ABSTRACT Many sessile marine invertebrates exhibit a common suite of adaptations, in¬ cluding life histories with both a sexual and an asexual phase, clonality and/or colo¬ niality, and the ability to distinguish between clonemates and non-clonemates. These characteristics, together with limited, or philopatric, dispersal patterns in free- swimming larvae, are all present in the common tunicate Botryllus schlosseri. This organism presents an ideal model for the study of the problems in population genet¬ ics posed by complex life cycles, because the genetics of its fusion-rejection reaction are well-understood, and it is relatively easy to collect and maintain in the labora¬ tory. I present here the results of experiments using both histocompatibility tests on live Botryllus oozooids in the laboratory, and arbitrarily-primed PCR, as tools for population genetics research. My work with Botryllus oozooids involved testing the rates of fusion and non-fusion between pairs of offspring obtained from wild colonies containing fertilized eggs. I compared fusibility rates for the offspring of near-neighbor colonies to those for offspring of non-neighboring colonies, and found that there was a significantly higher rate of fusion between offspring of neighbors than between offspring of non-neighbors. My work with arbitrarily¬ primed PCR was aimed at developing the technique for further work in Botryllus; was able to detect genetic diversity based on differences in the patterns of amplifica¬ tion products obtained with different DNA samples. I also describe some prelimi¬ nary conclusions as to the most effective conditions for arbitrarily-primed PCR reac¬ tions using Botryllus DNAS. INTRODUCTION Many types of organisms, both plant and animal, are characterized by life cy- cles including fixed adult phases, which grow asexually, and mobile, sexually pro¬ duced propagules. Specifically, this type of life history is common among sessile marine invertebrates belonging to very diverse phyla. Such life histories pose spe¬ cial problems in population structure and dynamics, since one must consider the fates of reproductive adults, asexual propagules, gametes, and larvae. Planktonic larvae produced by some sessile marine invertebrates are capable of dispersal over long distances, and may contribute to gene flow between popula¬ tions separated by considerable physical barriers. In other organisms, particularly many ascidian species, larvae travel only short distances-a few centimeters to a few meters-before settling. This pattern of philopatric larval dispersal may lead to pockets of inbreeding within what might appear to be a panmictic population (Grosberg, 1987). Another common feature of colonial species among several different phyla of sessile marine invertebrates is the capability for self/non-self discrimination. The adaptive value of such mechanisms is not always well-understood. They can, however, provide a useful tool as well as an interesting problem for the biologist. Often, tests of histocompatibility may serve, to a first approximation, as tests of relat¬ edness. Population genetics can also be explored through more conventional av¬ enues, particularly the use of diagnostic enzyme and DNA markers. One technique that has recently been developed for the quick and facile detection of DNA markers involves the use of arbitrarily primed polymerase chain reaction (PCR) amplifica¬ tion. The polymerase chain reaction itself is a technique of gene amplification, which employs primers specific to the beginning and end of the sequence of interest, and repeated cycles of primer annealing, DNA polymerization, and denaturation, to amplify the target sequence 10°-fold or more. The amplification products can be de¬ tected by gel electrophoresis, and are visible as distinct bands. In arbitrarily-primed PCR, the sequence of the primers used is known, but ar¬ bitrary with respect to the genome being probed. These primers are generally 10 to 30 nucleotides in length, with a G/C content of between 50 and 80% (Williams et al., 1990a; Welsh and McClelland, 1990). The underlying idea of the technique is that within any genome, a certain number of appropriate priming site for any primer of restricted length will exist, and some of these will be close enough together for the intervening sequence to be amplified under the right PCR conditions. The amplifi¬ cation products which result will give patterns of banding on a gel which are charac¬ teristic of the DNA sample used (Welsh and McClelland, 1990). In practice, some primers will fail to produce bands at all in certain species, while others will produce the same banding patterns in all DNA samples tested, presumably due to hybridization to highly-conserved regions in the genome. Öther primers, however, will produce bands which vary between individuals or strains within a species, and therefore allow very precise identification of strains once pre¬ liminary screening of primers has been carried out. Studies in which primers differ¬ ing only by one base-pair were compared suggest that the technique is able to detect differences as small as single point mutations between genomes (Williams et al., 1990a). Since the use of arbitrary primers requires no previous knowledge of the se- quences either of protein or of DNA, the method would seem to be ideally suited for rapid screening of DNA samples prepared from many individuals in a population. The sequences detected by arbitrarily-primed PCR have been shown in breeding ex¬ periments to be transmitted as dominant markers, and can be used in constructing genetic maps (Williams et al., 1990a; Martin et al. 1991). In my experiments, I attempted to approach problems of population genetics posed by limited larval dispersal in the colonial tunicate Botryllus schlosseri. This common sea squirt possesses a whole suite of adaptations which seem likely to bring about clustering of related colonies in limited geographic areas. Various studies have shown that larvae and sperm disperse only short distances, that crosses be¬ tween more distant colonies may be less likely to produce viable offspring than those between near-neighbor colonies, and that offspring have a tendency to settle near their parents or siblings (Grosberg 1986, 1987, 1991). Furthermore, Botryllus colonies not only are able to discriminate between related and unrelated neighbors, but can fuse with histocompatible partners to form chimeric colonies with a contin¬ uous circulatory system. It has been suggested that since nutrients, and also primor¬ dial germ cells, can be translocated through the circulatory system, this adaptation may allow Botryllus invading new territory to cooperate with relatives in compet¬ ing for space against unrelated colonies (Grosberg 1987). When two related colonies fused, each would effectively increases the space it occupied by a function of the de- gree of genetic identity between the partners in the chimera. I studied the relationship between colony relatedness and distance between colonies using two different methods. One the one hand, I used arbitrarily-primed PCR to identify DNA markers which could allow one to determine the relative ge¬ netic similarity of separate colonies. On the other hand, I set up assays to test the frequency of fusion between the offspring of neighboring colonies and offspring of non-neighboring colonies within the same general area. My goal was to find out whether near-neighbor colonies were more genetically similar than colonies distant from each other in the same population and colonies from populations separated by significant physical barriers. I hoped to confirm the results of earlier studies show¬ ing that pockets of inbreeding did exist within certain Botryllus populations (Grosberg 1987), and to break new ground in the application of genetic fingerprinting techniques to the species. MATERIALS AND METHODS Biology of Botryllus schlosseri: Botryllus schlosseri is a common colonial as cidian present throughout the northern hemisphere. It occurs in harbor fouling communities and other habitats characterized by relatively low water flow rates and constant immersion. In the Monterey marina, Botryllus colonizes algae and larger solitary ascidians growing under the floating docks, as well as available hard sur¬ faces. There are two genetically-determined color morphs, orange and blue. Botryllus colonies originate from sexually-produced larvae, which spend a brief period-on the order of a few minutes-swimming freely before settling on a hard surface and metamorphosing into oozooids. The oozooid in turn grows, then produces asexual buds which take over from the original oozooid, and produce fur- ther buds, so that eventually the colony takes on its mature form. The life span of a colony is approximately four months, although some colonies overwinter and may therefore live as long as ten months. The mature Botryllus colony consists of as many as five hundred or more ge¬ netically identical zooids, each a self-contained unit with respect to feeding and re- production, arranged in systems of 8-20 zooids sharing a common excurrent siphon. Each zooid is hermaphroditic, containing both testes and ovaries, and has its own apparatus for filter-feeding, a gut and excretory system, but only under unusual con¬ ditions do colonies produce physically separate ramets by asexual growth. Furthermore, since all zooids in a colony reach sexual maturity at the same time, and fertile eggs are produced a few days before sperm are released, selfing is thought to be rare in the wild. The colony as a whole is connected both by a continuous tunic and by a com¬ mon blood system. The blood carries nutrients, cells responsible for pigmentation and factors which control histocompatibility; it may also carry primordial germ cells. The vascular system includes blind sacs known as ampullae, which are present around the edges of the colony. Botryllus colonies which come into contact at their growing edges are capable of fusing: that is, when their ampullae touch, the vascular systems of the two colonies can merge, and with time the two colonies can become morphologically indistinguishable due to the movement of pigment cells between the two colonies. The product of such a fusion is termed a chimera. Fusion in Botryllus is controlled by a single genetic locus which is highly polymorphic in all populations studied to date. Fusibility genes segregate according to Mendelian predictions, and two colonies must share only one allele at the locus concerned to be able to fuse. If two colonies do not share a fusibility allele, they will eventually reject each other, and tissue necrosis will result at the site of rejection to form a permanent physical barrier between the colonies. Because the fusibility gene is so highly polymorphic-various populations have been estimated to contain 80- 100 different alleles-fusibility and relatedness are highly correlated. Study Site: The Monterey marina is located on the south side of Monterey Bay, California; it is enclosed by the breakwater of Monterey Harbor, and by seawalls and a pier. The various parts of the marina, near the seawall and near the shore, experience different rates of water flow and wave motion. Tidal currents can be strong (A. Carwile, 1989). Samples were collected from as many different parts of the marina as possible to avoid biasing the results in favor of phenomena specific to ar¬ eas of unusually high or low water movement. Arbitrary Primer Experiments Primers: Oligodeoxynucleotide primers obtained from Operon Technologies were used. 15 primers, each 10 nucleotides in length and each containing a restric- tion-enzyme site, were surveyed. DNA Isolation: DNA was isolated from colonies of Botryllus schlosseri col¬ lected from the Monterey marina, from Bodega Bay, and from the Santa Cruz ma¬ rina, and from colonies of known relatedness grown in tanks in the laboratory. Isolation was performed as follows: whole animals were frozen in liquid nitrogen, then powdered in liquid nitrogen using a mortar and pestle. Powdered samples were suspended in a buffer of 250 mM Tris-borate pH 8.3, 50 mM EDTA, 2% SDS, and 10 mM Nacl. 500 ug/ml proteinase K was added to each sample and the sam- ples were incubated at 55° C for 2 hours. Then each sample was extracted once with phenol:CHCl3: isoamyl alcohol (25: 4.8: 0.2, v/v), once with phenol: CHClz: isoamyl alcohol (25: 24: 1, v/v), and once with CHCl3: isoamyl alcohol (24: 1, v/v). Nucleic acids were precipitated with 0. 1 vol. 2M sodium acetate and 2 vol. ethanol and spooled immediately with a sterile Pasteur pipette, then resuspended in 1 to 3 ml of TE buffer (10 mM Tris-Cl pH 8.3, 10 mM EDTA). RNA was removed by incubation with DNase-free RNase A (50 ug/ml at 37° C for »2 hours). Samples were then ex¬ tracted twice with phenol: CHClz: isoamyl alcohol (25: 24: 1, v/v) and twice with 1 vol anhydrous ether. The DNA was again precipitated with 0. 1 vol 2M sodium ac etate and 2 vol ethanol, spooled with sterile Pasteur pipettes, washed in 70% ethanol, and resuspended in 1 ml TE (10 mM Tris-Cl pH 8.3, 1 mM EDTA). DNA concentrations and purity were determined by reading absorbances at 260 and 280 nm, and by electrophoresis on 0.8 % agarose gels (after Kumar et al., 1988 with modi¬ fications of Simona Bartl, pers. comm.). Amplification conditions: Amplification reactions were performed in vol¬ umes of 25 ul containing 10 mM Tris-Cl pH 8.3, 50 mM KCl, 2 mM MgCl, 200 uM each of dATP, dCTP, dGTP, and dTTP (Pharmacia), 15 ng primer, 25 ng of genomic DNA, and 1 unit of Taq DNA polymerase (Perkin-Elmer/Cetus). DNA was dena¬ tured at 94° C for approximately 5 min before amplification. Amplification was per¬ formed in a Perkin-Elmer/Cetus DNA Thermal Cycler programmed for 30 cycles of 1 min at 94°, 1 min at 36°, and 3 min at 72°, using the fastest available transitions be¬ tween each temperature. Amplification products were analyzed by electrophoresis in 1.4% agarose gels and detected by staining with ethidium bromide. Histocompatibility Assays Colonies of Botryllus schlosseri were collected in the Monterey marina; colonies were generally found growing on algae under the sides of the floating docks. Neighboring colonies were collected within 0-10 cm of each other along a horizontal transect. Non-neighboring colonies were collected at least 3 m apart. The colonies collected from the wild were tied to glass slides, placed in slide racks, and maintained in running raw seawater aquaria. Oozooids were collected from parent colonies by placing blank glass slides (settlement slides) in the racks approxi¬ mately 2.5 cm from the parent colony. Most of the larvae which hatched from the parent colonies settled on these slides shortly after release. Settlement slides were removed to another running raw seawater tank, then cleaned and transferred to standing filtered seawater prior to being used in the experiment. The oozooids collected were allowed to grow for 7-21 days and were set up in pairs on fresh slides to test for fusibility. Oozooids were removed from settlement slides by carefully scraping them off with a piece of razor blade, picked up in a drop¬ per and placed on a fresh slide, then checked to make sure they were right side up; excess water was removed with a piece of tissue. The paired oozooids were placed Imm or less apart, one to three pairs per slide. The transplanted oozooids were then either placed in a moisture chamber for 3-5 min or placed immediately in standing seawater. Äfter 5-10 days, pairs were scored for fusion or non-fusion. The oozooids used for these experiments were obtained from 12 separate par- ent colonies obtained from the field. Four pairs of parent colonies were neighbors (collected «10 cm apart), and four pairs were non-neighbors (collected »3m apart). Eight different colonies were used for the neighbor pairs; two of these colonies were also used as parents for the non-neighbor pairs, along with six other colonies. From each pair of parents, 20 pairs of oozooids were set up, for a total of 160 pairs of oozooids. Because some of the oozooids did not adhere well to their slides, died after transfer, or did not make tunic-to-tunic contact during the period of my obser¬ vations, sample sizes are smaller than the total number of oozooid pairs originally set up. RESULTS Arbitrary Primer Experiments Figure 1 shows the results of an experiment in which five primers were screened for amplification of segments of genomic DNA from two different Botryllus samples, one taken from Bodega Bay and one from the Santa Cruz marina. For some of the primers, most of the bands which were amplified from one sample were also amplified from the other, but for other primers several bands were pro¬ duced for one sample and none for the other. A total of 37 different bands were de¬ tected. All primers were also used in control reactions without genomic DNA; no primer artefacts were observed. Thus, it would appear that genetic differences can be distinguished using this technique. Ten other primers (data not shown) were also screened before these results were obtained; none gave any significant amplification products. After obtaining preliminary positive results, the next step was to refine the protocol to obtain sharper banding patterns and to test for reproducibility by repeat¬ ing the same amplification reactions. Williams et al. (1990) used a dNTP concentra¬ tion of 100 uM instead of 200 uM; the higher concentration was used by Welsh and McClelland (1990). Williams et al. (1990) also recommended reducing the concen- tration of either genomic DNA or polymerase in order to resolve a non-specific "smear" of amplification products into distinct bands. Figure 2 shows the amplifica¬ tion products of reactions were carried out using 100 uM dNTPs and varying con¬ centrations of DNA, MgCl, and Taq polymerase. The lower dNTP concentration produced much fewer bands than the higher concentration used at first. Welsh and McClelland, who used longer primers originally designed for other purposes, employed a similar temperature profile to the one described above, but ran the reactions for 45 cycles. Thus PCRs were allowed to run for 45 cycles of the standard temperature profile rather than 30 (see figure 3); this gave more ampli¬ 11 F-12 F-13 F-14 F.15 FGURE Figure 1: Lanes 1,3: molecular weight markers. Lanes 6, 9, 12, 15, 18: DNA sample isolated from a colony collected at Bodega Bay. Lanes 7, 10, 13, 16, 19: collected at Santa Cruz. Lanes 5, 8, 11, 14, 17: control lanes; no genomic DNA in reaction mixture. Five different primers were used as indicated. Figure 2: IdNTPsl = 100 uM Throughout. Lane 1: Molecular weight markers (lambda HindIII). Concentrations of magnesium, DNA, and Taq polymerase varied; two different primers and two different DNAs used. Note indistinct banding as compared to figure 1. Figure 3: 45 cycles of amplification. Lane 1: Molecular weight markers (lambda HindIII). Lanes 3-9: Primer F¬ 14; Lanes 10-16: Primer F-12. Lanes 4, 11: DNA from Bodega Bay (as in fig. 1) at 25ng. Lanes 5, 12: DNA from Santa Cruz (as in fig. 1) at 25ng. Lanes 6, 13: DNA from Bodega Bay (as in fig. 1) at 12.5ng. Lanes 7, 14: DNA from Santa Cruz (as in fig. 1) at 12.5ng. Lanes 8, 15: DNA from Bodega Bay (as in fig. 1) at 2.5ng. Lanes 9, 16: DNA from Santa Cruz (as in fig. 1) at 2.5ng. Lanes 3, 10: Genomic DNA omitted. -. markerg (XHindIIL) 3 DNAS; Rasa ap; 3 primers W32 3 686 Figure 4: Test PCR reactions. Lane 1: lambda HindIII molecular weight markers. Lane 2: empty. Lane 3: Ras 2 and 3 primers; genomic DNA omitted. Lanes 3-6: 3 different genomic DNAs; Ras 2 and 3 primers. Note the appearance of the expected 800 base¬ pair amplification product. fication, and therefore brighter bands, but seemed to give more non-specific amplifi¬ cation. In this experiment, DNA concentration was also varied; banding patterns, however, became more indistinct, and many bands disappeared entirely, as DNA concentration was reduced. Unfortunately, despite these encouraging preliminary results, the arbitrarily- primed PCR technique proved, under my conditions, to be unreliable, and it was very difficult to get reproducible results. I was unable to complete my experiments designed to refine the procedure to produce clear, reproducible banding patterns be¬ cause the primers simply ceased to amplify my DNA samples. Various experiments were performed to determine which, if any, of my reagents was responsible for the problem; test PCRs showed no difference between different stocks of dNTPs, primers, DNA and Taq polymerase (data not shown). Finally, a series of test reac¬ tions was performed using Ras 2 and Ras 3 primers designed to amplify a specific 800 base-pair product from Botryllus. Figure 4 shows that the DNAS I isolated and an¬ other DNA sample obtained from elsewhere all gave the expected product with the Ras 2 and Ras 3 primers. (Ras 2 and Ras 3 primers and DNA sample W32 were ob¬ tained from Simona Bartl.) Histocompatibility Assays Tables 1 and 2 show raw data from pairs of oozooids scored for fusion or rejec¬ tion. There was considerable variation in the frequency of successful transfer from one group of oozooids to another. Also, some of the data comes from oozooid pairs whose progress was followed for as long as 11 days, while other pairs were followed only for 9 days. Figure 6 shows frequencies of fusion and non-fusion for the oozooid pairs set up for each pair of parents. I used the Rx C test of independence to calculate G-values (Sokal and Rohlf 1981a, b) for the two groups of offspring, those from neighboring parents and those from non-neighbors. The frequencies for the rates of fusion between offspring of near neighbors varied significantly between dif- ferent pairs of parents (p«0.001), as did those for the offspring of non-neighboring parents (p20.001). My pooled data for all pairs of oozooids from neighboring parents and for all pairs from non-neighbors are compared in Figure 7. The frequency of fu¬ sion for offspring of neighbors varied significantly from the frequency of fusion be¬ tween offspring of non-neighbors, according to the G-test of independence (p«0.001) (Sokal and Rohlf 1981a, b). Parents Fusions Non-Fusions 6854685 687-688 10 7254726 664+665 Table 1: Fusion vs. Non-Fusion in Neighbors Fusions Parents Non-Fusions 6834708 6881669 6874693 10 6584682 14 Table 2: Fusion vs. Non-Fusion in Non-Neighbors Figure 6: Rates of Fusion and Non-Fusion Z Z V H Z Fused Pairs Non-Fused Pairs V Number of Oozoid Pairs Scored Figure 7: Rates of Fusion in Neighbors and Non-Neighbors 50 - 30 - Fusions Non-Fusions Neighbors Non-Neighbors DISCUSSION My results are clearly limited in their usefulness because of the small number of colonies I collected from the field and because of the difficulties I experienced in refining the arbitrarily-primed PCR techniques. Nevertheless, I was able to show that genetic diversity could be detected among genomic DNA samples from Botryllus schlosseri, and to come to some preliminary conclusions about the most effective techniques for identifying diagnostic DNA markers for population genetics studies, using arbitrarily primed PCR. The data from my tests of fusibility in Botryllus oozooids indicates a negative correlation between the distance between two colonies and their degree of relatedness. This implies that limited larval disper¬ sal and philopatry do indeed lead to grouping of related colonies and to inbreeding. In my experiments with the arbitrary primers, I found that the conditions recommended by Williams et al. (1990a) gave the best results, except that the dNTT concentration of 100uM which they used gave very poor results. Instead, I followed Welsh and McClelland (1990) in using 200uM dNTPs. I also found that 30 cycles of amplification gave more specific amplification and hence clearer banding patterns than 45 cycles. Furthermore, I found that decreasing the concentration of DNA and/or Taq polymerase, which Williams et al. (1990) recommend for resolving dis¬ crete amplification products with primers which give a smear under normal condi- tions, eliminated some bands and did not materially improve the specificity of am¬ plification. I obtained an average of 2.5 products per primer, as opposed to an aver¬ age of 4.3 products per primer reported by Martin et al. (1991), and found that only 20% of the primers I screened gave any amplification, as opposed to the 60%-90% success rate reported by Operon. Finally, I found the arbitrarily-primed PCR tech- nique to be very difficult to use for repeated tests and, under my conditions, quite unpredictable. It is hard to say what elements of my protocol were responsible for these problems; however, it is possible that the DNA samples I isolated were not pure enough or had suffered too much "nicking" to be useful for arbitrarily primed PCR, although they were perfectly sufficient for normal PCR reactions using specific primers. The data from my experiments with oozooid fusibility rates are also difficult to interpret for various reasons. First of all, the significant variance among the rates of fusion of offspring of pairs of neighbors and of non-neighbors makes it unclear whether the overall pattern I observed was actually due to the effects of philopatric dispersal and inbreeding or simply to chance. It would be necessary to repeat these tests for many more pairs of parents to be certain one way or the other. Secondly, when 1used the G- test to compare the results of the tests which used the offspring of one parent and those of a neighbor and a non-neighbor (see tables 1 and 2), the difference was not significant due to the small size of the sample. However, the pooled data gave me a large enough sample to determine that I had indeed found a significant difference in fusibility frequencies. Beyond problems of variability and small sample size, the oozooid fusion as¬ says present intrinsic difficulties of interpretation. Since it is impossible to deter¬ mine whether the fusibility alleles two oozooids share came from the maternal or the paternal parent, these data cannot be used to calculate directly the degree of re¬ latedness of the parent colonies I collected. Furthermore, the rate of fusibility over¬ all is much higher than would be expected from a population containing 80-100 alle¬ les in Hardy-Weinberg equilibrium. The overall rate of fusion for all my tests com¬ bined is 32%, and for the offspring of non-neighbor pairs the rate of fusion is 29%. This result reflects either a) a significantly smaller number of alleles than expected segregating in the Marina population; or b) a small number of alleles with very high frequencies, while the rest are present at such low frequencies that they did not con¬ tribute significantly to my data. A population in which 10 alleles were segregating at equal frequencies would have a total frequency of fusion of 32%. 100 alleles segre¬ gating in Hardy-Weinberg equilibrium would give a frequency of fusion of 3.9%. (See Appendix) Clearly, my data reflects not only philopatric dispersal but also a baseline con¬ dition of significant inbreeding or very low polymorphism at the fusibility locus as compared to other Botryllus populations studied. A possible explanation for this phenomenon is the fact that Botryllus colonies of a limited range of known histo¬ compatibility types are put out on slides in the marina as part of ongoing work with the phenomenon of self/non-self recognition. None of my samples were collected from the area where these introductions are made, but nevertheless this interfer- ence may have skewed allele frequencies throughout the Marina over time. More work needs to be done on the ecology and population genetics of Botryllus before its complex life cycle will be well-understood. In particular, further studies of the frequency of chimerism in the wild, and of the effects of different wa¬ ter-flow regimes on larval dispersal, would shed light on the issues I have raised. The use of DNA markers, whether through arbitrarily-primed PCR or otherwise, could help to elucidate these problems. Since arbitrarily-primed PCR detects mark- ers which are transmitted as dominants, chimeric colonies might well be distin¬ guishable because they would give greater than the normal number of amplification products. If the technique was used extensively, it might be possible to use it to con¬ struct family trees of individuals within a population and therefore discover its pre¬ cise genetic structure. Another possibility, more closely related to the issue of histo¬ compatibility genes in Botryllus, is that arbitrarily-primed PCR might be used to find genetic markers closely linked to fusibility genes in already-existing laboratory stocks of Botryllus. Such markers could then be used as a starting point for sequencing ef- forts aimed at isolating the genes responsible for self/non-self discrimination. Such work is already under way, using more conventional avenues of approach, but this technique might prove a useful alternative. ACKNOWLEDGEMENTS 1 would like to thank Simona Bartl for giving so generously of her time and of her expertise. Without her help and encouragement, which went above and be¬ yond the call of duty, this project would have been impossible for me to plan, let alone carry out. I would also like to thank Kathi Ishizuka and Karla Pateri, who gave me generous and patient assistance with my work on live Botryllus. Nanette Chadwick gave invaluable assistance with the design of my Botryllus oozooid exper iments and with statistical analysis. The entire Powers lab at Hopkins Marine sta¬ tion deserves my gratitude for putting up with my questions and difficulties. Finally, thanks are due to my adviser Dennis Powers, who was willing to put a rather complex and difficult project into the hands of a completely untried under¬ graduate and saw to it that I was never lacking for guidance. BIBLIOGRAPHY Carwile, A. 1989. Settlement of larvae, colony growth and longevity in three species of ascidians and the effect on the species composition of a marine fouling commu¬ nity. Ph.D. Diss., University of California, Los Angeles. Grosberg, R. K. 1987. Limited dispersal and proximity-dependent mating success in the colonial ascidian Botryllus schlosseri. Evolution. 41(2): 372-384. —— and J. F. Quinn. 1986. The genetic control and consequences of kin recognition by the larvae of a colonial marine invertebrate. Nature. 322: 456-459. Welsh, J. and M. McClelland. 1990. Fingerprinting genomes using PCR with arbi¬ trary primers. Nucleic Acids Res. 18(24): 7213-7218. Williams, J. G. K., A. R. Kubelik, K. J. Livak, J. A. Rafalski, and S. V. Tingey. 1990. DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Res. 18 (22): 6531-6535. Rinkevich, B. and I. L. Weissman. 1989. Variation in the outcomes following chimera formation in the colonial tunicate Botryllus schlosseri. Bull. Mar. Sci. 45(2): 213-227. Kumar, S., B. M. Degnan, I. L. Ross, C. J. Hawkins, and M. F. Lavin. 1988. Isolation of DNA and RNA from ascidians. Mar. Biol. 98: 95-100. Arnold, M. L., C. M. Buckner, and J. J. Robinson. 1991. Pollen-mediated introgres¬ sion and hybrid speciation in Louisiana irises. Proc. Natl. Acad. Sci. USA. 88: 1398- 1402. Martin, G. B., J. G. K. Williams and S. D. Tanksley. 1991. Rapid identification of markers linked to a Pseudomonas resistance gene in tomato by using random primers and near-isogenic lines. Proc. Natl. Acad. Sci. USA. 88:2336-2340. Appendix: Calculations of Fusibility Frequencies Based on Number of Fusibility Alleles Segregating in a Population Iused the equation x2+ ((Ex)-X) (2x-1) (ZX)2 to determine the frequency of fusion in a population with x alleles segregating at equal frequencies, based on the reasoning that the number of alleles squared is the number of fusion between homozygotes, there being one possible homozygote for each allele present, and that each homozygote can fuse with itself and x - 1 other genotypes; that the number of heterozygotes equals the summation of (1+ 2 + 3 +... +x - 1), and each heterozygote will fuse with (2x - 1) other genotypes; and that the to¬ tal number of different possible pairings of two different genotypes equals (1 + 2 + 3 +...+X)2.