Abstract The biochemical genetics of an intertidal bernacle, Tetreclite rubescens (Darwin, 1354), were examined at four sample sites along the coast of central California. Examination of the allelic frequencies at four polymorphic loci indicate that gene flow between subpopulations is either currentlg high or was so veru recentlu. Not all sites, however, appear to exchange migrants equallu. The samples from the Moss Landing site appear to be closelg related to each other. and other evidence suggests that Moss Landing may be quite isolated from the more southern populations. Point Conception does not appear to be a major barrier to gene flow, and migration between Pecific Grove and Pori Hueneme, three hundred miles awag, appears to be greater than migration between Pacific Grove and Hoss Landing, onlu twenty miles awag. A population oi Semibalanus cariasus (Pallas. 1768) was also studied, and found to be almost indistinquishable from the Tetrsclita populations at all four polymorphic loci scored. This indicates that strong selection pressures mag be maintaining these allelic frequencies, and the polumorphisms mag be older than the species themselves. Introduction Since the advent of starch gel electrophoresis in the late 1960s à great deal of genetic variation has been discovered in countless populations of orgenisms. Plants, insects, reptiles, mammals. invertebrstes, birds, and mang other groups have all been studjed and shown to harbor vast amounts of allelic variation within their populations. Discovering why this variation exists and its importance to evolutionary and population biology has been a more difficult and controversial task. Most explanations can be separated into two catégories: those that invoke natural selection as a force responsible for maintaining genetic variation, and those that believe most of the variation observed is selectivelg neutral. These positions are commonly known as the selectionist and neutralist theories. respectivelg. A great deal of time and effort has been spent on truing to prove or disprove both of these positions. Hany studies have sought to prove that environmental factors, and hence natural selection, are responsible for maintaining allelic polgmorphisms. Hang of these studies have been performed on marine organisms, and among the more significant are williams, Koehn and rtton's work on ihe American eel, Anguille restrsts (Williams et al 1972), Koehn, Hilkman, and Mitton's work on the blue mussel, Ftilus edulis (Koehn et al, 1975), and Power's group's work on a fish Fundulus beterochitus (Powers, 1985.) All three of these studies found signiticant changes in allele frequencies over geographic areas, and all three could correlate these changes with environmental factors, such as temperature in the case the American eel and Punoulus, and salinity in the case of vtilus In all three cases natural selection was shown to be a probable force in creating the clines, and therefore the polumorphisms. Thère have been considerably fewer cases where natural selection could be positivelg identified as a force for mainteining variation when no clines were present. Werd Wett's work on Colias butterflies is one such case (Watt, 1984), and Ronald Burton and Harcus Feldman's work on the copepod Tigriopus is another (Burton and Feldman, 1982.) Both of these groups have correlated the function of a specific enzume with differential fitness among individuals cerrying alternate alleles for the loci studied. These cases are dwarfed by the number studies in which variation is observed and the forces, if ang, maintaining it have get to be discovered. Among these are the mang geographic population studies in which no obvious patterns of allele frequencies were found. Hang of the projects involving marine organisms on the west coast of North America fall into this category (see Levinton et al, 1978. Dayis et al, 1931, Smith et al, 1987, Grant, 1988.) Altnough there still exists controversy over the meaning and importance of allelic variation to evolution, this variation has recentig proven itself useful in the studg of the population dunamics oi marine organisms. Traditional methods of studging population subdivision, migration and gene flow by direct observation are ditticult, if not impossible, to perform on small marine organisms. The problem is compounded when the species in question have microscopic, free swimming larva, or release their gametes directlu into the water. Both of these cases have a high potential for migration, even if the adult forms are stationarg. Unfortunatelu, it is impossible to directlg observe the extent of such migration. An alternative to direct observation is to studg the large scale population genetics of a species, and infer gene flow and migration rates from the genetic data. It is possible to estimate both the extent of population subdivision and migration from allelic frequencies (Slatkin. 1987) The data gained from biochemical population studies is also useful in that it allows examination of gene flow on a evolutionaru time scale, instead of simplg examining present trends. Jeff Hitton. Carl Berg and Katherine Orr used allozyme data to estimate the extent of gene flow between different subpopulations of the Caribbean Oueen conch Strombus giges(Mitton et al, 1989), and B.J. Davis et al used allelic frequencies to estimate gene flow between Pacific Coast populations of the fish Gylebius pictus (Davis et al 1981) Both studies found evidence of high gene flow between subpopulations, with signiticant but unpredicteble differences in local allele frequencies he purpose of the present studg was twofold: 1) to examine the biochemical genetics of a common intertidal barnacle, Tetrsclite rurescens, and use the genetic data to infer migration patterns and rates, and 2) to attempt to gain additional insight into the evolutionary importance of allelic variation by also examining the biocnemical genetics of a related barnacle species, Semibajanus ceriosus. The study area included the northern end of Tetrsclitss range and the southern end of Semibalanus s, and encompassed at least two potential barriers to gene flow: the Honterey Submarine Cangon and Point Conception, the largest point on the west coast of N. America. Haterials and Hethods Somple collection and storage Ouring the months of April and Hag, 1990, collected samples of Tetraclite rubescens from jetties in Port Heuneme, Horro Bag, Hoss Landing, and from a cerent wall in Pacitic Grove (figure 1). Samples oi Semibalanus csriosus (Pallas. 1738), originallg mistaken for Tetraclita, were collected from a jettu in Half Moon Bay. In all locations samples were collected throughout the species' vertical intertidal range. In Pacific Grove the sampled indiviquels ranged in size from verg small (2 mm basal diameter) to verg large (50 mm basal diameter). in all other locations ! took onlu medium to large individuals. I removed the soft tissue from all but the smallest individuals (Amm besal diemeter or less) with tweezers and stored each sample in an individual plastic centrifuge tube. The smallest individuals were simply placed shell and all into individual tubes. The samples from Half Hoon Bag, Horro Bag, and Port Huenere were trozen in a cooler of dry ice immediatelg after sampling, and then stored at -70 degrees C for the remainder of the project. The samples from Pacific Grove and Hoss Landing were kept on ice for un to three hours before being stored at -70 degrees C. I analuzed all the samples within four weeks of sampling time, and the number of thaw/freeze cycles was kept to a minimum. Sample pregeration and analysis All semples were diluted with one to three drops of distilled water, homogenized bi ultre sound, and then spun down for three to four minutes in a high speed microcentrifuge. The homogenates were absorbed into whatmans's *1 Tilter paper, and analgzed using horizontal starch gel electrophoresis. Gel preparation consisted of pouring a mixture of 35 g starch (Sigma) suspended in 50 ml of cold buffer solution into 220 ml of boiling butter solution. The final 3ooml volume was mixed vigorouslu. degassed, and poured into 17 cm by 17 em bg 1 em molds. The gels were cooled to 4 degrees C before loading the samples. Enzumatic activitg was detected at the following loci: phosphoglucoisomerase (Pgi-1 and Pgi-2), phosophoglucomutase (Pgm); tetrazolium oxidase (To), hexokinase (Hk-1 and Hk-2), malate dehgdrogenase (Mah-1 and fidh-2), malic enzyme (He), acid phosphotase (Acp), and xanthine dehydrogensse (Xdh). Onlg Pgi-1, Pgm, To, Hk-2, and Hoh-1 were scored. The other loci were either not resolved well engugh to be scorable,or in the case of Hk-1 and Fgi-2, were detectable onlu in verij small individuals. The buffer systems used are described in Agals et al (1973) Buffer "C"(Tris, citrate, EDTA,PH 70) was usedto resolve all scored loci. The enzyme stains used were all from Eichardson (1956), except that the Göpdh used as a linking enzume in the stains for Pgm, Pgi, and Hk was NADH, rather than NADEH dependent Dsts enslysis: For each sample site, Fst, an inbreeding coefficient which measures population subdivision, and Nm. an estimate of the number of migrants per generation was calculated (Hartl and Clark 1989, chapter 6). Fst-(Ht-Hst)/Ht, and Nm=17AFst-174, where Ht is the expected heterozugositu if all samples from all the sampling sites belong to a single panmictic population, and Hst is the expected heterozygosity of a subpopulation assuming random mating is occurring within the subpopulation. R. the coefficient of relation, wes calculated by a computer program (Jeff Mitton, unpublished) based on the following equation (Queller and Goodnight, 1939): R=2222(Pi(-j)-Pm)/EZZZ(Pijm-Pm), where i=groups j-individuals within a group, k-loci, and a-allelic positions within a locus. All of these statistics are useful in estimating the amount of subdivision within a population. Fst estimates the loss of heterozygositg due to population subdivision ( the Wahlund effect). Mm estimates the amount of migration between subpopulatjons, and R describes how closely related the individuals in a subpopulation are, in terms of the probability of shared alleles. Results Gecgrephic veriatien: Ui the five scored loci, Pgi-1, Pgm, To, and Hk were polumorphic in all populations studied. Näh-2 was monomorphic in all populations studied. Table shows the allele frequencies for all loci st each Tetreclite studg site. Alleles are referred to by letter, with A being the fastest moving allele, and F the slowest. For Pgi, each sample site has one verj common allele (75-858), one intermediate allele (5-155) and two or three rare alleles («53). All populations had the same common allele, but theg differed in the frequencies of the rarer alleles (see figure 2). In Horro Bay, allele C is the second most common allele, while in all other populations allele B is the second most: common. None of the Hoss Landing samples showed allele E. which had a frequency of 58 in Pacific Grove, only 20 miles awag. The Port Hueneme samples lacked allele A. The Pgi allele frequencies were found to differ significantli between sampling locations using a G test for independence (Sokal and Polhf, 1981.) All the Pgi heterozugotes displaged a three banded pattern on the gels tupical of a dimeric enzyme. There were six alleles at the Pam locus, and theu fall into a pattern similar to Pgi. All populations had one verq common allele one intermediate allele, and several rare alleles (figure 3). The gitferences in Pgm allele frequencies were not found to be significant (pe 1). Tetrazolium oxidase was scored only for Morro Bau and Port Hueneme, and both locations had verg similar allele frequencies (tigure 4). Hk was scored for Hoss Landing, Horro Bau and Port Hueneme, and there were no significent differences in allele frequencies between these locations (figure 5). Hk, To. and Pam neterozygotes all displaged a two banded pattern tupical of monomeric enzymes lable Il gives the observed and expected heterozugosities, the inbreeding coefficients, Fst, and Am for each locus. None of the observed heterozygosities are significantly different from what is expected under Hardg-weinburg conditions, except for Hk at Hoss Landing, where there is a significant excess of heterozugotes. The values for Nm varg considerably between loci, with Pgi, Pgm, and Hk all giving values between 15 and 25 migrants per generation, but To giving a value of over 200 migrants per generation. The value for To 13, however, derived from allele frequencies from Horro Bau and Port Hueneme onlg, while the other estimates are derived from frequencies from at least three locations. The average values for Nm are 73.2 using all four loci, and 20.2 if To is left out. Table IIl gives the coefficients of relation for the individuals sampled at each site. A value of zero indicates no relstionship, and a value of S would mean the sampled individuals were full sibs. Although standard errors could not be calculated, ran a computer Simulation with rändomlg chosen genotupes (expected relationshio equals zero) and got values ranging from -05 to 01 (average deviation from zero -03) using 15 individuals and three loci. All of the values calculated for the sample sites used at least 15 individuals and at east 3 loci, except for Pacific Grove, which used 40 individuals but onlg 2 loci. The R values varied considerably from site to site, with Port Hueneme having a negstive value (less related then would te expected bi chance) and Hoss Landing having the rather high value of 23. This means that on average ang two individuals sampled at Hoss Länding share two grandparents. The value of 008 at Pacific Groye indicates that those individuals are no more related then would be expect by chance, and the value of 13 at Horro Bay indicates that those individuals are slightig more related to each other then would be expected by chance. Interticsl verietion. At the Pacific Grove study site individuals were sampled speciticallg from both the verg bottom and the verg top of their intertidal range. In addition, verg small individuals were sampled. again both from the bottom and top of the specie's range. No signiticant differences in allele frequencies were found between high and low intertidal samples when only medium to large indiviquals were sampled, and there were no significant differences in allelic frequencies between large individuals and small individuals as a group (tables IVand V) However, there were significant differences in allele frequency at the Pgi locus between high intertidal and low intertidal samples among the small individuals alone (p005, table y1) he small barnacies were estimated to be about two months old (Kristi Hiller, personal communication.) Interspecitic veristion: able VIl gives the alleic frequencies at all four polumorphic loci scored for the population of Semibslanus cariosus sampled at Half rioon Bag. Each locus has the same alleles as the Tetraclite populations, and only at the To locus are the allelic frequencies signiticantig different from the averaged allelic frequencies of the letraclite populations. Nei's genetic identity (Nei 1975) between the pooled Tetraclits populations and the Semibalanus population equals 9903, a figure normally associsted with two populations within the same species. Discussion Geegrechic veristian he allele frequencies at each location sampled are clearlu veru simllar, and the estimated number of migrants per generation between sample sites is quite high, as might be expected for an intertidal species with a pelagic larval period. However, the individuals sampled probablg do not belong to a single panmictic population, and the results suggest ihat the dispersal patterns for barnscle larva are quite complex. The Pgi allele frequencies are significantlu different between subpopulations, indicating that either some subpopulations ere isolated endugh from each other for genetic drift to occur, or there are local selection pressures which favor different allelic frequencies in ditierent places. If local selection is occurring, it is probablu vert slight, since in general the allele frequencies at esch location sampled are quite similar. he coefficients of relation indicate that some subpopulations probably retain most of their larvae. Moss Landing, with an R of 23 seems to be particularly isolated from the other subpopulations. In addition to being the northernmost breeding group of Tetrsclits (Kristi Hiller, personal communication), the Moss Landing subpopulation is also located right at the mouth of the Montereu Bau Cangon, which mag act as a barrier to larval dispersal from the more southern populations. The Horro Bay samples were taken from calm water deep within a harbor, and their R of 13 suggests that theu mau also be a fairly isolated subpopulation. The samples from Pacific Grove were taken right at the southern edge of Honterey Bag, and the samples from Port Heuneme were taken from the open coast. Neither of these had large positive R values, indicating that these subpopulations may recruit individuals from a greater area then either the Horro Bay or Hoss Landing subpopulations do. it seems quite possible that the Pacific Srove subpopulstion trades more migrants witn Port Hueneme, three hundred miles awag, then it does with Hoss Langing, only twenty miles awag. This idea is further supported bu the Hoss Landing subpopulation's lack of Pgi allele E, which is present ai 4-53 at both Pacific Grove and Port Hueneme. Uf the two potential barriers to gene flow mentioned in the 12 introduction, Point Conception and the Monterey Bag Cangon, the latter seems to be a more likely candidete as a berrier to Tetractite laryee. This is not too surprising, considering Honterei Bag is the northern limit for large Tetraclita populations on the west cosst. It is not known whether Tetrachite populations do not appear north of Hontereu bay because they cannot get there, or because they are not capable of surviving north of the bag. Hy results indicate that Hoss Landing mau be a very isolated subpopulation, which supports the idea that Montereg Bay is a berrier to northern migration. The Moss Landing subpopulation mag be important es a stepping stone to the north, and mag supply larvae which will eventuallg form populations north of Monterey Bay. It seems likelg, however, that the barriers which isolate Hoss Landing from the south will also isolate it from the north, so this expansion, if it happens, could be a very slow process Evolutionsry implications and Semibalanus cariosus: My results have some relevance to the neutralistselectionist argument about the importance of genetic variation. In brief, the neutralist position is that most, if not all, natural genetic variation is selectively neutral and meaningless so far as the individual fitness of an organism is concerned. The selectionist argument maintains that much or at least some of the observed genetic variation is adaptive and maintained by natural selection. My results bear on this argument in two ways: 1) The values lobtained for Mm varied between loci. I all the variation observed was selectively neutral, the figation indices and estimates of Nm should have been nearly identical. Even leaving out the extremelg high To estimate, there may still be enough variation in the estimates derived from the other loci to implicate selection pressure as a force affecting allelic frequencies. 2) The population of Semibalsnus cariosus sampled at Half Hoon Bau was almost indistinguishable from the Tetraclits populations at all four polgmorphic loci studied. It seems unlikely that this is merelu a coincidence, and there are several alternative possibilities to explain this result. One possibility is that the two groups are reallu onlu one species, and there is extensive gene flow between them. This seems extremelg unlikelg, since there are important morphological differences between the two groups, and theg brood at opposite times of the geer (Horris, Abbot, Haderlie, 1980 p517,519). A second explanation is that there are strong selective forces operating on both species which maintain the allele frequencies at the same levels. This nas some plausibilitg; both species occupy the same ecological niche. the lower intertidal zone, and the species look superficiallu alike. Their geographical ranges meet up with each other with little overlan (tigure 1), and it seems likely that they face similar environmental conditions. In theory, genetic drift can be a powerful force in changing allele trequencies, even in large populations, so long as the alleles are selectively neutral and the time period in which change can occur is great (Hartl and Clark, chapter 2, 1939.) Tetraclits and Semibalanus nave probablg been reproductively separsted for tens of millions of gears (Kristi Hiller, personal communication), and this certainly would have been enough time for their allelic frequencies to have diverged considerably were the alleles involved selectivelu neutral. If, on the other hand, the polgmorphism was adaptive for the common ancestral species of both Tetraclits and Semibalsnus, and the forces maintsining the allele frequencies remained in place after the tyo groups diverged, then it is possible that the polumorphisms could remain and be maintained in the now separate species. Thus, the polgmorphisms shared by the two species may be older than the species themselves. Acknowledgements: would like to thank Denng Powers, Lunna Hereford, Tobe Cole, Lani West, Simone Sorger, Doug Crawford, Himi Brown, Rob Rowan, Linda Park, and most especiallu Jeff Mitton for their kindness, patience, and help with this project. References: Agala, F.J., D. Hedgecock, 6. Zumwalt, and J Valentine: Genetic variation in Tridscne maxims, an ecological analog of some unsuccessful evolutionary lineages. Evolution 27: 177-191 (1973) Burton, R.S., and H.W. Feldman: Phgsiological effects of an allozume polymorphism: glutamate-pyruvate transeminase and response to hyperosmotic stress in the copepod Tigriepus californicus. Biochem. Genetics 21 (3/4): 239-251 (1982) Davis, B.J., E.E. Deflartini, and K. HcGee: Gene flow among population of a teleost (Painted Breenling, Grylebius pictus from Puget Sound to Southern California. Harine Biologu 65: 17-23 (1981) Grant, W.S., and F.M. Utter: Genetic heterogeneitu on different geographic scales in Nucells lamellase (Prosobranchia, Thaididse) Halacologia 28 (1-2): 275-237 (1988) Hartl, D.L., and A.G. Clark: Principles of population genetics, 682 pp. Sunderland, Hassachusetts: Sinauer 1989 Koehn, R.K., R. Hilkman, and J.B. Mitton: Population genetics of marine pelecupods. IV. Selection, migration and genetic differentiation in the Blue Hussel Hytils edulia Evolution 30: 2-32 (1975) Levinton, J.S. and T.H. Suchanek: Geographic Varistion, Niche Breadth and Genetic Differentiation at Different Geographic Scales in the Mussels Mytilus califernienus and M edulis Harine Bioloqu 49; 363-375 (1978) Mitton, Jefferg B, C.J. Berg, and K.S. Drr: Population Structure, Larval Dispersal, and Gene Flow in the Queen Conch, Strambus giges, of the Caribbesn. Siol. Bull 177: 356-362 (1989) Morris, R. H., D.P. Abbott and E.C. Haderlie: Intertidal invertebrates of California, 690 pp. Stanford, Californis: Stanford Universitu Press 1950 Nei, H: Holecular population genetics and evolution, 25S pp. New York: Elesvier 1975 Powers, OA et al: Senetic variation in Fundulus heteractitus geographic distribution. Amer Zool 26: 131-144 (1986) Queller, D.C., and Goodnight, K.F.: Estimating relatedness using genetic markers. Evolution 43 (2): 258-275 (1939) Richardson, B.J., P.R. Baverstock, and M. Adams: Allozume electrophoresis, a handbook for animal systematics and population studies, 410 pp. Sudneg, Australia: Academic Press 1986 Slatkin, Hontgomerg: Sene flow and the geographic structure of natural populations. Science 236: 757-792 (1937) Smith, B.L. and D.C. Potts: Clonal and solitarg anemones (Anthepleura of western North America: population genetics and sustematics. Harine Biologg 94: 537-546 (1987) Sokal, R.R., and F.J. Rohlf: Biometry, 859 pp. New York: W.H. Freeman 1951 watt, W.B., P.A. Carter, and S.M. Blower: Adaptation at specific loci. IV Differential mating success among glycolgtic allozyme genotupes of Colias butterflies. Genetics 109: 157-175 (1985) Williams, G.C., R.K. Koehn, and J.B. Mitton: Genetic Differentiation without isolation in the American eel, Angwille rostrata Evolution 27: 192-204 (1972) Figure Captions Figure 1: Map of the study aree and ranges of Tetraclite and Semibalanus Sample sites are (4) Port Hueneme, (B) Horro Bag, (C) Pacific Grove, (D) Moss Landing, and (E) Half Hoon Bag. (Hodified from Morris, Abbot and Haderlie, 1980.) Figure 2: Geographic distribution of Pgi allele frequencies. Frequencies are not independent of locstion (P025). Figure 3: Geographic distribution of Pgm allele frequencies. There are no signiticent differences in allele frequency between locations. Figure 4: Tetrazolium oxidase allele frequencies. There are no significant differences between the two locations Figure 5: Hexokinase allele frequencies. There are no significant differences between the three locations. Table Captions lable l: Tetraclite allele frequencies at four polymorphic loci. Allele A is in all cases the fastest moving allele (most anodal), and allele f the slowest (most cathodal). A — indicates missing data, a means the allele was not found. äble II: Expected subpopulation heterozygositg, Hs; observed heterozygositg, Hi; expected population heterozygositu, Ht; the inbreeding coefficient, Fis; the subpopulation fixation index, Fst; and the estimated nümber of migrants per generation, Nm for each of the four Tetreclita sample sites. Table III: Coefficients of relation for the four Tetreclite subpopulations. able IV: Comparison of allelic frequencies between samples from high and low in the intertidal zone, medium to large barnacles onlu. lable V. Allele frequencies from samples small, recentlu settled Tetraclita (less than 4 mm) Table VI: Comparison of allelic frequencies between samples from high and low in the intertidel zone, verg small barnacles only. The Pai allele frequencies are significently different between high and low (05.05). lable VII: Allelic frequencies at four polgmorphic loci for a population of Semibolenus coriosus sampled at Half Hoon Bau. 05 — — — — — — — — — — — — — 0 0 u — o3 *. ole - OOO sssos HUESO u Aouenbad - n o00O OO SOSSS EUES bouenbe l 5 LTf IIIII u — oO QOO ooos HUES mu ummnn Kouenbedi -N OOO o0 osoe HUES unen unnnant Aouenbel 8 88 8 aa- 38 + — oo oo - - 5 a kataaa- 0 aa ktata- 2 aa- N 1- 5 - aa 8 o- D- 10 aataaaaaaa- a- tatakaa- 6 s. - kaataaaa sample site „ Moss Landing Pacific Grove Horro Bag Port Hueneme Table 11 1oc1 R Pgi,Pgm,Hk Pgi Pgm 0.008 Pgi,Pgm, Hk, To 0.13 Pgi,Pgm, Hk, To -0.12 3 a e staa- 55 5 kaata- 9 kaaa 0 53 8 — o