6to Simposio Centroamericano de Acuacultura Memoria: Sesiones de Tilapia DATE DUE .&. U Ju N lUU~ ' " . ¿¡¡~ Da ni --r---·--- ¡ ! -- -·- -----r--- - ·- - Fotos en la portada son todas de Honduras: 1. Estación Acuícola de Zamorano 2. Canales de producción, Aquacorporación 3. Jaulas con tilapia, lago de Yojoa 4. Ejemplares de la tilapia del Nilo 5. Ejemplar de la tilapia roja .. -.... Proceedings: Tilapia Sessions el E. Meyer Editor Cover photos are all from Honduras: 1. Zamorano Aquaculture Station 2. Production raceway, Aquacorporation 3. Cages with tilapia in lake Yojoa 4. Two Nile tilapia 5. A red tilapia ~ _r.;posio Centroamericano de Acuacultura ASOCIACION NACIONAL DE ACUICULTORES DE HONDURAS Asociación Nacional de Acuicultores de Honduras .-, 20G94• Global Aquaculture Alliance Escuela Agrícola Panamericana, Zamorano Pond Dynamics/Aquaculture Collaborative Reseach Support Program (PO/A CRSP) 6to Simposio Centroamericano de Acuacultura Program for the tilapia sessions (Programa de las sesiones sobre tilapia) ' lntroduction (Introducción) Contents (Contenido): Daniel E. Meyer, Panamerican Agriculture School, Zamorano, Honduras Tilapia genetics in Asia (Genética de la tilapia en Asia) Graham Mair, Asian lnstitute of Technology (AIT), Bangkok, Thailand Sex reversal: the directed control of gonadal development in tilapia (Reversión sexual: el control del desarrollo de las gónadas en tilapia) Ronald Phelps, Auburn University, Alabama, USA Nutrition and feeding of tilapia (Nutrición y alimentación de tilapia) Daniel Meyer, Escuela Agrícola Panamericana (Zamorano), Honduras Tilapia genetics: an American perspective (Genética de la tilapia: una perspectiva americana) Greg Lutz, Louisiana S tate University, USA Marketing tilapias in the Americas: 2001 and beyond (Mercadeo de tilapia en las Amércias: 2001 y más allá) Kevin Fitzsimmons, University of Arizona, USA Supermarket outlets for tilapia in Honduras: an overview of survey results (Supermercados para la comercialización de tilapia en Honduras) Nelson Ornar Funez, lvano Neira and Carole Engle, University of Arkansas at Pine Bluff, Arkansas, USA 8 9 35 61 71 72 82 Markets for tilapia ( Oreochromis sp.) in Nicaragua: 87 a descriptiva analysis of restaurants, supermarkets and stands in open markets (Mercados para tilapia (Oreochromis sp.) en Nicaragua: un análisis descriptivo de restaurantes, supermercados y puestos de venta en mercados públicos) lvano Neira and Carole Engle, University of Arkansas at Pine Bluff, Arkansas, USA Processing fresh tilapia fillets for export markets 92 (Procesando filetes frescos de tilapia para mercados de exportación) Jorge Maradiaga, Aqua Corporación de Honduras, Honduras Technology for for successful small-scale tilapia culture 97 (Técnicas para el cultivo exitoso de tilapia en fincas pequeñas) Daniel Meyer, Escuela Agrícola Panamericana (Zamorano) Production and marketing strategies used by small and 107 medium-scale fish farmers in Honduras (Estrategias de producción y mercadeo empleadas por productores de tilapia de escala pequeña y mediana, en Honduras) José Martínez, Escuela Agrícola Panamericana (Zamorano), Honduras, Joseph J. Molnar, Auburn University, USA, Freddy Arias, Escuela Agrícola Panamericana, Zamorano, Honduras, and Tom Popma, Auburn University, USA Levee pond design model 116 (Modelo para el diseño de estanques con diques) E.William Tollner, University of Georgia, USA Training and technical assistance in warm-water fish culture 118 (Capacitación y asistencia técnica en el cultivo de peces de aguas cálidas) Thomas Popma, Auburn University, USA and Daniel Meyer, Escuela Agrícola Panamericana, Zamorano, Honduras Web-based information delivery system for tilapia for sustainable 126 development of aquaculture in Honduras (Sistem de entrega de información por el internet para la tilapia y el desarrollo sostenible de la acuacultura en Honduras) Brahm Verma, University of Georgia, USA and Raquel lsaula, Sustainable Development Network in Honduras (RDS-Hn) Marine fish culture prospects in Latin America and 135 Carribean countries: review of candidate species and technological advances (Perspectivas para el cultivo de peces marinos en Latino América y el Caribe: reseña de especies con potencial y avances tecnológicos) Daniel Benetti, University of Miami, USA, Jorge Alarcón, University of Miami, Owen Stevens, Aquaculture Center of the Florida Keys, USA, Gill Banner­ Stevens, Aquaculture Center of the Florida Keys, Federico Rotman, University of Miami, Scott Zimmermann, University of Miami, Michael Feeley, University of Miami, William Matzie, Aquaculture Center of the Florida Keys, Refik Orhun, Mediafish Aquaculture and Seafood, USA, Brian O'Hanlon, Snapperfarm, USA and Loyal Eldridge, Aquaculture Center of the Florida Keys. Recirculating systems for fish culture 140 (Sistemas de recirculación para el cultivo de peces) Greg Lutz, Louisiana State University, USA Using aquaculture waste in diets for broilers and layer hens 141 (Empleando los desperdicios acuícolas en dietas para pollos y ponedoras) Abel Gernat, Escuela Agrícola Panamericana Zamorano, Honduras Sixth Central American Aquaculture Symposium: Proceedings of the Tilapia Sessions lntroduction Welcome to the Sixth Central American Aquaculture Symposium! We hope that the technical presentations and other activities planned for this week, will contribute to enhancing your knowledge of aquaculture and provide you with ample opportunities to interact with others attending this meeting. This publication, the proceedings for the tilapia sessions, has been organizad and printed with financia! support from the Pond Dynamics/Aquaculture Collaborative Research Support Program (PO/A CRSP), which in turn is funded primarily by USAID, Washington, OC, USA Our objective has been to provide all attendees with the appropriate written materials, to complement the oral presentations on tilapia culture to be given as part of the symposium. 1 want to thank each of the authors of these papers for their cooperation and hard work. 1 would also like to thank my secretary at Zamorano, Juana Espinosa de Ayestas, for her dedication and patience, to work with me through the process of compiling and editing these documents. 1 accept full responsibility for any errors or omissions in this document. Additional copies of the proceedings can be obtained by contacting me at my address in Honduras, or at the Zamorano booth in the commercial exhibition that is part of this symposium. Sincerely; Daniel E. Meyer, Editor P.O. Box 93, Tegucigalpa, Honduras dmeyer@zamorano.edu.hn telephone: 504-776-6140 ext. 2107 August2001 6to. Simposio Centroamericano de Acuacultura Tilapia Genetics in Asia 1 Graham C. Mair.2 School of Biological Sciences, University of Wales Swansea, Swansea SA2 BPP, Wales, U.K Abstract 9 This paper presents and discusses major issues in the application of genetics to tilapia, with emphasis on commercially important species used in aquaculture. The paper presents past and recent advances in the development and application of genetics based technologies to tilapia, from an applied perspective describing the existing and potential impacts on tilapia culture. Following on from a discussion of tilapia genetic resources and the impacts of domestication processes, the major emphasis of the paper is on the progress in the application of selection, hybridisation, sex control, chromosome set manipulation and transgenic technologies. The uptake and impact of these technologies is discussed mainly in the context of Asian aquaculture, where more than 85% of cultured tilapia are produced. Many of the tilapia genetic resources used in aquaculture today are little different from, or in many cases inferior to, wild caught stocks. Domesticated stocks have suffered from the consequences of poor broodstock management including inbreeding, genetic drift, unconscious negative selection and hybrid introgression. However, as the species has gained in importance as an international aquaculture commodity, so has there been a considerable increase in research effort to improve tilapia stocks, particularly over the past ten years. In terms of real applications and their uptake in aquaculture, recent advances in selective breeding and sex control technologies are having the greatest impact. A large volume of research work on hybridisation and chromosome set manipulations, 1 This article is modified from one that is to be published by INFOFISH in the procecdings of Tilapia 200 l. May 28-30. 2000. Kuala Lumpur, Malaysia 2 Presently based at the Aquaculture and Aquatic Resources Management Program (AARM), Asían Institute ofTechnology. PO Box 4. Klong Luang, Pathumthani 12120, Thailand. to where correspondence should be addressed. 1 O Tegucigalpa, Honduras whilst providing useful research tools, is having minimal impact upon aquaculture production. There have been a number of important recent advances in the application of transgenesis, which have been demonstrated to bring about substantial improvements in growth rates and yields, under experimental conditions. However, there are sorne technical and numerous socio-política\ and environmental constraints to be overcome before this technology can be widely adopted for aquaculture although it is likely that there will be isolated cases were transgenics might be approved for aquaculture in the near term . Future prospects are promising with the advent of molecular technologies including the identification of quantitative trait loci (QTL) and their application in marker assisted selection programmes that can enhance traditional breeding programmes, and it appears likely that significant further production gains will be achieved in the near future. There is thus considerable optimism that tilapia farmers will be among the first aquaculturists to benefit from the widespread applications of genetics based technology in tropical fish species. Tilapia culture in Asia Asian tilapia production dominates world production statistics with FAO data for 1999 (FAO, 2001) indicating that more approximately 82% of the world 's production of tilapia comes from Asia (although only 70% of total value) . Statistics provided from China indicate production levels in excess of 560,000 MT in 1999 indicating a growth rate in production of 400% per annum in the preceding decade (FAO, 2001 ). Other major producers in Asia include Thailand , the Philippines and Indonesia (Error! Reference source not found.) . The bulk of tilapia produced in Asia is for domestic consumption although Taiwan , China and Indonesia do export significant quantities, supplying 28%, 24% and 3% respectively of tilapia imports into the major import market of the U.S.A in 2000 (Fitzsimmons, 2000). Tilapia production is predominantly in freshwater ponds although there is sorne production in brackish water ponds and in cages in sorne locations. Tilapia are either cultured in monoculture or often in po\ycu\ture, mainly with carp species. Culture systems can be stand alone, as predominates in the Philippines for example, or integrated with other parts of farming systems, as commonly seen in Vietnam and parts of China There is sorne intensive production of tilapia, particularly in Taiwan . Unfortunately, whilst China dominates Asian tilapia production there is little available literature describing tilapia production systems in the country. Genetic resources for aquaculture Although there are a number of tilapia species of commercial interest for culture in Asia (see Table 1) Oreochromis niloticus, O. mossambicus and their various hybrids domínate in aquaculture. Recent growth in tilapia production from aquaculture has been in the culture of O. niloticus (Error! Reference source not found.) , which represents the most important tilapia species for freshwater aquaculture today. These figures should however, be considered with sorne caution and species wise production is often inaccurately recorded . The 6to. Simposio Centroamericano de Acuacultura 11 introduction of O. niloticus has commonly been preceded by the introduction of O. mossambicus. There have been severa! attempts to characterize and document indigenous genetic resources in A frica (Trewavas, 1983, Pullin et al., 1986 and Pullin et al., 1997) and to record the sequence of introductions and transfers of various species outside of their natural range (Pullin and Capili, 1988, Welcomme, 1988 and Agustín, 1999). Further to this, there have been numerous reintroductions of tilapias for aquaculture in the 1990s, especially in Asia and the Americas, to the extent that it would now be extremely difficult to document these adequately. However, the majority of tilapia cultured throughout the world (predominantly in Asia) is derived from the original introductions, in the 1950s for O. mossambicus and in the 1960s and early 1970s for O. niloticus. Genetic basis of introductions for aquaculture The original introduction of O. mossambicus to Asia can be traced to the discovery of five individuals (3d'and 2 !i') in Indonesia in the 1950s (Agustín, 1999). These individuals were bred (possible only a single pair mating!) and their progeny formed the basis of the aquaculture and feral stocks throughout Asia and probably further a field than this. Although it is still actively and deliberately cultured in a few areas, most commonly these are in brackish water areas or in regions where alternative species are not available. In general O. mossambicus is no longer popular for culture due to its poor growth performance (stunting is very common), early sexual maturation and high levels of recruitment under culture and it is regarded as a pest species in many countries. In a study of feral populations, Agustín (1999) used biochemical and molecular markers to characterize populations from a number of Asían countries and compared these to samples from a number of indigenous populations in southern Africa. She concluded that all the Asían populations most likely carne from a single source population in southern Africa, most probably vía Indonesia. She further determined that a number of rare alleles were missing in the feral Asian populations. In a study using five microsatellite DNA markers, she identified a six-fold reduction in allelic diversity and a two to three fold reduction in heterozygosity levels. Furthermore only one mitochondrial DNA haplotype was found in Asían stocks, compared to nine that were identified in indigenous African stocks. Thus, it is apparent that there have been profound bottleneck effects on these stocks, which may account for the relatively poor performance of domesticated stocks of this species in aquaculture. Furthermore, it is evident, that whilst the total size of the Asían populations may now exceed that of the indigenous African populations, little if any genetic variation has been regenerated through mutation or more recent introductions. The introduction of Oreochromis niloticus to Asia was traced by Pullin and Capili, ( 1988) to the transfer of approximately 120 fish from Egypt to Japan in 1962 and a shipment of 50 fish from Sudan to China in 1978. The majority of O. niloticus cultured in Asia today originate from these two introductions although this situation is rapidly changing following numerous reintroductions over the past 10- 15 years. Asian stocks have also been used to found stocks in other countries including the Americas. 12 Tegucigalpa, Honduras A substantial culture industry, producing in excess of 50,000 metric tones per annum (with sorne estimates being closer to 1 00,000) has built up in Thailand based on the introduction from Japan of 50 fish in 1965. lt is not firmly established how many of these fish were actually used to produce seed which formed the basis of the current industry but it seems likely that the effective population size (Ne) was not much greater than 30. However, compared to the apparent poor and declining culture performance of O. mossambicus, the culture performance of O. niloticus has been more robust with few reports of declining performance that could be associated with inbreeding depression. This may be a result of the slightly larger genetic base for the introduction of this species and inherently greater levels of genetic variation in O. niloticus. lt is evident, nevertheless, that the majority of cultured tilapia possess performance characteristics that are little different or in sorne cases worse than their wild relatives. Strain comparisons lt is currently accepted that the Nile tilapia O. niloticus is the best species for culture in the majority of inland, warm water aquaculture systems ranging from extensive, low input pond culture through to intensive recirculating systems. A number of studies in recent years have demonstrated that there are large differences in the relative culture performance of different populations and strains of tilapia across a range of different environments. In animal and plant breeding the term "strain" is normally applied to intra-specific sub-populations exhibiting distinctive traits, which are normally homozygous (i.e. true breeding). However, in fish and particularly in tilapia strains or isolates are normally loosely designated according to their location or origin and commonly have no distinctive traits, which can lead to considerable confusion. In the most comprehensive study of its kind, (Eknath et al., 1993) compared the performance of eight "strains" of O. niloticus, four Asian strains and four strains collected from the wild in Africa, across 13 different farm environments. This study demonstrated that, with the exception of a strain from Ghana, the wild caught "strains" had better culture performance than the domesticated stocks, with strains from Egypt and Kenya having the best performance. However, it should be noted here that previous genetic studies had indicated that the domesticated Philippine stocks of O. niloticus were introgressed with the slower growing O. mossambicus, which may have impacted their growth performance (Maca ranas et al., 1986). A further study on stocks established as pure O. niloticus was conducted by Capili, (1995). She compared the growth performance of 11 strains of tilapia from various African origins and found similar results in that the strains of Egyptian origin had the fastest growth rates, although the performance of three Kenya strains was relatively poor. In a further study, 6to. Simposio Centroamericano de Acuacultura 13 Capili (1995) also determined that the degree of sexual dimorphism varied between strains; with the Kenyan strains have the greatest difference in size between males and females. Oldorf et al., (1989) al so noted significant differences in the rate of sexual maturation between strains. In growth trials of tilapia species and strains, changes in ranking of strains between environments are common, indicating significant genotype x environment interactions in many cases (Dahilig, 1992, Elghobashy et al. 2000, Capili, 1995, Romana-Eguia and Doyle, 1992). However, in the Eknath et al., ( 1993) study, the analysis of the performance of the strains across environments led to the conclusion that the relative importance of genotype environment interaction was low compared to that of strain and sex differences. lt was as result of this finding that the researchers elected to go ahead with a large centralized breeding programme on the assumption that the improved fish would be superior across a wide range of culture systems. lmplications of domestication The effects of domestication which can include loss of genetic variation through genetic drift and inbreeding and unconscious selection, are felt less in tilapia than in sorne other species. This is largely due to the fact that individual tilapia have relatively low fecundity, necessitating the maintenance of large populations of brood stock, probably increasing the effective populations sizes and reducing the probability of mating close relatives, which would result in inbreeding. Furthermore, since brood stock are normally spawned randomly, the effects of unconscious selection are less than they might be in other species such as carp. However, genetic bottlenecks are common and this can result significant changes in gene frequencies through genetic drift. Also, due to its relatively short generation time (6-12 months) any selection forces acting on domesticated stocks can change their phenotypic characteristics over relatively short periods of time. lt is thus likely that local adaptation of strains does occur in sorne environments with resulting superior performance of these local adapted strains when evaluated in the particular environment under which they were domesticated. Another risk with the domestication of the tilapia, a species group in which many hybrid crosses produce fertile offspring, is that of the breakdown of species barriers through hybrid introgression. This was shown to have occurred on a wide scale among Philippine tilapia stocks (Macaranas et al. 1986). As more populations come under domestication, and attempts at genetic improvement move forward, more distinct strains of tilapia will be developed. lt is important that these strains be adequately characterized and documented. At present the only documentation of tilapia strains is within FishBase, a relational database being developed and maintained by ICLARM in cooperation with FAO (Froese and Pauly, 2000). Options for genetic improvement 14 Tegucigalpa, Honduras With the exception of sorne ornamental fish, common carp and salmonids, there has been relatively little application of genetic enhancement technologies to fish compared to that which has been achieved in other forms of agriculture. Given its relatively recent history of domestication and adoption for aquaculture, this is certainly true of the tilapia and it is only within the past 1 O years that significant attempts have been made to improve cultured stocks. In common with many fish species a number of approaches to improvement are available which are not possible in higher organisms. In the case of tilapia, its large effective population sizes, short generation time, ease of handling, stress and disease resistance and ease of both "natural" and artificial reproduction, make it highly suited to the application of a number of genetic enhancement technologies. With tilapia apparently destined to become a major international commodity, the development of these technologies has accelerated in recent years. Qualitative traits Qualitative traits are those which fall into discrete categories (i.e. are not continuously distributed) and their inheritance can usually be understood on the basis of basic or Mendelian genetics. New qualitative phenotypes arise from mutations. In tilapia the main qualitative traits that have been studied are body shape and colour traits. The inheritance of a number of body shape traits has been well described by Tave (1992). For aquaculture purposes body shape changes induced by mutation are invariably deleterious and thus their inheritance is of little interest to aquaculturists other than to know how to eliminate them if they arise in cultured stocks. Colour varieties, particularly the red tilapias, are however of considerable commercial interest, often securing a higher market price than the normal wild type coloration. There are a number of variants of red tilapia that have arisen independently in several different stocks and several of these variants, with variable expression of colour phenotypes, are being used in aquaculture around the world. lt is probable that the red colour first arose in O. mossambícus and was transferred to other stocks through hybridization with O. niloticus and in sorne cases also O. u. homorum. The situation regarding the precise origins of sorne of the red tilapias used in aquaculture today (e.g. Philippine, Thai, Florida and Taiwanese) is now somewhat confusing. However, it is clear that the nature of the inheritance of the various colour patterns in these strains is complex (see reviews by Wohlfarth, 1990 and McAndrew and Wohlfarth, in press). There are no known examples of a true breeding and homogenous red tilapia. The expression of the red phenotypes is highly heterogeneous and red tilapia can exhibit varying amounts of red, pink, orange and white pigmentation, often combined with degrees of black blotching. There is no doubt that there would be a considerable commercial demand for a true breeding homogenous red tilapia as many producers of red tilapia today have to grade their fish according to the colour and for different markets, reducing potential revenue and increasing labour costs. Efforts to select for homogeneity of colour have made sorne progress but would appear unlikely to yield true breeding homogenous phenotypes in the near future. 6to. Simposio Centroamericano de Acuacultura 15 There are a number of colour variants within pure species in which the inheritance is more easily understood and in which true breeding lines may be established. These include the red in O. mossambicus and "blond" and "pearl" in O. niloticus (Wohlfarth, et al. 1990, Scott et al., 1987, McAndrew et al., 1988) but commercial demand for these colour varieties has yet to develop. Selective breeding Tave (1988) reviewed studies on quantitative characters in tilapia up to that date, concluding that moderate heritabilities (0.15 - 0.5) are common for a number of commercia\\y important traits, indicating significant contributions of additive genetic variance to these traits. However, severa! early attempts to apply selective breeding to commercial stocks were disappointing indicating low heritabilities and producing low response to selection (Table 3). Since this time there have been a number of important attempts to apply techniques of traditional selective breeding to tilapia stocks as summarized in Table 3. Most studies have produced significant response to selection, usually for traits associated with growth rate. However, most of these published studies have been on an experimental basis and few of the benefits of selection have been passed on to the industry. A majar exception is the Genetically lmproved Farmed Tilapia (GIFT) programme coordinated by ICLARM in the Philippines. This comprehensive and well-funded selection programme has demonstrated significant responses to selection over multiple generations. Using a combined selection methodology on a synthetic base population developed from newly introduced strains from Africa and domesticated Asian strains, this programme achieved genetic gains averaging 13% over five generations providing an estimated cumulative increase of 85% in growth rate compared to the base population from which is was selected (Eknath and Acosta, 1998). Whilst the genetic gains are significant and clearly demonstrate benefits of well organized breeding programs, the accumulated response to selection appears not to be fully expressed in all culture environments and difficulties in identifying adequate controls have created difficulties in accurately assessing genetic gains. Since 1997, the benefits of this programme in the form of the GIFT tilapia have been widely disseminated into the tilapia culture industry in the Philippines. The GIFT strain has also been introduced toa number of other countries in the region for dissemination and as the base population for the establishment of a number of national tilapia breeding programmes. A number of issues have been highlighted in terms of sustainability of selection programmes and it is evident that unless they receive long-term financia! support (e.g. Government) for qualified personnel and running costs, it will be necessary to generate income via the dissemination process. This can have profound effects on the uptake and sustainability of genetic gains and on the social and economic impact of selection programmes on their intended beneficiaries. 16 Tegucigalpa, Honduras Molecular techniques for enhanced selective breeding There has been considerable effort in recent years directed at the construction of linkage maps of the tilapia genome using a range of different genetic markers particularly microsatellite DNA markers and amplified fragment length polymorphisms - AFLP (Lee and Kocher, 1996, Kocher et al. 1998). This mapping effort opens up the potential for identifying quantitative trait loci (QTL), gene loci that directly influence a trait, which can be identified through combining gene mapping, breeding and trait evaluation. One example of the identification of such a trait is the association between alleles ata microsatellite locus and cold tolerance in an F2 hybrid of O. mossambicus and O. aureus (Hallerman et al., personal communication). The identification of QTL associated with commercial important traits such as disease resistance, environmental tolerances, sex or even growth rate will create the possibility to carry out marker-assisted selection (MAS) for targeted traits. In MAS, DNA markers that are closely linked to one or more QTLs can be used to increase the response to selection in a population, increasing the efficiency of selective breeding programmes. Hybridization and crossbreeding Whilst the majority of research and development work on selective breeding of tilapia is relatively recent, the main early emphasis on the study of quantitative traits was through hybridization and crossbreeding. As is indicated in Table 1, tilapia species differ remarkably from one another in many traits of commercial importance. Differences between strains, within species, are less but can be significant as indicated above. These intra-strain differences have only be evaluated to any degree in the commercially important O. niloticus. Hybrids between tilapia species and even genera and have occurred in wild and feral populations, especially where translocations of species into new environments have occurred. There have also been numerous deliberate hybridizations between species (reviewed by Lovshin, 1982 and Schwartz, 1983) and between genera (reviewed by Rana et al., 1996). More than 60 different hybrids have been produced between and among the Oreochromis, Sarotherodon and Tilapia with the majority being between Oreochromis species. F1 hybrids are commonly produced with specific objectives in mind, usually in the hope of observing heterosis (hybrid vigour) for commercially important traits orto produce a particular desirable phenotypic feature (such as colour or environmental tolerances) in the hybrid or its subsequent generations. Virtually all reports of hybridization show that hybrids within and between the tilapia genera are viable indicating the speciation within the tilapiine fishes may be relatively recent. Furthermore, there are no reports of sterility, commonly found in hybrids of other species groups, among the tilapia hybrids. Despite the 6to. Simposio Centroamericano de Acuacultura 17 large number of reports of various hybridizations, there are few, if any, published studies that clearly demonstrate heterosis for any commercially important trait. In the vast majority of hybrids, the traits studied were intermediate between those of the parental species (McAndrew and Majumdar, 1988). Oreochromis hybrids are characterized by a surplus of males and the occurrence of all-male broods is relatively common and this is where the majar interest in hybridization lies. The first report of monosex hybrids created significant interest with the potential for mass production of all-male progeny to prevent the serious problem of unwanted reproduction in aquaculture. However, in most cases, sex ratios differ between reciproca! crosses and there are few hybrid combinations which consistently give monosex progeny, with perhaps those using male O. urolepis hornorum being the most reliable. Table 4 summarizes the hybrid combinations known to produce monosex male progeny. Attempts to commercialize monosex hybrids, usually with the O. niloticus x O. aureus cross or using male O. urolepis hornorum with females from O. niloticus or O. mossambicus, have been disappointing with females usually occurring in previously all male broods. Failure to sustain production of all-male tilapia hybrids is most likely due to insufficient care in keeping brood stock segregated by sex and species, and in preventing introduction of hybrids into the brood stock ponds (Wohlfarth, 1994). lmprovements in brood stock management may enable more effective utilization of monosex hybrids. However, with O. niloticus accepted as the best commercial species for the majority of tropical freshwater aquaculture environments, dilution of the O. niloticus genome with other species tends to reduce the performance potential in aquaculture, compared to pure O. niloticus except under specific circumstances, for example the benefits of cold tolerance of hybrids involves O. aureus in over wintering fish in seasonal sub­ tropical climates . Despite the lack of clearly demonstrated benefits, in terms of enhancement of commercially important traits, \there is a significant commercial production of hybrids in some parts of Asia, most notably the production of O. niloticus x O. aureus F1 hybrids in Taiwan and parts of China. lt is not clear whether these hybrids are produced primarily due to the high proportions of males or for their enhanced cold tolerance compared to pure O. niloticus, or possible for a combination of these and other factors. Also most red tilapia in commercial production have hybrid ancestry and are usually cultured for their marketability and/or their enhanced saline tolerance compared to pure O. niloticus. With the failure of hybridization to effectively salve the problem of early sexual maturation , unwanted reproduction and overpopulation in tilapia culture , alternative technologies were sought. One popular alternative is hormonal sex reversa! but this technique has a number of important technical , environment and social constraints. An alternative genetics based solution has been sought, founded on the current knowledge of the mechanisms of inheritance of sex in tilapia. 18 Tegucigalpa, Honduras Sex control Due to commercial interest in monosex populations for aquaculture a considerable amount of research has been conducted on the genetics of sex determination in tilapia (see review by Trombka and Avtalion , 1993). A number of theories have been proposed on the genetics of sex determination in tilapia ranging from a single gene model through to polygenic inheritance. The current consensus is that, in the commercially important Oreochromis species, sex determination is "predominantly" monofactorial, being controlled by sex chromosomes or primary sex determining gene(s) . Research using sex reversa! , progeny testing and chromosome set manipulation has revealed two alternative "sex chromosome" models. In O. niloticus the female is homogametic XX, the male being heterogametic XY (Mair et al., 1991 a) whilst in the closely related O. aureus the alternative model of heterogametic WZ females and homogametic ZZ males applies (Mair et al. , 1991 b). There is also substantial evidence for effects of one or more autosomal genes on sex ratio together with an increasingly well documented eftec\ of temperature. The role of temperature in influencing sex differentiation is becoming increasingly evident and should be considered when evaluating all past research on sex ratios. Elevated temperatures (-36°C) during the period of sex differentiation have been shown to increase the proportions of males in putative monosex female O. niloticus and , toa lesser degree, to increase the proportion of females in putative all male progeny (Baroiller et al., 1996, Abucay et al. , 1999 and Baras et al., 2001 ). Similarly high temperatures have been shown to push sex ratios in both directions in O. aureus (Mair et al. , 1990, Des prez and Melard, 1998, Baras et al. 2000). The only recorded effect of low temperatures was to increase the proportion of males in O. mossambicus (Mair et al. , 1990). Based on the theory of predominantly monofactorial sex determination, it has preved possible to manipulate sex ratio using a combination of sex reversa! and progeny testing to identify sex genotypes. In a majar breeding program in O. niloticus, Mair et al., (1997) were able to mass-produce novel YY "supermales". When crossed to normal females (XX) these YY males have the unique property of siring only male progeny. These progeny are termed genetically male tilapia (GMT) and are normal (XY) genetic males (although sorne can "naturally" revert to female , giving GMT an average sex ratio of >95% male) . The hormone treatments used as part of the process to produce YY males are two generations removed from the fish that are marketed to the consumer so neither the GMT or their YY male parents are hormone treated in any way. This makes the technology more user and environmentally friendly than the alternative of direct hormonal sex reversa!. Furthermore, the technology can be applied in a range of hatchery systems simply by replacing brood fish with YY males although good brood stock management is required to prevent contamination . On-station and on-farm trials indicated substantial increases in production (40% increases in yields) using GMT compared to normal mixed sex tilapia . Since the development of the original intra-strain GMT in an Egyptian strain of O. niloticus trials 6to. Simposio Centroamericano de Acuacultura 19 of various crossbred GMT have produced significantly faster growth rates. Recently trials involving GMT, created by crossing the original YY male to a selected female line, produced further gains in growth rate, compared to both original and crossbred GMT (Table 5). The selected female line was developed through three generations of within family selection for growth rate and selection for combining ability in GMT sex ratio, applied to a synthetic base population consisting of five fast growing strains of O. niloticus. The outputs of this is technology in the form of GMT and GMT producing brood stock are now being widely disseminated in the Philippines and Thailand through a network of accredited hatcheries. The technologies has also been adopted in up to 20 other countries worldwide. A similar although slightly simpler breeding programme has been developed to produce monosex male O. aureus through the production of sex reversed ZZ females (Mair et al., 1991b and Melard, 1995). Chromosome set-manipulation A number of chromosome set manipulations are possible including gynogenesis, androgenesis, production of clones, triploidy and tetraploidy (see review by Mair, 1993). Gynogenesis and androgenesis are forms of uniparental inheritance, induced by fertilizing one gamete with another in which the DNA has been denatured by irradiation, to produce a haploid zygote which can then be diploidized by application of physical shock such as temperature or pressure. Gynogenesis, where eggs are fertilized with denatured, UV irradiated, sperm, has been a useful research tool for studying the inheritance of sex, among other things. Where the diploid state is restored by disrupting first mitosis (as for mitotic gynogenesis and androgenesis) using physical shocks, the resulting individual is completely homozygous due to the retention of two sets of homologous chromosomes (i.e. 100% inbred). lnduction and survival rate of these homozygous fish are very low but all genes that are deleterious or lethal in their homozygous state will be selected out in the process. A second generation of gynogenesis or androgenesis will produce genetically identical individuals or homozygous clones. Crosses between homozygous gynogenetic or androgenetic fish will produce heterozygous clones. Both homozygous and heterozygous clones have been produced in tilapia (Hussain et al., 1998 and Jenneckens et al. 1999). There is the potential that heterosis may be realized in the crosses between these highly inbred clona! lines but evidence for this has yet to be found. The clona! lines, however, have considerable potential in research as interna! controls in communally stocked growth trials and through the elimination of genetic variance when studying the effects of other variables such as nutrition, sex determination or disease resistance. Triploidy has been induced in a number of tilapia species through the application of physical shock to eggs fertilized with normal sperm, at the stage of second meiosis, inducing retention of the second polar body (see review by Mair, 1993). Triploids are sterile and as such have sorne commercial potential in addressing 20 Tegucigalpa, Honduras the constraint of early sexual maturation and unwanted reproduction. In growth trials of induced triploids Bramick et al., (1996) demonstrated that, post-sexual maturation, growth of sterile triploids was significantly faster than diploids in a pond environment. Final yields of triploid fish after 25 weeks of grow-out were 56 to 123% greater than for diploids although these results were confounded by the presence of recruits following reproduction in ponds stocked with normal diploid fish. In a more recent study on triploid O. aureus (Byamungu et al. 2001) little difference was observed in the relative growth performance of diploids and triploids under normal feeding regimes, but triploids were found to have significantly higher yields under restricted feeding regimes. Despite these promising findings the potential for the application of triploidy in tilapia culture is very limited due to the requirement of artificial fertilization in order to apply physical shocks at precise intervals after fertilization. Due to the relatively low fecundity of individual fish and the difficulty in collecting ovulating eggs in this multiple spawning species, artificial fertilization in tilapia is impractical on a commercial scale. lt may be possible to mass-produce triploids in matings of diploid and tetraploid fish. Tetraploidy has been induced in tilapia, albeit at very low rates of induction (Mair, 1993) but the majority of tetraploid embryos have been characteristically deformed and inviable. A recent study produced a high incidence of tetraploidy (80%) in O. niloticus (EI-Gamal et al., 1999) although the viability beyond the early fry stages was not determined. Only a few viable and fertile tetraploids would be required to produce tetraploid lines, which would enable large-scale production of triploids in diploid x tetraploid matings. Transgenesis Transgenesis is one of the most promising technologies for generating relatively rapid genetic improvements. Transgenesis involves the introduction of an exogenous genes into a new organism to confer novel phenotypic characteristics on that organism. Typically, the desired genes are first identified and or constructed, commonly combining coding and prometer sequences from different donor sources, and then cloned. Multiple copies of the transgene are then introduced to the fertilized eggs, commonly by microinjection or electroporation. At a later stage of development, cells of the organism are tested to determine whether copies of the transgene has become incorporated into the genome and whether this incorporation is in all cells or only in the cells of sorne tissues (i.e. a mosaic). After incorporation is determined the organism can be evaluated to determine if the transgene product is being expressed and in what amounts. The next stage of development is to determine whether the transgene is inherited and expressed in the next generation via the germ cells. lnheritance of the transgene is required in order to develop true breeding lines of the transgenic organism. Severa! studies have made progress in developing methodologies to introduce and ensure or monitor incorporation, expression and inheritance of transgenes in tilapias (Brem et al., 1988, lndig & Moav, 1988, Maclean et al. 1992, Alam et al. 6to. Simposio Centroamericano de Acuacultura 21 1996, Rahman et al. 1997). However, only two major published programmes have produced true breeding, enhanced transgenic lines. The first of these programmes has focused on the introduction of a tilapia growth hormone cONA into a hybrid tilapia ( originating from O. urolepis homorum and O. aureus crosses) in Cuba (Martinez et al., 1999). This study has demonstrated that the transgene construct (incorporating a human cytomegalovirus regulatory sequence) has been incorporated, expressed and transmitted through four generations. Growth performance trials of the homozygous and hemizygous transgenic fish, communally stocked in ponds with non-transgenic fish, for a three-month grow-out period, were conducted. Transgenic fish were significantly larger (up to 82% with an average of 55%) than non-transgenic fish with indications of a transgene-dosage effect. The results indicate stable germ line transformation in this fast-growing transgenic tilapia line and it seems likely that this transgenic strain will be adopted for aquaculture in Cuba. The second published study in the U.K. involves the introduction of an all-fish construct of a chinook salman growth hormone gene with an ocean pout antifreeze regulatory sequence, into O. niloticus (Rahman and Maclean, 1999). The initial transmission rate from GO to G1 generation was observed to be less than 10% in these lines indicating a mosaic distribution of the transgene in the germ cells. However, transmission rates from the first to the second generation were found to follow the expected Mendelian ratios. The chinook salman growth hormone was produced in severa! generations of the transgenic tilapia indicating expression and transmission of the gene. This expression of the transgene resulted in dramatic growth enhancement with the average weight of the transgenic fish being three to four times that of their non-transgenic siblings and with equivalent food conversion efficiencies (Rahman and Maclean, 1999, Rahman et al., in press). Transgenesis would appear to offer very considerable potential for enhancement of yield in tilapia. However, the rate of genetic change in transgenics is such that their phenotypic and behavioural properties cannot easily be predicted and the introduction of these fish for commercial aquaculture faces many constraints. The risks to the environment posed by the uncontrolled introduction of transgenic fish needs to be adequately assessed and many governments are currently adopting cautious policies with regard to their introductions. Many of these constraints may be overcome if guaranteed sterile transgenic tilapia can be produced. This could be achieved through efficient methods of triploidization or, in the medium to longer term, through disruption of the physiological pathways of reproduction vía the introduction of new transgenes such as anti sense constructs, which are currently under development. Consumer response to genetically modified fish (GMOs) in sorne countries may be very negative to the extent that adoption by farmers may involve significant economic risks. This negative response may be lessened with regard to transgenic fish developed using con-specific gene constructs, which is now the trend in research. 22 Tegucigalpa, Honduras At the time of writing, negative reaction to the concept of genetically modified fish among the popular media across much of the world, particularly in Europe, is so strong, it seems unlikely that such fish would be approved for production or consumption in the developed world in the near future. The risk-benefit ratio is very different in developing countries where food security can be a majar issue and possibly we will see transgenic tilapia produced first in these countries. Conclusions and Future prospects lt is evident that significant progress has been made in the last decade in the application of genetic techniques to tilapia. The immediate applications of sorne current technologies such as hybridization, crossbreeding and chromosome set manipulation would appear to have limited potential for significant production gains. Substantial benefits in terms of growth rates and improved yields under culture have been demonstrated from breeding programmes for selection and sex control. The results of the successful applications of these breeding programmes need to be introduced to aquaculture through technically and economically sustainable dissemination programmes. Transgenesis would appear to offer great potential for genetic enhancement of tilapia under culture provided that the remaining technical constraints can be overcome and that an appropriate legislative environment can be created following the satisfactory completion of appropriate environmental and health risk assessments. Future development in the application of genetics are likely to include: • The use of hybrid introgression to breed desirable characteristics of some species/strains (such as saline and cold tolerance) into other faster growing species/strains. • The application of different genetic improvement technologies (such as selection and sex control) into combined breeding programmes. • The application of molecular markers to enhance selective breeding programmes through marker assisted selection. • The application of bioinformatics to enhance breeding programmes, enable strain labeling (e.g. for protection of IPR or breeding rights) and for disease diagnostics. • The application of sex specific markers, possibly combined with gynogenesis, to increase the efficiency of sex control breeding programmes (such as the YY male technology) and increase sex ratios up to 100% male. • The development of new strains of transgenic tilapia, most likely incorporating cloned tilapia genes identified through on-going gene mapping programmes. Whilst it can be said that the levels of genetic improvement present in aquaculture species is some distance behind those developed for other agricultura! species, at the current pace of development and with the options available in manipulating fish genomes, the gap may narrow appreciably in the 6to. Simposio Centroamericano de Acuacultura 23 coming years. The ease of handling and domestication of tilapia together with its reproductive characteristics and relatively short generation time, make this an ideal "model" species for research. When considered in the light of the rapidly growing worldwide commercial importance of tilapia as a cultured species, it appears likely that this will be the tropical species in which we see the most rapid developments of genetic technologies. 24 Tegucigalpa, Honduras Table 1 Summary of major characteristics of commercially important tila pías Common name Species Characteristics Nile tilapia Oreochromis Commonly the fastest growing of niloticus the tilapias in freshwater. Breeds readily in many types of hatchery system. Caudal fin bars a distinctive feature A number of colour varieties exist including "red" and "blond". Mozambique tilapia or Oreochromis High fecundity, overpopulation Black tilapia mossambicus and stunting common. High saline tolerance and well adapted to brackish water. Normally black coloration but red varieties exist. Blue tilapia Oreochromis Often sympatric with O. niloticus aureus but usually slower growth. Cold tolerant and used in hybridization for production of monosex. Blue colour, no other colour varieties known Non e Oreochromis U sed mainly for production of urolepis monosex hybrids. Similar hornorum characteristics to O. mossambicus although less fecund Non e Oreochromis Salt tolerant, u sed in seawater spilurus sp. cage culture. Zill's tilapia or redbelly Tilapia zillii Substrate spawner, tolerant of tilapia high salinities. Feeds on macrophytes Redbreast tilapia Tilapia renda/Ji Substrate spawner. Feeds on macrophytes Gallilee tilapia Sarotherodon Paternal mouthbrooder, saline ga/ileus tolerant. Slow growth Red tilapia Hybrid origins Commonly derived from crosses between O. mossambicus and O. niloticus but some also thought to include introgression from O. urolepis hornorum or O. aureus. Red colour seldom fixed with red/pink and black blotching common. Saline tolerant but sometimes exhibit low fecundity. 6to. Simposio Centroamericano de Acuacultura 25 Table 2 Various measures of genetic variability based on data from five microsatellite loci (N = 28-45 fish per sample) in feral populations of O. mossambicus from Asia compared to wild caught fish from Southeastern Africa (Source: Agustin, 1999). Population Mean no. of % polymorphic Mean observed alleles per locus loci heterozygosity Feral Malaysia 1.80 60 0.27 Ftii 2.20 80 0.40 Australia 1.80 80 0.28 Wild Bangu/a 14.20 100 0.86 Elephant Marsh 11.20 100 0.87 26 Table 3 Summary of the progress in the application of traditional methods of selective breeding to tilapia Species Trait Method Progress So urce O. niloticus Growth rate Mass No response to selection was detected after 2 generations. (Hu lata et al., 1986) selection O. niloticus Early growth Estimates of heritability were low producing negative realized Teichert-Coddington heritability and Smitherman (1988) O. niloticus Age at Within family Fish selected for early maturation matured 11-14 days earlier (Uraiwan, 1988) maturation selection after one generation of selection. lnconsistent results produced in 2nd generation O. niloticus Growth rate Mass No response was produced after one generation of selection (Huang and Liao, 1990) selection O. niloticus Growth rate Within-family Average response of 3% per generation over 8 generations (Bolívar et al., 1994) selection (recent gains have been greater, Bolívar pers. comm.) O. niloticus Growth rate Combined A genetic gain of 12-17% per generation has been recorded (Eknath & Acosta, selection over 5 generations of selection 1998) O. niloticus Late sexual Family Significant responses to selection were observed for stage of (Horstgen-Schwark and maturation selection maturation (-29%) in females and GSI (-39%) in males after two Langholz, 1998) generations O aureus Growth rate Mass High line fish were 49% heavier and 10% longer than random (Bondari et al., 1983) selection bred controls after one generation of selection. Low line fish were 52% lighter and 21% shorter than controls. r---------- ----- -- O. aureus O Cold Mass Realized heritabilities ranged from -1 to +1 indicating problems (Behrends et al., 1996) niloticus & toleran ce selection with methodology hybrids Red tilapia Growth rate Mass Realized h2 of 0.32 and 0.37 for weight and length respectively (Jarimopas, 1990) r selection after 5 generations ·------ --- ---------r---- ------ --·· Red tilapia Weight Mass Results confounded by correlated response. Realized heritability (Behrends et al., 1988) selection ranged from -0.75- +1.0 --------~ ------------- - 6to. Simposio Centroamericano de Acuacultura Table 4 Summary of different hybrid combinations that have been known to produce monosex male progeny Female parent Male parent Note O. niloticus O. aureus Applied commercially but results inconsistent ----·- O. niloticus O. macrochir - O. niloticus O. urolepis hornorum Majority of broods are all-male 1 Sorne commercial application ------- ·- O. niloticus O. variablis All progenies were monosex O. mossambicus O. aureus O. mossambicus O. urolepis hornorum All progenies were monosex O. spilurus niger O. macrochir O. spilurus niger O. urolepis hornorum All progenies were monosex O. aureus O. urolepis hornorum ---------- T. zillii O. andersonii All progenies were monosex Table 5 Summary of progress in the development of "new" genetically male tilapia (GMT) since its original development and release for culture in 1995 (Mair et al. unpublished data) Basis of Release no & Growth peñormance genetic date improvement Original GMT Release 1.0 30-35% faster growing than mixed sex (developed in (1995) tilapia of Philippine strains an Egyptian strain) Best crossbred Release 2.1 Rel. 2.1 15-25% faster growing than the GMT (Philippines - "original" GMT (Rel. 1) in the Philippines 1999) Release 2.2 (U.S.A. - 2000) ------ GMT from Release 3 (2001) 7.5 - 17.5% faster growing than best selected crossbred GMT (Rel. 2.1) female line GMT from Release 4 (2002?) Preliminary data indicates 5-15% crossbred yy advantage o ver Re l. 3. Still under mal e X development selected female line 27 28 Tegucigalpa, Honduras O North America \ O South Ame rica • Africa BRoW •China OThailand • Philippines O Indonesia , • Taivrcm \.RoA - ---··-·---- - -- ------------------- Figure 1 Pie chart illustrating the dominance of Asia in worldwide tilapia production in 1998 (Data Source: FAO 2000) ~--------- -- ---- ----,---,----------=-----=--=-=:----- -,_ ~ e: o :.¡:::; o :::1 , o .... a.. 1,200,000 1,000,000 800,000 600,000 400,000 200,000 o 1 : i • O niloticus , • other tilapia spp ; [;] O. mossambicus • 1 O O. aureus ~ ¡¿o?) ¡¿oro ~ ¡¿o'b ¡¿oOJ P.>\:) R>" P.>"' P.>'? P.>~ P.>?) P.>ro P.>'\ P.>co "<:!) "<:!) "Q) "Q) "Q) "Q) "Q) "Q) "Q) "Q) "Q) ~ "Q) "Q) "Q) Year Figure 2 Worldwide tilapia production according to species demonstrating that expansion in production in recent years has come from the expansion in culture of Oreochromis niloticus (Data source: FAO, 2000). 6to. Simposio Centroamericano de Acuacultura 29 References Abucay, J.S., Mair, G.C., Skibinski, D.O.F. and Beardmore, J.A. 1999. Environmental sex determination: the effect of temperature and salinity on sex ratio in Oreochromis niloticus L. Aquaculture 173: 219-234. Alam, M.S., Popplewell, A. and Maclean, N. 1996. Germ line transmission and expression of a lacZ containing transgene in tilapia (Oreochromis niloticus). Transgenic research 5: 87-95. Agustín, L. 1999. Effects of genetic bottlenecks on levels of genetic diversity and differentiation in feral populations or Oreochromis mossambicus. Ph.D. Thesis, Queensland University of Technology, Australia. Baras, E., Jacobs, B., and Mélard, C. 2001. Effect of water temperature on survival, growth and phenotypic sex of mixed (XX-XY) progenies of Nile tilapia Oreochromis niloticus, Aquaculture 192, 187-199. Baras, E., Prignon, C., Gohoungo, G. and Mélard, C. 2000. Phenotypic sex differentiation of blue tilapia under constant and fluctuating thermal regimes and its adaptive and evolutionary implications, Journal of Fish Biology 57: 210-223. Baroiller, J.F., Fostier, A., Caulty, A., Rognon, X. and Jalabert, B. 1996. Effects of high rearing temperatures on the sex ratio of progeny from sex reversed males of Oreochromis niloticus. In: Pullin, R.S.V., Lazard, J., Legendre, M., Kothias, J.B.A., Pauly, D. (Eds.), The Third lnternational Symposium on Tilapia in Aquaculture. ICLARM Conf. Proc. 41. lnternational Center for Living Aquatic Resources Management (ICLARM), Manila, Philippines. pp. 246 - 256. Behrends, L.L., Kingsley, J.B. and Bulls, M.J. 1996. Cold tolerance in maternal mouthbrooding tilapias: heritability estimates and correlated growth responses at sub optimal temperatures. In: Pullin, R.S.V., Lazard, J., Legendre, M., Kothias, J.B.A., Pauly, D. (Eds.), The Third lnternational Symposium on Tilapia in Aquaculture. ICLARM Conf. Proc. 41. lnternational Center for Living Aquatic Resources Management (ICLARM), Manila, Philippines., pp. 257-265. Behrends, L.L., Kingsley, J.B. and Price 111, A.H. 1988. Bi-directional­ backcross selection for body weight in a red tilapia. In: Pullin, R.S.V., Bukhaswan, T., Tonguthai, K., Maclean, J.L. (Eds.), The Second lnternational Symposium on Tilapia in Aquaculture. ICLARM Conference Proceedings. Department of Fisheries, Thailand and lnternational Center for Living Aquatic Resources Management, Bangkok, Thailand and Manila, Philippines, pp. 125-133. Bolívar, R.B., Bartolome, Z.P. and Newkirk, G.F. 1994. Response to within­ family selection for growth in Nile tilapia (Oreochromis niloticus L.). In: Chou, L.M., Munro, A.D., Lam, T., Chen, T.W., Cheong, L.K.K., Ding, J.K., Hooi, K.K., Khoo, H.W., Phang, V.P.E., Shim, K.F., Tan, C.H. (Eds.), The Third Asían Fisheries Forum. Asian Fisheries Society, Manila, Philippines, pp. 548-551. 30 Tegucigalpa, Honduras Bondari, K., Dunham, R.A., Smitherman, R.O., Joyce, J.A. and Castillo, S. 1983. Response to bi-directional selection for body weight in blue tilapia. In: Fishelson, L., Yaron, Z. (Eds.), Proceedings of the lnternational Symposium on Tilapia in Aquaculture, Nazareth, Israel. Tel Aviv University, Tel Aviv, Israel, pp. 302-312. Bramick, U., Puckhaber, B., Langholz, H.J. and Horstgen-Schwark, G. 1996. Testing of triploid tilapia (Oreochromis niloticus) under tropical pond conditions. In: Doyle, R.W., Herbinger, C.M., Ball, M., Gall, G.A.E. (Eds.), Genetics in Aquaculture V. Elsevier Science, Amsterdam, Netherlands, pp. 343-353. Brem, G., Brenig, B., Horstgen-Schwark, G. and Winnacker, E.L. 1988. Gene transfer in tilapia (Oreochromis ni/oticus). Aquaculture 68. 209-219. Byamungu, N., Darras, V.M. and Kühn, E. R. 2001. Growth of heat-shock induced triploids of blue tilapia, Oreochromis aureus, reared in tanks and in ponds in Eastern Congo: feeding regimes and compensatory gorwht response of triploid females. Aquaculture 198: 109-122. Capili, J.B. 1995. Growth and Sex Determination in the Nile Tilapia, Oreochromis ni/oticus (L.). Thesis, School of Biological Sciences, University of Wales Swansea, U. K., 271 pp. Dahilig, L.R. 1992. Genotype x Environment interaction in the genus Oreochromis: Growth under sex reversed male and mixed sex culture in ponds and tanks. Master of Science Thesis, Central Luzon State University, 106 pp. Desprez, D., and Melard, C. 1998. Effect of ambient water temperature on sex determinism in the blue tilapia Oreochromis aureus. Aquaculture, 162: 79-84. Eknath, A. E. and Acosta, B.O. 1998. Genetic improvement of farm' tilapia project final report (1988-1997). ICLARM. Manila, Philippines /5 pp. Eknath, A.E., Tayamen, M.M., Palada-de Vera, M.S., Danting, J.C., Reyes, R.A., Dionisia, E.E., Capili, J.B., Bolívar, H.L., Abella, T.A., Circa, A.V., Bentsen, H.B., Gjerde, B., Gjedrem, T. and Pullin, R.S.V. 1993. Genetic improvement of farmed tilapias: the growth performance of eight strains of Oreochromis niloticus tested in different farm environments. Aquaculture, 111: 171-188. EI-Gamal, ARA, Davis, K.B., Jenkins, J.A. and Torrans, E.L. 1999. lnduction of triploidy and tetraploidy in Nile tilapia Oreochromis ni/oticus L. J. World. Aqua. Soc. 30(2): 269-275. Elghobashy, H. A., Rahman, A., Gamal, A. E., Khater, A. M. 2000. Growth evaluation of four local Strains of Nile tilapia (Oreochromis niloticus) under different farming conditions in egypt. In Proceedings from the Fifth lnternational Symposium on Tilapia Aquaculture, (Eds. Fitzsimmons, K., Filho, J.C.). Rio de Janeiro, Brazil. Departmento de Pesco e Aquicultura do Ministerio da Agricultura): 346-351. FAO Fisheries Department, Fishery lnformation, Data and Statistics Unit. 2000. Fishstat Plus: Universal software for fishery statistical time series. Version 2.3. 2000. FAO Fisheries Department, Fishery lnformation, Data and Statistics Unit. 2001. Fishstat Plus: Universal software for fishery statistical time series. Version 2.3. 2001. 6to. Simposio Centroamericano de Acuacultura 31 Fitzsimmons, K. (2000). Tilapia: The Most lmportant Aquaculture Species of the 21st Century. In Proceedings from the Fifth lnternational Symposium on Tilapia Aquaculture, . September 2000 Fitzsimmons, K., Filho, J.C., (eds.) American Tilapia Association and ICLARM, Rio de Janeiro, Brazil pp. 3-8. Froese, R. and Pauly, D. (Eds). 2000. FishBase 2000: concepts, design and data sources. ICLARM, Los Baños, Laguna, Philippines. 344 p. Hórstgen-Schwark, G., Langholz, H.-J. 1998. Prospects of selecting for late maturity in tilapia (Oreochromis niloticus) : 111. A selection experiment under laboratory conditions. Aquaculture, 167: 123-133. Huang, C.-M. and Liao, l.-C. 1990. Response to mass selection for growth rate in Oreochromis niloticus. Aquaculture, 85: 199-205. Hulata, G., Wohlfarth, G.W. and Halevy, A. 1986. Mass selection for growth rate in the Nile tilapia (Oreochromis niloticus). Aquaculture, 57: 177- 184. Hussain, M.G., Penman, D.J., and McAndrew, B.J. 1998. Production of heterozygous and homozygous clones in Nile tilapia. Aquaculture international, 6: 197-205. lndig, F.E. and Moav, B. 1988. A prokaryotic gene is expressed in fish cells and persists in tilapia embryos following microinjection through the micropyle. Colloq. lnst Natl. Rech. Agron. 44: 221-225. Jarimopas, P. 1990. Realized response of Thai red tilapia to 5 generation of size-specific selection for growth. In: Hirano, R., Hanyu, l. (Eds.), The Second Asian Fisheries Forum (Proceedings of The Second Asian Fisheries Forum,Tokyo,Japan, 17-22 April 1989). Asian Fisheries Sor.iety, Manila, Philippines, pp. 519-522. JenneckE...1f J., Muller- Belecke, A, Hórstgen- Schwark, G. and Meyer, J.N. 199S roof of the successful development of Nile tilapia (Oreochromis niloticus) clones by DNA fingerprinting. Aquaculture173: 377-388 Kocher, T.D., Lee, W.-J., Sobolewska, H., Penman, D.J. and McAndrew, B. 1998. A genetic linkage map of a cichlid fish, the tilapia (Oreochromis niloticus). Genetics 148: 1225-1232. Lee, W. J. and Kocher, T. D. 1996. Microsatellite DNA markers for genetic mapping in Oreochromis niloticus, Journal of Fish Biology 49: 169-171. Lovshin, L.L. 1982. Tilapia Hybridization. In: Pullin, R.S.V. and Lowe- McConnell, R.H. (Eds.), The Biology and Culture of Tilapias. ICLARM Conference Proceedings. lnternational Center for Living Aquatic Resources Management, Manila, Philippines, pp. 279-308. Macaranas, J.M., Taniguchi, N., Pante, M.J.R., Capili, J.B. and Pullin, R.S.V. 1986. Electrophoretic evidence for extensive hybrid gene introgression into commercial Oreochromis niloticus (L.) stocks in the Philippines, Aquaculture and Fisheries Management 17: 249-258. Maclean , N., lyengar, A. Rahman, M.A., Sulaiman, Z. and Penman, D.J. 1992. Transgene transmission and expression in rainbow trout and tilapia. Mol. Mar. Biol. Biotech. 1: 355-365. Mair, G.C. 1993. Chromosome-set manipulation in tilapia-techniques, problems and prospects. In: Gall, G.A.E., Chen, H. (Eds.), Genetics in Aquaculture IV, Proceedings of the Fourth lnternational Symposium on Genetics in Aquaculture. Elsevier Science Publishers B. V., Amsterdam, Netherlands., pp. 227 - 244. 32 Tegucigalpa, Honduras Mair, G.C. (in preparation) Strategies for dissemination of improved fish breeds with a focus on small scale fish farmers. NAGA, ICLARM Quarterly. Mair, G.C., Abucay, J.S., Skibinski, D.O.F., Abella, T.A. and Beardmore, J.A 1997. Gene tic manipulation of sex ratio for the large-scale production of all-male tilapia, Oreochromis niloticus. Canadian Journal of Fisheries and Aquatic Science, 54: 396-404. Mair, G.C., Beardmore, J.A. and Skibinski, D.O.F. 1990. Experimental evidence for environmental sex determination in Oreochromis species. In: Hirano, R., Hanyu, l. (Eds.), The Second Asian Fisheries Forum. Asian Fisheries Society, Manila, Philippines, pp. 555-558. Mair, G.C., Scott, AG., Penman, D.J., Beardmore, J.A. and Skibinski, D.O.F. 1991 a. Sex determination in the genus Oreochromis 1. Sex reversa!, gynogenesis and triploidy in O. niloticus (L). Theoretical and Applied Genetics, 82: 144-152. Mair, G.C., Scott, AG., Penman, D.J., Skibinski, D.O.F. and Beardmore, J.A. 1991 b. Sex determination in the genus Oreochromis 2. Sex reversa!, hybridisation, gynogenesis and triploidy in O. aureus Steindachner. Theoretical and Applied Genetics, 82: 153-160. Martinez, R, Arenal, A, Estrada, M.P., Herrera, F., Huerta, V., Vazquez, J. Sanchez, T, and dela Fuente, J.N.A 1999. Mendelian transmission, transgene dosage and growth phenotype in transgenic tilapia (Oreochromis hornorum) showing ectopic expression of homologous growth hormone Aquaculture 173: 271-283 McAndrew, B.J. and Majumdar, K.C. 1988. Growth studies in juvenile tilapia using pure species, hormone treated and nine interspecific hybrids. Aquaculture and Fisheries Management 20: 35-47 McAndrew, B.J., Roubal, F.R., Roberts, R.J., Bullock, A.M. and McEwan, M. 1988. The genetics and histology of red, blond and associated colour . variants in Oreochromis niloticus, Genetika 76: 127-137. McAndrew and Wohfarth (in press) Qualitative Phenotypes - Colour Varieties. In Applied Genetics of Tilapias (Eds. Mair, G.C., Pullin, R.S.V and Hulata, G.). ICLARM Studies and Reviews. Melard, C., 1995. Production of a high percentage of male offspring with 17a­ethynylestradiol sex-reversed Oreochromis aureus. l. Estrogen sex­reversal and production of F2 pseudofemales. Aquaculture, 130: 25-34. Oldorf, W., Kronert, U., Balarin, J., Haller, R., Hórstgen-Schwark, G. and Langholz, H.-J. 1989. Prospects of selecting for late maturity in tilapia (Oreochromis niloticus) 11. Strain comparisons under laboratory and field conditions. Aquaculture, 77: 123-133. Pullin, R.S.V. (ed.) 1998. Tilapia Genetic Resources for Aquaculture. ICLARM Conf. Proc. 16, ICLARM, Manila, Philippines. p108. Pullin, R.S.V. and Capili, J.B. 1988. Genetic improvement of tilapias: problems and prospects. In: Pullin, R.S.V., Bhukaswan, T., Tonguthai, K., Maclean, J.L. (Eds.), The Second lnternational Symposium on Tilapia in Aquaculture. ICLARM Conference Proceedings. Department of Fisheries, Thailand and lnternational Center for Living Aquatic Resources Management, Bangkok, Thailand and Manila, Philippines, pp. 259-266. 6to . Simposio Centroamericano de Acuacultura 33 Pullin, R.S.V., Casal , C.M., Abban , E.K. and Falk, T.M. (Eds.) 1997. Characterization of Ghanaian tilapia genetic resources for use in fisheries and aquaculture . ICLARM Conf. Proc. ICLARM, Manila, Philippines., 58 pp. Pullin , R.S.V., Macaranas, J.M. and Taniguchi, N. 1986. Genetic resources for tilapia culture . Aquaculture, 57: 373. Rahman, M.A. , lyengar, A and Maclean, N. 1997. Co-injection strategy improves integration efficiency of a growth hormone gene construct, resulting in lines of transgenic tilapia (Oreochromis niloticus) expressing an exogenous growth hormone gene. Transgenic research . 6: 369-378. Rahman, M.A, and Maclean, N. 1999. Growth performance of transgenic tilapia containing an exogenous piscine growth hormone gene. Aquaculture 173: 333-346. Rahman, M.A. Ronyai , A , Engidaw, B.Z., Jauncey, K., Hwang, G-L. , Smith , A , Roderick, E.E. , Penman, D.J . Varadi , L. and Maclean, N. (in press). Growth and nutritional trials on transgenic Nile tilapia containing an exogenous fish growth hormone gene. J. Fish. Biol. Rana , K.J., McAndrew, B.J., Wohlfarth , G., and Macgowan, l. 1996. Observations on intergeneric hybrids in tilapias. In: Pullin , R.S.V., Lazard , J., Legendre, M., Kothias, J. B.A., Pauly, D. (Eds.) , The Third lnternational Symposium on Tilapia in Aquaculture. ICLARM Conf. Proc. 41 . lnternational Center for Living Aquatic Resources Management (ICLARM), Manila, Philippines., pp. 391 - 397. Romana-Eguia, M.R. and Doyle , R.W. 1992. Genotype-environment interaction in the response of three strains of Nile tilapia to poor nutrition, Aquaculture 108: 1-12. Schwartz, F.J. 1983. "Tilapia" hybrids: problems, value, use and world literature. In: Fishelson, L. , Yaron , Z. (Eds.) , Proceedings of the lnternational Symposium on Tilapia in Aquaculture, Nazareth, Israel. Tel Aviv University, Tel Aviv, Israel , pp. 611-622. Scott, AG., Mair, G.C., Skibinski, D.O.F. and Beardmore, J.A. 1987. 'Biond': a useful new genetic marker in the tilapia Oreochromis niloticus (L.) , Aquaculture and Fisheries Management 18, 159-165. Tave , D. 1988. Genetics and breeding of tilapia: a review. In: Pullin , R.S.V., Bhukaswan , T., Tonguthai , K. , Maclean, J.L. (Eds.), The Second lnternational Symposium on Tilapia in Aquaculture. ICLARM Conference Proceedings. Department of Fisheries, Thailand and lnternational Center for Living Aquatic Resources Management, Bangkok, Thailand and Manila, Philippines, pp. 285-293. Tave, D. 1992. Genetics for Fish Hatchery Managers. Chapman and Hall. 336p Teichert-Coddington , D.R. and Smitherman, R.O. 1988. Lack of response by Tilapia nilotica to mass selection for rapid early growth. Trans. Am. Fish. Soc .. 117(3): 297-300. Trewavas, E. 1983. Tilapiine fishes of the genera Sarotherodon, Oreochromis and Danakilia. Trustees of the British Museum, London, 583 pp. Trombka, D. and Avtalion, R. 1993. Sex determination in tilapia - a review. The Israelí Journal of Aquaculture-Bamidgeh , 45(1 ): 26-37. 34 Tegucigalpa, Honduras Uraiwan, S. 1988. Direct and indirect responses to selection for age at first maturation of Oreochromis niloticus. In: Pullin, R.S.V., Bhukaswan, T., Tonguthai, K., Maclean, J.L. (Eds.), The Second lnternational Symposium on Tilapia in Aquaculture. ICLARM Conference Proceedings. Department of Fisheries, Thalland and lnternational Center for Living Aquatic Resources Management, Bangkok, Thailand and Manila, Philippines, pp. 295-300. Welcomme, R. L. 1988. lnternational lntroductions of lnland Aquatic Species. FAO Fisheries Technical Paper 294. FAO, Rome, ltaly. 318p Wohlfarth, G.W. 1994. The unexploited potential of tilapia hybrids in aquaculture, Aquaculture and Fisheries Management 25, 781-788. Wohlfarth, G.W. Rothbard, S., Hulata, G. and Szweigman, D. 1990. lnheritance of red body colouration in Taiwanese tilapia and in Oreochromis mossambicus. Aquaculture 84: 219-234. 6to. Simposio Centroamericano de Acuacultura Sex Reversal: the directed control of gonadal development in tilapia Ronald P. Phelps Department of Fisheries and Allied Aquacultures Auburn University, Auburn, AL 36849 Abstract 35 Tilapia are becoming the most widely produced species of freshwater fish in the world. They can be produced in a variety of settings using a range of nutrient inputs. Males are the preferred sex to culture as they grow faster and divert less energy into reproduction. Males can be obtained using a variety of procedures but the most practica! is through controlling gonadal development. Recently hatched tilapia fry have gonads that have not differentiated into ovaries or testes. lt is possible to give such fish an exogenous source of hormone (androgen or estrogen) to control the development of the gonad. Fry less than 12 mm long can be harvested by seining along the edge of a spawning pond or from specialized spawning ponds where the pond is drained and harvested after 16-21 days. Proper size fish can also be obtained through a more intensive management approach where eggs are collected from the mouths of incubating females. Most commonly used approach to obtain male tilapia is to feed fry for 28 days or less a feed containing the androgen methyltestosterone. When fed properly the frequency of females in the population can be reduced to less than 5%. The short treatment duration very early in the fish's lite history and rapid metabolism of metyltestosterone helps insure that tilapia are free of MT befare fish reach the consumer. The production techniques associated with sex reversa! are efficient and stra\gnt forward enough so that sex reversa! has become the commercial procedure of choice to produce male tilapia fingerlings and has been a significant factor in the rapid growth of the tilapia industry lntroduction An aquaculturist is always looking as to how a product can be produced more economically and brought to the market sooner. The growth rate of the animal being cultured is one of the primary factors affecting the costs of production and how soon it reaches market size. Tilapia are a fast growing fish capable of going from egg to 1 kilo in a year under optimum conditions. Tilapia respond well to a variety of management practices and nutrient inputs making them a popular fish to culture. However, one of the primary disadvantages of tilapia is that they reach maturity early and are capable of reproducing befare reaching a marketable size. Most species of tilapia under favorable growth conditions will reach maturity within 6 to 8 months of birth. When mature fish are present in ponds they will reproduce and expend energy on reproduction that otherwise could be directed to growth. 36 Tegucigalpa, Honduras They will compete with their offspring for food, resulting in less food available, slower growth and typically unmarketable fish. Unless reproduction is controlled more than 75% of the fish biomass may be too small for public acceptance. Growth rates vary among species of fish and even among the sexes of the same species. The sexual dimorphism in growth can be significant. In Nile tilapia the growth rate of males continues to accelerate after they are more than 100 g. in weight. The growth rate of females often begins to slow once sexually maturity is reaches and they start to reproduce (Figure 1 ). Female tilapia are diverting energy from growth into egg development, forming new eggs to replace those lost each time it reproduced. When reproducing, feed intake by Nile tilapia- is limited due to the oral incubation of eggs. This diversion of energy to reproduction and limited feed intake can result in males being twice the size of females of the same age. For commercial tilapia culture, the issue is not only how to control reproduction during the production phase but also how to have only males for production. Adding a predator can control reproduction, but that does little to address the issue of the slower growth of female tilapia. To obtain maximum growth males are needed. Males can be hand selected but this is an inefficient system prone to errors where half the fingerling production has to be discarded. The genetics of tilapia are such that selective crosses using a male of one species crossed with a female of another has resulted in all male hybrids. This approach was at one time was a common method to produce males but was largely replaced by the mid- 1980's. Difficulties in maintaining two pure lines of brood fish and keeping them separate, and the space required contributed to the decline. In addition, the genetics of sex does not appear to be as straightforward as once thought, often less than 100% m ale progeny would be obtained in certain crosses e ven if pure lines were maintained. The most widely used technique to obtain male tilapia is commonly referred to as sex reversa l. However this term requires explanation and might be better described as directing of gonadal development. The term sex reversa! as used in this paper refers to the addition of exogenous steroids to override the genetically based control of gonadal development to produce a fish that is functionally the desired sex without altering the genetics of the fish. The production techniques associated with sex reversa! are efficient and straight forward enough so that sex reversa! has become the commercial procedure of choice to produce male tilapia fingerlings and has been a significant factor in the rapid growth of the tilapia industry. Why Sex Reversa! works The sex of fish is not permanently set at hatch and can be altered by a number of factors. At hatch distinct ovarian or testicular tissues are not present. The gonads develop from primordial germ cells with female differentiation occurring befare male differentiation. The point in time when differentiation begins differs among the different fish species. In tilapia and trout this is early in the life history while in grass carp and paddlefish it is months later. In newly hatched O. niloticus (Aivendia­Casauay and Carino 1988) and O. mossambicus (Nakamura and Takahashi 1985) primordial germ cells are found at the dorsal root of developing mesentery in the mesoderm, ventral to the gut and in the endoderm cells of gut. The germ cells eventually migrate to the gonadal region. Paired gonadal analgens are observed 9 to 10 days post hatching. The appearance of ovocoel and testocoel, indications of sex differentiation to female and male takes place at 16 to 20 days post hatching in O. mossambicus (Nakamura and Takahashi 1985) and perhaps as late as 30 to 33 days post hatching in O. níloticus (Aivendia-Casauay and Carino 1988). Hines et al (1999) found that in the first 7 days post fertilization when no gonadal structures are yet present, the level of androgen in tilapia begins to decline and when indifferent germ cells appear androgen and estrogen levels are lowest. From day-15 to 29 days post-fertilization, they found visible but undifferentiated gonadal tissue. lt is in this period when tilapia are vulnerable to exogenous steroids. We are fortunate in that tilapia fry are actively feeding at that point and an exogenous steroid can be added to the diet to influence gonadal development. lf tilapia are given exogenous steroids at the proper concentration and frequency from befare the start of gonadal differentiation through when it is complete, this will override the genetic control of gonadal differentiation and monosex populations can be obtained. The exact mechanisms and chemical pathways that control gonadal development in tilapia are not clear. Natural steroid production is not evident in tilapia until differentiation begins (Baroiller et al., 1988) suggesting that some precursor or other compound is involved in directing gonadal differentiation. However, from the numerous studies with tilapia, it is clear that, if exogenous steroids are given befare the start of gonadal differentiation and administered past differentiation, it is possible to alter the sex ratio. An understanding of the mechanics of exogenous direction of gonadal development in tilapia is further complicated by the success that has been obtained by short-term immersions in steroid solutions even though the treatment ended well befare gonadal differentiation is completed (Contreras et al 1997). Types of Steroids Steroids are a group of lipids with severa! unique properties affecting animal growth and development. Steroids are called androgens if they are able to induce male characteristics and estrogens if they induce female characteristics. Androgens have two physiological actions: (1) androgenic activity, promoting the development of male sex characteristics and (2) anabolic activity, stimulating protein biosynthesis. Androgens can be classified into two groups: androstane derivatives, having both androgenic and anabolic properties, and 19-nor­ androstane derivatives that have anabolic properties but only weakly androgenic ones (Camerino and Sciaky, 1975). From a sex reversa! perspective, androstane derivatives are of more value because of their potential to direct the sexual development of fish into males. When evaluating a steroid for sex reversa! by oral administration, three main criteria for selection should be considered: metabolic half-life, androgenic or estrogenic strength and solubility in water. Testosterone is the principal androgen secreted by the testis and the main androgenic steroid in the plasma of human males (Murad and Haynes, 1985). lt is often used as the standard to evaluate the androgenic properties of a steroid. lt is ineffective when given orally and has a short duration when given by injection due to rapid hepatic metabolism. Synthetic androgens are preferred over natural ones because some can be administered orally and withstand catabolism in the gut. The chemical structure, bonds and attached groups determine the effectiveness (Brueggemeier, 1986). lntroduction of a 3-ketone function ora 3a-OH group or reduction of the 4,5- double bond enhances androgenic activity. Alkylation of the 17a-position or the 1 a­ position allows for oral activity. Masculinization A number of synthetic androgens either applied as a bath or a feed additive have altered the sex ratio of tilapia. Clemens and lnslee (1968) produced all male populations of Oreochromis mossambicus incorporating 17 -a methyltestosterone into the diet at 1 O to 40 mg/kg. Methyltestosterone (MT) has sin ce become the most commonly used synthetic androgen to alter the sex ratio of fish. lt has proven to be effective in a number of different species of tilapia and under a variety of management scenarios. Other synthetic androgens have been incorporated into the diet of tilapia for sex reversa! are given in Table 1. A less widely used approach for producing male populations is through the use of non-steroidal compounds that interfere with steroid binding or metabolism. In the sequence of events associated with gonadal differentiation endogenous androgens are aromatized into estrogen by an aromatase enzyme. lt is possible to block this action by the addition of aromatase inhibitor. Kwon et al (2000) treated a genetically female population of O. ni!oticus fry with the aromatase inhibitor Fadrozole at 200 to 500 mg/kg diet and obtained 92.5 to 96.0% males. Guiguen et al (1999) was able to skew the sex ratio of an all-female O. niloticus population to 75.3% male using the aromatase inhibitor 1 ,4, 6-androstatriene-3-17-dione at 150 mg/kg of diet. Blocking of estrogen binding sites is another approach to production of males. Tamoxifen, an anti-estrogen when given as a feed additive to tilapia at 100 mg/kg of diet produced an all-male population (Hines and Watts, 1995). Feminization Female tilapia are not preferred for culture but feminization of genetically male Nile tilapia O. niloticus offers the possibility of all male tilapia through a YY breeding program. Likewise of interest is the feminization of a homogametic male O. aureus to produce functional females for mating with normal male O. aureus to produce all male offspring. Estrogens are those agents that induce feminization. Estrone and 17CJ-estradiol are two natural steroidal estrogens found in the ovary of tilapia (Katz et al. 1971 ). Synthetic estrogens are more potent than natural estrogens when given orally. This greater activity is due to their stability in the digestive tract and the liver (White et al. 1973). The most commonly used synthetic estrogens for sex reversa! are the non-steroidal estrogens, ethynylestradiol (EE) and diethylstibestrol (DES). DES is more potent and once was used as a growth promotant in livestock until banded by the U.S. Food and Drug Administration in 1979. Both are carcinogens. The effectiveness of DES and EE to feminize may be dependent on the species of tilapia and the management conditions. Hopkins et al. (1979) fed 100 mg DES/kg diet to O. aureus fry for 5 weeks and produced 64% females. Rosentein and Hulata (1993) obtained 98% and 100% females in two sets of O. aureus fed DES at 100 mg/kg for 30 days. Scott et al. (1989) fed two sets of genetically all male O. 6to. Simposio Centroamericano de Acuacultura 39 niloticus fry DES at 100 mg/kg and obtained 52% females in one set and 84% in the other. EE was given at 100 mg/kg to O. aureus for 40 days by Melard (1995) to obtain a 94% female population . Potts and Phelps (1995) fed O. niloticus fry EE at 100 mg/kg and obtained a 65% female population . Toxicity is an issue in estrogen treatments. Eckstein and Spira (1965) reported high mortality of O. aureus fry when given stilbestrol diphosphate baths at 400 to 1 000 ug/L. Fry Prod uction For successful sex reversa! it is critica! that the treatment begin with fish of an age where gonadal differentiation has not begun . Recently hatched fry less than 12 mm in total length are needed. They can be obtained by seining along the edge of spawning pond in the early morning for fry that tend to gather along the edge of a pond or tank. Large quantities of fry can be obtained from specialized spawning ponds where the pond is drained and harvested after 16-21 days. Proper size fish can also be obtained through a more intensive management approach where eggs are co\lected from the mouths of incubating females. The eggs are incubated in a hatchery and a more uniform age and size of fry obtained . The selection of a fry production technique is influenced by various factors including the number of fry needed at one time, labor availability, water resources, and facilities availability. Partial harvests Earthen ponds are stocked with up to 2000 to 3000 kg of brooders/ha at a sex ratio 2-3 females/male. Brooders are fed during the spawning period at approximately 1 to 2% body weight per day and the pond may be fertilized . Sorne brooders spawn within a few days after stocking and swimup fry can be expected witiÍin 1 O to 15 days after brood stocking. With once-weekly fry collections , Verdegem and McGinty (1989) obtained an average of 153,100 fry/ha per week (2 .2 fry/m2/d) over a 116-day period. Little (1989) averaged 2.5 sex reversible size fry/m2/d from ponds stocked with O. niloticus harvested every 5 days, six times/day, and 1.5 fry/m2/d of sex reversible size fry when harvested three times/day every 5 days. How long a spawning pond can be kept in production depends on how successful partial harvests are. ldeally all fry are harvested before they reacli a larger size. Those that escape capture soon prey on subsequent swim-up frY. Macintosh and De Silva (1984) found that even within fry of the same age, cannibalism contributed up to 35% of total fry mortality. Sorne fry will escape even with careful seining resulting in a progressive decrease in fry harvested due to cannibalism . Little (1989) found that the number of oversize fry harvested could be kept to a mínimum (0.015 fry/m2/d) if the pond was harvested six times/d every five days. Such careful seining is often not practica! and it is best to not leave a spawning pond in productiori for more than 8-1 O weeks befo re making a complete harvest. Partial harvesting of ponds to produce tilapia fry may be accepta.ble for locations where the production season is year-round and large quantities are not required at any one time. Fry yields from a pond are variable day-to-day therefore severa! ponds are needed to produce a constant production of fry. The technique is labor intensive but does not require highly skilled labor. 40 Tegucigalpa, Honduras Complete harvest of fry from ponds Complete harvest of fry can be made in a spawning pond with a catch basin or from a fine mesh net enclosure (hapa) stocked with brood fish. Spawning ponds are generally no larger than 2000 m2 and are designed to drain completely into a catch basin that is 1 O m2 or larger (at least 1% of pond area). The catch basin should be 30 to 40 cm deep with a firm bottom, ideally concrete. The spawning pond is prepared by lining the catch basin with large mesh netting that is about 20% larger than the catch basin. This net is used to remove brooders from the pond at harvest without removing the fry. The pond must be completely dry befare restocking or if puddles remain they are poisoned with chlorine or other toxicant to insure no fry remain from the last production cycle to cannibalize fry produced in the subsequent cycle. Tilapia fry can remain alive in small puddles for days if a special effort is not m a de to eliminate them. Brooders are stocked at a sex ratio of 1 male: 1.5 to 2 females, adding a total weight of fish up to 5000 kg/ha. Brooders are fed at approximately 1% body weight/day and the pond is not usually fertilized. The fish are allowed to spawn over a 2 to 4 week period befare the pond is harvested. The timing of the harvest is important to achieve maximum fry yields. Not all females will spawn at the same time but there will be a peak in the ·spawning activity and a point in time where there is a maximum number of fish of the desired size. lf a pond is harvested too soon, part of the reproduction will be eggs or sac fry that are generally lost when the brooders are removed. lf the pond is harvested too late, a portian of the fish will have started gonadal differentiation and cannot be sex reversed effectively. Green and Teichert-Coddington (1993) developed an equation to time the fry harvest to obtain the maximum number of fry of a sex reversible size. They found that a 195 to 220 degree-day (temperatureoc x no. days) period was optimum for best production of fry suitable for sex reversa l. Once the appropriate degree-days ha ve be en reached, fry are harvested by draining the pond into the catch basin early in the day. A screen with a fine mesh and large surface area is placed over the drain to prevent fry from being lost or imp.inged. Popma and Green (1990) recommend approximately 0.5 to 0.8 m2 of screen area to drain a 500-m2 pond over a 5 to 1 Oh period. Brooders are removed from the catch basin by lifting the netting previously placed in the basin. The brooders may be placed directly into another spawning pond or be separated by sex and be held for a few days in a recovery tan k. Fry are captured from the catch basin using fine mesh hand nets. lt is important to be organized and efficient during fry collection. Dissolved oxygen concentration in the catch basin often declines rapidly bringing the fry to the surface. Adequate labor should be on hand to catch all the fry and move them into fresh water in a few minutes. Tilapia fry are not as hardy as adults and extra care is needed to insure that healthy fry are harvested. Special care should be taken to prevent excessive turbidity in the catch basin. Fry should not be held in collecting buckets for more that a few minutes befare transfer into clean water. 6to. Simposio Centroamericano de Acuacultura 41 By following such a degree-hour guide, Green and Teichert-Coddington (1992) obtained 1,500 to 2,500 sex reversible size fry/kg of female brooder stocked with only a mínimum of over sized fish. The spawning pond was then prepared again and a new cycle of fry production begun. By making complete harvests and scheduling the timing of the harvest, it is possible to obtain 7.5 to 10 fry/m2/d, not counting down time between cycles. This method has the advantage of producing large numbers of fry at one time and giving the opportunity to rotate sets of brood stock. However, this approach requires an adequate water supply so the ponds can be drained and refilled frequently, and if the fry are not handled carefully during harvest a high mortality may occur. This approach also requires that the pond be held out of production for a period to insure no fry remain from the previous production cycle. Spawning in Net Enclosures The use of fine-mesh cages or net enclosures (hapas) is another alternative for producing fry for sex reversa!. Hapas have the advantage in that they can be placed in existing bodies of water where other fish species are present and do not require that the pond be drained befare the fry can be harvested. The downtime between reproductive cycles is mínimum. A complete harvest of spawning hapas also allows the collection of eggs or sac fry that may have been lost using techniques discussed earlier. Spawning hapas are typically rectangular in shape, ranging in size from 2 to >500 m2 and are constructed with 1.6 mm mesh netting (Figure. 2). The hapas are designed to allow the fish to be crowded to one end for collection. Once crowded together brooders can be removed and females examined for eggs or sac fry and any free-swimming fry in the hapa can be removed. Brooders are generally stocked at one male: <2 females ata density of 4 to 5 fish/m 2 of hapa or 0.2 to 0.6 kg/m2 of hapa. Sex reversa! is most successful when the initial age and size of fry being treated is tightly controlled. One advantage of hapas is that they can conveniently be harvested every 5 to 1 O days to obtain fertilized eggs. By using hapas, females can be collected with a mínimum of disturbance and each fish can be examined to determine which one is holding eggs. The eggs are rinsed from the mouth and the female returned to the hapa to spawn or placed in a conditioning hapa. As eggs are found in the mouth, the approximate age can be estimated by their color. Younger eggs are light yellow and older eggs a dark orange or brown. As the eggs are collected those of similar age can be pooled for incubation. Using an incubator system as described by Maclntosh and Little (1995) the sinking eggs of tilapia can be rolled vigorously in a round-bottomed incubator with a downward flow. A high hatch rate can be expected when older eggs are collected and incubated, younger eggs are more difficult to incubate. Fry that are collected right after they swim up and out of the incubator are ideal for sex reversa!. They are young and of a uniform size. 42 Tegucigalpa, Honduras Seed production/hapa can be improved particularly when the spawning units are harvested frequently and brood stock are replaced each cycle. The advantage of brood stock rotation is that the reproductive cycle of the brood females is more synchronized, permitting a higher percentage of females to spawn during the next cycle. Two or three sets of female brooders are maintained, one actively spawning and another one or two sets where the females have been separated from the males and are being fed to recover lost energy associated with spawning or from any physical damage. When harvesting every 1 O days without brood stock replacement, seed production averaged 106 seed/kg female/d but with female replacement increased seed production to 274 seed/kg/d. (Little et al. 1993). Broodstock replacement can double seed production, but this practice is more labor-intensive and requires additional facilities for brood stock maintenance. A disadvantage of weekly seed collection is that incubation facilities are needed, but short production cycles reduce fouling of nets (if air dried for a couple days between cycles), increase fry production per female brooder, and give uniformly smaller/younger fry. Extended spawning cycles of 21 days for fry p