Acknowledgments I thank my first reader, Dr. Lynna Hereford, for her patient guidance throughout the course of both the research and writing stages of this thesis. Your excitement about this project -- and your sense of humor! - have been invaluable. Many thanks to Dr. Carol Boggs, my second reader and my academic advisor, for her comments upon several drafts of this thesis and for offering her expertise in the population genetics field. I am also indebted to Dr. Dennis Hedgecock of UC Davis for providing the allozyme data from his 1979 study and for offering tools and insight to aid in the data analysis. Drs. Jeffrey Mitton and Patrick Gaffney were also influential in the data analysis. Many thanks to Dr. Dennis Powers and his laboratory for providing resources and work space and thanks to many others at the Hopkins Marine Station who assisted me in my time there. I would like to thank Gregory Morris for his contributions as my partner during the spring course at the Hopkins Marine Station. The present work is an extension and completion of our spring research. I have also appreciated the guidance, reassurance, and administrative help of those in the Human Biology office: Heidi Ballard, Jody Kamrowski, and Dr. Ellen Porzig. Finally, I thank my family, friends, and future husband for their love and support. Abstract The decline of natural stocks of the commercially harvested red abalone, Haliotis rufescens, has led to land-based cultivation. To date, there has been little research on the effects of domestication upon the genetic diversity of this species. In this study, protein gel electrophoresis was used to estimate the extent of inbreeding and/or genetic drift within cultured and natural red abalone populations. The present analysis of nine populations indicates that there are significant differences between the genetic compositions of the Santa Barbara 1992 population, believed to be an introduced population, and the natural populations, but only slight differences between the cultured and natural populations. This has implications for the success of the California abalone enhancement programs of 1979-1980. Additional results of this study have relevance for abalone aquaculture. The genetic drift analysis of one of the cultured populations indicates that not all of the abalone used for each hatchery spawn had actually contributed to the production of offspring. Table of Contents ACKNÖWLEDGMENTS ii ABSTRAC III TABLE OF CONTENIS I LISTOF TABLES LIST OF FIGURES VI INTRODUCTION MATERIALS & METHODS 4 Sample Collection 4 Sample Preparation. Horizontal Starch Gel Electrophoresis5 Interpreting Gels0 Data Analysis RESULS Genetic Variation Analysis Genetic Drift Analysis DISCUSSION O CONCLUSION4 Aquâculture14 Enhancement1 TABLES Z0 FIGURES 26 BIBLIÖGRAPH 30 List of Tables Page Table 1. Summary of Sampled Populations.. 20 2. Summary of Allozyme Analysis 21 3. Allele Frequency ChartZZ 4. Summary Statistics of Genetic Variation 23 5. Genetic Drift Analysis Summary 24 6. Comparison of Two Red Abalone Seedings.... 25 List of Figures Page Figure 1. California Red Abalone Landings (1931-1988) 20 2. Map of Collection SitesZ/ 3. D-values by LocusZ0 4. Sources of PMI Female and Male Broodstock29 Introduction California abalones have been harvested by humans for hundreds of years, but large-scale commercial harvesting did not begin until 1916 (Tegner et al., 1992). Red abalone, Haliotis rufescens, is preferred among the several native Californian abalone species due to its larger size and distinct flavor. However, the natural history of the abalone - a slow-growing, long-lived, fairly sessile creature - seems to facilitate overfishing (Tegner, 1989). Between 1931 and 1968, state-wide landings averaged over 900 metric tons wet weight per year. Since 1968, commercial catches have declined to 15% of that: current landings are less thun 136 metric tons per year (Tegner et al., 1992). (See Figure 1) This decline has occurred despite intensive effort by the California Department of Fish and Game to restore the fishery. Presently, commercial red abalone harvesting is centered in south-central California - most specifically near the Channel Islands off the coast of Santa Barbara. Loss of some formerly productive areas due to overfishing of natural stocks, expansion of the sea otter range, environmental degradation of some mainland sites, and competition from recreational divers have pushed the commercial divers to travel to the furthest Channel Islands (Tegner, 1989). According to John Colgate, the President of the Abalone Divers Association, San Miguel Island (see map Figure 2) is presently believed to have the largest bed of abalone in southern California (personal communication). Northern California, which has an even larger supply of abalone, is closed to commercial harvesting. As the supply of abalone diminishes, their per unit value continues to increase. Between 1973 and 1988, the price paid to divers has increased 800%, over twice the rate of inflation (Tegner et al., 1992). Abalone uquaculture is a logical response to the diminishing natural supply and continuing high demand that Californian, Japanese, and other seafood connoisseurs have placed upon this gastropod. Because the present market for abalone is twice as large as the current supply, cultivation of abalone is likely to continue to increase (Sietsema, 1992). Oysters, scallops and clams have been cultured for many years in the United States, and Japan has been ocean ranching shellfish, including abalone, even longer. A small laboratory constructed in 1964 at Morro Bay tested red abalone cultivation based on Japanese abalone hatchery methods and showed that mass cultivation of abalone was feasible for California industry (Ebert, 1984). Innovations in abalone culturing and the increase in demand and price have now made abalone farming profitable. However, this profit obviously relies on maintaining cost effective production. The most costly aspect of abalone production is the organism's slow growth rate: abalone take three years to reach market size. In addition to being slow, abalone growth also varies among individuals. One cohort in this study had a nearly 20-fold range in mass, from 0.3 grams to 5.85 grams. This variable growth rate forces most cultivators to sort their abalone into size classes and to discard the smallest 5-20%. Finally, the abalone’s high fecundity makes cultivated breeding possible, but tempts aquaculturists to use a small number of adults to produce thousands, and even tens of thousands of offspring. Over time, this continual loss of par of the population and these hatchery breeding techniques may alter the gene pool, potentially leading to inbreeding. Inbreeding often has devastating effects: homozygote appearance of deleterious rare alleles, inability to have varied responses to changes in the environment, loss of biodiversity, and a general loss of robustness. Ironically, the breeding practices of abalone cultivators could produce abalone with even slower growth rates than at present. As abalone cultivation is a fairly new industry in California, little research has been conducted upon the domestication of this species. However, several studies involving electrophoretic work have been done in Japan, examining cultivation and even ocean ranching of some other abalone species and of oysters (Fujino, 19781 & Il; Fujino and Nagaya, 19771 & II; Fujino and Sasaki, 1984; Wada, 1986; Saito, 1984). Additionally cultivation of other shellfish in the United States is used as a parallel to the more recent cultivation of abalone. Electrophoretic studies found genetic diversity within and/or between populations of the Coot Clam (Gaffney et al., 1990) and domesticated American oysters (Hedgecock and Okazaki, 1984; Vrijenhoek et al., 1990). In addition to simply comparing genetic diversity, several oyster studies pointed to the importance of measuring the temporal variance of allelic-frequency in order to estimate the effective size of the populations. These values provide a means to analyze genetic drift and can aid in determining the extent and nature of genetic change in reproductively isolated populations. Two studies of oysters and other marine shellfish estimated very low effective population sizes for these cultivated populations even when the traditional methods showed little or no significant inbreeding (Hedgecock and Sly, 1990; Hedgecock et al., 1992). A third study indicated an additional value of estimating the effective population size. Work by Gaffney et al. (1992) found that fewer oyster individuals than previously thought were actually contributing to the offspring produced in large group hatchery spawns. These studies show the importance of examining the population genetics of cultured California abalone by comparing genetic variation and by measuring genetic drift. This report analyzes nine populations of abalone: three separate spawns from a hatchery, samples from the male and female culture broodstock, natural abalone from two areas of the northern California coast, and samples of abalone collected in 1992 and in 1979 by commercial divers near Santa Barbara, California. The goals for the present study are to: determine the genetic composition of cultured and natural populations of the red abalone, Haliotis rufescens, and estimate the extent of inbreeding and genetic drift within these populations. Additionally, anecdotal evidence indicates that the 1992 Santa Barbara population analyzed in this study was the result of abalone seeding projects of the late 1970's and early 1980's (Westlotorn, personal communication). In addition to aquaculture, another response to the diminishing natural supply of abalone was artificial enhancement. Though testing by the California Department of Fish and Game officially labeled abalone seeding as unsuccessful, the present study provides some evidence to the contrary. A re-analysis of the California abalone enhancement projects is presented. Materials & Methods Sample Collection: Please refer to Figure 2 for a map of the collection sites and to Table 1 for a summary of the populations. A total of five populations of cultured red abalone, Haliotis rufescens, were collected from Pacific Mariculture, Incorporated (PMI) in Santa Cruz, California. Population 1 (N = 156), Population 2 (N = 120), and Population 3 (N = 122) were samples from three separate 1991 spawns that occurred on March 28, August 21, and July 25, respectively. For each spawning, PMI used between 25-50 gravid abalone. The females were chosen from a brood stock of 7-800 individuals while the males were chosen randomly from the production tanks. Population 4 (N = 43) was from the female broodstock and includes samples from 12, 5 and 6 tagged individuals that were known to have contributed eggs to spawns 1, 2 and 3, respectively. As PMI has no separate male broodstock, 2.5 inch (about 2.5 year old) male abalone from one production tank were sampled for Population 5 (N = 39). Based on size and growth rate estimations, most of the male abalone used in 1991 spawns had probably been sold before samples were collected for analysis. However, this population does offer an interesting comparison of allelic diversity. Further examination of the genetic history of the male and female abalone used in PMI's spawns will be explored in the Discussion. In addition to the culture populations, two abalone populations from northern California and two from southern California were studied. Seventy mantle and gill tissue samples were collected from abalone harvested by recreational skin divers near Gerstle Cove (population 6) and 46 samples were from the coastline around Fort Ross (population 7). Population 8 came from commercial divers who collected their abalone from the south side of San Miguel Island off the coast of Santa Barbara in May 1992 (Westlotorn, personal communication). Data from a 1979 Sea Grant study was used as a historical comparison. Dennis Hedgecock (1979) analyzed samples of natural red abalone populations collected from six sites in the Santa Barbara area, including the north and south sides of San Miguel Island, and found very little variation from one collection site to the next. Hedgecock's data were pooled to form Population 9 (N = 208) which represents the natural abalone population off the coast of Santa Barbara. The cultivated abalone populations in the study and in the other farms in California were ultimately derived from the abalone in the Santa Barbara area (McBride, Oakes, McMullen, and McCormick, personal communication). Sample Preparation Foot muscle, shell muscle, mantle, and/or gill tissue were analyzed (Table 2). Tissue samples were placed in 1.5 mL Eppendorff tubes, with 20-40 microliters of buffer and 5-10 mg of silicon beads. A buffer solution of O.1 MTris-HCl pH 7.0 buffer containing O.1% B-mercaptoethanol was used for better resolution (Gaffney et al., 1992). Samples were homogenized with an electric drill fitted with a glass grinding tip. To prevent any denaturing of the proteins, samples were stored at -70 degrees C and placed on ice during preparation and transport. Horizontal Starch Gel Electrophoresis The horizontal starch gel electrophoretic procedure was based on that described by Schaal and Anderson (1974). No enzymatic activity was detected for isocitrate dehydrogenase, lactate dehydrogenase, malic enzyme, and glucose-6-phosphate dehydrogenase. The following enzymes gave positive staining but were not resolved well enough or consistently enough to be scored: hexokinase, mannose-6-phosphate isomerase, esterase, 6-phosphogluconate dehydrogenase, and tetrazolium oxidase. Leucine amino peptidase was resolved, but it appeared monomorphic and required digestive tissue samples, so assaying for this enzyme was not continued. Four enzymes - glucosephosphate isomerase (PGI), phosphoglucomutase (PGM), glutamate-oxaloacetate transaminase (GOT), and malic dehydrogenase (MDH) - were consistently resolved using the Tris-citrate pH 5.8 buffer of Schaal and Anderson (1974). Additionally, these four enzymes were analyzed in Hedgecock's 1979 study. After centrifuging the samples for approximately five minutes, the solutions were absorbed into filter paper wicks, blotted, and placed onto the gel. Gels ran 4-6 hours at 160 volts (0.3-0.4 amps. per gel) and were sliced in preparation for staining. Interpreting Gels Two loci were observed for PGM and MDH, and the more quickly migrating isozyme was scored in both cases for it was more active and more polymorphic. These loci were labeled PGM-2 and MDH-2, in accordance with the analysis and labeling of the 1979 data (Hedgecock). For each locus, the most common allele was labeled 100 and the other alleles were designated by adding or subtracting the estimated number of millimeters they varied from the chosen standard. For the PGM-2 locus, the allele chosen as the standard differed between the 1979 data and the 1992 data, but this difference was accounted for in the data analysis. To reduce scoring errors, questionable samples were run more than once and samples from consecutive populations were loaded onto a single gel to ensure that the allele designations were consistent among the eight populations assayed. Additionally, the allelic designations were consistent with the 1979 data. Data Analysis Genotypes were tallied for each locus. In transferring data to table/graph form, the original numerical allele designations were replaced with alphabetic ones. The most common allele (originally 100) became A, the next common B, and so forth. Swofford and Selander's BIOSYS-1 program (1981) was used to compute: allele frequency, Hardy Weinberg equilibrium, mean number of alleles per locus (ni), observed and expected heterozygosity (Ho and He), and the differences between them (D- Ho-He/Ho). Another program (untitled, developed for Hedgecock et al., 1992) was used to analyze genetic drift variables: temporal variance of allelic frequencies (F), estimated per¬ generation effective population size (Ng), an independent measurement of the numbers of alleles expected to remain in each population (no, and a chi-square test of how ni compares with the actual number of alleles (na). The genetic drift analysis includes an additional variable, Np, the harmonic mean of the number of abalone spawned per hatchery-produced generation. As the determination of Np depended not only on estimating how many generations the abalone were removed from the wild, but also depended upon a compilation of old, inadequate hatchery and sales records of several different companies, a range was calculated using the lowest and then the highest possible variables. Results Genetic Variation Analysis Table 3 shows the allele frequencies at each locus for all studied populations. Note that all 60 samples collected in the 1992 Santa Barbara population were heterozygous at the PGI locus. Goodness-of-fit to Hardy-Weinberg equilibrium was tested at each loci for every population, using Levene’s correction for small sample size. For loci with several alleles, a chi-square test was performed with all but the most common alleles pooled together. Table 3 indicates that three culture populations contained one locus that was not in Hardy-Weinberg equilibrium at the p « 0.05 level. The Santa Barbara 1992 population had two such loci and the natural populations had none. To examine possibilities of inbreeding in the populations sampled, the average number of alleles per locus (ni) and the D-values for observed and expected heterozygosity D = (Ho-He)Hel were compared. As can be seen in Table 4, the cultured populations tended to have lower nj values than the natural populations, though the differences are not statistically significant using the t-test. Notice that ni for the 1992 Santa Barbara population, though not statistically significant, is much lower than the value for any of the cultured populations. Unlike all of the other populations, it was completely monomorphic at the MDH locus. Again, in comparing observed and expected heterozygosities and their differences (Ho, He, and D), there were no statistically significant differences or possible trends noticed between the cultured and natural populations. However, the 1992 Santa Barbara Population had the highest observed heterozygosity and the next to lowest expected heterozygosity, resulting in a large difference (D-.652). This difference is due mostly to the PGI locus, but also to the PGM locus (Figure 3). Genetic Drift Analysis In addition to measuring the previous variables that have traditionally been used to describe and compare population genetic structure, this study used a tool to analyze genetic drift more specifically. This genetic drift test uses the allele frequencies of the derived and progenitor populations to determine the allele frequency variance (F) and to estimate the effective size of each derived population based upon how many generations it has been separated from the progenitor (wild) population. In an independent test, Fis used to estimate how many alleles are expected to remain in the population (nt). This expected value is then compared with the actual number of alleles remaining (na) in a chi-square test. If the observed value, na, is determined to be not significantly different from the expected value, nt, then more validity is given to F and thus to the accuracy of the original estimated population size. However, the chi-square test is not very accurate if either the expected number of alleles remaining or lost is very small (£ 0.2). Table 5 summarizes the information found for each population in the genetic drift analysis. The 1979 Santa Barbara abalone population was used as the originating population to compare with all of the populations collected and analyzed in 1992. As it was very difficult to accurately determine how many generations separated the cultured abalone from the wild, the genetic drift analysis was carried out with two best estimates per population. The five culture populations from PMI did not come directly from natural populations, but were produced from broodstock, or are actually seedstock, that were purchased from several California abalone hatcheries. These other aquaculturists originally obtained their abalone primarily from souther California waters around Santa Barbara or the Channel Islands (McMullen, Oakes, and McCormick, personal communication), so it seems fairly safe to assume that the 1979 Santa Barbara population represents the source for the cultured abalone. The two northern California populations were analyzed as controls, to see if the 1979 Santa Barbara population could be assumed to have a genetic composition similar to a natural population from any location in California. The three populations spawned ut PMI show a trend toward having lower allele- frequency variance (F) with each subsequent sample: F = 0.130, 0.087, 0.027 for populations 1, 2 and 3, respectively. The population of 2.5 year old male abalone at PMI had a high allele-frequency variance (F = 0.136) and subsequently low effective population size estimates. The female PMI broodstock abalone had a low allele-frequency variance (F = 0.027) and subsequently had fairly high population size estimates, though the 95% confidence limits contained a wide range that extended to infinity. For all five of these populations, the independent comparison of na and ni found no significant differences between the observed and expected values, using a x2 test with one degree of freedom. For four of the five culture populations, the estimated effective population sizes are well within the expected range of Np, the average number of abalone used to spawn each hatchery-produced generation. However, the analysis of PMI spawn 1 does show a discrepancy between these two values. Nk is between 2.13 and 28.42 while Nb is between 61 and 66. For the three natural populations collected in 1992, the number of generations removed from the Santa Barbara 1979 population is based upon the size of the abalone collected in 1979 and in 1992 and upon the growth rates of abalone in the wild. It is evident that if the abalone collected from San Miguel Island in 1992 had been separated from the rest of the natural Santa Barbara Channel Island since 1979 (1-2 generations) then the allele-frequency variance was high (F = O.128) and the effective population size was very low, having a 95% chance of being between 1.09 and 21.39 individuals. The genetic drift analysis was as expected for the two northern California populations, with very low allele-frequency variances (0.019 and 0.014) and subsequently high estimated effective population sizes that include infinity in their 95% confidence limits. For two of these three natural populations, the independent comparison of na and nt found no significant differences, but a significant difference was found for the Fort Ross population. However, this large x2 value can be explained by the small value for the expected number of alleles lost (0.164). Discussion In the genotype variation analysis, three of the cultured populations had one locus with a genotype frequency below the 5% probability required to rule out Hardy-Weinberg equilibrium. The 1992 Santa Barbara population had two loci that were not in Hardy- Weinberg equilibrium. This is the first, but definitely not the last, indication that this Santa Barbara population is different than the other populations collected directly from the ocean. Intuitively, a farm would seem to have a smaller population size and less random mating than in nature. This could cause inbreeding, explaining the few deviations from Hardy-Weinberg equilibrium found in those populations. Simply defined, inbreeding is measured by a decrease in heterozygosity (Mitton, in press). However, though there was a slight trend towards lower values for the mean number of alleles per locus in the cultured populations as compared to the natural populations, the difference was not statistically significant. Only in the 1992 Santa Barbara population does the low mean number of alleles approach statistical significance. This loss of rare alleles and potential decrease in mean alleles per locus in the cultured and introduced populations as compared to wild populations could indicate a population bottleneck. In an initially perplexing conflict, this potential loss of alleles has not yet led to any decrease in heterozygosity. The cultured populations exhibited expected and observed levels of heterozygosity that do not differ from the natural populations and do not have any significant difference between them. However, the 1992 Santa Barbara population actually exhibited much more heterozygosity than predicted by Hardy-Weinberg equilibrium. This phenomenon - loss of alleles yet increase in heterozygosity - has been documented in other cultured marine invertebrates, including hard clams, Mercenaria mercenaria (Dillon and Manzi, 1987) and oysters, Crassostrea virginica (Vrijenhoek et al., 1990). The increases in heterozygosity in the 1992 Santa Barbara population could have resulted from a founder effect leading io genetic drift or from selection. While it is theoretically possible that there could be strong selection pressures in the San Miguel area for heterozygosity at the PGI locus, it seems very unlikely that a change this dramatic could occur in only 1-2 generations by this means alone. (The present analysis of genetic drift assumes that selection has not occurred.) The abalone that produced this population in Santa Barbara (whether the spawn occurred in nature or in a hatchery) could have, by chance, contained individuals with a higher concentration of the second-most common allele. While losing overall allelic richness, the allele evenness could increase. For example, the population from Santa Barbara in 1979 had allele frequencies for PGl of A:O.890, B:0.073, and C:0.037 while in 1992 the Santa Barbara population in question had PGI allele frequencies of A:0.500 and B:0.500. Though one allele was lost, heterozygosity was greatly increased as was allelic evenness. As the departure from Hardy-Weinberg equilibrium is signisicant, this situation is most likely with an extremely small number of parents. This founder effect can then be analyzed as genetic drift. A more accurate look at the changes in genetic diversity that could happen over time with aquaculture and/or enhancement programs is found by quantifying genetic drift. As discussed previously, decreases in heterozygosity may not be the only indication of significant changes in genetic composition resulting from reproductive isolation or small population size. In measurements of heterozygosity, an increase in the evenness of frequencies of the remaining alleles can compensate for the loss of rare alleles in a population bottleneck. Since population bottlenecks often affect overall allelic diversity more than they affect heterozygosity, it makes sense to analyze the shifts in allele frequency between progenitor and derived populations. In a study comparing hatchery and natural oyster populations and in a re-analysis of the data from several previously published electrophoretic studies (Hedgecock et al., 1992), it has been shown "that superficial similarities in levels of genetic diversity scan belie significant genetic changes in cultivated stocks." (Hedgecock and Sly, 1990) This is obviously the case with the 1992 Santa Barbara population, for although it exhibited an increase in heterozygosity, the genetic drift analysis indicates that this population most likely results from fewer than 22 parents. This is the strongest indication that these abalone were indeed the product of an enhancement program. First, the abalone hatcheries that produced abalone seed for the various abalone enhancement programs have been known to use as few as 2-4 highly fecund abalone to produce thousands of stock (Oakes, McMullen, McCormick, personal communication). Secondly, no other allozyme study of native Haliotis rufescens populations has ever indicated any type of localized genetic variation from one cove to the next. This present study shows very little genetic differentiation between the 1979 southern and 1992 northern California abalone populations, Hedgecock's 1979 data showed very little genetic differentiation between abalone populations on various Channel Islands, and a third study showed very little variation in abalone populations that were sampled from all along the California coast (McKean, 1980). As for the cultured abalone at PMI, the conclusions from the genetic drift analysis are not nearly as dramatic for these populations as they are for the Santa Barbara population. Perhaps what is most interesting to note is that by purchasing several different spawns of seed stock and brood stock from a variety of abalone hatcheries, PMI managed to inadvertently lessen any inbreeding that might have been present in any one hatchery- produced spawn. The female broodstock, which has been formed from a number of spawns from a variety of hatcheries, had a high estimated effective population size. The male abalone sampled, probably resulting from as few as one or two spawns, had a very low estimated effective population size. Figure 4 presents the origins of these two populations. The fairly low variance in allelic frequency found between the progenitor population and the female abalone and the fairly high variance of the male abalone are contrasted with the moderate shift in the three spawn populations. Though the three spawns were not produced from the males sampled, this could indicate that the males that were spawned have experienced a more intense population bottleneck than the female broodstock. However, Figure 4 demonstrates that the males used in the 1991 spawns were made up of abalone produced in a large number of hatchery spawns, just like the female broodstock. If these males had an effective population size comparable to that of the female broodstock, then the lower effective population sizes of the three spawns could indicate that an amount of genetic diversity has been lost with each subsequent generation. This effect could most likely be counteracted by mixing spawns together. Additionally, the discrepancy between Nb and Nk in PMI spawn 1 could indicate that not all of the abalone used for each hatchery spawn have contributed to the production of offspring. Overall, it appears as if little inbreeding can be detected in these culture populations after analyzing both the present genetic variation and the genetic drift. Conclusion Aquaculture The present study indicates that there are significant differences between the genetic compositions of the Santa Barbara 1992 population and the natural populations, but only slight differences between the cultured and natural populations. The short history of abalone aquaculture paired with the long abalone generation time has prevented much inbreeding and much genetic drift. However, it is always better to begin examining potential losses in genetic diversity too early rather than too late. During the course of this study, PMI realized that their present breeding practice: using a separate female broodstock and spawning gravid males pulled from the production tanks - could not remain permanent, for eventually the males pulled from the production tanks would be the progeny from some of PMI's earlier spawns and they could potentially be performing back crosses. Perhaps PMI could start using some wildstock male abalone in their breeding program, for using natural abalone along with a large brood stock would minimize inbreeding and genetic drift. The natural abalone would reintroduce rare alleles and the large brood stock would help to maintain genetic diversity. Additionally, evidence from PMI spawn 1 indicates a possible discrepancy between the number of abalone thought to be contributing to spawns and the number that actually are. Gaffney et al. (1992) found this phenomenon in oyster hatcheries and noted that this problem could be alleviated by replacing large group spawns with a greater quantity of spawns with fewer individuals. Abalone hatcheries could consider this breeding technique. Finally, as abalone aquaculturists have begun to perfect their hatcheries, the time might be appropriate for them to work together to begin to select for desired abalone traits and start to breed various strains. (Keeping in mind that an outbred natural or cultured population should remain as an antidote for any adverse inbreeding effects that could result from domestication.) (Newkirk, 1983; Lester, 1983) Enhancement In a striking example of how quickly genetic diversity can be altered, the genetic drift analysis shows dramatic changes in the allele frequencies of the 1979 and 1992 abalone populations from near San Miguel Island. This change is thought to be the result of human intervention. The present data show that for such a large change to occur in such a short time frame, then this population would have to be the result of fewer than 22 spawning adults. As explained previously, no other allozyme study of natural red abalone populations has ever found any information to indicate that any genetic differentiation exists between natural abalone populations, so it seems more plausible to assume that these could be hatchery produced abalone that were introduced to the area several years ago. Between the late 1970's and the early 1980's, a variety of groups, including the California Divers Association, the California Department of Fish and Game, Atlantic Richfield Company, and recreational divers groups, sponsored abalone enhancement programs which resulted in hundreds of thousands of hatchery-produced abalone being seeded in southern California water from Palos Verdes to many of the Channel Islands, including San Miguel. It seems as if a controversy was inadvertently unearthed during the course of this present study. The California Department of Fish and Game determined that enhancement was not successful and therefore not to be relied upon as a way to protect this valuable natural resource, but the professional abalone divers disagree. Many feel that the Department of Fish and Game studies were merely inconclusive - due to some poor site choices for abalone enhancement and to the great difficulties involved in being able to track and find abalone in natural environments months and years after introducing them. The divers experience with abalone has led them to believe that small and mid-sized abalone tend to hide in rock crevices and only reappear to be easily observed when they have grown much larger. Despite relating very low live seed recovery (£2.8%) for five large-scale seeding experiments with red and green abalones, one Fish and Game study allows that 16 swamping the system with miniseed (very small, inexpensive abalone) may be an economically feasible method of enhancement (Tegner and Butler, 1989). The most complete study of red abalone seeding was conducted on the Palos Verdes peninsula (Tegner and Butler, 1985). A total of 674 red abalone of two different sizes and originating from two different hatcheries were planted in the study site in May 1981. One year later, only 1% of the live seed was recovered. Seed mortality was estimated at 43% based upon the shells collected. Although over one half of the population was not accounted for, the 1% live recovery rate did not give much encouragement to the feasibility of future abalone enhancement. However, some evidence indicates that this study site, while accessible to the researchers, did not provide the best habitat for abalone and was already too damaged by human development. At the time of the planting, no emergent (visible to divers without turning over rocks) abalone were found in 700 m2. The average density of the other, cryptic (hidden) red abalone was very low. Density (1 S.E.) was 0.17 + 0.12 per square meter for 70 m2 area sampled. This was not the only test site used to examine the feasibility of red abalone enhancement. In the spring of 1979, several sites on the Channel Islands were examined to select a suitable site for planting seed (DeFelice, 1979). Tyler Bight, on the south coast of San Miguel Island, was selected as the planting site as it offered "excellent food supply physical habitats, and workable depths." (Butler, June 6-11 1979). The June 1979 survey of the habitat found both emergent and cryptic abalone. Additionally, the density of red abalone was calculated in three different areas where 46, 70, and 90 m2 samples were performed (Butler, June 19-25 1979). The density values were 1.09,0.70, and 1.00 abalone per m2, which is nearly a factor of ten greater than the abalone density in the Palos Verdes sample (Table 6). This indicates that San Miguel Island offered a much better habitat for red abalone than the mainland test site. In August 1979, 20,000 red abalone seed were planted in two sites at Tyler Bight (Butler, August 1-8 & 16-22 1979). However, bad ocean conditions prevented the survival rate from being determined. In October of 1980, the Kelp Bass vessel went to San Miguel Island, but "large ground swells prevented any sampling to be accomplished at San Miguel on two occasions during the week" (Butler). It appears that the sampling was never performed. While originally only anecdotal evidence from the abalone divers indicated that abalone seeding could be successful, the present study provides some new, albeit circumstantial, evidence that the 1979 seeding did provide abalone for the diving industry 1) According to John Colgate, the president of the California Abalone Divers Association, the south side of San Miguel Island contains the largest bed of abalone in southern California (personal communication). Anecdotal evidence among the divers indicates that part of the prosperity of abalone in this region is due to the 1979 seedings. Additionally, the divers who collected the abalone analyzed as the 1992 Santa Barbara population in this study claimed that these samples resulted from enhancement (Westlotorn, personal communication). 2) A study of red abalone growth rates determined that it took abalone in southern California an average of 15 years, with considerable variability, to reach the commercial minimum size of 197 mm (Tegner et al., 1992). The abalone planted in 1979 were around 30 mm (Butler, August 1-8 & 16-22 1979), indicating an age of approximately one year. As the Santa Barbara population was sampled in 1992, then any survivors from the seeding would be 14 years old. This is very comparable with the length of time determined for abalone growth to minimum legal size. 3) It is very likely that the abalone seed obtained for the 1979 seedings could have been produced from a very small number of parents. Unfortunately, they were bought from a company that is now defunct, Monterey Abalone Farms, so the exact hatchery records could not be obtained. However several of the present abalone hatcheries have indicated that, if necessary, they could (and have) obtained tens of thousands of abalone seed from as few as two large, fertile individuals (Oakes, MCMullen, McCormick, personal communication). If the abalone sampled were the seeded individuals, then a spawn of several females who were homozygous for A at the PGI allele and one male who was homozygous for B at the PGI allele could have produced offspring with allele frequencies like what was observed. In conclusion, there are several possible explanations for the unusual genetic variation found in the 1992 Santa Barbara population of abalone, including selection and genetic drift. Genetic drift could have occurred naturally at that site, or could be the result of a hatchery spawning only a very small number of individuals. Although an electrophoretic study of the Australian blacklip abalone, Haliotis rubra, found regional gene pools for these populations (Brown, 1991), no study has ever found any genetic variation by distance between different red abalone populations. The present study shows very little difference between samples collected in southern California in 1979 and in norther California in 1992 and between these natural populations and several cultivated populations. The most plausible explanation for the allele frequencies found in the 1992 Santa Barbara population is that these abalone were the result of the 1979 seeding. This study indicates that abalone enhancement could be successful in California, possibly helping to reverse the present decline in natural stocks. Further research could quantify the effectiveness of seeding. In the course of this present work, several rare alleles were found in individuals from the PMI broodstock that could be helpful to future studies. These animals could be spawned and their offspring would then contain allozyme markers that would indicate whether they were outplanted individuals with a greater than 90% assurance (Hereford et al., 1992). This study raises a final point. Although it is definitely possible to produce tens of thousands of seed from only a handful of the highly fecund adult abalone and although it appears to be possible for these hatchery-bred individuals to grow to commercial harvesting size in the wild, the researcher must consider the ecological effects of enhancement. Before any hatchery-bred abalone were outplanted, Dennis Hedgecock's study was conducted in the late 1970's to ensure that there was no distinct genetic composition that varied from one population to another. However, while great care was taken to ensure that no existing population diversity would be altered by human intervention, it appears that the genetic composition of the seeded individuals was not determined (until now). If enhancement programs do occur in the future, this measure of biological diversity must be considered. Population Table 1 Summary of Sampled Populations Description/Name Location PMI spawn 1 PMI, Santa Cruz PMI spawn 2 PMI, Santa Cruz PMI, Santa Cruz PMI spawn 3 broodstock/ PMI, Santa Cruz PMI females adult males/ PMI males PMI, Santa Cruz native N. Calif. abs./ Gerstle Cove Gerstle Cove native N. Calif. abs./ Fort Ross area Fort Ross Santa Barbara abs./ south side of SB 1992 San Miguel six sites in the Santa Barbara abs. SB area, including SB 1979 north and south sides of San Miguel Date Collected April 1992 May 1992 July 1992 July 1992 July 1992 May 1992 May 1992 May 1992 1978-1979 Population PMI 1 PMI 2 PMI 3 PMI females PMI males Gerstle Cove Fort Ross SB 199. SB 1979 Table 2 Summary of Allozyme Analysis Tissue sampled Locus foot PGI GOT PGM MDH PGI foot GOT PGM MDH PGI foot GOT PGM MDH mantle PGI GOT PGM MDH PGI mantle GOT PGM MDH PGI mantle GOT PGM MDH PGI mantle GOT PGM MDH shell muscle PGI GOT PGM MDH PGI mantle GOT PGM MDH Sample size 156 156 156 156 120 120 120 110 122 60 208 169 138 Table 3 Allele Frequency Chart Allelic frequencies. Asterisks indicate significant (pe0.05) deviations from Hardy- Weinberg equilibrium (using chi-square test with pooling of rare alleles when appropriate and using Levene’s correction for small sample size). PG GOT PGM MDH Population Allele 593 .990 .990 PMI 1 279 .010 215 .010 128 592 959 871 PMI 2 338 129 .041 .204 071 430* 914 987 PMI 3 .852 455 135 .086 .013 .115 012 95. 90 .884 .360 PMI Females 047 093 .488 105 152 012 PMI Males 195 692 205 .013 102 013 979 .307 893 Gerstle Cove 086 521 107 021 150 .007 014 007 967 326 848 946 Fort Ross 576 141 .054 033 098 011 21 500* SB 1992 .975 1.000 17 500 025 .008 SB 1979 989 266 856 968 .120 592 011 032 014 130 010 012 Table 4 Summary Statistics of Genetic Variation Mean number of individuals assayed per locus (S), mean number of alleles per locus (ni), mean heterozygosity observed per locus (Ho), mean heterozygosity expected per locus under Hardy-Weinberg equilibrium (He), and the difference in heterozygosities (calculated as D-Ho - He/He). 1979 PMI PMI PMI PMI PMI Gerstle Fort 1992 Statistics 3 females males Cove Ross SB SB Number of Loci Studied 4 4 4 4 4 4 4 4 4 46 60 148.3 156 117.5 121.3 43 (0) (2.5) (0.8) (0) (0) (0) (0) (0) (27.4) (S.E.) 3.0 2.5 2.0 3.0 2.3 2.5 2.5 2.5 2.3 (0.3) (0.3) (0.3) (0.3) (0.5) (0.7) (0.3) (0.4) (0.6) (S.E.) Ho 289 .267 340 .239 245 .375 .208 .255 .281 (141) (084) (.147) (101) (.120) (115) (.104) (231) (111) (S.E.) 259 .272 314 .255 247 227 225 He 233 .291 (.131) (095) (.122) (117) (101) (125) (112) (121) (.124) (S.E.) 094 -034 .116 -018 .083 -.063 -008 652 -.076 Table 5 Genetic Drift Analysis Summary Estimates of temporal variance in allelic frequencies (F) and variance effective population size (Np). St is the sample size of the derived populations (mean over loci); t is the number of generations separating the derived populations from the progenitor population, Santa Barbara 1979; Np is a range estimating the mean breeding number per generation; np is the number of alleles in the progenitor population, na is the number of alleles in St, and nt is the number of alleles expected to remain in a population of size Ny after t generations of random drift; goodness-of-fit of na vs nt is tested by x2 with one degree of freedom (note that this value is not reliable when the expected number of alleles remaining or lost is «0.2). Interval Nb np,na x2,1d.f. (95% c.l.s) range n navsnt SB 1979"-males PMI 4.16 12,10 0.136 0.924 5-21 (1.03 - 10.86) (7.9 8.33 (2.07 -21.72) 1 41.87 120-430 SB 1979-females PMI 43 0.027 12,10 1.571 (5.98 - inf.) (11.4) 83.74 (11.96 - inf.) SB 1979-PMI spawn 1 156 2 0.130 8.10 61-66 0.140 12,9 (2.13 - 18.95) (7.9 12.16 (3.19 -28.42) SB 1979-PMI spawn 2 117,5 2 0.087 12,9 0.000 12.51 48-51 (3.19 -30.85) (8.5) 18.76 (4.79 -46.28) 51.87 89-99 12,10 SB 1979-PMI spawn 3 121.3 2 0.027 0.071 (10.8) (11.06 - 208.99) 77.78 (16.58 -313.49) 60 1 4.30 12,8 0.082 SB 1979-SB 1992 0.128 (1.09 - 10.70) (8.0) 8.60 (2.18 -21.39) 70 0.332 SB 1979-Gerstle Cove 0.018 65.18 12,12 1 (10.05 - inf.) (11.8) 130.36 (20.09 - inf.) 11.038 46 0.014 12,10 SB 1979-Fort Ross (11.8) p «0.05 (13.55 - inf.) inf. (27.11 - inf.) * SB 1979 mean sample size is 148.3 Table 6 Comparison of Two Red Abalone Seedings* Palos Verdes Tyler Bight. San Miguel Density of Abalone before seeding (abalones/m2) 0.17 + 0.12 1.09, 0.70, 1.00 Emergent abalone visible before? yes no 46m2, 70m2, 90m2 70m2 Area surveyed f of abalones seeded 674 20,000 August 1979 May 1981 Date of seeding % of seed recovered 1% live, 43% of shells never sampled * data from Tegner & Butler, 1985 and cruise reports: Butler, 1979-80 Figure 1 California Red Abalone Landings (1931-1988) Commercial red abalone landings from California for 1931 to 1988. For 1951 to 1988, data is broken down into south-central California and the San Francisco area. (figure from Tegner et al., 1992) 1800- • SAN FRANCISCO AREA • SOUTH-CENTRAL CALIFORNIA 1600- a ALL CALIFORNIA 1400 1200- 8 000- 800 600- 400- 200 o+ 1970 1950 930 YEAR 26 1990 Figure 2 Map of Collection Sites ORICON ------------- CALIFORNIA Sonoma County San Francisco Bay Monterey Bay A-Pacific Mariculture Incorporated B -Gerstle Cove C - Fort Ross D - San Miguel Island Santa Barbard P. CHANNEL ISLANDS PACIFIC OCEAN mies 20 0 attias 80 —— --— — BAJA MALIFORNIA 27 1.0 0.8 - 0.6: 0.4 0.2 0.0- 0.2 + ka Figure 3 D-values by Locus gtt —4 Sample Populations — PG —— O —n— PO —0— MH 28 Bibliography Brown, L.D. (1991). Genetic variation and population structure in the Blacklip Abalone, Haliotis rubra. Australian Journal of Marine and Freshwater Research 42: 77-90. Butler, Robert. (June 6-11, 1979). Cruise report 79-KB. Experiemental Abalone Enhancement Program California Department of Fish and Game Operations Research Branch, Long Beach. Butler, Robert. (June 19-25, 1979). Cruise report 79-X-4. Experiemental Abalone Enhancement Program California Department of Fish and Game Operations Research Branch, Long Beach. Butler, Robert. (August 1-8, 1979). Cruise report 79-A-5. Experiemental Abalone Enhancement Program California Department of Fish and Game Operations Research Branch, Long Beach. Bulter, Robert. (August 16-22, 1979). Cruise report 79-A-6. Experiemental Abalone Enhancement Program California Department of Fish and Game Operations Research Branch, Long Beach. Butler, Robert. (October 6-10, 1980). Cruise report 80-KB-2. Experiemental Abalone Enhancement Program California Department of Fish and Game Operations Research Branch, Long Beach. Colgate, John. President of the California Abalone Divers' Association. Santa Barbara. personal communication, September 1992. DeFelice, Richard. (March 26-30, 1979). Cruise report 79-X-7. Experiemental Abalone Enhancement Program California Department of Fish and Game Operations Research Branch, Long Beach. Dillon, R. and Manzi, J. (1987). Hard clam, Mercenaria mercenaria, broodstocks: genetic drift and loss of rare alleles without reduction in heterozygosity. Aquaculture 60: 99-105. Ebert, E. and Houk, J. (1984). Elements and innovations in the cultivation of red abalone, Haliotis rufescens. Aquaculture 39: 375-392. Fujino, K. (1978). Genetic studies on the Pacific Abalone (Haliotis discus hannai ) - I: Inbreeding and overdominance as evidenced by biochemical polymorphism in a wild population. Bulletin of the Japanese Society of Scientific Fisheries 44(4): 357-361. 30 Fujino, K. (1978). Genetic studies on the Pacific Abalone (Haliotis discus hannoi Ino) - II: Excessive homozygosity in deficient animals. Bulletin of the Japanese Society of Scientific Fisheries 44(7): 767-770. Fujino, K. and Nagaya, N. (1977). Biochemical polymorphisms in the Pacific Oyster (Crassostre gigas ) -I: Variants in myogen and esterases. Bulletin of the Japanese Society of Scientific Fisheries 43(8): 983-988. Fujino, K. and Nagaya, N. (1977). Biochemical polymorphisms in the Pacific Oyster (Crassostre gigas) -II: Variants in Tetrazolium Oxidase and Leucine Aminopeptidase. Bulletin of the Japanese Society of Scientific Fisheries 43(12): 1455-1459. Fujino, K. and Sasaki, K. (1984). Age association of genotypic proportions of isozymes in the Pacific Abalone (Haliotis discus hannai). Bulletin of the Japanese Society of Scientific Fisheries 50(1): 11-15. Gaffney, P.M., Scott, T., Koehn, R., and Diehl, W. (1990). Interrelationships of heterozygosity, growth rate and heterozygote deficiencies in the Coot Clam, Mulinia lateralis. Genetics 124: 687-699. Gaffney, P.M., Davis, C.V. and Hawes, R.O. (1992). Assessment of drift and selection in hatchery populations of oysters (Crassostrea virginica). Aquaculture 105: 1-20. Hedgecock, D., Chow, V. and Waples, R.S. (1992). Effective population numbers of shellfish brood stocks estimated from temporal variance in allelic frequencies. Aquaculture 108: 215-232. Hedgecock, D. and Okazaki, N.B. (1984). Genetic diversity within and between populations of American oysters (Crassostrea). Malacologia 25(2): 535- 549. Hedgecock, D. and Sly, F. (1990). Genetic drift and effective population sizes of hatchery-propagated stocks of the Pacific oyster, Crassostrea gigas. Aquaculture 88: 21-38. Hereford, L., Powers, D., Kirby, V., and Scriviani, P. (1992). A proposal to the California Abalone Divers' Association. Lester, L. James. (1983). Developing a selective breeding program for penaeid shrimp mariculture. Aquaculture 33: 41-50. McBride, Susan. (April-September 1992). Director of Pacific Mariculture, Inc. hatchery. Davenport, CA. personal communication. McCormick, Thomas. (September 1992). President of McCormick & Assoc. Ojai, CA. personal communication. Mckean, M. (1980). Electrophoretic analysis of four California red abalone (Haliotis rufescens ) populations. for Animal Science 119 at UCDavis, taught by Dennis Hedgecock. McMullen, John. (September 1992). President of Ablab. Port Hueneme, CA. personal communication. Mitton, J. (1992-3, in press). Theory and data pertinent to the relationship between heterozygosity and fitness. In: The Natural History of Inbreeding and Outbreeding, ed. by N. Thornhill. University of Chicago Press, Chicago, IL, 1-45. Newkirk, Gary F. (1983). Applied breeding of commercially important molluscs: A summary of discussion. Aquaculture 33: 415-422. Oakes, Frank. (September 1992). President of Abfarm. Cayucas, CA. personal communication. Saito, K. (1984). Ocean ranching of abalones (Haliotis discus hannai) and scallops (Patinopecten yessoensis ) in northern Japan. Aquaculture 39: 361-373. Schaal, B.A. and Anderson, W. (1974). An outline of techniques for starch gel electrophoresis of enzymes from the American oyster Crassostrea virginica Gmelin. Georgia. Marine Sciences Technology Report Ser. 74. Sietsema, T. (April 22, 1992). California’s abalone industry is thriving. San Francisco Chronicle. Swofford, D.L. and Selander, R.B. (1981). BIOSYS-1: A FORTRAN Program for the comprehensive analysis of electrophoretic data in population genetics and systematics. Journal of Heredity 72(4): 281-283. Tegner, Mia J., DeMartini, John D., and Karpov, Konstantini. (1992). The California red abalone fishery: A case study in complexity. In: Abalone of the World: Biology, Fisheries, and Culture, ed. by S.A. Shepherd, M.J. Tegner, and S. Guzman-del Proo. Fishing News Books, Oxford, 370-383. Tegner, Mia J. (1989). The California abalone fishery: Production, ecological interactions, and prospects for the future. In: Offprints from Marine Invertebrate Fisheries: Their Assessment and Management, ed. by John F. Caddy. John Wiley & Sons, Inc., 401-420. Tegner, Mia J. and Butler, Robert A. (1989). Abalone seeding. In: Handbook of Culture of Abalone and Öther Marine Gastropods. Kirk O. Hahn (ed.), CRC Press, Inc., Boca Raton, Florida, 157-181. Tegner, Mia J. and Butler, Robert A. (1985) The survival and mortality of seeded and native red abalones, Haliotis rufescens, on the Palos Verdes peninsula. California Fish and Game 71(3): 150-163. Vrijenhoek, R.C., Ford, S.E. and Haskin, H.H. (1990). Maintenance of heterozygosity during selective breeding of oysters (Crassostrea virginica ) for resistance to MSX (Haplosporidium nelsoni ) disease. Journal of Heredity 81: 418-423. Wada, K.T. (1986). Genetic variability at four polymorphic loci in Japanes pearl oysters, Pinctada fucata martensii, selected for six generations. Aquaculture 59: 139-146. Westlotorn, Scott. (May 23, 1992). Processor of commercially harvested abalone. Santa Barbara. personal communication.