GENETIC VARIATION IN CULTURED STOCKS OF TIGER SHRIMP ( Penoeus monodon ) IN INDONESIA

Three stocks of tiger shrimps, Penaeus monndon, obtained from brackish water pond culture in Aceh (Sumatera Island), Cilacap (Java Island) and Sumbawa (West Nusatenggara) were assayed for allozyme variation at 9 enzyme loci from muscle biopsies.


INTRODUCTION
Tiger shrimp, Penaeus nt.onodon, is an important commercial shrimp species in Indonesia.Although most of the annual domestic yield of penaeid shrimp comes from the harvest of wild shrimp, aquaculture industries of this species have been rapidly growing.Oonsequently, the need of fry as a vital production component has also been increasing.
At the beginning, the supply of fry relied on natural capture whose availability was seasonal.
Fortunately, developments in the sciences and technologies, especially in reproduction, have allowed controlled mass production of fry in hatcheries (Primavera, 1978).Supply of brood.stocks for mass production, however, still relies on natural population.Hatchery owners prefer to use natural broodstocks because they are larger, more fecund and produce more viable nauplii (Sugama et al., 1988).Realizing the importance of natural broodstock population and as a starting point toward genetic improvement, Sugama e, ol. (1998)have studied the distribution of genetic variability and population structure of some natural penaeid shrimp populations in Indonesia.
Although up to now supply ofbroodstocks for fry mass production is stillpredominantly derived from natural sources, efforts of supplying broodstocks from rearing ponds should not be ignored.
The ability of doing this is very useful in antici- pating the decreasing supply of natural brmdstocks caused by fishing pressure or by uncontrolled exploitation, and to keep the availability of broodstocks in appropriate quantity, quality and continuity.It is also useful in providing greater possibility to carry out genetic improvement.The weakness of cultured broodstocks especially in genetic point of view is the possibility of reduction in genetic variation due to both inbreeding and genetic drift.A multidiscipline approach is needed to avoid any loss of genetic variation.
Genetic variability data are useful as starting point for genetic improvement.In addition to genetic variability data of natural population, information on those of cultured stocks will be of great assistance in determining breeding pro€fram and genetic improvement strategies.

MATERIALS AND METHODS
Three samples of cultured stocks of tiger shrimp obtained from rearing ponds in Langsa @ast Aceh), Indramayu @est Java) and Negara @ali) were used @igure 1).The parents of the three stocks were natural broodstocks from Aceh's, Oilacap's, and  Horizontal starch gel electrophoresis was used following the procedures of Taniguchi & Sugama (1990).Seven enzyme systems were used to suwey genetic variation.These enzymes systems were: a-glycerophosphate dehydrogenase (a-GPD, E. C. Locus and allele nomenclature used were as described by Shaklee et ol. (1990).Multiple loci encoding I enzyme separated by a hyphen G) at the back of loeus name, indicating relative migration of those loci.The fastest anodal loeus was designated locus 1. Allele nomenclature was based on migration distance allele against the common allele, usually designated by allele 100.
The data obtained were used to calculate some parameters of genetic structure of population including degree of polymorphism (Leary & Booke, 1990), average heterozygosity, genetic distance (Nei,19?8), genetic similarity or dissimilarity (Nei, 1972) and genetie variation among populations (Ilartl, 1980).Calculation of degree of polymorphism and average hetero- zygosity was based on the assumption that the number of enzymes and detectable loci were the same as obtained by Sugama et ol. (1996).This was aimed to eliminate bias in comparing with genetic variation of natural population.Chisquare test was used to detect both the deviation of genotype frequency from Hardy-Weinberg equilibrium and heterogeneity of allele frequ- encies among stocks.Grouping of stocks based on similarity or dissimilarity of allelefrequencies was carried out by cluster analysis and presented in dendrogram using Unweighted Pair Group Method with Arithmetic averages (UPGMA) method (Sokal & Rohlf, 1981).

RESULTS
Electrophoretic analysis results as expressed by banding patterns of zymograms showed that the seven enzymes used in this study detected 9 Ioci, with three of them were polymorphic at least in one of the samples, namely: Idh, Gpi, and Pgm. 1.The observed phenotype, Hardy-Weinberg expectation, and allelie frequencies of the poly- morphic loci in the three stocks are presented in

T2
Genotype frequencies of the three polymorphic loci (Iable 1) showed that the genotype proportion was in Hardy-Weinberg equilibrium (P<0.05).There were allele frequency differences in the three stoeks especially in ldh-100 and Gpi'100 alleles (table 2).Allele frequencies of Idh.100 were significantly different between the Aceh and the Sumbawa stocks and between the Cilacap and the Sumbawa stocks.Allele frequencies of (]pi'100 were significantly different between each of paired stocks.There were differences in degree of poly' morphism (Iable 3).Degree of polymorphism in the Aceh and Oilacap stoeks (14.29%) were higher than that of the Sumbawa stock (4.76%).The only polymorphic locus in the Sumbawa stock was Pgm' l.The same phenomenon was also seen in the average of number of alleles per locus, where that of Sumbawa stock (1.05) was slightly lower than those of the Aceh and Oilacap stocks (1.19).
Degrees of polymorphism of all three stocks in this study were lower than those of the natural populations (Iable 4).Many of the loci, which were polymorphic in natural population, became mono' morphic in cultured stocks.In natural population of the Aceh and Sumbawa waters, for instance, 6 polymorphie loci were found, namely: Est, Ldh, Mdh,Idh, Pgm-1and 6-Gpd (Sugama e,taI.,1988)Using the same enzymes but, Est, the Aceh eultured stock showed only three polymorphic loci, namely Pgm-1, Idh, and (lpi.In this study, comparison with the natural population suggested that the cultured stocks had undergone reduction in degree of polymorphism and number of alleles per locus of 67.6% and24.1%,respectively.Heterozygosity per locus (he) and average hete' rozygosity of each stock (fable ;J) indicated that the Gpi contributed most to the number of hetero' zygotes, followed by Pgm-l and Idh in Cilacap stock, While in the Aceh and Sumbawa stocks, the loci contributed most to the number of hetero' zygote consecutively in decreasing order: Pgm'1, Idh and Gpi.Average heterozygosity was highest in the Cilacap stock, followed by the Aceh and Sumbawa stocks.
Index of fixation (Fst) 0.06 among the three stocks (Iable 5) indicated that genetic variability in the cultured stocks of tiger shrimp, 67o, was caused by genetic variability among the stocks, and 94% was caused by genetic variability of individuals within the stock. (]enetic distance (D)  between stocks ranged from 0.011 to 0.02;l with average D 0.017. (ienetic identity (I) ranged from 0.9?? to 0.989 with average I 0.09811 (Iable :]    Gpi (tre)   Pgm-l (he)  Dendrogram showing the genetic relationship among three stocks of cul.tured tiger shrimp using Nei genetic distance (lg?8) as the metric and UPGMA clustering method.These values of genetic variability among the cultured stocks were slightly higher than those of the natural population (Iable 4).Dendrogram of genetic distance between stocks (Figure 2) indicated that there were genetic differentiations in the three stocks.The Aceh stock and Oilacap stock were relatively close compared to the Sumbawa stock.

DISCUSSION
In general, the low genetic variability exhibited by the level of polymorphism, number of alleles per locus and average heterozygosity in the cultured stocks suggested that cultured tiger shrimp have undergone a reduction in genetic variability.Factors that could cause this reduction are the founder effect that can take place through broodstock selection, inbreeding or a bottleneck effect.Since all of the hatcheries used natural broodstocks without selection activities, then it is unlikely that selection process and inbreeding have caused the reduction in genetic variability.Rather, random genetic drift or bottleneck effect caused by a small founding population in the hatcheries is probably the main factor causing the reduetion of genetic variability in the cultured stocks.According to Allendorf & Phelps (1980)   there are three major ehanges in genetic variability of population caused by random genetic drift: reduction in polymorphic loci, number of alleles per locus and average heterozygnsity.
Those changes were clearly seen in this study (fable 4).
The number of broodstocks usrirllly used for fry production in hatcheries is based on the production target.The larger the production target, the larger the number of broodstocks used.
In terms of genetic variability, it is clear that the small number of broodstock used in the hatcheries resulted in the high probability of founder effect random genetic drift, because the number of broodstocks taken as spawners in hatcheries did not fully represent the gene pool of the natural population.The smaller the number of founder population, the larger the probability of genetie drift.These were observed in all of the cultured stocks studied.The loss of some polymorphic loci in the cultured stocks was probably an indicator of the occurrence of random genetic drift associated with the unrepresentative samples.
The loss in genetic variation which might be due to small number of effective parents was also already observed in hatchery stock ofbrown trout @yman & Stahl, 1980) In addition to the small number of broodstocks, it was estimated that fry production management at hatcheries also contributed to the decline of genetic variation in cultured broodstocks.Fry production management practiced by a hatchery in Aceh as will be described below for example, perhaps can explain the low genetic variation in cultured stocks.A hatchery owner uses about forty broodstocks to obtain fry production target of 4 million fry per cycle.Each broodstock is placed in 5-ton conerete tanks which functioned as egg releasing tanks, larva rearing tanks as well as post larva rearing tanks.Generally all of the developmental stages from the hatching to the ready to stock post larua take place in the same tank and there has no mixing process with offspring from another broodstocks during that period.Because of the high fecundity, usually one hundred thousands of ready to stock fry can be producedfrom each broodstock.The fry harvested from that post larva rearing tanks usuallv then are stocked in a growing pond without any mixing with fry from another broodstocks, so the genetic variation of shrimp at that pond comes from just one pair of broodstocks.As a result, although the total numbers of broodstocks used at the hatchery are of significant number, the actual parent of the sample shrimps used in this study might be just one or two broodstocks.Fry production mana.gement sueh as this, of course, can considerably reduce the genetic variation in cultured stocks.
Ryman & Utter (1987) had discussed widely the management praetices of hatchery stocks resulting in the reduction of genetic variation.
Reduction of genetic variability in the cultured stocks in this study was similar to that reported by Taniguchi e.t aI.(l98il) for the hatchery stock of black sea bream, Stahl (l98il) for the hatchery stock of Atlantic salmon, and Sunden & Davis   (1991) for the cultured stock of tiger shrimp.
Genetic variability in the cultured stocks, to some extent reflected genetic variability of the natural population from which the parents of the eultured stocks had been taken.The causes of genetic differentiation among stocks, therefore, can be explained to some extent by factors causing genetic variability differences in the natural 16 population.The low degree of polymorphism in Sumbawa stock, for instance, may be traced to the genetic variability of the natural population of broodstock in Sumbawa waters.The natural population of shrimp in Sumbawa waters, from which the broodstocks of Sumbawa hatcheries were taken, had frequencies of homozygote genotype at major allele (allele-100) which were higher in all of polymorphic loci detected compared to those of the natural population of Aceh waters (Sugama e.t al., 1996)..ds a result, the probability of oecurrence for individuals of homozygote genotype at any locus in Sumbawa waters shrimp chosen as broodstock in hatchery were higher than those of Aceh waters.f)ifferences in allele frequency existing among the three stocks indicated that they did not come from homogenous gene pools.This might be eaused by geographic barrier in reproduction, i.e. individuals tend to mate with those from the same geographic region.ln addition, allele frequency variation was also related to adaptive patterns to the environment.A study carried out by Powers & Place (1978) reported the presence ofan allele cline from the north to the south in Funclulucl he.te.ro c litu s p op u lation.This ph en omenon se em s to be associated with an environmental gradient.
They found the Ldh-B allele in Fun.rlulus he.te.roclituswhich had two primer alleles.varierl in frequency according to the distribution ofthat species.The highest frequency of Ldh-tlb allele was found in the north Atlantic coast and the lowest was in the south Atlantic.This hieh frequeney of allele seerned to be associated with temperature.This allele more often appeared in warm region and rarely in cold region.This prediction was supported by Yardley e.t aI. (1914)  who found that gene frequency varied between lentic and lotic environments in mosquito fish, (iambusia affinis and largemouth bass, Micropterus salm.oides.Whether the previous results can explain the phenomenon emerging in this researeh.however, need further assessment.

Oloseness in genetic distance of Aceh and
Cilacap cultured stocks might be caused by genetic introgression.Many hatcheries in Java use broodstocks from Aceh waters.Likewise in aquaculture industries, many farmers in Java prefer to use tiger shrimp fry from Aceh brcnd- stocks..ds a result, there is a hrgh probability that offspring of Aceh broodstocks, through various ways, have entered Java Island waters and developed to broodstocks, then are chosen as broodstock for hatcheries in Java.
Difference in average heterozygosity (ID might be caused by difference in adaptive patterns to different environments.According to Nelson & Hadgecock (1980), maximum heterozygosity at Gpi locus positively correlated with the waters productivity and trophic instability.Heterozygosity at the Pgm locus was positively correlated with opportunism and negatively with trophic levels.
Average hetorozigosity positively correlated with productivity and trophic stability.The values of heterozygosity ranging from 0.017 to 0.05 were similar or slightly higher than previous results, both in natural population and cultured stocks.Heterozygosity in population of P. monodan of Mexico, Panama and Equador waters were 0.01?, 0.017 and 0.021, respectively, while that of the cultured stocks in Equador was 0.011 (Sunden & Davis, l99l).Basedon Nelson & Hadgecock (1980) on the eorrelation between heterozygosity and waters productivity, it may be predicted that there were differences in waters productivity among localities in Indonesian waters as well as other countries.Whether this conclusion is correct.however, still need more study.
suggested that the three stocks were local popula- tions of the same species (Avise, 1978).Dendro- gram of cultured stocks constructed based on genetic distance showed the same pattern as that from natural population (Sugama et al., 1988).

CONCLUSION
Based on some indicators of biochemical variability, the Cilacap stock had the highest variability followed by the Aceh and Sumbawa stocks.The three stocks were the local populations of the same species.Genetic variability of the cultured tiger shrimp stock had undergone re- duction in degree of polymorphism, number of alleles per loeus and average heterozygosity compared with the natural population.It is recommended to increase the number of brood.
stocks and apply appropriate fry production management strategies to minimize the decline of genetic variability in cultured stocks.

Table 3 .
Summary of genetic variability of cultured tiger shrimp based on 2l enzyme loci and 1% polymorphism criterion.

Table 4 .
()omparisonof average genetic variability between natural population and cultured stock of P, m.onod,on.Both natural population and cultured stock were represented by Aceh and Sumbawa samples.Natural population data were from Sugamaetol.(1996).

Table 5 .
Values of F", based on alleles frequency in the three cultured stocks of P. monodon,