Blazyk1 ABSTRACT Potential for gene flow in marine environments is often much higher than what is actually observed, due to various barriers to dispersal. Understanding the divisions between populations of organisms is important to fishery regulations and designing effective marine reserves. Tegula funebralis have a pelagic larval phase, lasting 5-14 days, giving them a moderate potential for dispersal. Samples ranging from Oregon to Santa Barbara were collected from seven geographic locations. A 700 base pair region of the mitochondrial cytochrome oxidase I gene (COI) was sequenced for an average of 9 individuals from each location. Comparing these sequences, there is no clear overall population structure, although there seem to be some slight patterns corresponding to geographic location. In particular, snails from Oregon are more similar to each other and are more different from snails from other populations, suggesting a lower level of dispersal between T. funebralis from Oregon and the more southern locations. These smaller signals could not be shown to be significant with the current sample sizes, using Fsr values or chi-squared tests, and had p values of between 0.07 and 0.10. Overall, there appears to be a fairly high level of gene flow across the range studied, which makes it unlikely that there are major barriers to dispersal in this range. However, there may be subtle population structure in T. funebralis along the west coast of North America that will emerge when larger sample sizes from each location are used, a larger portion of the gene is sequenced, and populations further to the north and south are sampled. Blazyk 2 INTRODUCTION Many marine invertebrates are sessile and may live their entire adult lives within a few square meters of space. Therefore, the genetic makeup of populations is largely due the nature of larval dispersal and development (Hoskin 1996). The pelagic larval phase of some invertebrates is the one chance that animal has to spread its genes to another (possibly quite distant) location. It is then to be expected that invertebrates that lay egg cases would have genetically distinct populations on a much smaller scale than species that have pelagic larvae. Pelagic larvae of invertebrates can drift in the oceans for anywhere between a few days to several months. Off the west coast of North America, where the California Current moves at an average rate of 10 cm/s (Debenham et al. 2000), larvae have the potential to move from tens to thousands of kilometers. However, in many cases, species exhibit a more limited dispersal than is predicted by the duration of their pelagic larval phase. This can be explained by natural selection, larval behavior, including vertical migration and swimming, eddies and other counteracting currents, and by possible physical boundaries to dispersal. Because it is very difficult to directly measure larval dispersal distances, other, indirect measures are used. One such tool is analyzing genetic differences between populations. This evaluation, when compared to genetic variation within populations, gives a good idea of the amount of larval exchange between populations. One migrant per generation prevents the accretion of large genetic differences, and ten migrants per generation stops all but very minor differences from appearing. Possible barriers to dispersal include Point Conception, where the California coast turns sharply eastward, moving away from the stream of the California Current, Monterey Bay, and San Francisco Bay. However, these features seem to not always impede dispersal. Blazyk 3 Miner (2002) found that although there appear to be two distinct physiological races of Pollicipes polymerus (gooseneck barnacle, with a pelagic larval phase of up to 40 days), which divide at Point Conception, there were no differences in the genetic makeup of individuals spanning this range. Tetraclita squamosa rubescens (pink barnacle) appear to have high gene flow along a range from San Francisco to south of Point Conception. However, the population at Moss Landing, at the center of Monterey Bay, does seem to have a higher level of isolation than the rest, which could indicate more larval retention in some populations than others (Ford and Mitton 1993). Debenham et al. (2000) shows that in Strongylocentrotus franciscanus (red sea urchin, with a pelagic larval phase of 61 to 131 days), there is high gene flow throughout the species range from Alaska to Baja California. Samples from Alaska did seem to deviate slightly from the rest, possibly indicating a more limited exchange of larvae, selection, or evidence of selective mating. In a study of Haliotis cracherodii (black abalone, with a pelagic larval phase of 5 to 15 days), Hamm and Burton (2000) conclude that there are significant differences between populations of abalone between Santa Cruz and Santa Barbara, although genetic difference does not correlate with geographic distance. This indicates that black abalone might be fairly poor in dispersing larvae even moderate distances, which is consistent with observations of pink and green abalone dispersal. Buonaccorsi et al. (2002) studied Sebastes caurinus (copper rockfish, with a pelagic phase of several months) from Canada to south of Point Conception. Among these samples, there was evidence of significant divisions between populations, as well as a positive correlation between geographic and genetic distance. Blazyk 4 Discovering the population structure of various species may be important for its own sake, but it has implications for fisheries policy and designing marine reserves as well. In many areas, for example, black abalone populations are severely depleted and are being protected from human harvesting. However, because they have very limited dispersal (and hence have few incoming larvae from other, less depleted areas), these populations may not be able to recover on their own (Hamm and Burton 2000) Marine reserves are designed to have positive effects beyond their boundaries, especially when meant to restore commercial species, where the economic value of the reserve depends on export from nearby areas where fishing is allowed. If there is little dispersal between a reserve and the area outside of it, this will not improve the overall fishery. However, if a high percentage of larvae or adults migrate out of the reserve, the reserve loses its effectiveness. A successful self-seeding reserve needs to be as big as the mean larval dispersal of the targeted species. This is not always possible (especially since some species have potential dispersal of hundreds of kilometers), so in some cases, a better alternative is a series of stepping stone reserves (Palumbi 2003). Tegula funebralis has a pelagic larval period of about 5 to 14 days (Moran 1997), which gives it a moderate potential for dispersal (about 40 to 120 kilometers). It ranges from Vancouver Island south along the west coast to central Baja California. Since it spans several geographic features proposed to be barriers to dispersal and has a relatively short pelagic period, it seemed that Tegula funebralis had the potential for interesting population structure. Here, I compared a region of the mitochondrial cytochrome oxidase 1 (COl) gene of seven populations of Tegula funebralis between Oregon and southern California. I found no significant differences between populations, although there was a significant positive correlation between genetic and Blazyk 5 geographic difference, which indicates that more structure may resolve with larger sample sizes and additional populations further to the north and south. MATERIALS AND METHODS Populations sampled T. funebralis were collected from seven sites along the west coast of the United States (Table 1; Fig. 1). These areas include Strawberry Hill, Oregon (44°15’42'’N, 124°07’36°W), Bodega Bay, California (38°19’N, 123°04’W), just outside the breakwater at Pillar Point, near Half Moon Bay, California (37°29’42"’N, 122°29’42"’W), exposed and protected regions of the rocky intertidal at Hopkins Marine Station, Pacific Grove, California (36°37’N, 121°54’W), the rocky intertidal at Soberanes Point, California (36°27’092’N, 121°55’45’’W), San Luis Harbor, San Luis Obispo, California (35°10’36'’N 120°45’36’’W), and Lompoc, California (34°43’N, 120°34’) DNA isolation To isolate DNA, all snails were relaxed in a 1:3 mixture of isotonic MgClz and seawater. Once relaxed, small bits of tissue were snipped from the edge of the snails' feet. This tissue was then processed with the Nucleospin kit, extracting relatively clean DNA. The products were run on a 2% agarose gel, showing bright bands corresponding to DNA. The extracted DNA was diluted with water to a 1:10 solution, which was used in PCR. mtDNA amplification 1 ul of the 1:10 DNA was used as a template in a 25 ul polymerase chain reaction comprising 2.5 ul 1Ox AmpliTaq buffer, 2.5 ul 1Ox dNTPs, 0.30 ul AmpliTaq polymerase, and 1.25 ul each of the forward and reverse primers. Amplification conditions were as in Hellberg Blazyk 6 (1998)—50 sec at 94°C, 90 sec at 50’C, and 90 sec at 72°C, but for 30 cycles instead of 25 and 700 without the 8 min extension time at /z’C. The Folmer et al. (1994) primers: HCO2198 (5’- TAACTTCAGGGTGACCAAAAAATCA-3’) and LCO1490 (S’GGTCAACAAATCATA AGATATTGG-3’) were used to amplify a section of cytochrome c oxidase I (COI). Some other primers were tried, without success, including LCO/HCO (which are the same as 1490/2198 but are not degenerate), COla/COlf, and 2198/conch2f. PCR products were run in 2% agarose gels to identify the samples that successfully amplified, producing a distinct band, about 700bp in length. DNA sequencing To remove excess dNTPs and primers, 5 ul of PCR products of successfully amplified samples were then added to 2 ul shrimp alkaline phosphotase (SAP), 1 ul exonuclease 1 (EXÖ), and 0.5 ul SAP buffer. The samples are then held at 37°C for 30 minutes and 80°C for 15 minutes. To add fluorescent markers for sequencing, 2 ul of the SAP/EXO products were added to 6 ul dd H20, 1 ul big dye, 1.5 ul 5X buffer, and 0.5 ul of either 1490 or 2198 (for forward and reverse sequences). The samples were then cycle sequenced with 25 cycles of heating to 96°C for 10 seconds, 50°C for 5 seconds, and 60°C for 4 minutes. After the samples cycle sequenced, 40 ul 75% isopropanol was added to each, and they were let to stand for 15 minutes. The samples were then spun at room temperature at about 3000 rpm for an hour to form DNA pellets. The tubes were then uncapped, turned over, and spun for two minutes at 700 rpm to remove excess fluid. To resuspend the DNA, 20 ul of hi-di was added to each tube. The tubes were then vortexed to ensure the DNA mixed in the hi-di and then briefly spun to bring all the liquid to the bottom of the tubes. The samples were heated at 96°C Blazyk 7 for two minutes to denature the DNA. Äfter cooling to room temperature, the samples were transferred to a sequencing plate and sequencing using an ABI3 100 Genetic Analyzer. Post-sequencing analysis Äfter sequencing, the files were cleaned and sorted, using Sequencher 4.1. The forward and reverse files for each snail were combined to create one consensus sequence per snail. These consensuses were accumulated into a master contig, which was used in making trees and in statistical analysis. PAUP*4.Ob10 (PPC) was used to make parsimony trees with branch lengths in terms of numbers of nucleotide changes. Structure among clades was analyzed using contingency tables and chi-squared values with the program ChiSquare v1.0. To compare diversity within sequences from one location to diversity of sequences among two or more locations, Fsr values were calculated. This was done using heap big Power Mac, programmed by Stephen Palumbi. Pairwise Fsr values (between each pair of locations) were then correlated with distance between populations to examine whether values increased with distance, as would be expected by the isolation by distance model. RESULTS Amplification and Sequencing The first several PCR conditions tried were unsuccessful, producing no visible bands of products, including the positive control (Fig. 2). In these unsuccessful reactions, I was using 0.15 ul of enzyme (both AmpliTag and DNAzymeEXT), as well as a number of different primer pairs, including 1490/2198, LCO/HCO, COla/COIf, and conch 2F/2198. After doubling the amount of enzyme per reaction to 0.30 ul, and using the degenerate primers 1490 and 2198, I was finally able to consistently produce successful reactions (Fig. 3). In this figure, bands of Blazyk 8 amplified DNA about 700bp long are visible. About 90% of the PCR products sequenced produced clean enough data to be used in analysis. In total, there were 67 sequences from snails from seven locations, giving an average of about ten sequences per location. Parsimony and Minimum Spanning Trees I first created a parsimony tree with the 67 sequences I had produced, plus the published sequence for Tegula funebralis COI found on GENBANK, which represents an individual from Baja California (Fig. 4). This tree is one of a hundred randomly generated trees, all of which contained the same major clades, including matching groupings of snails. These trees represent the most parsimonious relationship between the snails, meaning that each base change occurs as few times as possible (ideally, just once). This same data can be viewed in another way as a minimum spanning tree (Fig. 5). Here, identical sequences are grouped together in circles, with divergent sequences radiating out from those circles on branches proportional in length to the number of nucleotide changes that occur in the sequence. Statistical Analysis First, I selected three major clades from the parsimony tree (Fig. 4). I chose these clades because they seemed to be the largest groupings of sequences. Most of the other branches on the tree led to only one or two individual snail sequences. I then examined the makeup of these major clades by creating a contingency table (Table 2). To test whether the distribution of clades in any location was significantly different from the total, I used a chi-squared test. The true value of chi-square was 9.828, with p = 0.08 + 0.05 (not significant). To compare diversity between locations to diversity within locations, l used the measure of Fsr. Including all seven populations as separate demes, Fsr was negative (although not significantly different from zero). I then calculated pairwise Fsr values for all pairs of locations Blazyk 9 (Tables 3 and 4) and graphed these values over distance (Fig. 6). Here, the linear regression line had an R“ of 0.5035 and a slope of 1x10“, with p = 0.000. Changing negative values of Fsr to zero (Fig. 7), the linear regression had an R* of 0.5894 and a slope of 5x10%, still with p - 0.000. DISCUSSION From the parsimony tree, no striking structure or clades corresponding to locations are evident. There appear to be individuals from each location spread throughout the tree. However, upon closer inspection, some subtle patterns are noticeable. For example, Clade II is made up of eight individuals, four of which are from the Oregon population. None of the other four members of this clade is from south of Soberanes. With current sample sizes, this is not conclusive, yet may suggest that structure is not truly lacking. The results of the chi-squared test show that there is not a significant difference between the distribution of snails from Oregon, Lompoc, and the remaining samples across the three clades specified. The test was not too far from significant, with p - 0.08, giving further reason to think that a larger sample size could resolve some hints of differences in structure between populations. Using the principles of the island model, in which all populations have an equal chance to exchange individuals, Wright and Malécot first developed analyses of the relatedness of populations in the 1940s. Wright, in fact, derived the formula for Fsr, relating it to population size and migration rate (Fsr = 1/(4Nm+1)). The “st“ stands for the relationship between the subpopulation and the total population, and is a type of "inbreeding coefficient" (Templeton 2003). This has now become one of the most standard measures in population genetics to show population structure. Blazyk 10 To examine the relationship between genetic variation within and between populations, calculated Fsr values. To obtain an Fsr for two or more populations, the ratio between average genetic variation within populations and between populations is subtracted from 1. If there is a large difference between populations compared to within, the Fsr will approach 1, while if two populations are not genetically different, Fsr will be approximately 0. In this study, comparing all seven populations produced a negative Fsr (not statistically different from zero), which indicates that there is a high level of gene flow between the populations. 1 thought that this might be a result of some populations obscuring differences in others, so I decided to calculate pairwise Fsr values for all sets of locations. The majority of these were negative, except the values between the Oregon populations and the other locations. This led me to believe that there might be a correlation between Fsr and geographic distance between populations. Strawberry Hill, Oregon is much farther from the other locations than they are from one another. Indeed, when I graphed pairwise Fsr values against distance and ran a linear regression, there was a significant positive slope, both when 1 included negative values of Fsr and when I changed the negative values to zero (which should be the minimum value of Fsr). By examining the slope of the regression lines on the plots of Fsr versus distance, it is possible to estimate the average dispersal distance of that species. In both Fig. 6 and Fig. 7, the Fsr is about 0.04 at a distance of 1000km. I compared this value to a calibration graph (Palumbi 2003), which predicted a mean larval dispersal of about 40 kilometers. This value is much less than the hypothetical maximum (over 120 km), which may suggest that factors other than the California current affect dispersal. The graphs of Fsr versus distance indicate that although there may be a fairly high level of gene flow along the American west coast (based on the fairly homogenous phylogenetic trees Blazyk 11 and the mostly small/insignificant Fsr values), there is probably an isolation by distance effect acting as well. Isolation by distance is based on the stepping stone population model, which is the theory that gene flow is restricted with increasing geographic distance, resulting in a genetic structure. This is likely more realistic than the island model, since populations very far apart would be expected to exchange fewer individuals than closer populations, even if the species has a moderate or high dispersal potential. I then calculated values of Nm for each of the pairwise combinations including Oregon that had a positive Fsr (using the formula Fsr = 1/(4Nm+1)). The Oregon and Lompoc populations gave an Nm of about 3, which represents the number of migrants per generation between those two populations (Bohonak 1999). Exchange rates between Oregon and the other locations had an average of about 7 migrants per generation, although one site suggested an Nm as high as 40. Even with a rate as low as 3-7 migrants per generation, any major genetic differences between populations would be prevented. Numerous studies mentioned previously, as well as my results, support the likelihood that many species with pelagic larvae have relatively high gene flow along this coast. Ford and Mitton (1993) and Debenham et al. (2000) showed that Point Conception and Monterey Bay do not obstruct dispersal of the pink barnacle or the red sea urchin, respectively. In the present study, suggested barriers such as Monterey Bay and San Francisco Bay appear to not impede dispersal very significantly, at least not to an extent detectable with the statistical power of my data. The connection between Tegula population structure and implications for marine reserves may not be immediately apparent, since it is unlikely a reserve would ever be created for the purpose of protecting these snails. However, many other marine species (including several Blazyk 12 crustaceans, algae, mollusks, and many fish) have a pelagic larval phase of similar duration or have comparable dispersal distances (Shanks 2003). This means that although Tegula funebralis may not be a target species for a reserve, their population structure may be representative of many other species that do need to be protected. Most species with pelagic larvae appear to have dispersal distances greater than 20 kilometers, with many on the order of hundreds of kilometers (Shanks et al. 2003). This means that it may often be impractical to create individual reserves of ideal size, remembering that the ideal is equal to the average larval dispersal distance for targeted species. Currently, there is no single no-take zone in the United States that is as large as 20-50 km (Palumbi 2001), let alone as large as would be required for species with long-lived pelagic larvae. An alternative design is a network of smaller reserves, although these must be close enough to allow significant dispersal and recruitment between reserves (Airamé et al. 2003). Another difficulty is that fact that delineating habitat as no-take reserves usually means directly taking away that area from fishermen. At the moment, no-take reserves only comprise about 0.1% of marine habitat in the US. In order to have controlled fish population sizes, it has been estimated that this area should increase to as much as 50%. This value is impractical, at least at this time, since it would result in serious declines in fishery harvests, despite growth of overall fish populations (Palumbi 2001). It is clear that some equilibrium must be met, for the current trend is that of fishing populations until they all but disappear and then moving on to another species. It is still unclear, however, whether setting aside 20% of marine habitat as reserves would provide a happy solution to both fishermen and declining fish populations, or if it would merely hurt the economic value of the fisheries while not being enough to regenerate declining marine populations. Blazyk 13 To increase the usefulness of this study, I would like to increase sample sizes from each location, while adding additional locations to the north and south, to investigate further the apparent effect of isolation by distance. Also, I plan to sequence additional sections of COI, or a section of another gene to isolate a greater number of polymorphic sites that would add power to fürther resolve genetic differences between populations. With current populations and sample sizes, there are no apparent significant differences, suggesting fairly high levels of gene flow. However, Fsr does increase significantly with distance, which shows that there is not unlimited exchange of larvae between all locations studied. Rather, there is a balance between fairly high larval exchange and isolation by distance in Tegula funebralis between Oregon and southern California. Blazyk 14 ACKNOWLEDGEMENTS I would like to thank Stephen Palumbi for being a great advisor and knowing the level of involvement that allowed me to learn and explore things myself, without ever feeling completely lost. Also, thanks to Emily, who also worked on snail genetics, and who knows everything about scientific labs, having spent many summers in such thrilling environments. I also owe a lot to everyone in the Palumbi lab—Vollmer, Julie, Laura, Cathy, Roxanna, Tom, and Adam. You guys always helped me out, answered my questions, and even took the time to speculate with me about Tegula genetics. I would also like to thank the other 175H professors—George Somero for being a wonderful advisor and helping me get snails from various locations, and Mark Denny and Jim Watanabe for general advice and input. Lastly, thanks to all the sea kids for making the first five weeks of the quarter raucous and fun, and to all the other 175H/176H students for the lovely, although somewhat quieter, times the rest of the quarter. Blazyk 15 LITERATURE CITED Airamé, Satie, Jenifer E. Dugan, Kevin D. Lafferty, Heather Leslie, Deborah A. McArdle, and Robert R. Warner. 2003. Applying ecological criteria to marine reserve design: a case study from the California Channel Islands. Ecological Applications. 13: S170-S184. Bohonak, Andrew J. 1999. Dispersal, gene flow, and population structure. The Quarterly Review of Biology. 74:21-45. Buonaccorsi, Vincent P., Carol A. Kimbrell, Eric A. Lynn, and Russell D. Vetter. 2002. Population structure of copper rockfish (Sebastes caurinus) reflects postglacial colonization and contemporary patterns of larval dispersal. Can. J. Fish. Aquat. Sci. 59: 1374-1384. Debenham, Patty, Mark Brzezinski, Kathy Foltz, and Steven Gaines. 2000. Genetic structure of populations of the red sea urchin, Strongylocentrotus franciscanus. J. of Experimental Marine Biology and Ecology. 253: 49-62. Folmer, O., M. Black, W. Hoeh, R. Lutz, and R. Vrijenhoek. 1994. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol. Mar. Biol. Biotech. 3: 294-299 Ford, Michael J. and Jeffry B. Mitton. 1993. Population structure of the pink barnacle, Tetraclita squamosa rubescens, along the California coast. Mol. Mar. Biol. Biotech. 2: 147-153. Hamm, D. E. and R. S. Burton. 2000. Population genetics of black abalone, Haliotis cracherodii, along the central California coast. J. of Experimental Marine Biology and Ecology. 254: 235-247. Hellberg, Michael E. 1998. Sympatric sea shells along the sea’s shore: the geography of speciation in the marine gastropod Tegula. Evolution. 52: 1311-1324. Blazyk 16 Hoskin, M. G. 1997. Effects of contrasting modes of larval development on the genetic structures of populations of three species of prosobranch gastropods. Marine Biology. 127: 647-656. Miner, Benjamin G. 2002. Are the two physiological races of Pollicipes polymerus (Cirripedia) genetically divided along the California coast? Invertebrate Biology. 121: 158-162. Moran, A. L. 1997. Spawning and larval development of the black turban snail Tegula funebralis (Prosobranchia: Trochidae). Marine Biology. 128: 107-114. Palumbi, Stephen R. 2001. The Ecology of Marine Protected Areas. pp. 509-530 in Mark D. Bertness, Steven D. Gaines, and Mark E. Hay, eds. Marine Community Ecology, Sinauer Associates, Inc., Sunderland, Massachusetts. Palumbi, Stephen R. 2003. Population genetics, demographic connectivity, and the design of marine reserves. Ecological Applications. 13: S146-S158. Shanks, Alan L., Brian A. Grantham, and Mark H. Carr. 2003. Propagule dispersal distance and the size and spacing of marine reserves. Ecological Applications. 13: S159-S169. Templeton, A. 2003. Gene flow and population subdivision. pp.6-1 to 6-34 in A. Templeton Population Genetics and Microevolutionary Theory, Wiley & Sons, in press. Blazyk 17 APPENDIX 1: TABLES Table 1: This table organizes data about the samples used in this project. It includes the "snail 1D" used in labeling samples in various trees and tests. Also listed are the collection locations, with latitude and longitude. Finally, the number of snails from each site is recorded, as well as the number of those that were successfully sequenced. Snail ID Collection Location Latitude Longitude sequenced 124°07°36”W ORI-ORIS 44°15°42”N Strawberry Hill, OR 11 38°19'N 123°04’W BBII-BB20 Bodega Bay, CA 10 37°29’42'’N 122°29°42°’W 10 HMBI-HMB10 Half Moon Bay, CA Hopkins Marine HPI-HP10; 36°37'N 20 121°54’W 15 Station, Pacific Grove, HEI-HE10 CA 36°27°09°N 121°5545°W Soberanes Point, CA 11 10 SPI-SP11 35°10’36”N 120°45°36”W 10 SLOI-SLO10 San Luis Obispo, CA LMII-LM20 34°43'N 120°34’ Lompoc, CA Blazyk 18 Table 2: This table shows the contingency table used in analysis. The phylogenetic tree (Fig. 4) was broken into 3 major clades—I, II, and III, with everything else as "Rest." The samples from Oregon (OR), Lompoc (LM), and everywhere else grouped together (Rest) are listed below with appropriate distributions between clades. Clade I II III Rest 4 OR 2 2 4 5 Rest 15 24 4 LM Blazyk 19 Table 3: This is a list of average nucleotide distance (diversity) within each location. Average diversity within location Location Code Location OR 0.00430 Strawberry Hill, OR BB 0.00530 Bodega Bay, CA HMB 0.00556 Pillar Point, Half Moon Bay, CA Hopkins Marine Station, Pacific HE/HP 0.00458 Grove, CA SP 0.00408 Soberanes Point, CA San Luis Harbor, San Luis SLO 0.00504 Obispo, CA 0.00259 LM Lompoc, CA Blazyk 20 Table 4: Below are average nucleotide differences within and between pairs of locations, which are used to calculate Fsr values, also listed. In addition, distances between the pairs of locations are given in kilometers. Diversity between Distance between Avg. diversity Locations FST locations locations (km) within locations 0.003445 0.00368 1103.4 0.06556 ORLM 0.00485 0.05675 ORSLO 0.00507 1049.5 0.00419 ORSP 0.00411 -0.01946 887.4 ORH 0.00444 870.0 0.03986 0.00462 ORHMB 0.00493 0.00496 0.00605 764.3 0.00502 ORBB 0.00480 666.5 0.04382 0.00383 -0.03003 0.003945 458.1 BB/LM -0.07000 0.00535 0.00539 404.8 BB/SLO -0.05157 0.00469 BB/SP 0.00446 230.2 -0.04000 0.00494 0.00475 BB/H 215.1 -0.09919 0.00494 BB/HMB 0.00543 104.2 0.00379 353.9 -0.07520 HMB/LM 0.004075 0.00539 -0.01670 HMBSLO 0.00548 300.8 -0.08559 HMB/SP 0.00444 126.3 0.00482 0.00475 110.9 -0.06737 HMB/H 0.00507 -0.01558 0.00353 0.003585 243.0 HLM 0.00499 190.1 0.00500 H/SLO 0.00200 -0.05353 H/SP 0.00411 18.4 0.00433 -0.01988 0.00327 0.003335 228.8 SP/LM 0.01660 SP/SLO 0.00474 0.00482 176.6 0.01396 0.003995 0.00394 54.1 SLO/LM Blazyk 21 APPENDIX 2: FIGURES Fig. 1: Map of the west coast of the United States with collection sites marked in red. Fig. 2: Example of an unsuccessful PCR, with products run on a 2% agarose gel. There are no clear bands, even for the positive control. The lane of many bands is the 1kB standard. Fig. 3: Example of a successful PRC, with products run on a 2% agarose gel. There are clear bands at about 700bp for most samples, including the positive control, and an empty lane for the negative control. The 1kB standard is in the first lane. Fig. 4: Parsimony tree produced with all 68 sequences. Horizontal branch lengths represent numbers of nucleotide changes. Included are three major clades used in analysis. Fig. 5: Minimum spanning tree with identical sequences grouped in circles and divergent sequences radiating outwards on branches proportional to the number of nucleotide changes. Fig. 6: Plot of pairwise Fsr versus distance (km), including negative values of FsT. Fig. 7: Plot of pairwise Fsr versus distance (km), with negative values of Fsr changed to zero. Fig. 1 Strawberry Hille Bodega Bay Half Moon Baye HMS Soberanes San Luis Obispoy Lompoc Blazyk 22 Blazyk 23 Blazyk 24 0 Fig. 4 B817 HMBIO 8512 ORI3 HEE BB19 ORI s ORI SBI8 — S820 SLO9 SP9 0.5 changes Tegula funebralis CO — HMBS BB13 BB14 B16 HEG — EMBS OR4 SB14 HE9 2 5105 ORII BB18 Hr3 — HMBS — HP10 S10 910 -8p11 Clade 1 Clade II Clade III Blazyk 25 SLOA¬ Fig. 5 SPI1 S105 BBI4 NO S0 HEIO HMB7 BBI? HPS ORi4 SB13 SB159819 HE3 )SPI SP2 SPS SP10 BBI3 SP6 SP8 o HMB8 Tegulh furebraks c0 — HP7 EIORBBIS 9917590 S109 HMB6 HE8 () SLO1 HP10 HEA HMB4 Blazyk 26 -HP3 AMESBBI8 sio2 Fig. 6 y- 1E-04x- 0.0604 Pairwise Fsr v. Distance 8=0.5035 0.08 0.06 0.04 0.02 -0.02 0 —200 —600800 • 10001200 -0.04 -0.06 0 -0.08 -0.1 - -0.12 Distance between locations (km) Blazyk 27 Fig. 7 Pairwise Fsr versus Distance y = 5E-O5X - 0.0083 (without negative values) R/=0.5894 0.08 0.06 0.04 0.02 oee 400 600 800 1000 1200 200 -0.02 Distance between locations (km) Blazyk 28