al. (1991). The higher D E concentration in the present study could be attributed to the higher crude lipid content of the high-fibre sunflower cake which was 9% while in the studies by Anderson et al. (1991) it was only 2%. 3 . 3 . 6 Apparent digestibility coefficient for organic matter (ADC-OM) The trend in apparent organic matter digestibility was similar to that observed for energy digestibility. The fishmeals had the highest organic matter digestibility. Among the two sunflower cakes, organic matter digestibility was higher in the fibre-reduced cake than in the high-fibre cake, but the differences were not significant. The high-fibre cake and wheat bran had a similar A D C - O M of 31 and 30%, respectively. Kirchgessner et al. (1986) reported that, in general, ingredients with higher protein content had higher digestibility values for organic matter and energy in tilapia. The authors also noted a negative correlation between ADF content and the digestibility of organic matter and energy. In the present study, anchovy and omena fishmeals had higher organic matter digestibility than the sunflower cakes and wheat bran, both of which were high in ADF. ADF in the fibre-reduced sunflower cake and wheat bran, was lower than in the high-fibre sunflower cake, but, despite this, the differences in organic matter digestibility between the three ingredients were not significant 78 3.4 Conclusions The results from this study indicate that the protein from sunflower cake, regardless of its fibre content, is well digested by tilapia (O. niloticus) at a level not markedly different from that of high quality fishmeal. Similarly, the digestibility of protein in the locally produced omena fishmeal was equal to that of the high-quality anchovy fishmeal. Reducing the fibre content in sunflower cake increased the digestibility of nutrients and energy, thereby increasing its feeding value. The greatest effect of fibre reduction was observed in the digestibility of energy, with the value increasing from 30% to 42%. The digestibility of omena fishmeal in tilapia had not been studied before. This study showed that digestibility of nutrients in omena fishmeal was similar to that of anchovy fishmeal. Compared to fishmeal and soybean meal, sunflower cake is inexpensive and the digestibility values suggest that the fibre-reduced sunflower cake could be used at high levels of replacement in tilapia diets, with little need for adjustment in dietary crude protein level. Additional lipid may be necessary to adjust the D E concentrations, but increasing voluntary feed intake by using appetite enhancers could at least partially counteract the dilution in dietary DE. This is an area that needs further research. The results suggest that it may be possible to use omena fishmeal, which is locally available, to replace the imported expensive fishmeals, in diets for tilapia and perhaps other fish species. Al l the plant sources tested" had high levels of crude fibre. They were incorporated at a level of 30% in the diets. The subject of the additional studies described in this thesis was to investigate the effect of increasing the levels of these ingredients in the diets on tilapia performance. 79 3.5 References Anderson, J., Capper, B.S., and Bromage, N R . , 1991. Measurement and prediction of digestible energy value in feedstuffs for the herbivorous fish tilapia (Oreochromis niloticus Linn). British Journal of Nutrition, 66: 37 - 48. Anderson, J., Jackson, A.J., Matty, A.J. , and Capper, B.S., 1984. Effects of dietary carbohydrates and fibre on the tilapia (O. niloticus). Aquaculture, 37: 303 - 314. Anderson, J.S., 1996. Dietary protein quality and quantity for Atlantic Salmon (Salmo salar) reared in sea water. PhD thesis, The University of British Columbia. 156 pp A O A C , 1984. Association of Official Analytical Chemists. Official methods of analysis. Animal Feed Section. 1141 pp Bendi, A. and Spandorf, A , 1953. The activity of digestion enzymes of the carp. Bamidgeh, 5: 116-130. Cho, C.Y., and Slinger S.J., 1979. Apparent digestibility measurements in feedstuffs for rainbow trout. In: Finfish Nutrition and Fish Food Technology. Proceedings of a World symposium. Hamburg, 20 - 23 June 1978. Cho, C.Y., Slinger, S.J., and Bayley H.S., 1982. Bioenergetics of Salmonid fishes: Energy intake, expenditure, and productivity. Comp. Biochem. Physiol., 73B: 2 5 - 4 1 . Davadasan, K. , Nair, P.G.V., and Antony, P.D., 1985. Effect of oxidation of dietary fish lipids on the quality of proteins in the diet. Fishery Technology. Society of Fisheries Technologists, (India) 22: (1) 70-73. Degani, G., and Revach, A., 1991. Digestive capabilities of three commensal fish species: carp (Cyprinus carpio), L.tilapia (Oreochromis aureus x O. niloticus) and African catfish (Clarias gariepinus). Aquaculture and Fisheries M anagement, 22: 397 - 403. Degani, G., Viola, S., and Yehuda, Y. , 1997. Digestibility of protein and carbohydrates in feed ingredients for adult tilapia (Oreochromis aureus x O. niloticus). The Israeli Journal of Aquaculture - Bamidgeh, 49: (3) 115 - 123. Eid, A.E. , and Matty, A.J., 1989. A simple in-vitro method for measuring protein digestibility. Aquaculture, 77: 111-119. Farukawa, A., and Tsukahara, H. , 1966. Acid digestion method for the determination of chromic oxide as an index substance in the study of digestibility of fish feed. Bull. Jpn. Soc. Sci. Fish., 32: 502 - 504. 80 Forster, I., 1999. A note on the method of calculating digestibility coefficients of nutrients provided by single ingredients to feeds of aquatic animals. Aquaculture Nutrition. Vol 5, no. 2. 143 - 145. Hanley, F., 1987. The digestibility of feedstuff's and the effects of selectivity on digestibility determinations in tilapia (Oreochromis niloticus). Aquaculture, 66: 163 — 179. Hossain, M.A . , and Jauncey, K. , 1989. Studies on the protein, energy, and amino acid digestibility of fishmeal, mustard oil cake, linseed and sesame meal for common carp (Cyprinus carpio L ) . Aquaculture, 83: 59-72 . Kirchgessner, M . , Kurzinger, H. , and Scwarz, F. J. 1986. Digestibility of crude nutrients in different feeds and estimation of their energy content for carp (Cyprinus carpio L.). Aquaculture, 58: 185 - 194. Maynard, L .A . , Loosli, J.K., Hintz, H.F., and Warner, R.G., 1979. Animal Nutrition. 7 t h Edition. McGraw Hill. 602pp. Nose, T., 1960. Digestion of food protein by goldfish (Carassius auratus) and rainbow trout (Salmo gairdneri). Bull. Freshwater Fish Res. Lab., 10: (1) 12 - 22. Opstvedt, J., 1974. Influence of lipids on the nutritive value of fishmeal. Effects of fat addition to diets high in fishmeal to fatty acid composition and flavor in broiler meat. Acta. Agric. Scand., 24: 62 - 67. Popma, JT., 1982. Digestibility of selected feedstuffs and naturally occurring algae by tilapia. Ph.D. Thesis, Auburn University, Alabama. 80pp. Sanz, A., Morales, A.E. , De LaHiguera, M . , and Cardenete, G , 1994. Sunflower meal compared with soybean meal as partial substitutes for fish meal in rainbow trout (Oncorhyncus mykiss) diets: protein and energy utilization. Aquaculture, 128: 287-300. Shiau, S.Y., and Chen, M.J. , 1993. Carbohydrates utilization by tilapia (Oreochromis niloticus XO. aureus) as influenced by different chromium sources. Journal of Nutrition, 123: 1747- 1753. Shiau, S.Y., and Liang, H.S., 1995. Carbohydrates utilization and digestibility by tilapia, (Oreochromis niloticus x O. aureus), are affected by chromic oxide inclusion in the diet. Journal of Nutrition, 125: 976 - 982. Sintayehu, A. , Mathies, E., Mayer, Burgdorff, K - H . , RosenowH., and Gunther, K-D. , 1996. Apparent digestibilities and growth experiments with tilapia (Oreochromis niloticus) fed soybean meal, cottonseed meal, and sunflower seed meal. Journal of Applied Ichthyology, 13: (2) 125 - 130. 81 Smith, B.W., and R.T. Lovell., 1971. Digestibility of nutrients in semi purified rations by channel catfish in stainless troughs. Proc. Annu. Conf. Southeast Asia Association of Game Fish. Commun., 25: 425 - 459. Smith, R.R., 1971. A method for determination of digestibility and metabolizable energy of feedstuffs for finfish. Prog, fish Cult., 33: 132-134. Smith, R.R., Peterson, M.C. , and Allred, A.C. , 1980. The effect of leaching on apparent digestion coefficients in determining digestibility and metabolizable energy of feedstuffs for salmonids. Prog. Fish - Cult., 42: 195 - 199. Talbot, C , 1985. Laboratory methods in fish feeding and nutrition studies. In: Fish Energetics, New Perspectives. P. Tyler, P. Calow, and H . Croom (Eds), John Hopkins Univ. Press. Baltimore, M D . 349 pp Viola, S., and Arieli, Y . , 1983. Nutrition studies with tilapia hybrids. The effect of oil supplements to practical diets for intensive culture. Bamidgeh, 35: 44 - 52. Waldern, D.E., 1971. A rapid micro digestion procedure for neutral and acid detergent fibre. Canadian Journal of Animal Science, 51: 67-69 . Watanabe, T., Takeuchi, T., Satoh, S., and Kiron, V. , (1996). Digestible crude protein in various feedstuffs determined with four freshwater fish species. Fisheries Science, 62: (2) 278 - 282. Windell, J.T., Foltz J.W., and Sarokon J.A., 1978. Methods of fecal collection and nutrient leaching in digestibility studies. Prog. Fish-Cult., 40: (2) 51 - 55. 82 Chapter 4: Experiment 2: The feeding value and protein quality in high-fibre and fibre-reduced sunflower cakes and Kenya's "omena" fishmeal for tilapia {Oreochromis niloticus) 4.0 Abstract This study was undertaken to assess the nutritive values of some locally available protein sources in Kenya, as replacements for anchovy fishmeal in tilapia diets. The test protein sources included were omena fishmeal made from Rastrineobola argentea, anchovy fishmeal, as well as fibre-reduced and high-fibre sunflower cakes. The four protein sources were each tested at two protein concentrations. O. niloticus fingerlings with an initial weight of 16 g. were used for the study. Water temperature was maintained above 26°C throughout the trial and dissolved oxygen concentration in the tanks was maintained above 5.5 mg/litre. Eight experimental diets, four based on fishmeal and four on sunflower cake were formulated. In the diets based on the anchovy and omena fishmeals, the fishmeals provided most of the the dietary protein, while in diets based on the sunflower cakes, 50% of the protein was provided by each of the cakes while the remaining 50% was provided mainly by anchovy fishmeal. Each diet contained one of two levels of protein, viz., approximately 20% (low-protein) and 30%> (high-protein). Further, each diet was fed to triplicate groups of fish for 78 days. Diets based on the fibre-reduced cake had higher levels of all amino acids than the ones based on the high-fibre cake. Lysine and threonine concentrations were lower in diets based on the sunflower cakes than the ones based on the fishmeals. Dietary protein level had a significant effect on growth rate and weight gain. Fish fed diets with 20% protein gained less weight and had higher feed:gain ratios than those fed diets with 30% protein. The source of protein had a significant effect on weight gain. Fish fed diets based 83 on anchovy fishmeal had higher weight gains than those fed diets based on the high-fibre sunflower cake. Reducing the fibre content of sunflower cake improved growth rate and weight gain, but the improvements in both of the parameters were not significant. Diets based on omena fishmeal had lower protein concentrations than noted for the other diets due to an error during mixing. Despite this, the growth rates and weight gains of fish fed these diets were not significantly different from those of fish fed the anchovy fishmeal diets. 84 4.1. Introduction and objectives Tilapia (Oreochromis niloticus) are herbivorous fish that possess morphological and physiological adaptations for utilization of diets high in fibre content. This aspect of its feeding habits has not been fully exploited in commercial aquaculture. Most formulated feeds for tilapia resemble those for omnivorous fish in that they contain significant levels of animal proteins (Hughes and Handwerker, 1993). Much research has been done to evaluate new protein sources to partially or wholly replace fishmeal in diets for all types of fish. Among the plant protein sources, soybean meal has been used widely because of its good amino acid profile, which, as the main protein source, supports the growth of most fish species (Tacon et al., 1984; Wilson and Poe, 1985; Shiau et al., 1987; Viola and Arieli, 1983). Soybeans, however, are not suitable for growing in many countries; hence the need to evaluate other plant proteins sources. Sunflower cake contains little or no known anti-nutritional factors. Sunflower is cultivated extensively due to its adaptability to a wide range of climatic and soil conditions (Ravindran and Blair, 1992). Its seeds are inexpensive to process, and the cake remaining after oil extraction is used as a protein supplement in animal diets (Daghir et al., 1980). The crude protein content of the cake ranges from 25 to 45% (air-dry basis) depending on the extent of dehulling and the efficiency of the oil extraction process. The crude fibre level in the cake generally varies between 14% and 39% (air-dry basis) (Villamide and San Juan, 1998). Protein concentration in sunflower cake is inversely proportional to the fibre content. Kenyan sunflower cakes contain 25% to 40% crude protein, and 24% to 40% crude fibre (air-dry basis) (Jacobs, 1998). Differences in these 85 components among the various cakes tested were caused mainly by differences in seed types and in the processing methods used. The potential use of sunflower cake in fish diets is limited by its high fibre content. Crude fibre not only has no known dietary value for fish, but it also dilutes digestible nutrient densities, thus increasing the release of polluting wastes into the environment. In view of the above, a fibre-reduced, high-fat sunflower cake was tested in the present experiment as a replacement for fishmeal in tilapia feeds. A high-fibre sunflower cake was also tested to determine the extent to which dietary fibre level would influence fish performance. In addition to the sunflower cakes, Kenya's omena fishmeal was also evaluated as a source of protein. From the first experiment (Chapter 3), it was established that the digestibility of energy and nutrients in omena fishmeal produced values comparable to those obtained with LT Anchovy meal. Furthermore, it was also demonstrated that, at low levels of dietary inclusion (30% of the diet), there were no significant differences in fish performance between fish fed the omena fishmeal diets and those fed anchovy meal diets. In the present experiment, the feeding value of omena fishmeal, and low-fibre and high-fibre sunflower cakes were evaluated at a higher level of inclusion (providing 50% of the dietary protein), and over a longer period than in Experiment 1 (Chapter 3). The objectives of this study were to compare the nutritional values and protein qualities of diets based on high-fibre and low-fibre sunflower cakes, and omena and anchovy fishmeals when fed to tilapia (O. niloticus) at each of two levels of dietary protein. 86 4.2 Materials and Methods 4.2.1 Experimental diets and design The fibre-reduced cake used in this experiment was prepared as described in Experiment 1, Chapter 3. The high-fibre cake was from the same batch as used in Experiment 1. The omena fishmeal was prepared by sun-drying "omena" fish (Rastrineobola argentea), and then grinding them in a Wiley mill. Eight diets whose compositions are shown in Table 4.2 were formulated. LT anchovy fishmeal, omena fishmeal, high-fibre and fibre-reduced sunflower cakes were used as sources of protein. In the diets based on anchovy and omena fishmeals, the fishmeals provided most of the dietary protein, while in the diets based on sunflower cakes, only 5 0 % of the dietary protein was provided by the cake, while the remaining 5 0 % was provided by anchovy fishmeal, as shown in Table 4.2. Each diet contained one of two levels of protein, approximately 2 0 % or 30%. At each protein level, the diets were formulated to contain similar levels of digestible energy by varying the level of corn oil. In calculating the digestible energy contents of the diets, the apparent digestibility coefficients for energy that were used were as follows; anchovy fishmeal, 8 6 % (Anderson et al, 1991; Hanley, 1987), omena fishmeal, 7 8 % (present study), low-fibre sunflower cake 44%o (present study), and high-fibre cake, 3 0 % (present study). The digestible energy concentrations of cornstarch and corn oil were taken as 2700 kcal/kg D M (NRC 1993, for channel catfish), and 8100 kcal/kg D M (Santiago and Reyes, 1993) respectively. 87 I i '3 '•3 i-0 0 o on o o o o *c3 o B s a fe cu oo i i cu cu C J S3 fe CU I i x> cu fe U ri '53 •<-> o cu -a 2 fe o 00 00 C N CD C X CD 3 CD '-3 — «r> O N t-~ CN OONCNf^-vo'oOi—< i—'• O N CO CN — — CN „ o o o o ~ T, °" ^ — t~- — o r- c-ON CS CS 1—1 i—' o o NO Tf O N — O ^ - O O O O O — ^ v o v o O f O O N v o m - - o o c o c N r ~ - c N v o - - — ON CS CN >— i—i o o o o o o o ^ m c o c o o o o o o N N O . i o \ c sodr~ : - -No- - — ON CS CS — CO o — u - l O O O O O O o ^ O N O c N c N O O N N q o c A O C f t r i c i K - i — > ON CS CS — CN P g Q Q -3 P-l Q CM O ' (air-dry basis). Calculated D E concentrations in low-protein diets ranged from 2751 kcal/kg ( D M basis) in the diet based on high-fibre sunflower cake (HFSC-20), to 2965 kcal/kg ( D M basis) in the diet based on anchovy fishmeal. The calculated D E concentrations in the high-protein diets ranged from 2797 kcal/kg to 3077 kcal/kg (DM basis). The lower D E concentrations of the diets based on the high-fibre sunflower cake were lower than in the other diets due to an over-estimation of the D E concentration of the cake during formulation of the diets. The D E concentrations of the sunflower cakes and omena fishmeal were calculated based on the results of the digestibility study in chapter 3. Later on, it became necessary to recalculate the digestibility coefficients for energy and the digestible energy concentrations of the ingredients (Experiment 1, Chapter 3). The high-fibre sunflower cake had lower levels of digestible energy than had been assumed during formulation of the diets. The low protein diets were formulated to contain less D E concentration than the high-protein diets in order to minimize differences in energy:protein ratios between the diets at the two protein levels. The stipulated D E requirement for tilapia (O. niloticus) is 3000 kcal/kg D M (NRC, 1993). The calculated D E concentrations in most of the diets (except the ones based on anchovy fishmeal) were slightly below this level. 93 The determined crude protein levels in the diets based on omena fishmeal at both high and low protein levels (O-20 and 0-3 0), were slightly lower than the values found for the other diets. It is not clear why this situation arose. It may have been caused by an error during the mixing of the diets. The fibre-reduced cake was rich in oil and consequently diets made from this cake (LFSC-20 and LFSC-30) had a high oil content. In diets based on the fishmeals, the ADF and NDF values reflected the amount of a-cellulose added to the diets. Despite the fact that the high-fibre sunflower cake had higher ADF and NDF contents than the fibre-reduced cake, this was not clearly reflected in the diets at the low protein level because a-cellulose was added to the diet based on the fibre-reduced cake, thus increasing its ADF and NDF contents. In the diets where fishmeal was partially replaced by the sunflower cakes, phosphorus was balanced by the addition of dicalcium phosphate. Extensive replacement of animal proteins that contain high levels of phosphorus by plant proteins may result in phosphorus deficiency (Viola et al, 1988). This is because plant proteins contain phytate phosphorus which is not available to tilapia. Gur (1996) stated that a phosphorus level of 0.6%- 0.7%) of the diet (air-dry basis), supplied from dicalcium phosphate, animal by-products, or a combination of both is adequate to meet the minimum demand of tilapia for inorganic phosphorus. Most of the phosphorus in the diets based on sunflower cake was supplied by anchovy fishmeal and dicalcium phosphate. Determined levels of phosphorus in the diets were all above 1%> (DM basis). The amino acid profiles of the diets are shown in Table 4.3. When the amino acids were expressed as a percentage of the dietary protein, diets based on the tw co O N O N s CD o & . O " ? o 0s-o CN .a o pi o o CN O oo ^ 0 0 O ( N « O N C N O O C N f- — C O — V O m m o r- f- oo TJ* i—i C O r O »iO CS I C O C O C N V O cs cs Tt CO O N CS p T f i n o 00 CO cs T f 00 ' ' d d ' 1 d • ' 1 d 1 I 1 1 • vo O O N r--O N r-00 T o o CS 00 CS cs »n 00 CO r-N O ro Tt CS - — ' T f ••—' T f ^ c i c i ro r-V O O N CS O N Tt cs CO m T, CO T f O N o O N r- T f cs T f O N ro N O • — d d ^ d d d d T f d o v- O N r- 00 o r- N O CS T, T f CO oo o CS 00 T f - — CO T\ T f CN CO CO CO CS i n vo N O f -p •n r--O N Tt o f- T f CS T f o O N CS T f T f ro N O p ro T o ro d • _ < ~ d d -—< CS T f ' - ' '-' '-' vo CO o oq o CO CS 00 T f CS cs T f CS T f T f ro O N O N T T •-' CO Tt N O CN - -< ro ro CO T f T f O N «*o O N o cs O N N O O o 00 T f CO T f o o r-co CO vo CN ro c--p r-T ro T *~ d CN d d CS T f O N O N 00 00 r--CO m T f O N O T f CN Ti r-00 00 o T f T f CO N O t--- cs -- ro CO T f CO T f m o O O N T o N O r-N O CO O N o CO 00 CS ro Tt 00 o o T f r-N O Tt C S O O --> i CS T f I-H --I f-T f rs O N vo p r-p O N CN O N N O >n CO N O o t--O N T f r--T f *~ ro Tt T f CO cs ro 00 O N CN T f f -N O T f O N r-ro r-CN r-N O T N O CN 00 N O N O O 00 vo O N T f 00 d d d d d d d d d ro d d d O N r-00 o N O CN T CN O N T f O N O N r-N O CS CN O N o o O N T f CN ro Tt T f •-' •-• ro ro ro ro O N T f r-r- T f CN N O o r-T f N O ro N O f -ro r-CS O N CS r-; ro O N f-; T f o CS O N d d — d d d d d r-H ro d r-' d O N T f 00 o r- r-cs CS N O O N T f T f CN o T f oo t-; T f T f T f CN ro N O N O CN •-< ro ro T f O O N T f CO 00 T f O N T f V O T 00 CN N O r-U~i 00 o o O N O N CN ro oo O N T f d d ~ d d d d ro d — o r-T f 00 CO N O r- CN V O T f T CO OO Tt 00 00 T f •-' CO N O N O CS ro ro T f T f O N r-ro CO r-T f CN CO ro ro T f r-CS cs r-vo T f 00 N O 00 r- O N O r-O N d d d -—< -—i d d d d d -—1 cs d --" d <5 S 43 -3 F> =i >^ CO < C3 do O ^ •a CJ o I I O 2 t x 1 1 I I respectively, and observed that dehulling increased the concentration (% DM) of amino acids in the cake. The findings in the present experiment are consistent with these observations made by Green and Kiener (1989). When the dietary amino acids were expressed as a percentage of the diet, the diets with a crude protein content of 20% did not meet tilapia requirements for most of the essential amino acids except leucine and valine. It must be cautioned at this point that strict comparisons between the requirements as stipulated in N R C (1993) may not be appropriate because the N R C figures quoted are for juvenile tilapia weighing less than 1 gram. The fish used in this study had an initial weight of about 16 g. In fish, as in all animals, protein and consequently amino acid requirements (as percentage of the diet) decrease as the animal grows. At a protein level of 30%, diets based on omena fishmeal, anchovy fishmeal, and fibre-reduced sunflower cake met all the essential amino acid requirements for tilapia (NRC, 1993). The lysine levels were lower in the sunflower 96 cake diets than in the diets based on fishmeal at both protein levels. Lysine has been identified as one of the limiting amino acids in sunflower cake (McGinnis et al, 1948; Klain et al., 1956). Methionine and cystine levels in the fibre-reduced sunflower cake were almost similar to the levels in the fishmeal diets. Jackson et al. (1982) observed that sunflower cake had high levels of methionine and cystine compared to other plant proteins. 4.3.2 Fish performance, PER and PPV The effects of the dietary protein level and protein source on absolute weights, weight gains, growth rates, and feed and protein utilizations after 78 days of feeding on the various diets are shown in Tables 4.4 and 4.5, respectively. The effects of both protein level and protein source on the various parameters are shown in Table 4.6. At the start of the study, the mean weights of fish fish in all the groups were not significantly different (P > 0.05). After 78 days, the fish fed the diets with high protein contents had higher weight gains and growth rates (P < 0.05) than those fed the diets with low protein contents (Table 4.6). The interaction between protein level and protein source was not significant for any of the parameters (Table 4.6). The growth rates of the fish increased in direct relation to the dietary protein level. Feed intake was not significantly affected by dietary protein level, but feed utilization was better for fish fed the high protein diets than the low protein diets. Protein utilization (PER and PPV) decreased at the higher level of protein intake. The analyzed protein content of the low-protein diets (DM basis) ranged from 20% to 23.6%, while in the high protein diets, the range was 29% to 33.8% (DM basis). Protein intake (feed intake x protein concentration) was therefore lower for fish receiving the 97 Table 4.4: Effect of protein level on fish performance Protein level Low-protein High-protein SEM Final weight (g/fish)1 51.40b 57.10" 0.96 Wt gain (g/fish)2 35.00b 40.40" 0.90 Specific growth rate (% per day)2 1.47b 1.58a 0.03 Feed intake (g/fish)2 76.703 75.30a 0.47 FCR (Feed:gain ratio)2 2.20a 1.87b 0.04 PER 2 2.16a 1.69b 0.02 PPV 2 39.44a 32.21b 0.36 Means with a different superscript for each factor in a row are significantly different (P < 0.05) 1 Means (n = 300) (Individual fish weights were used, 4 diets x 75 fish/diet) 2 Means (n = 12) PER = protein efficiency ratio PPV = productive protein value 98 Table 4.5: Effect of source of protein on fish performance. OM ANC Fibre-reduced High-fibre SEM Diets fishmeal fishmeal sunfl. cake sunfl.cake Final weight (g/fish)1 52.40b 57.70" 54.90ab 52.00b 1.34 Weight gain (g/fish)2 35.92ab 40.86" 38.73ab 35.39b 1.27 Sp. growth (% per day)2 1.48 1.59 1.57 1.47 0.04 Feed consumption (g/fish)2 77.10 FCR (feed: gain ratio)2 2.17 75.90 76.10 74.90 0.66 1.87 1.98 2.12 0.07 PER 2 2.01 1.98 1.86 1.84 0.06 PPV 2 38.90" 37.00ab 35.14ab 32.30b 1.27 Means that do not have a superscript or share a common superscript letter for each factor within a row, are not significantly different (P > 0.05) 'Means (n = 150) (Individual fish weights used, 2 diets x 75 fish/diet) 2Means (n= 6) OM - Omena ANC - Anchovy PER = protein efficiency ratio; PPV = productive protein value 99 Cd C O td c 3 CO -o CU C J p "T3 CU u cu « s= cd CD t-H X ) 60 a •s tn 'cd O C J 00 +-> O J '-a T 3 C M s C J 0 M-H O CU O c CO cd >% C cd c L H rfo: 0 0 r-a> u. PH , 0 vb CO TT *3 CU cu 5 cd CO H cd O 2 0) C S PH o 2 PH ^? O C N 8 CM S * OH 00 OH HJ l H O OH 00 2* 8?S o &o Pi o BH OO 2S o o 3 o 00 •S 00 00 00 00 00 00 00 2; Z z z Z z z 1 i n i n i n i n i n i n O p 0 0 0 0 O 0 O d O 0 V V V 00 V V V OH OH OH z OH OH PH 1 < n i n m 0 0 p O d 0 cb d V V 00 00 V 00 V OH OH Z z OH Z OH 1 O O OV Ov 00 r~ ov 00 H OV O O d d O O 1 O 0 00 0 CN CO 0 CO CO O i n i n CO CO vb CO CN r l m CO r~ CN 0 0 CN 0 CN 0 CO CO VO ^H 00 VO CO CO i n CO O 0 VO O Ov 0 CO r~ 0 vq CO vq 00 CO 0 ^ i n r-i CN r-H VO T r~ CO O O i n 0 CO 0 r~ Ov O i n OV 00 —1 vo 0 ,-H r—- CN CN d d CO -3 !G HIH s ? H i •—'oh- ; > OH — ' T3 OH OO PH PH OH o 2 o o fl 1 CJ xi +H o Cjv O < n -a •8 i l H a. CJ •s CJ tH I s i3 xi s •i-H CH CJ fl CJ CH O CO T 3 O CN fl O s .fl '53 I CH 1 O o > PH PH O l -& fl CJ • 1 CJ m CJ l H CH II PH CJ i n CJ Q low-protein diets, compared to the ones ingesting the high-protein diets. In tilapia (O. niloticus), Twibell and Brown (1998) and Bowen et al. (1995) observed that at low protein intake, increasing the protein level of the low protein diets resulted in proportional improvements in weight gain and feed conversion ratios up to 30% (DM basis) level in the diet. A wide range of estimates have been reported for the optimal dietary protein content of tilapia feeds. Winfree and Stickney (1981) reported that 56% dietary crude protein level promoted maximum weight gain of tilapia (0. aureus) weighing 2.5g, while in those weighing 7.5 g., 34% dietary crude protein was adequate. Shiau and Huang (1989) reported 24% crude protein as the optimum for tilapia (0.niloticus x O. aureus) weighing 2.9 g., and fed on purified diets with dietary protein levels ranging from 0% to 56%. Weight gain was proportional to the dietary protein content up to a protein level of 24%. At higher protein levels, there was no increase in weight gain or protein gain. Luquet (1991) reviewed several studies and recommended a protein content of 30 - 35% as the optimum for tilapia. This recommendation was based on studies that utilized good protein sources such as fishmeal and casein. N R C (1993) gives the protein requirement as 30% of the diet ( D M basis). Estimates of optimal dietary crude protein concentrations have typically been within the range of 30-35% for tilapia weighing less than 5 g (Mazid et al., 1979; Jauncey, 1982; Siddiqui et al., 1988). The stated requirements for protein show a wide variation, reflecting the different environmental conditions in which the experiments were done, and the different fish and feed factors involved. There is evidence that the optimal protein requirement for tilapia is inversely related to fish size. In studies by Twibell and Brown (1998) using (O. niloticus x O. aureus) fingerlings with 101 an initial weight of 21 g., there was no improvement in growth at dietary protein levels higher than 28%. The initial weight of the fish used in this study was 16 g., which was close to the starting weight of the fish employed in the latter study. The source of protein (omena fishmeal, anchovy fishmeal, fibre-reduced and high-fibre sunflower cakes) had a significant effect on the final fish weights, weight gains, and PPV values (Table 4.5), but not on growth rates, feed intakes, feed utilization and PER values. Fish fed diets based on anchovy fishmeal and the fibre-reduced sunflower cake had an 8% and 7% improvement in growth over those fed diets based on the high-fibre sunflower cake, but the differences were not significant (P < 0.05). Fish fed diets based on the high-fibre cake gained less weight (P < 0.05) over the whole experiment than fish receiving the diets based on anchovy fishmeal (Table 4.5). As noted above, the growth rate was also lower for these fish than for those fed diets based on omena fishmeal and the fibre-reduced sunflower cake, but the differences were not significant. Fish fed diets based on the high-fibre cake tended to have lower feed intake, and a higher feed:gain ratio compared to those fed on the other three diets, but again the differences were not significant. Tilapia, like other fish, consume organoleptically acceptable diets in an attempt to satisfy energy demands. The D E concentration in all the diets was almost similar, which may explain the similarity in feed intake. Diets based on the high-fibre cake had a slightly lower D E concentration compared to the other diets. Despite this, the fish fed on these diets did not increase their feed intakes to compensate for the lower dietary D E content. Residual hulls in the high-fibre sunflower cake diets (FTF-SC20 and FTF-SC30) may have hindered this compensatory increase in feed intake. The percentages of high-102 fibre sunflower cake were 36% and 54% for the low-protein and high-protein diets, respectively. The weight gain of the fish fed the diets containing the high-fibre sunflower cake was significantly lower than for the fish fed on the anchovy fishmeal-based diets. This may be attributed to the lower digestibility of the diet, or the fact that essential amino acids lysine, phenylalanine and threonine in that diet did not meet tilapia requirements for these amino acids as stipulated in NRC (1993). a-Cellulose was used as an inert filler in the diets based on omena and anchovy fishmeal at both protein levels. Different components of dietary fibre vary in their chemical and physiological properties and thus have different effects on physiological functions. Studies on the utilization of fibre in fish have yielded varying results. In sea bream, Morita et al. (1982), studied the effect of carboxymethylcellulose (CMC) on carbohydrate utilization by adding 0 to 12% C M C to diets containing 10%, 20%, or 30% dextrin. They noted that C M C supplementation improved weight gain and feed conversion ratio. It was not clear from the study what the mechanism of improvement was. It could be postulated that C M C , which is a water-soluble fibre, formed a highly viscous solution, that slowed the flow of the digesta, leading to a lower gastric emptying time, and thus more time for absorption of the nutrients in the intestine. Cellulose, unlike C M C , is an insoluble fibre which has been reported to increase gastric emptying time in rainbow trout (Hilton et al., 1983). In channel catfish (Ictalurus punctatus) reared on purified diets, Dupree and Sneed (1966) reported growth improvement when 21% a-cellulose was added to these diets. Indeed, N R C (1977) indicates that it is desirable to have some fibre in semi-purified test-diets for catfish to add structural integrity to pelleted diets. The fibre level however, should not exceed 8% of the diet. 103 Contrary to these findings, Anderson et al. (1984), working with tilapia (O. niloticus), observed that growth was depressed when diets contained more than 10% a-cellulose. In the study by Anderson and co-workers, a-cellulose was added at 10% to 40% of the diet at the expense of glucose, sucrose, dextrin or starch and no attempt was made to balance the digestible energy content in the diets. The fish were also fed at 3% of their body weight daily. Since cellulose is not digestible by fish, diets that contained the cellulose had less digestible energy content than the diets that were based on carbohydrates. In studies by Dioundick and Storm (1990) with tilapia (O. mossambicus), the best growth rates and feed utilization values were obtained with diets containing 2% to 5% supplemental fibre as a-cellulose. Fish fed on cellulose-free diets or on diets that contained 10% a-cellulose demonstrated reduced growth. In that study, a-cellulose was added at the expense of cornstarch, and the fish fed at 6% of their body weight twice a day. Feed intake was higher for fish fed diets containing 7.5% and 10% a-cellulose compared to those fed on diets containing 0, 2.5% and 5% a-cellulose. Despite the higher feed intake, the fish fed on diets containing 10% a-cellulose had reduced growth rate compared to those fed diets containing 2.5% and 5% a-cellulose. Hence, the fish responded differently to the inclusion of dietary fibre in the various studies quoted above. In most of the studies, increasing the cellulose levels in the diet resulted in a reduction of digestible energy concentration. The other explanations for the different observations are likely related to the amount of feed offered to the fish and the number of feedings per day. As explained earlier, fish respond to low dietary energy intake by increasing feed intake if given the opportunity. In studies by Hilton et al. (1983) with trout, fish fed ad libitum with diets containing 10% and 20% a-cellulose, 104 were able to increase feed intake so that the nutrient levels consumed were similar to those of the control fish (no cellulose added). On the other hand, when they were fed on a restricted feeding regime (3% of body weight), feed intake was markedly less than that of the control fish. The other factor that may have affected the utilization of fibre in the experiments quoted above may have been the dissimilar fish sizes that were used. Jackson et al. (1982), used fish sizes ranging from 13 g to 50 g to test utilization of various plant proteins (copra, groundnut, soya, sunflower, rapeseed, cottonseed and leucaena meals) in complete diets for O. mossamhicus, and showed that tilapia could effectively utilize diets containing relatively high levels of fibre, compared to other fish. In that study, fish fed diets containing 13% crude fibre (from natural sources) performed as well as those fed the control diet. The initial weight of the fish in the present study was about 16 g, while the final average weight, after 78 days, was approximately 50g. The digestible energy concentration was not appreciably different between the different diets, and the fish were fed ad libitum. Feed intake did not vary significantly between fish fed the various diets. Fish on the high-fibre sunflower cake diets showed a trend to a reduced feed intake. PER was significantly affected by dietary protein level, but not by protein source (table 4.5). Fish fed diets containing 20% protein had higher PERs than those fed on diets containing 30%> protein. Protein source did not significantly affect PER, although fish fed diets based on the two fishmeals tended to have higher PER values. PER is a measure of protein quality that is presented as a ratio of gain/protein intake, and is affected by the level of protein in the diet, the digestibility of the protein, and the levels of essential amino acids, particularly the first limiting amino acid. Within each protein level, dietary 105 were able to increase feed intake so that the nutrient levels consumed were similar to those of the control fish (no cellulose added). On the other hand, when they were fed on a restricted feeding regime (3% of body weight), feed intake was markedly less than that of the control fish. The other factor that may have affected the utilization of fibre in the experiments quoted above may have been the dissimilar fish sizes that were used. Jackson et al. (1982), used fish sizes ranging from 13 g to 50 g to test utilization of various plant proteins (copra, groundnut, soya, sunflower, rapeseed, cottonseed and leucaena meals) in complete diets for O. mossambicus, and showed that tilapia could effectively utilize diets containing relatively high levels of fibre, compared to other fish. In that study, fish fed diets containing 13% crude fibre (from natural sources) performed as well as those fed the control diet. The initial weight of the fish in the present study was about 16 g, while the final average weight, after 78 days, was approximately 50g. The digestible energy concentration was not appreciably different between the different diets, and the fish were fed ad libitum. Feed intake did not vary significantly between fish fed the various diets. Fish on the high-fibre sunflower cake diets showed a trend to a reduced feed intake. PER was significantly affected by dietary protein level, but not by protein source (table 4.5). Fish fed diets containing 20% protein had higher PERs than those fed on diets containing 30% protein. Protein source did not significantly affect PER, although fish fed diets based on the two fishmeals tended to have higher PER values. PER is a measure of protein quality that is presented as a ratio of gain/protein intake, and is affected by the level of protein in the diet, the digestibility of the protein, and the levels of essential amino acids, particularly the first limiting amino acid. Within each protein level, dietary 105 protein content was not appreciably different between diets, while the digestibility of protein in fishmeal and sunflower cakes was similar (Experiment 1, Chapter 3). Consequently, the observed trends for PER for fish ingesting the diets containing the various protein sources likely resulted from differences in amino acid levels (particularly lysine) between the diets based on fishmeals and those based on sunflower cakes. The trend observed for PPV in relation to diet treatment was similar to PER. Fish fed the diets containing the 20% protein level had significantly higher PPV values than those fed the diets containing 30% protein (Table 4.4). Protein source also had a significant effect on PPV. Fish fed on the high-fibre sunflower cake diets had significantly (P < 0.05) lower values than those fish fed on the diets based on the omena fishmeal. PPV, like PER, is sensitive to dietary protein level. Diets based on omena fishmeal had slightly lower protein levels compared to the other diets and this may have contributed to the observed differences. The differences would also have resulted from differences in the dietary levels of essential amino acids, particulary lysine, which was low in the diets based on the high-fibre sunflower cake. 4.3.3 Effect of diets on whole body composition. Dietary treatment had no significant effect on whole body proximate composition (Table 4.7). There was a trend for fish fed the diets with anchovy fishmeal at the 20%> and 30%> dietary protein level to have higher moisture and lower lipid contents than fish fed the other diets. Generally, fish fed diets high in fat showed a trend to higher lipid levels, 106 protein content was not appreciably different between diets, while the digestibility of protein in fishmeal and sunflower cakes was similar (Experiment 1, Chapter 3). Consequently, the observed trends for PER for fish ingesting the diets containing the various protein sources likely resulted from differences in amino acid levels (particularly lysine) between the diets based on fishmeals and those based on sunflower cakes. The trend observed for PPV in relation to diet treatment was similar to PER. Fish fed the diets containing the 20% protein level had significantly higher PPV values than those fed the diets containing 30% protein (Table 4.4). Protein source also had a significant effect on PPV. Fish fed on the high-fibre sunflower cake diets had significantly (P < 0.05) lower values than those fish fed on the diets based on the omena fishmeal. PPV, like PER, is sensitive to dietary protein level. Diets based on omena fishmeal had slightly lower protein levels compared to the other diets and this may have contributed to the observed differences. The differences would also have resulted from differences in the dietary levels of essential amino acids, particulary lysine, which was low in the diets based on the high-fibre sunflower cake. 4.3.3 Effect of diets on whole body composition. Dietary treatment had no significant effect on whole body proximate composition (Table 4.7). There was a trend for fish fed the diets with anchovy fishmeal at the 20% and 30% dietary protein level to have higher moisture and lower lipid contents than fish fed the other diets. Generally, fish fed diets high in fat showed a trend to higher lipid levels, 106 o cd cd a S3 cd O JS o S3 cd H co" CD cd o -o (U <0 o cj I f c o > C cj PH J CJ * i l H O PH C/3 « c3 >H CJ PH hJ W m CJ T3 00 C/3 C/3 C/3 Z Z 2 Z in in ui Tt m m m m 0 0 V J CS CS (N \C VO -H' © © O o q=l S3 C O -o C U o 3 -o cu X) tC T3 S3 cd (U t-H M3 -S3 •J3 S3 o -a Ui H) <« cd a 3 ^ M-H o S3 o o a o CJ t - O CU O N o CO g '53 N ? o CN O CO CJ C/3 O _ / CO Pf O PH CO O CN CJ on o °r u PH C/3 © CN vq os ov t— oo vi vo -H O OS CN| H oo r-- oo' wj VO '—' Vi Tt C N J O © r- vo' vi c\ r- r-_ >/->' oo' oo vi VO l—< H o J 3 cj S oo r- oo vi vo -H O OV VO ^H oo' r~ oo' vi VO l-H o o © o CO UTI vq ov c-' vo vi VO l - H © >o vo oo oo r~ r-' vi VO -H to cj TJ | s 2 PH O © A •s CJ ^ | l g CO •IP a C/3 1 cj CJ CH g '53 o l H Si O O CN CJ T3 CJ cd PH CJ TJ 3 a CJ s o © CN •5 o CJ OH NT © CO •a a ed © CN CJ S3 •a CJ o 1 © CO © CN CJ (=1 o < © CO TJ e CO © CN CJ I CJ *H CJ o TJ CJ I P Eg © CO I CJ C/3 Pi TJ © CN o C/3 PH I PH > , CJ c Q . <-> CH © CO TJ © CN CJ T3 CJ •a o >H I -S3 6 0 a © CO CJ C/3 TJ S ca © CN CJ C/3 PH I HH © o cd CD cd C CD -13 cd O o cd H o us OJ •a CD g OO OH J 1/1 W W M oo oo oo oo z z z z oo oo oo oo z z z z CD cd o T3 CD CD oo 00 oo >n CN cs CM vo vq -* • - H d d d o e - i d CD O "O CD l— CD u -mi -o c cd XS _C C o -a CD cd X> CO H-> CD i f CD •vD t+H <+H O o & TT JD cd H c/i -o 00 CD cd CO a R cj C+H o c _o 'cn O. I" O o >-. o X JD *o X I O co co , 2 PH = X O CN O CO O oo o « C J & H O0 fe o CN U OO o a> "N oo O CN VO OS o\' r-- oo' >/->' o os CNi i—j oo' r- oo' > -8 .5 g 2 8 X3 " i n O N? P H OJ _ O H O ^ £ ^ © OJ NT OS o C O T3 NT O CN e d I CJ o CO « CS o CN r-o but the results were not significant. Protein and ash percentages were fairly constant between fish fed on the different diets, and were not affected by the diets. In his review on factors that affect composition of farmed fish, Shearer (1994) stated that the levels of body protein are life cycle and size dependent. He further stated that data on whole body proximate composition should be subjected to an analysis of covariance to remove the effect of size. The average fish weights in the current study ranged from 47.82 grams to 60.66 grams at the end of the experiment. Van der Meer. (1995) observed that there were significant differences in whole body protein percentage between fish weighing 5 grams and those weighing 50 grams, but the differences were not significant between fish weighing 50 grams and 150 grams. It could be postulated from the study above that there is a certain size where fish attain "chemical maturity" and where further increases in size do not affect whole body protein composition. Trends in whole body lipid concentration followed trends in dietary lipid content. Fish fed high fat diets tended to accumulate more body lipid compared to ones fed on low fat diets, regardless of the D E concentrations of the diets. The differences between groups however, were not statistically significant. Shearer (1994) stated that whole body lipid stores are more influenced by energy intake than by dietary lipid levels. Recently, the utilization of dietary lipids by tilapia has been the subject of study by various authors. De Silva et al. (1991) for example, observed that the addition of lipids to diets of tilapia fry (mean weight, 1 g.), increased growth rate, and that the best growth response was elicited with diets containing 18% lipid. They also observed that whole body fat concentration was directly proportional to dietary lipid content. The diets used in their study were isocaloric. Hanley (1991) on the other hand, using O. niloticus 108 fingerlings (average initial weight, 42 g) noted that increasing dietary lipid level above 9% did not result in any increase in growth. Also, the fish had significantly higher carcass lipid concentrations than those fed the diets containing 5% fat. The lack of fish growth response to an increase in dietary lipid content in the latter study could be attributed to the large size of the fish used and the fact that the experiment was carried out in outdoor ponds where there was substantial natural productivity. It could also be postulated that tilapia can effectively utilize both fat and carbohydrates as energy sources. In most of the studies quoted above, fat was added as a replacement for carbohydrates in isocaloric diets. Chou and Shiau (1996) fed hybrid tilapia (O. niloticus x O. aureus) on isocaloric and isonitrogenous diets containing 0% to 20% fat and observed that fish fed diets with 5%, 10% and 15% lipid, had similar weight gains, although whole body lipid concentrations increased as dietary lipid content was increased. Fish that were fed high dietary lipid levels had higher percentages of body lipid and lower percentages of body moisture than fish fed the low lipid diets. An interesting point in lipid nutrition in tilapia is the fat distribution in the body. Despite the fact that tilapia have a tendency to store much of the dietary fat, they have been described as lean fish compared to carp (Viola et al., 1988). Viola et al. (1988) and Hanley (1991) observed that 40% of the body fat in tilapia was distributed around the viscera while the muscles contained only 8% of the total body fat. This observation is opposite to the pattern of fat distribution in carp, where fat distribution between the viscera and the muscle are almost equal. 109 4.4 Conclusions The low protein diets were formulated to contain a CP level of approximately 20% (DM basis) and a D E concentration of 2800 kcal/kg (DM), while the high protein diets were formulated to contain a CP level of 30% (DM basis) and a D E concentration of approximately 3000 kcal/kg D M . The D E concentrations in the diets based on the high-fibre sunflower cake at both protein concentrations were lower than observed in the other diets. This occured due to an over-estimation of the D E energy concentration in that cake during the formulation of the diets. Diets based on the omena fishmeal at the 30% and the 20% protein levels had a low determined protein concentration compared to the other diets, which may have been caused by an error during the formulation or mixing of the diets. There was no significant interaction between protein level and protein source. Protein level had a significant effect on weight gains and growth rates. Generally, fish fed diets at the 20% protein level gained less weight and had higher feed:gain values compared to the ones fed diets at the 30% protein level. The source of protein had a significant effect on weight gain, but not on specific growth rate, feed intake, or feed utilization. Fish fed diets based on anchovy fishmeal had higher weight gains than those fed diets based on omena fishmeal, fibre-reduced cake and the high-fibre cake respectively, but the differences were only significant relative to the fish fed the diets based on the high-fibre sunflower cake. Results from this study showed that omena fishmeal could supply all the protein in tilapia diets, while the fibre-reduced sunflower cake could provide up to 50% of the dietary protein. 110 Fish fed diets based on the fibre-reduced cake had a mean weight gain of 38.75 g, and a specific growth rate of 1.57% per day, while those fed on the high-fibre sunflower cake had an average weight gain of 35.4 g and a specific growth rate of 1.47% per day. Reducing the fibre content of sunflower cake improved weight gain and growth rate, but the improvement was not significant in any of the parameters tested (P < 0.05). The failure of the statistical test used to detect significant differences despite the relatively large absolute differences may have been caused by the small sample size used in the tests. Most of the parameters were tested using a mean of 3 replicates (n = 3), except for the final fish weights where individual fish weights were used (n = 75). The effect of sample size is illustrated in Table 4.5, where there was a significant difference in absolute weights (n = 75) between fish fed the anchovy and the omena fishmeal diets, yet weight gain (n = 3) was not significantly different, despite the fish having a similar initial weight. Diets with a crude protein content of 20% did not meet tilapia requirements for most of the essential amino acids except leucine and valine. At a protein level of 30%, diets based on omena fishmeal and anchovy fishmeal met all the essential amino acid requirements for tilapia (NRC, 1993). The diet based on the fibre-reduced sunflower cake at this protein level had a slightly lower level of methionine, while the diet based on the high-fibre cake had lower levels of lysine, methionine and threonine. Lysine and threonine levels were low in diets based on the sunflower cakes, compared to the fishmeal diets. Levels of the other essential amino acids in the fibre-reduced sunflower cake were comparable to those in the fishmeal diets. Ill PER was significantly affected by dietary protein level, while PPV was affected by both dietary protein level and protein source. Fish fed diets based on the high-fibre sunflower cake had lower PPV values compared to those fed the omena fishmeal diets. 112 4.5 References Anderson, J., Capper, A J . , Matty, A.J., and Capper, B.S., 1984. Effects of dietary carbohydrate and fibre on the tilapia (Oreochromis niloticus - Linn). Aquaculture, 37: 303-314. Anderson, J., Capper, B.S., and Bromage, N R . , 1991. Measurement and prediction of digestible energy values in feedstuffs for herbivorous fish tilapia (Oreochromis niloticus-Linn). British Journal of Nutrition, 66: 37-48. AO A C , 1984. Association of Official Analytical Chemists. Official Methods of Analyses Animal Feeds section. 1141 pp. A O AC, 1998. Official Methods of Analyses of AO A C International, 16 th Edition. AO A C Official Method 994.12. Amino acids in Feeds (CD ROM) Bowen, S.H., Lutz, E .V. and Ahlgren, M.O., 1995. Dietary protein and energy as determinants of food quality. Ecology, vol 76: 3. Chou, Ben-Shan., and Shiau, Shi-Yen., 1996. Optimal dietary lipid level for growth of juvenile hybrid tilapia, (Oreochromis niloticus X O. aureus). Aquaculture, 143: 185-195. Daghir, N.J. , Ras, M.A . , and Wwayjan M . , 1980. Studies on the utilization of full fat sunflower seed in broiler rations. Poultry Science, 59: 2273-2278. Desilva, Sena S., Rasanihi, M.G. , and Shim, K.F. , 1991. Interactions of varying dietary protein and lipid levels in young red tilapia. Evidence of protein sparing. Aquaculture, 95: 305-318. Dioundick, O. B. , and Stom, D.L., 1990. Effects of dietary a cellulose levels on the juvenile tilapia, (O. mossambicus, Peters). Aquaculture, 91: 311-315. Dupree, H.K., and Sneed, K .E . , 1966. Responses of channel catfish to different levels of major nutrients in purified diets. U.S. Bull of Sport Fish. Wildlife Technical paper 9: 21pp Green, S., and Kiener, T., 1989. Digestibilities of nitrogen and amino acids in soya-bean, sunflower and meat and rapeseed meals measured with pigs and poultry. Anim. Prod., 48: 157-179. Gur, N . , 1996. Minerals in tilapia nutrition - the protein phosphorus connection. In Proceedings of the intensive aquaculture workshop. Shefayim Israel March 9 t h - April 16 th 1996. 113 Hanley, F., 1987. The digestibility of feedstuffs and the effect of selectivity on digestibility determinations in tilapia (Oreochromis niloticus). Aquaculture, 66: 163-179. Hanley, F., 1991. Effect of feeding supplementary diets containing varying levels of lipids on growth, food conversion and body composition of Nile tilapia (Oreochromis niloticus). Aquaculture, 93: 323-334. Hilton, J.W., Atkinson, J.L., and Slinger, S.J., 1983. Effect of increased dietary fibre on the growth of rainbow trout Salmo gairdneri. Can. J. Fish Aquacult, Sci., 40: 81-85. Hughes, S.G., and Handwerker, T.S., 1993. Formulating for tilapia: Al l vegetable protein feeds. Feed International, September, 1993. 55 -60 . Jacob, J.P., 1993. The feeding value of Kenyan sorghum, sunflower seed cake, and sesame seed cake for poultry. Ph.D. Thesis, The University of British Columbia. Jackson, A.J., Capper, B.S., and Matty, A.J., 1982. Evaluation of some plant proteins in complete diets for the tilapia (Oreochromis mossamhicus). Aquaculture, 27: 97-109. Jauncey, K. , 1982. The effects of varying dietary protein level on the growth, food conversion, protein utilization and body composition of juvenile tilapias (Oreochromis mossamhicus). Aquaculture, 27: 43-54. Klain, G.J., Hill , D . C , Branion, H.D., and Gray, J.A., 1956. The value of rapeseed oil meal and sunflower oil meal in chick starter rations. Poultry Science 35: 1315-1326. Luquet, P., 1991. Tilapia Oreochromis sp. In: Handbook of nutrient requirements of finfish. R P Wilson (Ed). CRC Press. Boca Raton, Florida, USA. 196pp Mazid, M.A . , Tanaka, Y. , Katayama, T., Rahaman, M.A. , Simpson, K . L . , and Chichester, C O . , 1979. Growth response of tilapia zilli fingerlings fed isocaloric diets with variable protein levels. Aquaculture, 18: 115-122. McGinnis, J., Hsu, P.T., and Carver, J.S., 1948. Nutritional deficiencies of sunflower oil meal for chicks. Poultry Science, 27: 389-393. Morita, K. , Furuichi, M . , and Yone, Y. , 1982. Effects of carboxymethylcellulose supplemented to dextrin containing diets on the growth and feed efficiency of red sea bream. Bull. Jap. Soc. Sci. Fish., 48: 1617-1620. NRC, 1977. Nutrient Requirements of Warmwater Fish. In: Nutrient Requirements of Domestic Animals Series, National Academy Press, Washington, D . C , 78pp. NRC, 1993. Nutrient Requirements. In: Nutrient Requirements of Fish. National Academy Press. Washington, D.C. pp 114. 114 Ravindran, V. , and Blair, R., 1992. Feed resources for poultry production in Asia and the Pacific. Plant protein sources. World's Poult Sci. J., 48: 205-231. Santiago, C.B., and Reyes, S.O., 1993. Effects of dietary lipid source on the reproductive performance and tissue lipid levels of Nile tilapia (Oreochromis niloticus) (linnaeus) broodstock, J. Appl. Ichthyology, 9: 33-40. SAS Users Guide: Statistics, Version 5 Edition. 1985. SAS Inst., Inc., Caary, NC. Shiau, S.Y., and Huang, S.L., (1989) Optimal dietary protein level for hybrid tilapia (O. niloticus x O. aureus), reared in seawater, Aquaculture, 81: 119-Shiau, S.Y., Kwok, C C , Chen, C I , Hong, H.T., and Hsieh, H.B., 1989. Effects of dietary fibre on the intestinal absorption of dextrin, blood sugar level and growth of tilapia (Oreochromis niloticus x O. aureus). J. Fish Biol., 34: 929-935. Shiau, Shi-Yen., Chuang, Jan-Lung., and Sun, Chan-Lan , 1987. Inclusion of soybean meal in tilapia (Oreochromis niloticus x O. aureus) at two protein levels. Aqaculture, 65: 251-261. Shearer, K .D. , 1994. Factors affecting proximate composition of cultured fishes with emphasis on salmonids. Aquaculture, 119: 63-88. Shiau, S. Y. , and Huang, S.L., 1989. Optimal dietary protein level for hybrid tilapia (Oreochromis niloticus X O. aureus) reared in sea-water. Aquaculture, 81: 119-127. Siddiqui, A.Q., Howlader, M.S., and Adam, A. A., 1988. Effect of dietary protein levels on growth, feed conversion, and protein utilization in fry and young Nile tilapias (Oreochromis niloticus). Aquaculture, 70: 63-73. Tacon, A.G.J. , Webster, J.L., and Martinez, C.A., 1984. Use of solvent extracted sunflower seed meal in complete diets for fingerling trout (Salmo gairdneri). Aquaculture, 43: 381-389. Twibell, R.G., and Brown P.B., 1998. Optimal dietary protein concentration for hybrid tilapia (Oreochromis niloticus x 0. aureus) fed an all plant diet. World Aquaculture Society, Vol. 29 1 9-16. Van der Meer, M B . , 1995. The effect of dietary protein levels on growth, protein utilization, and body composition of Colossoma macropomum (Cuvier). Aquaculture Research, 26: 901-909. Villamide, M.J . , and San Juan L.D. , 1998. Effect of chemical composition of sunflower seed meal on its true metabolizable energy and amino acid digestibility. Poultry Science, 77: 1884-1892. 115 Viola, S., and Arieli, Y . , 1983. Nutrition studies with tilapia. Replacement of fishmeal by soybean meal in feeds for intensive tilapia culture. Bamidgeh, 35. 1 9-17. Viola, S., Arieli Y . , and Zohar, G., 1988. Animal protein free feeds for hybrid tilapia (Oreochromis niloticus x O. aureus) in intensive aquaculture. Aquaculture, 75 (1-2) 115-125. Wilson, R.P. and Poe, W.E., 1985. Apparent digestible protein and energy coefficients of common feed ingredients for channel catfish. Prog. Fish Cult., 47: 154-158. Winfree, R.D., and Stickney, R.R., 1981. Effects of dietary protein and energy on growth, feed conversion efficiency and body composition of tilapia aureus. Journal of Nutrition, 111: 1001-1012. 116 Chapter 5 Experiment 3: Partial replacement of fishmeal with high-fibre and low-fibre sunflower cakes in diets for tilapia (O. niloticus): Effect on fish performance and whole body fatty acids 5.0 Abstract The objectives of the study were to determine the effects of replacing fishmeal with high-fibre and low-fibre sunflower cakes (HFSC and LFSC) on fish performance, body proximate composition, and whole body fatty acid, and to determine the upper limit of sunflower cake inclusion that would not compromise fish performance. Sex-reversed O. niloticus males with an initial weight of 15.65g + 0.95 (SD) were used. They were stocked in circular tanks with a base circumference of 1 metre and filled with water to a depth of 0.3 metres. Stocking density for the heaviest fish at the end of the trial was 0.013 kg per litre of water. Water temperature was maintained at 27°C ± 2 °C throughout the trial, and dissolved oxygen concentration in the tanks was above 5.5 mg/litre. A control diet based on herring meal and soyabean meal was formulated. Six test diets were formulated such that low fibre (LF) and high-fibre (FTF) sunflower cakes (SC) contributed 30%, 60% and 80% of the dietary protein and the diets were designated as LFSC-30, LFSC-60, LFSC-80, HFSC-30, FTFSC-60, and HFSC-80 respectively. The rest of the protein in each diet was supplied by herring meal. Fish were fed on the experimental diets for a period of 70 days. At the end of this period, they were starved for 24 hours and weighed. Five fish representing the average weight of each replicated group (n = 3 per diet treatment) were 117 frozen in plastic bags at -5 °C for determination of body composition and fatty acid composition. At 70 days, the absolute weights and weight gains were similar between fish fed the control diet, and those fed the LFSC-30, LFSC-60, LFSC-80 and HFSC-30 diets. Specific growth rates were 1.60%, 1.60%, 1.51%, and 1.32% for the control fish, and those fed the LFSC-30, LFSC-60, and LFSC-80 diets, respectively, and 1.57%, 1.38%, and 1.26% for the HFSC-30, HFSC-60, and HFSC-80 diets respectively. Feed intake decreased with increasing levels of sunflower cake in the diet. Fish fed diets LFSC-80, HFSC-30, HFSC-60 and HFSC-80 had significantly lower feed intakes than those fed the control diet. Protein efficiency ratio (PER) and productive protein value (PPV) did not differ significantly between fish fed the control, LFSC-30 and HFSC-30 diets. Fatty acid levels (%) in the whole body were significantly influenced by diet. Linoleic (18:2 co 6), oleic (18:1 co 9), and palmitic (16:0) acids were the most abundant fatty acids in the diets and in the fish bodies. Percentages of the long chain poly-unsaturated acids, of the co3 family viz., docosahexaenoic (22:6 co 3) and eicosapentaenoic (20:5 co 3) acid, were low in the diets and in the fish bodies. The low-fibre sunflower cake could replace 60% of the dietary protein without compromising fish performance. By contrast, the high fibre cake could only constitute 30% of the protein in the fishmeal/soybean control diet without adversely affecting fish performance. At higher levels of inclusion, the fish did not consume enough feed to meet their nutrient requirements, and hence their growth was reduced compared to those fed the control diet. 118 5.1 Introduction and objectives In the first experiment (Chapter 3), it was established that the digestibility of protein in the high-fibre and low-fibre sunflower cake diets by tilapia (O. niloticus) was high, and that despite the low level of lysine in these diets, the fish fed the diets where low-fibre sunflower cake supplied 50% of the dietary protein performed as well as those fed diets based on anchovy fishmeal as the major protein source. Digestibility of the energy in low-fibre and high-fibre sunflower cakes was found to be low (Experiment 1, Chapter 3). Consequently, the digestible energy levels in the diets based on sunflower cakes were also low, and they were increased by the addition of corn-oil, which also served as a source of linoleic acid. Kanazawa et al. (1980), and Tekeuchi et al. (1983), established that the only essential fatty acid required in the diet of tilapia is linoleic acid (18:2 eo 6). Moreover, it was shown that tilapia possess enzymes that desaturate and elongate fatty acids of the co6 series to provide sufficient levels of the long chain unsaturated fatty acids necessary for membrane fluidity and function. For that reason no attempt was made to add these fatty acids to the diets. There is evidence that suggests that consumption of fish containing high levels of highly unsaturated ©3 fatty acids is favorable for human health (Higgs, 1986; Bates et al, 1989; Thais and Stahl, 1987). Generally, marine fish oils are characterised by low levels of linoleic (18:2 oo 6) and linolenic acids (18:3 oo 3), and high levels of the long chain ©3 polyunsaturated fatty acids, with eicosapentaenoic (20:5 oo 3) and docosahexaenoic acid (22:6 co 3) being the predominant oo 3 fatty acids (Hilditch and Williams, 1964; Yamada and Hayashi, 1975). Many studies have been done to assess the effect of the fatty acid composition of the diet on the fatty acid fatty acid composition of fish. Body fatty acid 119 composition to a large extent has been found to reflect the dietary fatty acid composition (Toyomizu etal, 1963; Braekhan etal, 1971; Yu and Sinnhuber, 1972). The objectives of this experiment were: a) To establish the upper limit to which high-fibre and low-fibre sunflower cakes could replace fishmeal in the diets of tilapia (O. niloticus) without affecting growth rate, c) To evaluate the effect of substituting sunflower cake for fishmeal on whole body fatty acid composition. 5.2 Materials and Methods 5.2.1 Experimental diets and design Seven diets whose compositions are shown in Table 5.1 were formulated. They contained between 2600 and 3297 kcal/kg of D E (DM basis), and approximately 31% crude protein ( D M basis). A low-fibre sunflower cake (CF = 10%) supplied 30, 60, or 80% of the dietary protein in the first three test diets. A high fibre cake (CF = 24%) was used in a similar manner for the other test diets. Danish herring meal and soybean meal were the protein sources in the control diet. The low-fibre sunflower cake was processed from a high-oil hybrid variety (Kenya Fedha) as explained in Experiment 1. Each diet was fed to three groups of 17 fish, three times daily, for a period of 70 days. The D E concentration of the diets was calculated as explained in Experiment 2, chapter 4. The digestible energy concentration in herring meal was estimated as 4000 kcal/kg (DM basis) (Degani et al, 1997). Digestible energy for wheat flour was estimated at 4169 kcal/kg ( D M basis) (Degani et al, 1997) and the digestibility coefficient for energy in soybean meal was estimated as 70% (Popma, 1982). 120 Table 5.1: Compositions of the diets used in Experiment 3 (g/100 g air-dry). Low-fibre SFC 1 High-fibre SFC Control % protein supplied by sunflower cake Diets 30 2 LFSC-30 60 LFSC-60 80 LFSC-80 30 3HFSC-30 60 HFSC-60 80 HFSC-80 0 Control Diets Herring meal 28.4 15.8 7.5 28.3 16.3 7.3 31.5 SFC (low- fibre) 25.0 49.0 65.0 - - - -SFC (high- fibre) - - - 26.0 50.0 68.0 -Soybean meal - - - - - - 16.0 Corn starch 31.0 16.1 5.9 29.6 10.7 - 38.9 Whole wheat 8.0 8.0 8.0 8.0 8.0 8.0 8.0 Corn oil 3.5 5.9 7.7 4.0 10.0 11.3 1.5 Dicalcium phosphate 1.5 2.6 3.3 1.5 2.4 2.9 1.5 *Vit/min premix 1.0 1.0 1.0 1.0 1.0 1.0 1.0 Iodized salt 0.5 0.5 0.5 0.5 0.5 0.5 0.5 Ascorbic acid 1.0 1.0 1.0 1.0 1.0 1.0 1.0 Choline chloride 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Chemical composition (DM basis)4 DM 92.8 91.2 92.3 90.9 90.2 91.3 89.0 DE (kcal/kgDM) 3150 2974 2870 29871 2850 2600 3057 Crude protein (%) 30.0 31.8 31.4 30.8 31.0 30.0 32.6 Crude fat (%) 5.3 13.5 20.3 7.4 8.5 13.9 4.7 Crude fibre (%) 3.2 5.9 7.6 6.6 12.3 17.5 1.2 ADF (%) 4.7 14.5 17.4 8.1 15.0 20.0 3.3 NDF (%) 6.5 13.8 17.4 11.7 22.7 30.5 10.5 Calcium (%) 1.6 1.8 1.2 1.5 1.7 1.0 1.6 Phosphorus. (%) 1.4 1.5 1.1 1.4 1.5 1.2 1.4 'SFC Sunflower cake. 2LFSC Low-fibre sunflower cake 3HFSC High-fibre sunflower cake 4 All values were determined by analysis except for DE, which was estimated from published data (see text). Vitamin/mineral premix contained the following per kg: Vitamin A, 6000 IU; Vitamin D 3 ) 600 IU; Vitamin E, 100 mg; Vitamin K 3 , 3 mg; Vitamin B,, 10 mg; Vitamin B 2 , 20 mg; niacin, 150 mg; D-pantothenic acid, 50 mg; Vitamin B 6 , 10 mg; Vitamin B 1 2 , 0.03 mg; folic acid, 4 mg; biotin, 0.8 mg; choline, 600 mg; Vitamin C, 600 mg; inositol, 300 mg; manganese, 192 mg; iron 51.2 mg; copper, 6.4 mg; zinc, 57.6 mg; selenium, 0.15 mg; traces of cobalt and iodine. 121 5.2.2 Fish sampling O. niloticus sex-reversed males weighing 15.65g + 0.95 (at the start of the experiment) were used for this trial. They were bought at Sagana Farm in Kenya (Sagana, Kenya) and transported to the research facilities in Nairobi. They were acclimated to laboratory conditions for a period of 2 weeks before the onset of the trial. At the end of the second week, they were weighed in groups of 17 fish and these were randomly allocated to the experimental tanks. Stocking density for the heaviest fish at the end of the trial was maintained below 0.013 kg per litre of water. Each tank was fitted with an AS8-1 (3 inch) diffuser. Dissolved oxygen concentration in the tanks was maintained above 5 mg/litre. Water in the tanks was completely exchanged every 48 hours, or when the dissolved oxygen concentration in the tanks fell below 5 mg/litre. Feed intake for each group of fish was recorded daily, while water temperatures were taken 3 times a day. Water temperature was maintained between 25°C and 28°C throughout the experimental period by the use of thermostatically-controlled heaters. Fish were weighed 3 times - namely on day 0, 32 and 70. Before weighing time, the fish were starved for 24 hours. At the end of the experimental period, 5 fish representing the average weight of fish in the tank from which they were taken were selected and killed with an overdose of MS-222. They were frozen at -5 °C and stored in plastic bags pending analysis of moisture, protein, crude fat, ash and fatty acids. Before analysis, the fish were partly thawed and chopped into small pieces. Then they were homogenized in a food blender. The homogenized samples were divided into two portions. One part was used for the analyses of moisture, protein and ash at the University of Nairobi, while the other portion was used for fatty acid analyses in Canada. 122 5.2.3 Data collection and analytical procedures The following parameters were used to assess fish growth and performance: absolute weights, weight gain, specific growth rate, feed intake, and feed conversion. Specific growth rate (% per day) was calculated as: 100*(ln final wt (g) - ln (initial wt (g))/number of experimental days. Feed conversion ratio was calculated as ingested dry feed (g)/wet weight gain (g). PER was calculated as wet weight gain/ingested protein and PPV as 100 (gain in body protein/protein intake). 5.2.4 Chemical analyses All ingredients and diets were analyzed in duplicate for their contents (%) of dry matter, ash, protein, lipid, fibre, calcium and phosphorus according to standard procedures (AOAC, 1984). ADF and NDF were analyzed according to the method of Waldern (1971), using an Ankom technoloanalyzer (Ankom Technology, 140 Turk Hill Park, Fairport, N Y 14450). Analyses of total lipids in fish and diets were done at the Department of Food Science, U B C . Total lipids were extracted in 50 ml of chloroform:methanol mixture (2:1) according to the procedure of Folch et al. (1957). The fatty acid compositions of the diets and fish samples were measured after methylation of the samples by gas chromatography (Shimadzu GC-17A). In methylation, 10 mg of the lipid material was saponified with 2.5 ml of 0.5 N CH3 -OH-KOH. This was achieved by first neutralizing with 0.4N H C L , and then adding 5 ml of boron trifluoride. The mixture was heated for 15 to 20 minutes to achieve complete methylation. The fatty acid methyl esters were extracted 3 times with hexane and concentrated. In the analysis for fatty acids, a fused silica capillary column (Omegawax 320, Supleco Park, Bellefonte, PA, USA) was used. 123 Temperature of injection was 150 °C. It was increased by 2 °C per minute to 170 °C. It was then increased by 3 °C per minute to 210 °C, and maintained at that temperature for 9.5 minutes. The detector temperature was set at 220 °C. An auto injector was used, and helium was used as the carrier gas at 1 ml per min. Peak areas were quantified using a Shimadzu Class VP chromatography data system, Version 4.2, and by reference to an internal standard (heptadecanoic acid, C 17:0). Amino acid compositions were determined by H P L C after performic acid oxidation (AOAC 1998) (See Chapter 4). 5.2.5 Statistical analyses Data for absolute weights, weight gain, specific growth rate, feed intake, feed conversion, body composition and fatty acid composition were subjected to statistical analyses using PROC G L M of the Statistical Analysis Systems (SAS 1985). An analysis of covariance was done on all the fish performance data using the initial weight of the fish as the covariate. In the analyses of the body composition data, the final weight of the fish for each group was used as the covariate. After the initial comparison of the seven diets as a 7 x 1 completely randomized design, the performance of fish fed on the low-fibre and high-fibre sunflower cakes was compared using a 2 x 3 factorial design (2 sunflower cakes x 3 levels of dietary inclusion). Treatment means were separated using Tukey's multiple range test. The level of significance was set at P < 0.05. 124 5.3 R e s u l t s a n d d i s c u s s i o n 5.3.1 Chemical composition of the diets The chemical compositions of the diets used in this experiment are presented in Table 5.1. The low digestible energy content in the HFSC-80 diet is consistent with the low digestible energy level in that cake. Despite the addition of corn oil, the energy level of this diet was still below the levels of the other diets. The D E concentration in most of the diets based on sunflower cake was slightly below the 3000 kcal/kg (DM) stipulated by N R C (1993) for tilapia (O. niloticus). Crude protein percentages of the diets ranged from approximately 30% to 32.6%> (DM basis). As in Experiment 2, crude fat levels (%) in the LFSC-60 and LFSC-80 diets were high due to the high residual oil content in the low-fibre cake and the high amounts of the cake needed to provide 80% of the dietary protein. Extraction of oil from dehulled seeds using a conventional screw press is more difficult than from seeds that contain husks, which explains why the low-fibre cake diets had consistently higher oil levels than the diets based on the high-fibre cake. ADF and NDF percentages increased with increasing levels of sunflower cake in the diets, and they were higher in diets based on the high-fibre sunflower cake. Phosphorus was maintained above 0.6 %» (air-dry basis). Dietary levels of essential amino acids are shown in Table 5.2. The lysine levels in the diets containing the highest levels of the sunflower cakes were 1.34 % and 1.00% for the LFSC-80 and HFSC-80 diets, respectively, which were lower than the stipulated requirement for juvenile tilapia (O. niloticus) (Santiago and Lovell, 1988). Methionine level was also low in the diets based on the high-fibre sunflower cake. 125 C O os as (,_, '-1 o U ^ g 53 +-» o >H Q. o 00 HH C O o vo o V O u G O o C O (N t r . CO a ° O i 90 O HH O 00 J C O o i vo O HH U V O H J C O C O <=> H J U C O - C O HH =1 HH U CO 6 & g '53 £ Q O CN i-H OS CN 00 CN l-H CO i-H vo m v-> o r-~ «-» oo • t H n n i/i I C O C O C N vo C N co cs Tt o CNI in Os o m oo >-< © © i-H ^H' © ' C O C O Tt i-H i-H 00 -H' d CN| r--vi i tr,i coi Os v i os C N co V O ^H C O V } ' CO' l-H i-H co co Tt v o - H V O T t © C ~ - v i 0 0 v o O s O s c N © v i c o _j, oq vt © vo_ © v i Tt ^H OS C N Tt C O <-H Tt ^H •< r H O ' H H H O C i H O H f v i l f i H ' H ' H Z C N C O O O O S O S T t ^ H O S C O C N vo C N co' vi Tt CN C N co' co' Tt C O © t ^ T t - H C O W - > C N - H © T t V 0 ' / - > C N 0 0 ov r~ i-H oo vt r - vo C N © C O p C N C N co < I - H O' i-H I—I i—I ©' O' l-H I—I I—I C O vi l-H - H i-H Z I-H p vq Tt r»_ co p vq vq co vi CN* co vo vi C N C N co co Tt r - - H o r - i / - ) - H o o - H C N r - O T r c N o c «Ti voi I-H ov c-. r-. vq i-< I - H COI V O C N -H V O C O < , l—I O i-H i-H i-H © i O i-i l-H l-H CN Tt i-H i-I i-I Z C N C O Vt 00 C O Vt C N C N C O vo vo C N co' vi Tt C N C N Tt co' Tt T t - H O S l - H T t O S O C N C O T t T t C O O V O r ~ n Ov r~- O 001 co f ~ ; coi p Tt; © CN co t— CO <; ^H © i—< I—I I—I O O i-H I—I I—I CO vci i—I i—i i—I Z os ov os oo vi CN co vi Tt CO' Tt CO' O - H 0 s r ^ C N 0 S O - H 0 v 0 0 T t ' - < O V 0 t ^ . r- ov, •—i oo vt f~ o co o Tt r- ^ ^ vo C N < ^H O ^H rH ^H O ^H ^H ^H ^H CN Tt r~i H^* H^* Z oo co I - H vq C N vo I - H t C N o vi C N Tt vo' vo' C N C N C O Tt vi T t O s c N O O v O O v c N O v r ^ o s O S T t O S i - H O -H r ~ V 0 C N 0 s 0 0 t ^ V 0 - H C N T t O t ^ C N 0 0 V - l - < ^ H O r H ^ H ^ H d O - H ^ H r - I C O * T t ^ H l - H l - H Z c N o s r ~ T t r ~ v o ^ H o o o s i o V } ' i-H* CO* VO* V O CN* i-H CO CO Tt O O H M O V h O M O ^ h O O i f l O V O O M t ^ VCi VC; C N I O i - | i f ; ; COI CN CN Tt Ov V I CO VO VO H ? i-I ©' i-I CN* CN* © ' O i-I I-H i- i CS Tt i-I i-I I-H m 'a '5b < s cj •5 CJ j a 3 s C J cj . g - g c 5 .5 ^ J S o £ P > < . g '53 CH VO CN T3 C J N o Threonine levels in the LFSC-80, HFSC-60, and HFSC-80 diets were 1.03%, 1.01%, 0.96% (DM % ( D M basis), which were also lower than the stated requirement of 1.05% of diet ( D M basis) (NRC, 1993). All the other diets met the requirements for all the amino acids. The percentages of fatty acids in the diets are presented in Table 5.3, and the fatty acid compositions of corn oil and sunflower oil are shown in Table 5.4. The control diet, LFSC-30 and HFSC-30 diets had high levels of palmitic acid (16:0) compared to the other diets. These diets contained more herring meal, which has a higher level of that fatty acid than corn oil and sunflower oil (Table 5.4). The fatty acid compositions of the diets based on low-fibre sunflower cake reflected the fatty acid composition of sunflower oil, which was high in these diets, while the fatty acid composition of the high-fibre cake diets reflected the composition of both corn-oil and sunflower oil. Both sunflower oil and corn oil have high levels of oleic (18:1) and linoleic acids (18:2 co 6) which were reflected in the diets. Linolenic (18:3 co 3), eicosapentaenoic (20:5 co 3) and docosahexaenoic acids (22:6 co 3) were all very low in the diets. The control diet had a slightly higher level of docosahexaenoic acid than the other diets. 5.3.2 Fish performance, PER, PPV, body and fatty acids composition Data on absolute weight, weight gain, specific growth rate, feed intake, and feed and protein utilization are shown in Tables 5.5 and 5.6, while data on whole body proximate composition are shown in Table 5.7. Specific growth rates were not significantly different (P > 0.05) for fish fed the control diet, and those fed the LFSC-30, LFSC-60, and HFSC-30 diets. Fish fed the HFSC-60, LFSC-80 and HFSC-80 diets had 127 CO 12 cd C+H "cd +-» o +-> H-H o o x 22 CD X ! CO 13 cd M-H C+H o co CD CD o CD PH CO in' CD U oo O CJ OO O CO — U C/j a CJ OO. P o on fe P o CJ O fe OO o <+H g '53 +J o O H ID h M t— V O V O O t"- T f T f •-< T f ' o c o T f T3 CD 3 O O O cd Tf' vb oo CD td k H i C/5 "cd +-> o H IT) vo CO CO VO T f T f 0 0 CS ON •—1 CO CO ON CN CN CO o o iri Tf ' o © -—< CN r-' © ' P p l> 0 0 CN T f T i - r o CO ON T f VO ON CTA T f r o vo T f ON o ON o 0 0 T f 0 0 r o CN •—i CO CO T f CN o o VO CO o © ' 00 ON ON o O o O VO CN CO CO CO T f ON VO ON T f CO t~- IT) C-- rt i n 00 o CN I—1 ir> 00 CN © o r o ON © CO o CN CN T f 0 0 -—i CO iri TT O O r-i VO r o CN CN CO r o r o CO CO I/O VO T f CO T f r- vo r o CO CO T f •—1 CO o o IT) CN _ _ 0 0 O o iri iri -—i o co Tf ' T f O P P Tf ' 0 0 '—' CN r o r o T f ^< T f r- CN r- VO t-» r-- T f r- VO T f TT ,_. VO T f 00 t-- ON ir> CN CN CO T f o o t > iri CO o CN CO I—H o O O CN vo CN CO CO TT T f VO r - T f T f 0 0 vo CN 0 0 '—_ •—< CN ir> VO ON IT) r o _ IT) 0 0 o ON iri iri i—i T f VO t--' p © ' P l> r o 1 — 1 CN CO CO r o CO T3 CD td Ui I c/l t~- ON -a CD Id UH 3 ts CO a i o a o o 3 3 «VO 00 o - - H H ON ON 0 0 r- CN ON O o T f IT) ON T f CO T f iri ON CN i—i CN CO CN CN CN T 3 CD td k H I 11 I T l ON ON CN T3 CD td UH 3 td CO e 3 >N o co r o r o 3 3 3 ^ _>N CN CO I T ) vp ^ 2 O 0 0 0 0 © CN P PH I-H <-H CN CN H CD cd o WH CD O 3 CO o P I CD cd o <-< CD O 3 CO CD on ffl I CXI fe P 00 cs T3 CD +-» CJ CD +J /-> 1— 1 Tf CO o r -ON d d d d © Tf r -o NO CN T) /-> o 00 r o T3 u CO Tf o ON *« Tf O CN CO o CN Ti ON r o —' Tf' CN NO V O CN CN CN r o Tf •8 T o NO CN & CO CO •8 CN CO -8 > fe fe PH PH g u y cd oo OJ • i > '8 -*-» s CH OJ *+H CJ -§ p > PH PH O •i OJ 'o OJ g '53 H-t o *H CH II PH W PH .2 ca s >H 1 o CJ T3 OJ ,1> H-H Pi o fe C/3 o c/3 a (H ° CJ •s HI Q. If 8 CJ o l H CJ IPH ° c o CN CO CN ON CO CO ON O ON CN O O r-© © © © © ' d d J3 o ON o ON CN o 00 o co 00 X i © CO NO CN ON CO rt co' CN NO NO CN rt CN CN •n NO Si V Tt O v, o ON CN NO X i W] CO X i ON Tt r—1 NO CN oo' NO CN rt CO CN CS o ON co r-f- o t-O Tt CN ON CN CS O Tt o o o o o o o I—I O O Tt V> ON 00 ON Tt NO O NO CO Tt d i-< NO' r— CN* I-H co* Tt CN NO CN ON 00 Vt V) 00 Tt NO C- Tt ON ON Tt Tt Tt O CN i™< Vt r-~ CN H ^ 0X •3 S ^ a " - . a l l 5 PH S CJ O PH PH I> C/3 !> PH pL, CM PH 03 60 PH K.OO^IH Table 5.7: Percentages of body proximate constituents viz., moisture, protein, fat and ash (Air-dry basis) at 70 days in relation to diet treatments. % Protein from SFC 'Moisture 'Protein 'Fat 'Ash Control 0 69.1 16.60 7.40 5.40 LFSC 30 68.0 17.00 8.30 5.10 60 68.7 16.20 7.90 5.40 80 68.4 16.30 8.50 5.30 HFSC 30 67.9 16.00 8.70 5.30 60 69.6 16.00 7.00 5.60 80 69.4 15.80 7.70 5.40 SEM 0.53 0.28 0.36 0.12 2NS NS NS NS ^eans (n = 3) 2 Not significant 132 significantly lower growth rates than those receiving the control diet. Weight gain was significantly influenced by diet (P < 0.05). Generally, the fish fed on the diets made from the low-fibre cake (LFSC), gained more weight than those fish fed the diets made from the high-fibre cake (HFSC). Among the fish fed the diets based on the LFSC, the weight gains for those fed the LFSC-30 and LFSC-60 diets were not significantly different from that of the control fish (P > 0.05), while fish fed the LFSC-80 diet gained significantly less weight (P < 0.05) than those fed the control diet. For fish fed the diets based on the HFSC, only those fed the HFSC-30 diet had an average weight gain comparable to those fed the control diet (P > 0.05). Generally, the weight gain for fish fed the HFSC-30 diet was higher than that noted for fish receiving diets with higher levels of the HFSC, but the differences were only significant relative to the fish fed the HFSC-80 diet. Mean feed intakes for the fish fed the LFSC-30 and LFSC-60 diets were not significantly different from that of the control fish. The feed intakes of fish fed the LFSC-80 diet and all of the diets based on the HFSC were lower than observed for the control fish. The type of cake (low-fibre and high-fibre), and the level of cake in the diet (30% of protein, 60% of protein and 80% of protein) significantly affected feed intake (P < 0.05) (Table 5.6). Generally, fish fed diets based on the LFSC had a higher feed intake than those fed the HFSC diets. The interaction between the type of cake and the level of cake in the diet was significant for feed intake. Feed intakes decreased as the level of sunflower cake in the diets was increased (Table 5.6), but the decline was higher for fish fed diets based on the HFSC. Differences in feed intake between fish fed the LFSC and HFSC diets and between those fed the diets where sunflower cakes supplied 30%, 60%, and 80% of the protein may have been caused by the amount of fibre in the diets. 133 Dietary fibre dilutes nutrient density and increases bulkiness in feeds. Fish respond to the diluting effects of fibre by increasing feed intake, thereby consuming enough nutrients to grow at a rate comparable to that of the control fish. This compensatory increase in feed consumption may however be hindered by physical limitations in the ability of the gut to extend, in which case feed consumption may not increase enough to satisfy nutrient requirements. Tilapia have a relatively small stomach (Balarin, 1979), which may limit the extent to which the gut can distend to compensate for the increased dietary fibre level. Reduced feed intake may also be due to the presence of phenolic compounds in sunflower cake. The most important of these compounds is chlorogenic acid (Sabir et al, 1974). Sunflower cake contains 1% to 3% of chlorogenic acid (Sosulski and McCleary, 1972), which has been shown to reduce feed intake and weight gain in rats (Liener, 1980). In fish, no deleterious effect of chlorogenic acid has been reported. The concentration of the acid is highest in the hulls (Bau et al, 1983), which may explain why the reduction in feed intake of the fish in the current study was more prominent when they ingested diets containing the high-fibre sunflower cake. In trout, Stickney (1996) observed that feed intake was depressed when sunflower protein concentrate was fed at 35% of the diet. The depressed feed intake was attributed to low palatability of the protein concentrate. Tacon et al (1984) also noted that rainbow trout fed diets containing 36% sunflower cake (25% fibre) had reduced feed intake compared to those fed the control diet. In tilapia, Jackson et al. (1982) fed diets where sunflower cake provided 75% of the protein and observed a growth rate similar to that of fish fed the control diet based on fishmeal. The sunflower cake used in their study had a fibre content of 14%, which is comparable to the low-fibre sunflower cake used in the present study. However, 134 the growth rates of tilapia attained in the study by Jackson et al. (1982) were low, even for those fed the control diet, indicating that there could have been other factors affecting fish performance. In the present experiment, feed intake was reduced in fish fed diets containing the highest level of sunflower cake, especially those containing HFSC. The type of sunflower cake (low-fibre and high-fibre) in the diets also affected weight gain and PPV, but not specific growth rate, FCR or PER. In this regard, fish fed diets based on the low-fibre cake gained more weight than those fed the high-fibre cake diets (Table 5.6). Although the type of cake had no significant effect on PER, the level of cake in the diet did significantly affect values for PER (Table 5.6). Fish fed diets where sunflower cake provided 30% of the protein had significantly higher PER values than those fed diets where sunflower cake provided 60% and 80% of the protein. In a review of factors affecting dietary protein utilization, Steffens (1981) lists fish species, fish size, environmental and feed factors as some of the main factors that determine protein utilization. In the current experiment, the same source of fish was used and they were maintained under similar environmental conditions, regardless of diet treatment. Thus the only difference in the current study was in diet formulations. Some of the dietary factors that could have contributed to the observed differences were; digestibility, levels of essential amino acids, lysine to total protein ratio, digestible energy concentration and palatability of the diets. In relation to these, the digestibility of protein in the low-fibre and high-fibre cakes was high (Experiment 1, Chapter 3), and the determined levels of essential amino acids did not differ markedly among the diets, except for lysine, which decreased as the level of sunflower cake in the diet was raised. The other probable cause of the observed differences was the dietary D E concentration which was lower in the 135 diets containing high levels of sunflower cake (LFSC-80, HFSC-60 and HFSC-80) than the other diets. Therefore, it is plausible that some of the dietary protein in the latter diets could have been used for energy production. PPV (protein gain/protein intake) is a measure of protein quality that takes into account elaboration of new tissue protein in relation to dietary protein intake. It is an indicator of the amount of dietary protein required to gain a unit weight of protein in fish (Steffens, 1981), and it is affected by the same factors that influence PER. The type of sunflower cake and the level of sunflower cake in the diets significantly affected PPV. Fish fed diets based on low-fibre sunflower cake had higher PPV values than those fed diets based on the high-fibre sunflower cake. As with PER, PPV values decreased with increasing levels of sunflower cake in the diet. Fish fed diets where sunflower cake provided 30% of the protein had significantly higher PPV values compared to those fed diets where sunflower cake provided 60% and 80% of the dietary protein. Diet did not have a significant effect on the whole body proximate composition of the fish (Table 5.7). Morever, there were no specific trends in the contents (%) of moisture, lipid, protein or ash in relation to diet treatment. In salmonids, body lipid content is influenced more by energy intake, rather than lipid intake (Shearer, 1994), whereas in tilapia, De Silva et al. (1991), Hanley (1991) and Chou and Shiau. (1996) observed that fish fed high-fat diets had more body fat than those fed low-fat diets. In the current experiment, fat deposition did not show any particular pattern. Fatty acid levels in the whole body of fish are presented in Tables 5.8 and 5.9. Whole body fatty acid composition closely resembled dietary fatty acid composition (Table 5.8). 136 c cd o •a op o cd cd -*-» o M-H O c IU I CO CU '-3 o +H c o "5 I-I .£ cn M-H O T3 O X> J U "o cu Si cu o cd cd PH 00 m 2 o VO O oo o hJ oo O CO PH O >-> VO O co PH O i—1 c o O v O v T f O c o T f f v o O r-I r-n' CS O m CN v o c o r; N o oo oo CO T f O T f CS CS CO v o v o f v o v n f o c o o o o T f O C~" '—' Tf" r-H , - H T f f Ov f VO Ov f T f c s vq T f i n ' vo r~ (S CO i n CO ^3 OV > n c s CO •8 T f CO o T f 00 CO T f O v XI o r- 00 00 CN T f CO CS T f > n c s v o CS T f o o ^ H i n o v c s « n o O v o ^ H co T f o T f d S i o o o o ^ c s T f vb od r- v o T f i n d d T f r- ^ H > n o o T f vo ov i n i n o vq c o ov c s c s c o f i n c s VO Ov vo i n o v T f CO f T-H '—' CS T f CO CO t v o CO Ov 00 00 I - H d c s CO CO CS T f v o r - - i n c s Ov CS —< CS OO r-H i n i n ' ^ H c s c o 1 I H cs o o o o f • r— c o o o v o c o r- _ o t— a d d B H H Ov I-H VO T f c s c s O T f T f T f OV CO d d d d ^ H c o CS CS Ov VO Ov Ov Ov CS VO CO CO i—i o f rn d d d -.o T f o o v o c o v n v o f c s v o O o o i n •^r d d d d \6 CO CO •S O I— VO l - H oo c o c o i n o T f i n H O CO O — CO CO i n v o oo o v ^ H o v r— co cs o oq cs d d d ~ o\ cs cs f CS OV ^ H CS T—l o v v o c s T f c s i n c o d d d c o o o .>> c s • ! T f m v o o oo 2 6 d i s PH — CS CS CS o P H 3 o H O a CD Ui cd O t-H cd C o M 3 Si o cd 1) i -cS u-cu H-» CN CU Table 5.9: Effect of type of sunflower cake and level in the diet on percentages of whole body fatty acids. Main effects Type of sunflower cake Percentage protein from SFC ! L F S C 2FfFSC S E M 30 60 80 S E M Fatty acid 3 2 2 a b 12:0 2.92 3.40 0.24 2.23b 4.05a 0.29 14:0 4.18 6.26 0.93 4.63 7.23 3.80 1.14 16:0 20.19 20.88 1.08 24.22 18.59 18.80 1.33 18:0 6.16 7.62 1.62 6.48 8.95 5.25 1.99 16:1 2.01 2.59 0.29 2.69 2.76 1.44 0.35 18:1 31.39 31.97 0.34 31.74ab 2997b 33.33a 0.41 18:2 30.93a 24.48b 1.60 25.22 25.69 32.21 1.96 18:3 0.31 0.37 0.01 0.36 0.35 0.29 0.017 20:4 0.48 0.54 0.05 0.33b 0.48ab 0.71a 0.06 20:5 0.06 0.21 0.10 0.14 0.25 0.013 0.13 22:6 1.36b 1.68a 0.05 1.95a 1.66b 0.94° 0.06 Means that do not have a superscript or share a common superscript letter for the same factor in a row within a main effect are not significantly different ( P > 0.05) T J S C Low fibre sunflower cake 2 HFSC High fibre sunflower cake 3Means (n = 6) 4Means (n = 4) 138 Palmitic (16:0), oleic (18:1 co 9) and linoleic (18:2 co 6) acids were the most abundant in both the diets and fish. The levels of these fatty acids in the diets ranged from 16.0% to 31.0% for palmitic acid (16:0), 25.5% to 41.9% for oleic acid (18:1 co 9), and 23% to 45.0 % for linoleic acid (18:2 co 6). Palmitic acid (16:0) was the most abundant fatty acid in the control diet, which was reflected in fish fed this diet. Indeed the control fish had a significantly higher content of palmitic acid (16:0) than fish fed diets with high levels of sunflower cake (LFSC-60, LFSC-80 and HFSC-80). The higher level of 16:0 in the control diet was due to the higher percentage of herring fishmeal, which has a higher content of palmitic acid (16:0) than present in sunflower oil and corn oil (Table 5.4). Oleic acid (18:1 co 9) was the most abundant of the mono-unsaturated fatty acids in both the diets and the fish. Fish fed the control diet had significantly lower levels of this fatty acid than those fed diets based on the two sunflower cakes. There was a trend to increasing levels of oleic acid (18:1 co 9) in fish fed diets with high levels of sunflower cake, especially the high-fibre cake. Diets based on sunflower cake had high levels of corn oil and sunflower oil, while the control diet was rich in herring oil from the herring fishmeal. Corn oil and sunflower oil have higher levels of oleic acid (18:1 co 9) than herring oil (Table 5.4). Linoleic acid (18:2 co 6) was the most abundant fatty acid in diets based on sunflower cake. The level of this fatty acid was also significantly higher in fish fed diet with a high level of sunflower cake (LFSC-60, LFSC-80 and HFSC-80) than in those fed the control diet. It is also worth noting that the levels of linoleic acid (18:2 co 6) were much lower in the fish bodies than those determined in the diets. In tilapia (O. niloticus), 139 desaturation and elongation enzymes efficiently convert C 18 P U F A to longer chain P U F A (Kanazawa et al, 1980; Olsen et al, 1990; Tekeuchi et al, 1983). It is plausible that some of the dietary linoleic acid was converted into the long chain highly unsaturated fatty acids, especially arachidonic acid (20:4 co 6), which was not detected the diets, and yet was present in the fish. The fish had low levels of eicosapentaenoic acid (20:5 co 3) and docosahexaenoic acid (22:6 co3). This may have been caused by several factors. First, the fish were frozen at -5 °C for a period of about one year before they were analyzed for fatty acids. The highly unsaturated fatty acids are more susceptible to oxidation and may have undergone some degree of oxidation during storage, leading to the low observed values. This notwithstanding, levels of these fatty acids are generally low in fresh-water fish compared to marine fish. Ackman (1967) and Hilditch and Williams (1964) noted that fatty acids of fish could be altered by manipulating temperatures. At low temperatures, there was an increase in the long chain highly unsaturated fatty acids which are necessary for membrane fluidity at these temperatures. It is possible that the high temperatures prevailing during the experimental period may also have contributed to the low levels of these (20:5 co 3 and 22:6 co 3) fatty acids observed. Besides, the diet had very low levels of the 20:5 co 3 and 22:6 co 3 fatty acids, as well as 18:3 co 3 precusor needed to make 20:5 co 3 and 22:6 co 3 fatty acids. This was reflected in the fish bodies. The addition of fish oils to fish diets has been shown to increase the body and tissue contents of co3 unsaturated fatty acids in a number of fish species. This observation has been made in various species such as rainbow trout (Oncorhynchus mykiss) (Yu et al, 1977), channel catfish (Ictalurus punctatus) (Stickney and Andrews, 1972), grass carp 140 (Ctenopharyngodon idella) (Tekeuchi et al., 1991) and hybrid tilapia (Oreochromis niloticus x Oreochromis aureus) (Chou and Shiau, 1999). In tilapia, recent studies (Chou and Shiau, 1999) have shown significant increases in the percentages of ©-3 fatty acids in the muscle and liver by feeding diets supplemented with cod liver oil. In the same study, it was also observed that eicosapentaenoic acid (20:5 co 3) is not well retained by tilapia. The levels of docosahexaenoic acid (22:6 co 3) were not appreciably different from those determined by Huang et al. (1998) in the muscles of hybrid tilapia (O. niloticus x O. aureus) fed diets fortified with soy oil. The levels of docosahexaenoic acid reported in the above study were 2.1% for the fish fed a lipid-free diet (0.02% lipid) and 3.8% for fish fed diets containing soy oil. In addition to dietary lipid composition and level, the fatty acid composition of fish is affected by other factors such as temperature, fish size, section analyzed, sex, and physiological status (Kinsella et al, 1977; Stephens, 1997). In the present experiment, all these factors were similar for all treatments. The only difference between treatments was in the final weights but this was taken into account when analyzing the fatty acids composition data by analysis of covariance, with the final weight of the fish as the covariate. Whole body percentages of fatty acids did not differ appreciably between the two types of sunflower cakes mainly because of the similarity in fatty acid composition of sunflower oil and corn oil. Diets based on the LFSC had higher level of linoleic (18:2 co 6) acid than those made from the HFSC. This was due to the higher content of residual sunflower oil in the LFSC. Sunflower oil contains approximately 66% linoleic acid (Table 5.4). 141 5.4 Conclusions Al l diets contained adequate levels of essential amino acids, except for lysine and threonine, which were low in the diets containing high levels of sunflower cakes. There was little variation in dietary fatty acid percentages. The long chain polyunsaturated fatty acids, eicosapentaenoic and docosahexaenoic were not detected in most of the diets. Generally, the fish performed well on all diets even when they contained high levels of both the low-fibre and the high-fibre sunflower cakes. Absolute weights, weight gains, feed intakes and PPV values were higher for fish fed diets based on the low-fibre cake than those fed diets containing the high-fibre cake. Feed intake decreased with increasing levels of sunflower cake in the diet, but the decline was higher for fish fed diets based on the high-fibre cake. The reduced feed intake was reflected in the growth rates of fish, which also decreased with increasing levels of sunflower. The low-fibre cake could comprise up to 60% of the dietary protein without compromising the performance of the fish. At higher dietary inclusion levels, feed intake was depressed, leading to a lower growth rate than that of the control fish. The high-fibre sunflower cake could supply up to 30% of the dietary protein with the same performance as that obtained with fish fed the control diet despite a reduction in feed intake. The inclusion of higher levels of the cake in the diet led to a drastic reduction in feed intake which in turn resulted in lower weight gains of the fish fed these diets, compared to those fed the control diet. The level of sunflower cake in the diet had a significant effect on most of the parameters assessed except for whole body proximate composition. Growth rates and feed intakes of the fish decreased with increasing levels of the sunflower cakes in the diets (Table 5.6). FCR values were better for fish fed diets where sunflower cake 142 contributed 30% of the protein than for those fed diets where it contributed 80% of the protein. PER and PPV values were significantly lower for fish fed diets containing a high level of the sunflower cake (60% and 80%), than those with a low level of sunflower cake (30%) Whole body fatty acid percentages were significantly influenced by diet. Palmitic, oleic and linoleic acids were the most abundant fatty acids both in the diets and the fish. Eicosapentaenoic and docosahexaenoic acids were low in the fish reflecting the low level in the diets and perhaps other factors. 143 5.5 References Ackman, R.G., 1967. Characteristics of the fatty acid composition and biochemistry of some fresh-water fish oils and lipids in comparison to marine oils and lipids. Comp. Biochem. Physiol., 22: 907-922. A O A C , 1984. Association of Official Analytical Chemists. Official methods of analysis. Animal Feed Section Balarin J.D., 1979. Tilapia; A guide to their biology and culture in Africa JP . Hatton (Ed). University of Stirling. 174pp. Bates, D., Cartlidge, N . , French, J.M., Jackson, M.J. , Nightingale, S., Shaw, D A , Smith, S., Woo, E., Hawkins, S.A., Millar, J.H.D., Berlin, J., Conroy, D M . , Gill, S.K., Sidey, M . , Smith, A.D. , Thompson, R.H.S., Zilka, K., Gale, M . , and Sinclair, H . M . , 1989. A double-blinded controlled trial of long chain n-3 polyunsaturated fatty acids in the treatment of multiple sclerosis. J. Neurol. Neurosurg. Psychiatr., 52: 18-22. Bau, H . M . , Mohtadi-Nia, D.J., Mejean, L. , and Debry, G., 1983. Preparation of coloress sunflower protein products: Effect of processing on physiological and nutritional properties. Journal of American Oil Chemists Society, 60 (6) 1141-1146. Braekhan, O R . , Lamberstein, G., and Andresen, J., 1971. Influence of dietary fat on the fatty acid patterns of muscle and liver lipids in rainbow trout. SKR. Fiskeridir, 5 (8) 1-12. Chou, B.S., and Shiau, Shi Yen, 1999. Both n-6 and n-3 fatty acids are required for Maximal Growth of juvenile Hybrid Tilapia. North American Journal of Aquaculture, 61: 13-20. Chou, B.S., and Shiau, Shi-Yen, 1996. Optimal dietary lipid level for growth of juvenile hybrid (Oreochromis niloticus x Oreochromis aureus). Aquaculture, 143: 185-195. De-Silva, S.S., Rosanthi M . , Gunasekera and Shim, K.F. , 1991. Interaction of varying dietary protein and lipid levels in young red tilapia. Evidence of protein sparing. Aquaculture, 95: 305-318. Degani, G., Viola, S., and Yehuda, Y. , 1997. Apparent digestibility of proteins and carbohydrates in feed ingredients for adult hybrid tilapia (O. niloticus x O. aureus). Israeli Journal of Aquaculture Bamidgeh, 49: 3 115-123. Folch, J., Lee, M . , and Sloane-Stanely G.H., 1957. A simple method for isolation and purification of total lipids from animal tissues. Journal of Biological Chemistry, 226: 497-509. Hanley, F., 1991. Effects of feeding supplementary diets containing varying levels of lipid on growth, feed conversion and body composition of Nile tilapia (Oreochromis niloticus). Aquaculture, 93: 323-334. 144 Higgs, G.A., 1986. The role of eicosanoids in inflammation. Prog. Lipid Res., 25: 555-561. Hilditch, T. P., and Williams, P.N., 1964. The chemical constitution of natural fats. 4 t h Edition. Chapman and Hill (Eds). London. Huang, Chen-Huei., Huang, Ming-Chi., and Hou, Ping-Chun, 1998. Effect of dietary fatty acid composition and lipid peroxidation in sarcoplasmic reticulum of hybrid tilapia, (Oreochromis niloticus x O. aureus). Comp. Biochem. Physiol., PartB 120: 331-336. Jackson, A.J. , Capper, B.S., and Matty, A.J., 1982. Evaluation of some plant proteins in complete diets for tilapia (O. mossambicus). Aquaculture, 27: 97 - 109. Kanazawa, A , Teshima, S. I., Sakamoto, M . and Awal, M.A. , 1980. Requirements of tilapia zilli for essential fatty acids. Bull. Jap. Soc. Sci. Fish., 46: 1353-1356. Kinsella, J.E., Shimp, J.L., Mai J., and Weihrauch, J., 1977. Fatty acid content and composition of freshwater finfish. Journal of American Oil Chemists Society. 424-429 Liener, L .E . , 1980. Toxic constituents of plant feedstuffs. Academic Press London and New York. 502. NRC, National Research Council, 1993. Nutrient Requirements of Fish. National Academy Press, Washington, D.C. 144 pp. Olsen, F.E., Henderson, R.J., and McAndrew, B.J., 1990. The conversion of linoleic acid to longer chain polyunsaturated fatty acids by Tilapia (Oreochromis nilotica) in vivo. Fish Physiology and Biochemistry, Vol. 8: (3) 261-270. Popma T.J., 1982. Digestibilities of selected feedstuffs and naturally occurring algae by tilapia. Ph.D. Dissertation, Auburn University, Alabama. Sabir, M . A., Sosulski, F.W., and Kernan, J.A., 1974. Phenolic constituents in sunflower flour. J. Agric. Food Chem., 22: 572-574. Santiago, C.B., and Lovell, R.T., 1988. Amino acid requirements for growth of Nile tilapia. Journal of Nutrition, 118: 1540-1546. Santiago, C.B., and Reyes, OS . , 1993. Effect of dietary lipid source on reproductive performance and tissue lipid levels of Nile tilapia (Oreochromis niloticus (L.)) broodstock. J. Appl. Ichthyol, 9: 33-40. SAS Institute, 1985. General Linear Models Procedure. SAS Institute, Cary, North Carolina. 145 Shearer, K .D. , 1994. Factors affecting the proximate composition of cultured fishes with emphasis on salmonids. Aquaculture, 119: 63 -88. Sosulski, F. W., and McCleary, C. W., 1972. Diffusion extraction of chlorogenic acid from sunflower kernels. Journal of Food Science, Vol. 37: 253-257. Steffens, W., 1981. Protein utlization by rainbow trout {Salmo gairdneri) and carp (Cyprinus carpio). A brief review. Aquaculture, 23: 337-345. Stephens, W., 1997. Effects of variation in essential fatty acids in fish feeds on nutritive value of freshwater fish for humans. Aquaculture, 151, 97-119. Stickney, R.R., and J.W. Andrews, 1972. Effects of dietary lipid on growth, food conversion, lipid and fatty acid composition of channel catfish. Journal of Nutrition, 102: 249-258. Stickney, R.R., 1996. The effects of substituting selected oilseed protein concentrate for fishmeal in rainbow trout (Salmo gairdneri) and Carp (Cyprinus Carpio). A brief review. Aquaculture, 23: 337 - 345. Tacon, A.G.J. , Webster, J.L., and Martinez, C.A., 1984. Use of solvent extracted sunflower seed meal in complete diets for fingerling rainbow trout (Salmo gairdneri Richardson). Aquaculture, 43: (4) 381-389. Tekeuchi, T., Watanabe, K. , Yong, W.Y., and Watanabe, T., 1991. Essential fatty acids of grass carp (Cteno-pharyngodon idella). Nippon Suisan Gakkaishi, 57: 467-473. Tekeuchi, T., Sitoh, S., and Watanabe T., 1983. Requirements of tilapia nilotica for essential fatty acids. Bull. Jap. Soc. Sci. Fish., 49: 1127 - 1134. Thais, F., and Stahl, R.A.K. , 1987. Effect of dietary fish oil on renal function in immune mediated glomerular injury. In: W.E.M. Lands (Ed.) Proceedings of A O A C short course on polyunsaturated fatty acids and eicosanoids. American Oil Chemists Society., Champaign, Illinois, pp 123-126. Toyomizu, M . , Kawasaki, K. , and Tomiyasu, Y. , 1963. Effect of dietary oil on fatty acids composition of rainbow trout oil. Bull. Jap. Soc. Sci. Fish., 29: 957 - 961. Viola, S., Arieli, Y . , and Zohar, G., 1988. Animal protein free feeds for hybrid tilapia (O. niloticus xO. aureus) in intensive culture. Aquaculture, 75: 115 - 125. Viola, S., Mokady, S., Behar, D., and Cogan,U, 1988. Effects of polyunsaturated fatty acids in feeds of tilapia and carp. Body composition and fatty acid profiles at different environmental temperatures. Aquaculture, 75: 127-137. 146 Waldern, D.E., 1971. A rapid micro digestion procedure for neutral and acid detergent fibre. Canadian Journal of Animal Science, 51: 67-69 . Yamada, M . and Hayashi, K. , 1975. Fatty acids composition of lipids from 22 species of fish and molluscs. Bull. Jap. Soc. Sci. Fish., 41: 1143-1152. Yu, T C , and Sinnhuber, R.O., 1972. Effect of linolenic acid and docosahexaenoic acid on growth and fatty acid composition of rainbow trout. Lipids, 7: 450-454. Yu, T C , Sinnhuber, R.O., and Putnam, G.B., 1977. Effect of dietary lipid on fatty acid composition of body lipid in rainbow trout (Salmo gairdneri). Lipids, 12: 495- 499. 147 Chapter 6 Experiment 4: Evaluation of the most limiting amino acids in diets based on sunflower cake fed to tilapia (Oreochromis niloticus) 6.0 Abstract The objective of this study was to determine the effects of supplementing diets based on sunflower cake with lysine, threonine and methionine on the performance of tilapia (O. niloticus). A basal diet in which a fibre-reduced sunflower cake provided 80% of the dietary protein was formulated. The levels of lysine, threonine and methionine in the basal diet expressed as a percent of the diet (DM basis) were 1.17%, 1.05% and 0.75%» respectively, while the stipulated requiments (NRC, 1993) are 1.54%, 1.2% and 0.8% respectively. The amino acids were added to the basal diet singly or in various combinations. A positive control diet based on herring fishmeal and soybean meal was also formulated. All diets were isonitrogenous and isocaloric and were fed to triplicate groups of fish with an initial weight of 24 g + 0.59 (+ Sd). Further, all groups were held at 27 °C for a period of 39 days. There was a trend to improved growth rate in fish fed diets supplemented with lysine and threonine, but the improvement was not significant. There was no response to methionine added alone or together with threonine, but fish fed diets in which methionine and lysine were added together had a 12% increase in growth rate over those fed the basal diet, while those fed diets in which threonine was added together with lysine had a 10%> increase. There was a trend to improved FCR with the addition of lysine, threonine, lysine and methionine, lysine and threonine, and lysine, methionine and threonine. Growth rates and FCR's for the fish supplemented with the above amino acids were not significantly different from those of fish fed the basal diet, 148 but the values observed were also not significantly different from those obtained with the positive control fish. 6.1 Introduction and objectives. Feed accounts for about 50 % of the total cost of production in an intensive fish farm, with protein being the most expensive dietary component (El-Sayed, 1999). Fishmeal has traditionally been the most widely used protein source for many cultured fish species (Tacon, 1993). Due to its high cost, many attempts have been made to reduce its proportion in fish diets (Fagbenro and Jauncey, 1994; Mansour, 1998; El-Sayed, 1998). Many protein sources have been tested as complete or partial replacements for fishmeal. Jackson et al. (1982) demonstrated that certain plant protein sources could be used to meet much of the protein requirements of tilapia (O. mossambicus). In their study, one of the limiting factors when high dietary inclusion levels of plant proteins were tested related to a deficiency of certain essential amino acids, particularly lysine and methionine. Information on limiting amino acids in plant protein sources for tilapia species is inadequate, making it difficult to use free amino acids to supplement comparatively poor protein sources. Sunflower seeds are important sources of edible oils and protein for inclusion in animal feeds. The meal resulting from the oil extraction process is a valuable source of protein, and has been tested in many animal diets. In studies with pigs and poultry, Green and Kiener (1989) found lysine to be the most limiting amino acid. In their study, the highest growth response was obtained when lysine and methionine were supplemented in diets based on grain and sunflower cake. In ducks, Attia et al. (1998) found lysine to be the most limiting amino acid in diets containing sunflower cake, while in broilers, 149 Mohme et al. ( 1 9 9 7 ) established that partly-dehulled and solvent-extracted sunflower meal supplemented with lysine and methionine could be included up to 3 0 % in broiler diets without compromising their performance. In pigs, Jorgensen and Sauer ( 1 9 8 2 ) reported that the apparent availability of lysine in sunflower cake was 7 1 % , while the digestibilities of methionine and threonine in the same cake were 8 1 % and 69%o, respectively. In fish, studies on supplementation of diets based on sunflower meal with crystalline amino acids have yielded conflicting results. In the European eel, (Anguila anguila), Hinguera et al. ( 1 9 9 9 ) noted that the inclusion of sunflower meal as the only source of dietary protein resulted in poor growth, and that growth could be improved when sunflower meal was mixed with fishmeal or supplemented with essential amino acids. Contrary to these findings, Sanz et al. ( 1 9 9 4 ) found no improvement in dietary protein utilization of rainbow trout when diets that contained 3 9 % sunflower meal were supplemented with lysine, leucine, and methionine. Similarly, Tacon et al. ( 1 9 8 4 ) , did not observe any beneficial effects on growth and feed utilization when rainbow trout were fed diets containing 3 6 % sunflower meal, supplemented with 0 . 2 % methionine. The goals of this experiment were to: 1. determine the most limiting amino acids in diets based on sunflower cake fed to O. niloticus. 2. evaluate the effect of these diets with the foregoing limiting amino acids on fish performance. 150 6.2 Materials and Methods 6.2.1 Experimental diets and design The sunflower used in this study was dehulled using a manual "Cecolo" dehuller (Ibraki, Osaka, 567, Japan) as described in Experiment 1, and the oil was extracted using a laboratory Komet screw press (model 80B/2Q FDR, Germany). The compositions of the basal, control, and the test diets are shown in Table 6.1. Test diets were supplemented with the synthetic amino acids, L-lysine H C L , D L -methionine, and L-threonine singly or in various combinations. Herring meal and soybean meal were the main protein sources in the control diet. The calculated level of methionine in the control diet was considered to be low, and it was added to bring it up to 0.8% of the dry diet. It was not possible to determine the amino acid contents of the diets before commencement of the trial. These were calculated from the values of Scott et al. (1982). L-lysine H C L , and DL-methionine, and L-threonine were added to the diets at levels which were estimated to be approximately 10% higher than the respective requirements of tilapia as stated by N R C (1993). The diets were as follows: Diet 1. Basal diet with no amino acid supplementation Diet 2. Basal diet + lysine (1.7% total in diet D M basis) Diet 3. Basal diet + methionine (0.9% total in diet D M basis) Diet 4. Basal diet + threonine (1.25% total in diet D M basis) Diet 5. Basal diet + lysine and methionine (1.7% and 0.9% of diet D M basis) respectively. Diet 6. Basal diet + lysine and threonine (1.7% and 1.25%) of diet Dm basis) respectively. 151 Table 6.1: Compositions and chemical analyses of diets used in Experiment 4 (air-dry basis). Diets Amino acids 1 Basal 2 ] L 3 M 4 Th 5 L+M 6 L+T 7 M+Th 8 L,M,T 9 Control Sunflower cake 61.0 61.0 61.0 61.0 61.0 61.0 61.0 61.0 Soybean cake - - - - - - - - 30.0 Herring meal 7.4 7.4 7.4 7.4 7.4 7.4 7.4 7.4 24.0 Corn starch 7.4 6.2 6.7 7.0 5.5 5.9 6.29 5.2 29.8 Wheat 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 Corn oil 8.6 9.0 8.9 8.7 9.2 9.1 9.0 9.3 0.5 Iodized salt 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 Dicalcium phospate 2.8 2.8 2.8 2.8 2.8 2.8 2.8 2.8 2.1 2Vit/mineral premix 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.3 Ascorbic acid 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.2 Choline chloride 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 L-lysine HCL - 0.8 - - 0.8 0.8 - 0.8 -DL-methionine - - 0.5 - 0.5 - 0.5 0.5 0.3 Threonine - - - 0.2 - 0.2 0.2 0.2 -Chemical analyses (DM basis3 Dry matter 92.7 92.09 92.05 92.04 92.1 92.1 92.07 92.14 90.75 4 DE (kcal/kg D M 2951 2951 2958 2949 2949 2951 2954 2949 3182 Protein, % 33.0 33.1 33.4 33.4 33.1 32.9 33.3 33.1 33.9 Crude fat, 19.6 20.2 20.0 19.9 20.4 20.3 20.1 20.5 4.7 Crude fibre, % 9.0 9.0 9.0 9.0 9.0 9.0 9.0 9.0 2.1 Calcium, % 1.2 1.3 1.2 1.3 1.3 1.3 1.3 1.3 1.3 Phosphorus, % 1.6 1.6 1.7 1.6 1.7 1.7 1.6 1.6 1.5 1 L = lysine, M = methionine, T = threonine. (For diet description, see text). 2 Composition of vitamin/mineral premix is as described in Table 5.1 (composition of diets used in Experiment 3). 3 All values were determined by analysis, except D E concentrations which were calculated as described in Chapter 4, Experiment 2. 152 Diet 7. Basal diet + methionine and threonine (0.9% and 1.25% of diet Dm basis) respectively. Diet 8. Basal diet + lysine, methionine and threonine (1.7%, 0.9%, and 1.25% of diet Dm basis) respectively. Diet 9. Positive control diet. 6.2.2 Fish sampling Sex-reversed tilapia (O. niloticus) male fingerlings weighing 23.9 + 0.60g were used. They were bought from the Sagana Government Fisheries Farm, in Sagana, Kenya, and transported to the University of Nairobi. They were acclimated to laboratory conditions for a period of two weeks before the onset of the trial. Thereafter, they were weighed in groups of 16 fish that were selected at random, allocated to the experimental tanks, and managed as in Experiment 3. Fish were weighed at the beginning of the trial and 39 days later. They were starved for a period of 24 hours before each weighing. Water temperatures in each tank were maintained between 25 °C and 28 °C, and dissolved Dissolved oxygen concentration in the tanks was maintained above 5.5 mg/L during the experimental period. 6.2.3 Data collection and analytical procedures Fish growth and performance were assessed by absolute weights, weight gain, specific growth rate, feed intake, and feed conversion ratio. Specific growth rates, feed intakes and feed conversion ratios were determined as described in Experiment 3. 6.2.4 Chemical analyses The ingredients and diets were analyzed in duplicate for percentages of dry matter, ash, protein, lipid, calcium and phosphorus according to standard procedures (AOAC, 1984). 153 Dietary amino acid compositions were determined by H P L C after performic acid oxidation (see Chapter 4) 6.2.5 Statistical analyses Data on absolute weight, weight gain, specific growth rates, feed intake and feed utilization were subjected to statistical analysis using PROC G L M of the SAS statistical analysis systems (SAS, 1985). The model for a 9 x 1 CRD was used. An analysis of covariance was done using the initial weights as the covariate. The covariate was not significant in any of the parameters. Treatment means were compared using Tukey's multiple range test, with the level of significance set at P < 0.05. 154 6.3 Results and Discussion 6.3.1 Chemical composition of the diets The low-fibre sunflower cake used in this experiment had a dry matter content of 93%, while crude protein, crude fibre, fat and ash contents were 42%>, 14%>, 15%>. and 7.5% ( D M basis) respectively. All diets were formulated to contain 30% protein (as fed), and in the test diets, 80% of the protein originated from dehulled sunflower cake, and 20% from LT herring meal. The chemical compositions of the diets are presented in Table 6.1. The determined protein levels in the diets were similar in each case. Due to the low digestible energy content of sunflower cake, corn oil was used to increase the energy level in the diets. This resulted in the crude fat levels in the diets based on sunflower cake being higher than in the control diet. The crude protein content of the diets was approximately 30%o (AD basis). Estimates of the protein requirements of young tilapia range from 25% to 35%> (AD basis) depending mainly on the protein source and fish size (Santiago et al, 1982; Siddiqui et al, 1988). In the current experiment, the estimated energy to protein ratio in the diets based on sunflower cake was 9.0 kcal of digestible energy per gram of protein (113 mg protein/kcal), while in the control diet it was 9.4 kcal of digestible energy per gram (107 mg per kcal). Kubaryk (1980) reported that an energy to protein ratio of 8 - 10 kcal digestible energy per gram of protein (100 - 125 mg protein/kcal) was adequate for young tilapia fingerlings, while El-Sayed and Teshima (1992) reported an optimum protein to energy ratio of 103 mg of protein per kcal DE. The calcium content of the diets was approximately 1.3%, while the total phosphorus content was approximately 1.6% for the diets based on sunflower cake and 1.5% for the control diet 155 (AD basis). In tilapia, the calcium requirement is largely met by absorption of calcium through the gills and skin, while phosphorus has to be supplied in the diet. Tilapia has been reported to require 0.6 to 0.7% phosphorus in the diet (AD basis) (Viola and Arieli, 1988), which can be met using animal by-products, dicalcium phosphate or a combination of both. The quality of protein is influenced by its amino acid profile. The amino acid profiles of the diets expressed as a percentage of the diet, and as a percentage of the dietary protein, are presented in Tables 6.2 and 6.3 respectively. The basal diet met or exceeded the requirements for most of the essential amino acids except lysine, methionine and threonine. The stipulated requirements for these amino acids are 1.54%, 0.8%, and 1.12% of the dry diet respectively (NRC, 1993) while their levels in the basal diet were 1.17, 0.75 and 1.05% (DM basis) respectively. Diet 2 was deficient in methionine and threonine, while diets 3, 4, 5, 6 and 7 were deficient in lysine and threonine, lysine and methionine, threonine, methionine, and lysine, respectively. Due to budgetary constraints, it was not possible to determine tryptophan and tyrosine levels in all of the diets. This was only done for the basal and control diets. Determined amino acid levels were not appreciably different among the different diets. 6.3.2 F i sh performance Data on fish performance are presented in Table 6.4. The specific growth rate for fish fed the positive control diet was higher than the values found for fish fed the other diets, but the differences were only significant (P< 0.05) relative to those fed the basal diet, the basal + methionine, and the basal diet + methionine and threonine. Fish fed the positive control diet had a growth rate of 2.28% per day, followed by those fed the diet 156 cd o o c S •a -o CO c« cfl H-» CO '-3 CO <+H o c« c o o I o O -a 'a a o c cd -a CO S3 l-H CO -t-» CO Q CN o 3 cd H 5^ 1—1 g 00 3 • 5 3 o O H i -H r-vq Tt O • • i i i i i o ,—i 00 oo o o ,—i vi I o\ l-H C-; vn vq co vi Ov H^ <—i CN CN o 1—1 i—i o OV d Tt vi VI Ov CO TT CS 00 CS co TT CO n 00 CO vq Tf; CO TT TT vq r~ r- o\ ^ Q Q 5 Q Q f 00 OV H> H cd H O co -j-j <5 Ov -3 "3 - 1 1 3 er "J cj •3 b M-H O M-H H-j o o N? S ? ov o x VI V i ""I d o CJ CJ _ g 'T-H, v-i c5> r 0 CD U i=l > on 3 •5 cj 5 2 5 i n 00 O N ro o p 1 1 1 1 1 1 o p 00 N O co 1 1 1 • 1 1 1 CS 1 i n i n oo ro i n ro i n T f i n m 00 O N CN ro i n ro ro i n i n T f i n in T T CS O N CN o o i n i n ro ro ro T f oo ON ro i n i n N O i n i n i n i n i n T f r o N O ro i n ro CO 00 i n N O O N CN cs T f i n T f r - CN O N r—1 1—1 ro ro ro ro ro ro ro ro CN ro N O f ^ 00 o CN CN T f o o O N ro o CN O O o r-o m ro T T T f T f ro T f T f T f T f CO i n i n o o T f ro o CN O N ro T f T f i n o i n o 00 i n i n m' T f T f T f T f i n T f CN r -cs N O i n N O T f i n N O 00 N O r-N D ro r-o O CN N O N O •S N O N O N O N O N O N O T f r-CN CN T T ro oo 00 CN oo ro CN T f O N T f i n CN f~ O 00 NO CN CN ro CN ro CN ro T f CO CN o o i n p CN CN T f p i n p N O o o 00 o o CN CN CN CN CN CN CN CN r s • OO T f T f o T f O o 00 o ro in ro ro ro T f CO T f T f T f T f CO ro r -i n p CN CN r-r -N O o o CN o CN N O O N T f m' i n T f i n i n i n T f 1 o i n O N O o 00 o O N o N O i n o T f r s CS ! 1 CN CN CN CN CN CN CN cs O N O N N O O ro 00 N O N O o 1> N O N O T f 00 cs cs ro ro T f ro ro ro ro CO T f i n i n T f CN 00 i n i n O N oo N O 00 oo 00 N O 1> N O O N CN O N o o O N i n O N l> N O O N i n N O r-. T f O N 00 t< O N 00 00 00 00 00 O N • + f + + 1 >-H + p § + p 1 + + ! o CD 6 — _ C N C O T f i n N O oo O N CD CU ^ "cu "fD "cD "cD "cu Q Q Q Q Q Q Q Q P H C O O N O N g '55 V H C H OJ <4-H o N ? o V H C H OJ •3 T f in © OJ c '53 C H OJ N T O N OJ a • i-H O oo in co O t§ u C O ti OJ E-o o -o cd - C CO H 3 C+H O CD O a cd a •g CD OH T f V O CD cd CD CD O - D cd M t-H PH o PH C '3 f CD <+H CD T 3 CD cd PH .S l -H CD O H 00 '3 M-H cd co +3 « •a ^0 CD P X I T f Os •3 0 X I VO ON -s IT) O •9 r- •8 T f X I r -ON •8 00 0 C3 00 CN vo 0 '-' CN CN CN CN *-* CN CN 0 ' o o CN V O O N 00 ON VO CN 00 00 CO 00 ir> 00 o os 00 co m CN m CN VT) OS CN IT) O ^ CO IT) CO m CO i n 00 CN ITS CO m T f CN X I x> Os X I CS T f •i T f •s ON •9 CO •8 IT) ccj r - H m CN vd CN O ' CO vd CN Os CN CO O CO 00' CN O ' CO T f CO X I T f "8 VO X I CO •8 CO "In •8 VO •8 00 •8 00 C3 ON i n CN O ' i n T f IT) 0 ' i n CO IT) iri m T f vO CN i n T f i n r - ' m t - l O T f O N vo i n CO 00 co CN T f CN CO CN CO CN CO CN T f CN T f CN T f CN CO CN CD '-3 CO cd m CD C •a 0 CD .2 -4-> CD ' c O X > a + + CN CO CD CD CD C - a o CD J3 J3 +-> CD a T3 C cd CD .£ 'co >> + + CD CH •a o CD +-» -o c cd CD _ C ' « >s + CD c •a - . o o CD CD M M fi cd CD C •a o t3 a + T 3 +-> CD 1/1 CD '-a o o in vo P P P CD CD CD P P P + 00 ON H + H + H CD CD CD 5 S p i n o 0 ' A PH C CD l-H > N "a c d o M 3 • i-H C 60 O c CD l _ cd l-H CD +-> + H CD OO CO CO c cd CD 13 > '+-> o CD O H CO CD l -H CD G -a o CD J3 +-> T3 C cd CD C - a o +-» CD CD _o 'co O +•> <+H CD l -H O CD V H - C +-» T 3 C cd x f + H CD a CO t-1 Os supplemented with both lysine and methionine (diet 5) which had a growth rate of 2.17% per day, while those fed the diet supplemented with lysine and threonine (diet 6) had a growth rate of 2.14% per day. Fish fed the diet supplemented with lysine, methionine and threonine (diet 8) had a growth rate of 2.08%>. There were no significant differences (P > 0.05) in growth rates between fish fed the basal diet and those fed diets supplemented with amino acids. However, the addition of lysine and methionine (diet 5) to the basal diet improved growth rate by approximately 12%, while the addition of lysine and threonine (diet 6), and lysine, methionine and threonine (diet 8) improved growth rate by 10%) and 7% respectively. Lysine, added alone improved the growth rate of tilapia by approximately 6% over that noted for fish fed the basal diet. Although the improvement in growth rate for fish fed the above diets was not significant relative to those ingesting the basal diet, the growth rates attained were also not significantly different from that of fish fed the the positive control diet (P > 0.05). Fish fed the basal diet alone or supplemented with methionine had the lowest growth rates of 1.94%, and 1.96% per day, and as indicated above, these values were lower than those observed for the fish fed the control diet (P < 0. 05). There were no significant differences in feed intake between fish fed the various diets, but there was a trend to improved feed conversion ratios for the fish fed the diets supplemented with amino acids compared to those fed the basal diet. FCR values of fish fed diets 2, 4, 5, 6, 7, and 8 were not significantly different from that of fish fed the control diet. Lysine has been identified as the first limiting amino acid in sunflower cake (Senkoylu and Dale, 1999). Besides its effect on growth, it has been shown to chemically 160 enhance feed intake together with glutamic, aspartic, citric and malic acids (Adams et al., 1988). In the current experiment, there was no evidence of higher feed intake in fish fed diets supplemented with lysine, but there was a trend to improved FCR in fish fed diets where lysine was added alone (diet 2) or together with methionine (diet 5), threonine (diet 6) or both methionine and threonine (diet 8). The values for FCR of fish fed these diets were not significantly different from that of the control fish (P > 0.05). Data on the response of tilapia to dietary lysine supplementation are inconsistent. In studies by Sintayehu et al. (1996) with sunflower cake, the addition of lysine and methionine did not enhance fish performance. The initial weight of the fish used in the above study was 93 g, which may explain the lack of response. Furthermore, the lysine level in the basal diet was 2.3% (air-dry basis), which was higher than the stipulated requirement (NRC, 1993; Jackson and Capper, 1982). The lysine to protein ratios in the diets used in the current experiment are shown in Table 6.3. In the basal diet and in diets 3 (basal + methionine) and 4 (basal + threonine), the ratio was below 3.9%, while in diet 7 (basal + methionine and threonine), the ratio was 4.53%. The ratio was above 5.5%> in all the other diets. Viola et al. (1994) found that in hybrid tilapia weighing 80, 190 and 203 g respectively, supplementing diets with lysine where the lysine to protein ratio was above 4.9%, did not enhance fish performance, whereas lysine supplementation of diets where the ratio was below 4.4%o improved growth rate. In the present experiment, the lysine to protein ratios in the basal diet and diets which were not supplemented with lysine were below 4.4%. There was a trend to increased growth rate and weight gain in fish fed diets supplemented with lysine, to a level where they were not significantly different from those of fish fed the control 161 diet (P > 0.05). The lack of significant differences in fish performance over fish fed the basal diet may have been due to the small sample sizes (n = 3) used to test for statistical significance. Supplemention of the basal diet with methionine alone did not have any effect on fish performance. Methionine levels (% of protein) in the control diet and the other diets supplemented with this amino acid were 3.07%, 3.83%, 3.38%, 3.49% and 4.25% in the control diet, diet 3 (basal + methionine), diet 5 (basal +lysine and methionine), diet 7 (basal + methionine and lysine) and diet 8 (basal + lysine, methionine and threonine), respectively. The lack of response was presumably because the methionine level in the basal diet was adequate for fish growth. It is also plausible that diet 3 (basal + methionine) and diet 8 (basal + lysine, methionine and threonine) contained excessive levels of this amino acid. As explained earlier in the text, it was not possible to determine amino acid composition of the diets before the onset of the experiment. One of the signs of amino acid imbalances in animals is depression in feed intake (Harper et al., 1970; D'Mello, 1994). There were no significant differences in feed intake between fish fed diet 3 and diet 8 and those fed the other diets, indicating that the levels of methionine in these diets were not sufficiently excessive to create an amino acid imbalance. Different responses have been reported in tilapia fed diets supplemented with methionine. In studies by El-Dahhar and El-Shazly (1993) with tilapia (O.niloticus) (initial weight 16 g), the addition of methionine and lysine to diets where soybean meal and cotton seed cake supplied 100%» of the dietary protein improved growth and FCR slightly, but the values obtained were still lower than those attained with the fish fed the fishmeal control diet. Shiau et al. (1987) observed a significant increase in weight gain 162 of tilapia when they were fed diets in which fishmeal and soybean meal supplied 70% and 30% of the protein, supplemented with methionine. The methionine level in the unsupplemented diets in the above study was 0.84% of diet (DM basis), and the average initial weight of the fish was 1.24 g. Teshima and Kanazawa (1988) did not observe any improvement in growth rate in tilapia fry weighing about 0.4g, when diets in which soybean protein was the sole protein source, were supplemented with crystalline methionine. However, when methionine was added as methionine-enriched soybean plastein, the growth rate of tilapia was improved significantly to a level comparable to that of fish fed the fishmeal control diet. The authors postulated that methionine in soybean plastein is more effectively utilized by tilapia fry than crystalline methionine. The growth responses of tilapia to methionine addition to diets varied in the studies quoted above. The variation may have been caused by the different sizes of fish used in the studies. It is also worth noting that in all of the studies quoted above, soybean meal provided part or all of the dietary protein. Sunflower cake contains a higher level of methionine than soybean meal (Jackson et al, 1982; NRC, 1993; Scott et al, (1982). Santiago and Lovell (1988) reported that the dietary methionine requirement of juvenile tilapia was 0.8% in diets containing 0.15% cystine, while Jackson and Capper (1982) observed that the minimum methionine requirement for tilapia (O. mossamhicus) (1.7 g.) was below 0.53% of the diet (DM basis), when the cystine level was 0.74%. The diets used by Jackson and Capper (1982) had a high level of cystine, and the low methionine requirement observed may indicate a high level of methionine sparing. In tilapia (O. mossamhicus) weighing approximately 12.5g., Jackson et al (1982) reported that the minimum requirement for methionine was 0.7%> of diet (DM basis), in the presense of 163 0.45% cystine ( D M basis). The methionine and cystine levels in the basal diet used in the current experiment were 0.75% and 0.66%> of the diet (DM basis). The methionine level was lower than that recommended by Santiago and Lovell. (1988), but it was higher than the value reported by Jackson and Capper, (1982) and Jackson et al. (1982). The lack of response was presumably because the methionine level in the basal diet was adequate, especially in the in the presense of the high levels of cystine in the diets based on sunflower cake (> 0.45%). Methionine added with lysine elicited a positive response with respect to the growth rate and FCR of fish in the current study. The values observed for growth rate and FCR were not significantly different from those obtained with fish fed the basal diet, but nevertheless they were higher than those observed when methionine or lysine were added alone. It is also possible that methionine may have been marginally deficient, but it was not the first limiting amino acid. The threonine level in the basal diet was 1.05 % of the diet, while the stated requirement for juvenile tilapia (O. niloticus) (Santiago and Lovell, 1988) is 1.12% (DM). The basal diet may have been marginally deficient in this amino acid. Adding threonine to the basal diet improved growth rate by approximately 6%, while adding it together with methionine and lysine (diet 8), improved the growth rate by 7 %. The growth rates were not significantly different from that of fish fed the basal diet (P > 0.05), but they were also not significantly less than that of fish fed the control diet. In chicken broilers fed diets containing sunflower meal as the only protein source, threonine was identified as the second most limiting amino acid in high protein diets (37-43%) CP) (DM), while lysine was the first limiting amino acid. Information on the dietary 164 threonine requirement of tilapia species has only been reported in one study (Santiago and Lovell, 1988), for juvenile fish weighing less than 1 g. In the current study, it is plausible that lack of response to dietary threonine supplementation may have been due to the fact that lysine may also have been inadequate. 165 6.4 Conclusions There was no response in growth rate or FCR in fish fed diets supplemented with methionine alone, but there was some improvement when it was added together with lysine, to a level that was not significantly different from that of fish fed the control diet. N R C (1993) recommends a methionine level of 0.8% in diets containing 0.16% cystine (DM basis), while Jackson and Capper (1982) determined the optimum methionine level in the diet as 0.53% in diets with a cystine level of 0.73% of diet (DM basis). The methionine level in the unsupplemented diets in the current study,was 0.75%, while the cystine level was 0.66% of the diet. The level of sulfur amino acids in the diets was higher than that recommended by Jackson and Capper (1982). In addition, the high level of cystine in the diets may have spared methionine for growth, leading to a lower requirement than stipulated by N R C (1993). It is therefore likely that methionine was not the first limiting amino acid in the diets fed to tilapia in the current study. Threonine added alone or together with lysine to the basal diet improved growth rate and FCR of tilapia, but the values attained were not significantly different from those obtained for fish fed the basal diet. There was very little response in weight gain or FCR in fish fed diets supplemented with both threonine and methionine, suggesting that they were not the first limiting amino acids. From the observations presented above, it it is probable that lysine was the most limiting amino acid in the basal diet, and possibly threonine may have been marginally deficient in the diet as well. 166 6.5 References Adams, M.A . , Johnsen, P.B., and Zhou, Hongi - Qi, 1988. Chemical enhancement of feeding for a herbivorous fish, tilapia zilli. Aquaculture, 72: (1-2) 95-107. A O A C , 1984. Association of Official Analytical Chemists. Official methods of analyses. Animal Feed Section. Attia, Y . A . , El-Deek, A . A , and Osman, M . , 1998. Evaluation of sunflower meal a a feedstuff in diets for ducks. Archiv. Fur Geflugelkunde, 62: (6) 273 - 282. D'Mello, J.P.F., 1994. Amino acid imbalances, antagonisms, and toxicities. In: Amino acids in Farm Animal Nutrition. LP.D'Mello (Ed.) 418 pp. El-Dahhar, A. A., and El-Shazly, K. , 1993. Effect of essential amino acids (methionine and lysine) and treated oil in fish diet on growth performance and feed utilization of Nile tilapia, Tilapia nilotica (L). Aquaculture and Fisheries Management, 24: 731-739. El Sayed, A . F . M . , 1999. Alternative dietary protein sources for farmed tilapia, Oreochromis spp. Aquaculture, 179: 149-168. El-Sayed, A .F .M. , 1998. Total replacement of fishmeal with animal protein sources in Nile tilapia (Oreochromis niloticus (L.) feeds. Aquaculture Research, 29 (4) 275-280. El-Sayed, A . F . M . , and Teshima, S., 1992. Protein and energy requirements of Nile tilapia (Oreochromis niloticus) fry•. Aquaculture, 103: 55-63. Fagbenro, O.A., and Jauncey, K. , 1994. Chemical and nutritional quality of dried fermented silage and their nutritive value for tilapia (Oreochromis niloticus). Animal Feed Sci. Technol., 45: (2) 167-176. Green, S., and Kiener, T., 1989. Digestibilites of nitrogen and amino acids in soya bean, sunflower, meat and rapeseed meals measured with pigs and poultry. Animal Production, 48: (1) 157- 179. Harper, A.E. , Benevenga, N.J., and Wohlhueter, R .M. , 1970. Effects of ingestion of disproportionate amounts of amino acids. Physiological Reviews, 50: 428-558. Hinguera, D.L. , Akharbach, M . , Hidalgo, H , Peragon, M C , Lupanez, J., and Garcia Gallago M . , 1999. Liver and white muscle turn-over rates in the European eel (Anguila Anguila) effects of dietary protein quality. Aquaculture, 179: (1 - 4) 203 - 216. Jackson, A.J., and Capper, B.S., 1982. Investigations into the requirement of the tilapia (Oreochromis mossambicus) for dietary methionine, lysine and arginine in semi synthetic diets. Aquaculture, 29: 289 - 297. 167 Jackson, A.J. , Capper, B.S., and Matty, A.J. , 1982. Evaluation of some plant proteins in complete diets for tilapia (O. mossamhicus). Aquaculture, 27: 97 - 109. Jorgensen, H. , and Sauer, C.N., 1982. Amino acid availabilities in soybean meal, sunflower meal, fishmeal and meat and bone meal fed to growing pigs. Journal of Animal Science, 58: 926-934. Kubaryk, J.M., 1980. Effects of diet, feeding schedule, and sex on food consumption, growth and retention of protein and energy by tilapia. Ph.D. dissertation, Auburn University, Auburn, A L . Mansour, C.R., 1998. Nutrient requirements of red tilapia fingerlings. MSc. Thesis, Faculty of Science, University of Alexandria, Egypt 121 pp. Mohme, H. , Toska, M . , and Gunther, K.D. , 1997. Sunflower seed meal as a protein source in nutrition of broilers. Fett-Lipid, 99: (3) 78-80. National Research Council, 1993. Nutrient Requirements of Fish. National Academy Press, Washington, D . C , 144pp. Santiago, C.B., Aldaba, M.B . , and Laron, M.A. , 1982. Dietary crude protein requirements of tilapia nilotica fry. Kalikasan, Phil. J. Biol., 11: 61 - 71. Santiago, C.B., and Lovell, R.T., 1988. Amino acid requirements for growth of Nile tilapia. Journal of Nutrition, 111: 46-52 Sanz, A., Morales, A.E. , De La Higuera M . , and Carenete, G., 1994. Sunflower meal compared with soybean meal as a partial substitute for fishmeal in rainbow trout (Orchorhynchus mykiss) diets. Protein and energy utilization. Aquaculture, 128: 287 -289. SAS Users Guide: Statistics, Version 5 t h Edition. 1985. SAS Inst., Inc., Caary, NC. Scott, M L . , Nesheim, M . C and Young, R.J., 1982. Nutrition of the chicken. 3 r d Edition, M L . Scott and Associates, Ithaca, New York. Senkoylu, N . , and Dale N . , 1999. Sunflower in poultry diets. A review. World's Poultry Science Journal Vol. , 55: 153 - 174. Shiau, Shi-Yen, Chuang, Jan-Lung, and Sun, Chan-Lan., 1987. Inclusion of soybean meal in tilapia (Oreochromis niloticus x O. aureus) diets at two protein levels. Aquaculture, 65: 251-261. Siddiqui, A.Q., Howlander, M.S., and Adams, A. A., 1988. Effects of dietary protein levels on growth, feed conversion and protein utilization in fry and young Nile tilapia (Oreochromis niloticus). Aquaculture, 70: 63 - 7 3 . 168 Sintayehu, A. , Mathies, E. , Mayer- Burgdorff, K . H . , Rosenow, H. , and Guenter, K .D. , 1996. Apparent digestibility and growth experiments with tilapia (Oreochromis niloticus), fed soybean meal, cotton seed meal and sunflower seed meal. Journal of Applied Ichthyology, 12: (2) 125 - 130. Tacon, A.G.J. , Webster, J.L., and Martinez, C.A., 1984. Use of solvent extracted sunflower seed meal in complete diets for fingerling rainbow trout (Salmo gairdneri Richardson). Aquaculture, 43: (4) 381-389. Tacon, A.G.J. , 1993. Feed ingredients for warm water fish. Fishmeal and other processed feedstuffs, FAO Fish. Circ , No. 856, FAO, Rome, Italy, 64pp. Teshima, S., and Kanazawa, A., 1988. Nutritive value of methionine enriched plastein for Oreochromis niloticus fry. 2 n d Intl. Symp. on Tilapia in Aquaculture, 16 -20 March 1987. Bangkok, Thailand. 393 - 399. Viola, S., Angeoni, H. , Gur, N . , and Lahav, E., 1994. Growth performance, protein and energy balances of hybrid tilapia fed two levels of lysine at three levels of protein. Israeli Journal of Aquaculture/ Bamidgeh, 46: (4) 212 - 222. Viola, S., Arieli, Y . , and Zohar, G., 1988. Animal protein free feeds for hybrid tilapia (Oreochromis niloticus x O. aureus) in intensive culture. Aquaculture, 75: 115- 125. 169 Chapter 7 7.0 General discussion, conclusions and recommendations. Sunflower cake, a by-product of the oil extraction industry is an inexpensive protein source with potential for use in fish diets. In addition to its high protein content, it also has a high fibre level, which reduces its digestible energy concentration. Fish respond to low-energy diets by increasing their feed intake in an attempt to meet their energy requirements. However, in some cases, this adjustment in feed intake is hindered by the bulkiness of the diet. This is especially true for tilapia, which have a small stomach size (Balarin, 1979). Based on this information from the literature, four experiments were designed to investigate the effect of reducing the amount of fibre in sunflower cake on nutrient digestibility and feed utilization, and to compare this fibre-reduced cake with the commercially available high-fibre cake. The extent to which protein from both the high-fibre and the fibre-reduced cake could replace fishmeal protein in tilapia diets was investigated. The effects of supplementing diets made from a fibre-reduced sunflower cake with the amino acids lysine, methionine and threonine singly or in combination on growth, feed intake and FCR were also investigated. Protein from both high-fibre and fibre-reduced sunflower cake was well digested by tilapia. ADC-P of protein was not appreciably different between the two types of cakes. A D C - E of energy improved with fibre-reduction from 30% in the high-fibre cake to 42% in the fibre-reduced sunflower cake. Despite the low fibre content of the fibre-reduced cake (11% Air-dry basis), the digestibility of energy was still poor compared to the fishmeals. In rainbow trout, Sanz et al. (1994) found poor (40%) digestibility of carbohydrates (NFE and fibre) in sunflower cake compared to soybean meal 170 carbohydrates (50%). Similarly, in carp, Bendi and Spandorf (1953) observed very low digestibility of carbohydrates in sunflower cake (26%). It was not possible to determine the digestibility of all nutrients in this study. It is plausible that the digestibility of carbohydrates in sunflower cake by tilapia (O. niloticus) may be low, which could have contributed to the observed results. In this study, the digestible energy density of the diets was increased by the addition of corn oil. Research has shown that adult tilapia (over lOOg) have a limited ability to use high amounts of fat in the diet (Degani et al., 1997). Al l the fish used in this study were below 90 g. There was no evidence of poor utilization of oil. The fibre-reduced cakes used in all of the experiments had higher levels of crude protein and all of the amino acids that were measured than the high-fibre cake. Sunflower cake is low in lysine and consequently diets based on this cake were relatively low in lysine. They were also low in threonine. When protein from both the fibre-reduced sunflower cake and the high-fibre cake replaced 50% of the fishmeal protein, the weight gain of the fish receiving the fibre-reduced cake was not significantly different from that of those fed the anchovy fishmeal diets. Fish receiving the fibre-reduced cake also had a trend towards a higher growth rate, feed intake, and a better FCR than those on the high-fibre diets. From Experiment 2, it was established that the protein from the fibre-reduced sunflower cake could effectively replace up to 50% of the fishmeal without compromising the performance of the fish. In Experiment 3, the fibre-reduced and high-fibre sunflower cakes were tested over a wide range of dietary inclusion, i.e., each supplied 30%, 60% and 80% of the 171 dietary protein. It was found that the low-fibre cake could supply up to 60% of the dietary protein without compromising fish performance. Feed intake and weight gain were reduced when the cake was incorporated into the diet at higher levels. The high-fibre sunflower cake could satisfactorily supply only 30% of the dietary protein. Higher levels of this cake caused a reduction in feed intake and weight gain. Experiment 4 was designed to investigate the effect of supplementing diets containing high levels of the low-fibre sunflower cake with the amino acids, lysine, methionine, and threonine. The levels of these amino acids were found to be low in diets containing high levels of sunflower cake in Experiment 3. Lysine levels were low in diets containing high levels (80%>) of sunflower cake. Besides its direct effect on growth, lysine has been shown to enhance feed intake in tilapia (O. niloticus) (Adams et al., 1988). There was a trend to improved weight gain and FCR with dietary lysine supplementation, so that the values obtained were not significantly different from those of the control fish. Methionine, added singly had no effect, but there was a trend to improved growth rate and FCR when it was added together with lysine, to a level comparable to that of the control fish. Methionine and cystine levels in the unsupplemented diets were 0.75%> and 0.66% ( D M basis). N R C (1993) recommends a methionine level of 0.8% in diets with a cystine level of 0.16% (DM basis), based on a study with juvenile tilapia (O. niloticus) weighing less than 1 g. It is plausible that with the larger size of fish used in the current study, and the higher level of cystine in the diets decreased the requirement tilapia for methionine. Moreover, it appears that methionine was not the first limiting dietary amino 172 acid for the fish in this study. This may account for the observed trend of improved growth rate and FCR when lysine was added together with methionine. Threonine, added alone or together with lysine improved growth rate and FCR by 6% and 10% respectively, but there was no response when it was added together with methionine. It is not clear why this situation arose. The threonine level in the basal diet was 1.07 % (DM basis), while the stipulated requirement (NRC, 1993) is 1.12% of the diet (DM basis). There is only one study on the threonine requirement of tilapia (O. niloticus) (Santiago and Lovell, 1988), which was done with tilapia fry weighing less than 1 g. It is plausible that the threonine level in the diet was adequate for the fish size used in the present study, or that threonine was not the first limiting amino acid in the diets. The fatty acid compositions of the diets were determined and compared with the whole body fatty acid compositions of the fish. Earlier studies (Kanazawa et al., 1980; Tekeuchi et al., 1983) established that the only essential fatty acid required by tilapia is linoleic acid (18:2 co 6), and that they possess the enzymes that are required to desaturate and elongate fatty acids of the co 6 and co 3 series to make the long chain highly unsaturated fatty acids viz., 20: 4 co 6, 20: 5 co 3 and 22: 6 co3 of nutritional significance. For that reason, many formulated diets for tilapia do not contain a source of the long chain poly-unsaturated fatty acids. Recently, Chou and Shau ( 1999) observed significant increases in the co 3 fatty acids in tilapia muscle and liver by feeding cod liver oil. In this study, the fatty acid composition of the fish reflected the dietary fatty acid composition. In view of the recent studies that have shown the beneficial effect of the long chain co 3 fatty acids for humans, it may be worthwhile to consider adding these fatty acids to the tilapia diets to improve the nutritional quality of the flesh. 173 One of the primary objectives of this work was to evaluate locally produced raw materials (high-fibre sunflower cake, fibre-reduced sunflower cake and omena fishmeal) as replacements for imported fishmeal in the diets of tilapia (O. niloticus). Omena fishmeal, made from a cyprinid fish Rastrineobola argentea is used widely in the animal feed industry in Kenya, but there is no documented information on its quality and feeding value. In this study, the digestibility of nutrients and energy and the feeding value of Omena fishmeal were compared to that of prime quality anchovy meal at two levels of protein intake. Apparent digestibility coefficients for protein, energy and organic matter in omena fishmeal were comparable to those of anchovy fishmeal. In Experiment 2, the determined CP level for the diets based on omena fishmeal at both protein levels were slightly lower than the calculated (expected) values, which may have been caused by an error during the mixing of the diets. Despite this, there were no significant differences in the growth rate and FCR in fish fed diets based on the two fishmeals. Generally, omena fishmeal did show some promising results. There is a need for a more detailed study of this fishmeal, especially the processing methods that are cost effective and that produce a high quality fishmeal. The fish did well on most of the ingredients tested. Reducing fibre in sunflower cake improved the growth rate and FCR of the fish to levels that were not significantly different from those of the control fish. In Kenya, labour costs as a percentage of the total cost of production are relatively low, and the feed cost is the major component of the variable costs. FCR is therefore a useful measure of feed quality. Replacing high- cost herring meal and soybean meal with low cost ingredients such as omena fishmeal, fibre-reduced sunflower cake and high-fibre sunflower cake would reduce the importation of 174 expensive dietary ingredients, and this in turn, will reduce the cost of tilapia culture Kenya 175 7.1 References Adams, M.A . , Johnsen, P.B., and Zhou, Hongi - Qi, 1988. Chemical enhancement of feeding for a herbivorous fish, tilapia zilli. Aquaculture, 72: (1-2) 95-107. Balarin, J.D., 1979. Biological characteristics of tilapia In: Tilapia, A guide to their biology and culture in Africa. J.D. Balarin and J.P Hatton (Eds.), University of Stirling, U.K. 174pp. Bendi, A. , and Spandorf, A., 1953. The activity of digestion enzymes of carp. Bamidgeh, 5: 116-130. Chou, B.S., and Shiau, Shi-Yen., 1996. Optimal dietary lipid level for the growth of juvenile hybrid (Oreochromis niloticus x Oreochromis aureus). Aquaculture, 143: 185-195. Degani, G., Viola, S., Yehuda, Y. , 1997. Apparent digestibility of carbohydrates in feed ingredients for adult tilapia (O. aureus x O. niloticus). Israel Journal of Aquaculture/Bamidgeh, 49: 115-123. Kanazawa, A., Teshima, S. I , Sakamoto, M . and Awal, M.A. , 1980. Requirements of tilapia zilli for essential fatty acids. Bull. Jap. Soc. Sci. Fish., 46: 1353-1356. N R C (National Research Council) 1993. Nutrient requirements for fish. National Academy Press. Washington, D.C., U.S.A. 114pp. Santiago, C.B., and Lovell, R.T., 1988. Amino acid requirements for growth of Nile tilapia. Journal of Nutrition, 118: 1540-1546. Sanz, A., Morales, A.E. , HigueraM., Cardenete, G., 1994. Sunflower meal compared with soybean meal as partial substitute for fishmeal in rainbow trout (Onchorhyncus mykiss) diets: protein and energy utilization. Aquaculture, 128: 287-300. Tekeuchi, T., Sitoh, S., and Watanabe T., 1983. Requirements of tilapia nilotica for essential fatty acids. Bull. Jpn. Soc. Sci. Fish., 49: 1127 - 1134. 176 i Appendix 1. Kabete water quality parameters. Parameters Unit of measurement Results pH pH scale 7.2 Color mg pt/1 Less than 5 Turbidity N . T . U 7 Manganese mg/1 Less than 0.1 Calcium mg/1 16.8 Magnesium mg/1 4.86 Sodium mg/1 30 Potassium mg/1 8.8 Aluminium mg/1 -Chlorides mg/1 29 Fluoride mg/1 0.2 Nitrates mg/1 -Nitrites mg/1 Less than 0.01 Ammonia mg/1 -Total nitrogen mg/1 -Sulfates mg/1 5.33 Orthophospate mg/1 Less than 0.01 Total suspended solids mg/1 -Free carbon dioxide mg/1 12 Total dissolved solids mg/1 210 Residual chlorine mg/1 Less than 0.1 177