Reference: Biol. Bull. 175: 218—229.(October, 1988)

Geographically Widespread, Non-hybridizing, Sympatric Strains of the Hermaphroditic, Brooding Clam Lasaca in the Northeastern Pacific

DIARMAID O FOIGHIL* AND DOUGLAS J. EERNISSE** Friday Harbor Laboratories, University of Washington, Friday Harbor, Washington 98250

Abstract. We have studied phenotypic variation in six Victoria, British Columbia, populations on the basis of enzymes of , a taxonomically complex genus of shell color phenotype is to some degree possible, but is small brooding clams, from nine northeastern Pacific often equivocal. Electrophoretic analysis (especially of sites. Each ofthe individuals examined produced one of reddish specimens) is necessary for their reliable identi five combinations ofelectromorph patterns. Lasaea phe fication. notypes could be differentiated into two main types, one containing two and the other three phenotype combina Introduction tions. Samples from each population contained from one to three phenotype combinations and there was no Members of the taxonomically complex bivalve ge evidence for crossbreeding among phenotypes. These re nus, Lasaea, are reproductively specialized, minute, in sults are strongly at variance with random mating expec tertidal, crevice dwellers (Keen, 1938; Oldfield, 1964; tations and indicate that the phenotype combinations Glynn, 1965; Ponder, 197 1) with a near-cosmopolitan represent reproductively isolated strains. This is substan distribution (Chavan, 1969). A prominent reproductive! tiated by a more detailed study ofthe McNeill Bay, Brit developmental dichotomy exists within the genus. La ish Columbia, population where both main strains co saea australis (Lamarck, 18 18) engages in random mat exist. Electrophoretic characterization of Lasaea from ing and broods its young to a straight-hinged, plankto individual 100 cm2 samples of barnacle cover revealed trophic stage of development (0 Foighil, 1988). that strains are not spatially segregated. Progeny of (1) All congeners studied to date, release their young as pair mating experiments, (2) brooding field individuals, crawl-away juveniles (Pelseneer, 1903; Oldfield, 1964; and (3) specimens that reproduced in isolation, all per Glynn, 1965; Rosewater, 1975; Booth, 1979; Kay, 1979). petuated the maternal electromorphs. Data from previ There is as yet no evidence for cross-fertilization in La ous studies of reproduction in northeastern Pacific La saea that have crawl-away juveniles (Crisp et al., 1983; saea suggest that the formation of non-hybridizing 0 Foighil, l986a, Crisp and Standen, 1988), which to strains has resulted either from the predominance of self gether form a complex assembledge of nominal fertilization or as a result ofpseudogamy combined with and subspecies ofunknown phylogenetic affinity. meiotic parthenogenesis. We currently favor the former Molluscan systematists have conventionally relied hypothesis. The two main strains are largely conserved heavily on shell morphology to distinguish between spe between geographically distant sites, despite the lack of a cies. There is great individual variation in Lasaea valves planktonic larval stage. Separation ofthe main strains in (Dall, 1899; Ponder, 197 1; Roberts, 1984; 0 Foighil, 1986a), even among those collected from any particular site, and this poses a difficult taxonomic dilemma. Keen Received 24 February 1988; accepted 29 July 1988. (1938) lists >40 species of Lasaea, distinguished from * Present Address: Biology Department, University ofVictoria, P.O. Box 1700, Victoria, British Columbia, Canada V8W 2Y2. each other on the basis of slight differences in shell mor ** Present Address: Museum ofZoology and Department of Biology, phology and color. However, subsequent workers, have University ofMichigan, Ann Arbor, Michigan 48109. been unable to separate many of these nominal species

218 NON-HYBRIDIZING, SYMPATRIC, LASAEA STRAINS 219

(Soot-Ryen, 1960; Barnard, 1964; Dell, 1964; Ponder, mating L. australis is an order of magnitude greater (ap 197 1; Haderlie and Abbott, 1980; Beauchamp, 1985). In proximately 50% ofgonadal volume) than in northeast the northeastern Pacific, Keen recognized two species, L. ern Pacific congeners (0 Foighil, 1988). Several observa subviridis Dall, 1899 and L. cistula Keen, 1938. In a tions imply that cross-fertilization might be a relatively more recent worldwide review ofLasaea, Ponder (1971) rare phenomenon in northeastern Pacific Lasaea popu concluded that most ofthe nominal species, including L. lations. Sperm are present only in small numbers and subviridis and L. cistula, are merely regional subspecies have reduced motility (0 Foighil, l985a), and there is an or ecotypes ofthe type species L. rubra (Montagu, 1803). apparent absence of specialized sperm transfer mecha Population genetic studies of European Lasaea have nisms typically found in cross-fertilizing brooding bi @ revealed the existence of a variety of non-hybridizing, valves, such as spermatophores/spermatozeugmata sympatric, genetic strains capable of reproducing in iso (Coe, 1931; 0 Foighil, 1985b), dwarf/complemental lation (Crisp et al., 1983, Crisp and Standen, 1988). males (Turner and Yakovlev, 1983; 0 Foighil, l985c), These results have important implications for under and pseudocopulation (Townsley et al., 1965). standing morphological variation and systematic rela In this study, we examined isozyme variation of four tionships within the genus. Crisp et al. (1983) concluded British Columbia (B. C.), Canada, and five California, that the populations studied were composed of female, U. S. A. populations ofLasaea. The F1progeny of adults apomictic (i.e., ameiotic) clones. However, Oldfield from a B. C. site, either resulting from pairs or single mdi (196 1) described European Lasaea as simultaneous her viduals that reproduced when placed in isolation, or maphrodites with minute sperm production, recently from broods collected from field individuals, were also confirmed by McGrath and 0 Foighil ( 1986) and by characterized electrophoretically. We used isozyme data D. J. Crisp and associates (pers. comm.). Crisp and Stan from adults and juveniles (1) to help identify the breed den (1988) proposed that European Lasaea reproduce ing system ofLasaea by comparing observed phenotype by a combination of pseudogamy and apomixis. As yet frequencies to random mating expectations and (2) to there is no available data on spawning and gamete inter assess the systematic status of Lasaea populations using action in European Lasaea. Thiriot-Quiévreux et al. shell and protein phenotypes. (1988) found an unusually large chromosome comple ment (100 2n 120) in Kerguelen Lasaea and pro Materials and Methods posed that this resulted from an apomictic mode of re production. However, re-examination oftheir histologi After collecting, adult nonbrooding were ei cal material has revealed that Kerguelen Lasaea are ther analyzed electrophoretically within 2 days following simultaneous hermaphrodites with disproportionately storage in seawater tables or placed at —¿70°Cuntil pro tiny testicular versus ovarian tissue (0 Foighil, unpub.). cessed. Electrophoresis was performed at Friday Harbor The presence of sperm in European and Kerguelen La Laboratories using 13% starch (Sigma hydrolyzed potato saea leaves open the possibility of reproductive modes starch) gels, standard power supplies, and horizontal besides apomixis, including self-fertilization. electrophoretic apparatus. Whole animals were individ Considerable data are available on the reproduction of ually homogenized with glass rods in an approximately northeastern Pacific Lasaea. They brood their young to equal volume ofgel buffer and gels were run, not exceed a crawl-away juvenile developmental stage (Keen, 1938; ing 200 volts, until the front had reached a preset “¿des Glynn, 1965; Beauchamp, 1986; 0 Foighil, 1986b), are tiny― 80 mm from the origin. A single discontinuous simultaneous hermaphrodites (Glynn, 1965; 0 Foighil, Tris-citrate buffer system (electrode: 18.55 g boric acid/I 1985a; Beauchamp, 1986), and can reproduce in isola and 2.4 g sodium hydroxide/l, pH 8.2; gel: 9.21 g tris/l tion (0 Foighil, 1986b, 1987). Isolated individuals simul and 1.05 g monohydrate citric acid/l, pH 8.7) was used taneously spawn sperm and eggs into the suprabranchial for the following enzymes: esterase with a-naphtyl ace chamber, sperm attach to eggs by an acrosomal reaction, tate substrate (EST; nonspecific), (leucine) aminopepti the male pronucleus is incorporated into the egg cyto dase (LAP; E.C. 3.4. 11. 1), peptidase with glycyl-leucine plasm and theoocyte produces two polar bodies before substrate (PEP-GL), peptidase with leucyl-glycyl-glycine first cleavage (0 Foighil, 1987). A relatively tiny amount substrate (PEP-LGG), and peptidase with leucyl-valine of testis is produced in the ovotestis, sperm occupy ap and leucyl-tyrosine substrate (PEP-LVLT). In addition, proximately 5% of gonadal volume, the rest being de phosphoglucomutase (PGM; E.C. 2.7.5. 1) and glucose voted to oogenesis (0 Foighil, 1985a). This pattern of phosphate isomerase (GPI; E.C. 5.3. 1.9) were investi reduced sperm production is theoretically consistent gated using Crisp et al.'s (1983) discontinuous Tris-ci with the hypothesis that self-fertilization is common trate buffer system. Most individuals showed no activity (Heath, 1979; Fischer, 198 1; Charlesworth and Charles for PGM and the results for this enzyme are not pre worth, 198 1). Indeed, sperm production in the randomly sented. Enzyme staining assays for EST, LAP, and PGM 220 D. O FOIGHIL AND D. J.EERNISSE

containing one white and one red shelled individual, were placed in separate culture vials containing 20 mls of 1 @zmfiltered seawater. Lasaea are positively thigmo tactic (Morton, 1960) and each pair was positioned in close physical contact to ensure mutual attachment with byssal threads to facilitate potential cross-fertilization. The culture vials were maintained at 18°Cand the clams were fed on cultured Thalassiosira pseudonana (strain 3H) and had salt water changes twice weekly. All individ uals were checked weekly for brooding activity using a dissecting microscope (broods are visable through the translucent shells). Once a brooding individual was de tected, the non-brooding partner was preserved at —¿70°C,as was the brooding parent when all the F1juve niles had been released from the brood chamber. The F1 juveniles were cultured under the same conditions until HALF September 1987, when parents, non-brooding partners, and the F1 progenies ofpairs containing adults of differ CAVUCOS ent enzyme phenotypes were electrophoretically ana lyzed. Originally, it was hoped to type the offspring (0.7-. PALOS VERDE@ MONTEREY 1.2 mm in valve lengths) using both LAP and PEP-GL enzyme assays, however, results were obtained only for @Okm the latter. Ten red-shelled non-brooding adults collected from Figure 1. Sampling sites for northeastern Pacific Lasaea used in the McNeill Bay in February 1987 were placed individually electrophoretic study. in culture vials and maintained in the laboratory, as pre viously described, until they released juveniles. In May were as described by Ayala et al. ( 1972) and for GPI by 1987, 30 Lasaea adults brooding early embryos were se Tracey et al. (1975). The PEP-GL, PEP-PP and PEP lected from pooled McNeill Bay samples. They were also LVLT staining assays consisted of 30 ml of a 2% agar individually raised in laboratory conditions until juve solution (60°C)added to 20 mg of peptide substrate, nile release. Both sets of parents were then frozen at 2000 units of peroxidase, 10 mg of O-Dianisidine, 5 mg —¿70°Cand electrophoretically characterized together of Corotalus adamanteus toxin, 0.5 ml ofO. 1 M MnC12 with their offspring in September 1987 for the PEP-GL and 20 ml ofO. 1M Na2HPO4 buffer. enzyme phenotypes. At least one McNeill Bay individual ofknown protein Twenty five random 100 cm2 samples of barnacle phenotype was included per gel to provide a standard cover containing Lasaea were removed from the Mc electromorph, and specimens from all nine populations Neill Bay site in October 1987 to investigate the degree of were repeatedly run together on the same gels to verify spatial overlap between strains. Samples were collected electromorph scoring. The right valves of 73 McNeill from the mid to high intertidal along an 80 m stretch Bay specimens were retained and analyzed for possible of shore and were representative of the various crevice correlations of shell morphology and color to protein habitats available to Lasaea at the study site. All individ phenotype. uals in each sample were characterized for the PEP-GL Subsequent to the initial electrophoretic survey of La phenotype after embryos had been dissected from brood saea from the nine study sites (Fig. 1), a more detailed ing individuals. investigation of the McNeill Bay population was per formed. This involved electrophoretic characterization Results ofthe progeny of(l) pair mating experiments, (2) speci mens that reproduced in isolation, (3) individuals that The electromorphs obtained from the survey of nine reproduced in the field. The degree ofspatial overlap be populations for EST, LAP, PEP-GL, and PEP-LGG are tween strains in the McNeill Bay study site was also elec presented diagramatically in Figure 2. GPI was mono trophoretically assessed. morphic in all individuals, PEP-LVLT phenotypes for In February 1987, non-brooding Lasaea adults were each specimen were identical to the combined PEP-GL sampled from McNeill Bay and specimens were sorted and PEP-LGG electromorphs for that individual and according to shell color phenotype. Thirty pairs, each only one esterase presumed locus (EST) was consistently NON-HYBRIDIZING, SYMPATRIC, LASAEA STRAINS 221

TableI + A @ Composite electromorph composition ofthefive Lasaea strains 50 observedin B.C. and Californiapopulations. Reftr to Figure 2 —¿I2@ —¿ 4 for diagrammatic representationofelectromorphpatterns @40 0 .L.L B 30 —¿â€”¿ patternPEP-GLPEP-LGGLAPEST“,, electromorph IL' zC.)20 4 @ StrainsComposite 10

AI PEP—GL PEP—LGG LAP EST 1“B―3 21 21 11 21 Figure 2. Diagrammatic representation of the total electromorph variation observed over four polymorphic gene-enzyme systems for 4 3 2 4 2 northeastern Pacific Lasaea. Each electromorph pattern is individually 53 32 33 42 2 numbered. For instance, the three PEP-GL protein phenotypes are la belied PEP-GLI, PEP-GL2, and PEP-GL3, respectively.

resolved in this study. Each Lasaea individual examined gous or homozygous phenotypes from polymorphic pop produced one of only five different composite electro ulations. For example, populations containing both “¿A― morph patterns observed in this study (Table I). These and “¿B―EST phenotypes (Table II) lacked obvious het patterns could be differentiated into two main sets (“A,― erozygotes. Because all specimens had a single GPI band “¿B―)on the basis ofsimilarity ofthe protein phenotypes. of identical mobility, we assume that all were homozy Patterns 1, 2 are of the “¿A―set and share identical PEP gous at a single GPI coding locus. Genetic interpretation LGG and EST phenotypes; “¿B―patterns 3, 4, 5 produce was more difficult for other, multiple-banded patterns identical PEP-GL and EST phenotypes. Table II shows and are thus provisional in nature. The three-banded the distribution ofthe five composite electromorph pat PEP-GL 3 pattern (Fig. 2) probably does not represent a terns in the nine populations studied. The Bamfield, Pt. heterozygous phenotype for a dimer protein because the Pinos, and Palos Verdes samples were fixed for patterns middle band is neither equidistant from, nor twice as 1, 4, and 2 respectively, while the remaining samples prominent as, the two outlying bands (see Fig. 6). PEP were polymorphic. The two main electromorph patterns GL 1,2 electromorphs share identical staining character (“A,―“¿B―)were respectively represented by patterns istics with the fastest of the three PEP-GL 3 bands and (1, 4) in both B. C. and California populations. they may represent three alleles ofa single locus (Fig. 2). Genetic interpretation of these complex protein phe The two slower PEP-GL 3 bands may represent a hetero notypes is affected by the absence of expected heterozy zygous monomer or two homozygous loci. PEP-LGG patterns 1, 2, and 3 all share an identical fast band which may represent a distinct locus (Fig. 2). It is unclear TableII whether the slower PEP-LGG bands are products of one Distribution ofihefive compositeelectromorphpatterns(strains) or more loci. Similarly, LAP 1 may represent a heterozy in the nine Lasaeapopulationsstudied.n = number gous monomer condition or else these individuals may ofindividuals analyzedperpopulation be homozygous at two LAP loci (Fig. 2). The other LAP patterns 2, 3, and 4 do not appear to be consistent with ofLasaea with ea compositern“A―electromorph patte a uniform heterozygous condition. Three potential loci per populationch may be represented by LAP 3 and 4; the three-banded patterns do not have the characteristics ofa heterozygous “¿B―1 dimer, the two slower bands are composed ofa relatively 45Bamfield841PopulationnFrequency 2 3 weak faster band and a more pronounced slower one and probably do not represent a heterozygous monomer con —¿â€”SaxePoint780.15 .0 —¿ —¿ dition (Fig. 2). 0.14—McNeill —¿ 0.71 None ofthe five composite electromorphs appeared to 0.08—TenBay1400.29 —¿ 0.63 be the result of matings among electromorphs in any of 0.03—HalfMoonMilePoint790.56 —¿ 0.42 the individuals characterized from the six populations 0.630.37PointPinos50—Bay19— —¿ —¿

1.0—CypressPoint25— —¿ —¿ that contained > 1 strain. Assuming that these electro 0.640.36Cayucos540.04 —¿ —¿ morph patterns have a similar genetic basis to that ob 0.720.24Palos —¿ —¿ served in other organisms, this is markedly at variance Verdes14— 1 .0 —¿ —¿â€” with random mating expectations. The observed and 222 D. O FOIGHIL AND D. J. EERNISSE difference in the maximum sizes attained by either pro 2.9 0 tein phenotype or shell color groupings.

0 The relationship between shell color and protein phe 2.7 notype was also investigated for Lasaea from Victoria's McNeill Bay population. Figures 5 and 6 show the right 2.5 0 0 E 0 •¿ •¿o 000 valves and the respective PEP-GL electromorphs for

E 0 0S0@ 0 2.3 0 0 oD. 0 forty specimens (20 white; 20 predominantly red). The I. 0 @•¿ I •¿0 20 white-shelled individuals, sorted before electropho 0 00 00 0 2.1 retic analysis, all exhibited the PEP-GL 3 electromorph. 0 000 I 0000 0 Four hundred and forty-one (441) adult specimens from 1.9 0 McNeill Bay were typed for the PEP-GL 1 enzyme dur i&a 0 0 > COO ing the initial electrophoretic survey, and the subsequent @1 1.7 investigation of spatial overlap between the two main > strains. Of these, 272 and 169, respectively, expressed PEP-GL 3 and PEP-GL 1 phenotypes. If shell color is 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 U independent of PEP-GL protein phenotype, a ratio of @ V A LV E I EN G I H mm 12.34 PEPGL 3 to 7.66 PEP@GL 1 for the 20 white speci . . . . mens in Figure 5 is expected (when analyzed by Chi Figure 3. Plot of valve height against valve length for 73 McNeil Bay Lasaea characterized accordingto protein phenotype. Squares rep- square test. P < 0.001). The 19 predominantly red resent strain 1animals(n = 27), solid squaresare double scores, correla- shelled animals for whom electromorphs were obtained, tion coefficient (r) = 0.976. Circles represent strain 3 animals (n = 46), contained both PEP-GL 1 (1 1 specimens) and PEP-GL solid circles are double scores, r = 0.921 (0.01 < P < 0.02). When the 3 (8 specimens) patterns (expected ratios: 1 1.72 PEP-GL @ large strain 1 individual (3.6 mm in length) is excluded, strain 1 r 7.27 PEP-GL 1; Chi square test: 0.25 < P < 0.50). Of = 0.967 is not significantly different from 0.921 (0.05 < P < 0.10). these 19 predominently. red animals,. 11were uniform. in color and 8 had some white patches. Both of these two Hardy-Weinberg-Castle expected genotype frequencies color subgroupings, however, were heterogenous in pro for the EST locus of 140 McNeill Bay individuals charac- tein phenotype expression. Eight ofthe totally red speci terized during the initial electrophoretic survey (Table II) mens had the PEP-GL 1pattern and four were type PEP are very different (P < 0.001). In addition, the apparent GL 3. Four of the eight mixed color individuals were gametic disequilibrium between sympatric strains mdi- PEP-GL 1 and the rest PEP-GL 3. Similar results were cates that they are reproductively isolated. obtained when shell coloration and protein phenotypes The height and length of the right valves were posi- of the Ten Mile Pt. population were investigated; white tively correlated for 73 individuals from McNeill Bay, characterized according to protein phenotype and shell

coloration. Although the height/length dimensions of 2.9 strains 1, 3 overlap substantially (Fig. 3), their respective correlation coefficients (0.976 and 0.921) are signifi 2.7 cantly different (0.01 < P < 0.02) due to the presence of E 2.5 an unusually large (3.7 mm in length) strain 1 specimen. E When this specimen is removed from the calculation, a I- 0 O€@@ 0 I 2.3 0 0 OU 0 correlation coefficient of 0.967 is obtained for strain 1, 0 0

which is not significantly different from that of strain 3 ‘¿U 2.1 U 00 (0.05 < P < 0. 10). Individuals varied in shell coloration I from totally white to totally red. Intermediate forms ‘¿U U a 0 > commonly occurred which were dorso-laterally red with -a •¿oo 4 1.7 white ventral patches. Some individuals appeared to >

have switched abruptly from forming a red shell to se Ic creting a lighter zone ofwhite growth. The height/length 2.0 2.2 2.4 2.6 2.8 3.0 32 3.4 3.6 ii 4.0 correlation coefficients do not differ significantly (P VALVE LENGTH mm > 0.50) among white and predominantly red-shelled in dividuals(0.942 and 0.953 respectively)(Fig. 4). Anterio Figure 4. Plot of valve height against valve length for 73 McNeil Bay Lasaea characterized according to shell color. Squares represent dorsal shell margins varied considerably in shape within predominently red individuals(n = 48), solid squaresaredouble scores, each strain and shell color group, e.g., see specimens 2, correlation coefficient (r) = 0.953. Circles represent white shells (n 4, 5 and 3 1, 37, 38 (Fig. 5). There was no consistent = 25), solid circles are double scores, r = 0.942, (P > 0.50). NON-HYBRIDIZING, SYMPATRIC LASAEA STRAINS 223

I®

@.,

,10 @! PL@ 4 @ *@,*@@@:. ,@.‘(i@

Figure 5. Right valves of4O Lasaea individuals from McNeill Bay, Victoria. Specimens l-20(top two rows) have white shells and specimens 21—40(bottom two rows) are predominantly or totally red. Scale = 5 mm.

individuals all expressed the PEP-GL 3 phenotype (15 GL 1 and PEP-GL 3 electromorphs. All 146 F1 progeny specimens) and totally red or mixed color specimens pro (mean brood size of 14.6 ±5.0 S.E.) perpetuated the pa duced either the PEP-GL 1or PEP-GL 3 phenotypes (64 rental PEP-GL phenotypes, including the 26 progeny of specimens). the 2 PEP-GL 3 parents. Assuming that the isolated PEP No mortality of adult McNeil! Bay Lasaea occurred GL 3 phenotype parents reproduced by self-fertilization, during laboratory culture. Less than 5% of F1 progeny as previous cytological evidence would suggest (0 died in culture, mainly as a result of premature release Foighil, 1987), the result for the 26 F1 progeny is signifi from the brood chamber before the development of valve cantly different from the 1:2: 1phenotype ratios expected opposition. Four ofthe 30 Lasaea pairs used in the pair if the PEP-GL 3 electromorph represented a heterozy mating experiments did not spawn while in the labora gous dimer protein, or if the 2 slow bands represented a tory. One member of all remaining pairs released F1 heterozygous monomer protein (Chi square test, P progeny, however, in eight ofthese pairs both adult speci <0.001). mens expressed the PEP-GL 3 electromorph. The re Eleven of the 30 brooding adults sampled in McNeill maining 18 pairs were composed of individuals exhibit Bay in May 1987 expressed a PEP-GL 3 phenotype and ing different PEP-GL phenotypes and in 17 ofthese cases the remainder exhibited the PEP-GL 1 electromorph. the individual that initiated brooding expressed a PEP Altogether, 435 F1 progeny (mean brood size = 14.5 GL 1 electromorph. When analyzed by Chi square test, ±8.4 S.E.) were typed and all perpetuated the parental this result is significantly different (0.005 < P < 0.01) phenotypes, with a single exception. This exceptional in from an expected ratio of9 PEP-GL 1to 9 PEP-GL 3, if dividual occurred in a brood of2 1juveniles, 20 of which precedence in the onset ofbrooding were independent of expressed the maternal PEP-GL 3 electromorph, the PEP-GL phenotype. Mean brood size was 17.8 ±6.4 S.E. other produced the PEP-GL 1 phenotype. It is more and a total of330 F1progeny were typed for the PEP-GL likely that this individual resulted from inadvertent enzyme. In all cases, the F1 PEP-GL phenotypes were transfer between cultures, rather than from cross-fertil identical to those ofthe confirmed parents (brooding in ization between the strains, because it totally lacked the dividuals) and did not reveal any evidence of cross-fertil maternal phenotype for this brood. ization by the potential sperm donors (non-brooding Three hundred and one (301) individuals were charac partners) (see Fig. 7). terized for the PEP-OL enzyme from the 25 samples of The 10adult McNeill Bay Lasaea maintained in isola barnacle cover (1002 cm2) taken from McNeill Bay in tion during laboratory culture reproduced successfully. October 1987. Both main Lasaea strains co-occurred in Eight and 2 individuals, respectively, expressed the PEP 18 of these samples (Figs. 8, 9) and there was no signifi 224 D. O FOIGHIL AND D. J. EERNISSE

6

@,4ss1 40

8

@ . S •¿ @ *

•¿â€¢ 0 .•. •¿S• ....,•@ [email protected] P NB

Figure 6. PEP-GL electromorphs ofthe same 40 Lasaea individuals in Figure 5. Specimens 1—20,27, 28, 30, 33, 34, 36, 39, and 40 produced a PEP-GL 3 electromorph pattern. Individuals 2 1—23,25, 26, 29, 31, 32, 35, 37, and 38 gave a PEP-GL 1electromorph pattern. Specimen 24 gave no detectable result and specimen 25 was too faint to photograph well. Figure 7. PEP-GL phenotypes of McNeill Bay Lasaea pair mating experiment showing parent (P), non-broodingpartner (NB)and the 13Fl progeny.Note that all offspringperpetuate the electromorphof the confirmed parent. Figure 8. PEP-GL phenotypes of 14 McNeill Bay Lasaea recovered in a 100 cm2 sample of barnacle cover. Both PEP-GL 1(8) and PEP-GL 3 (6) phenotypes are expressed.

cant correlation between the distributions of the main Discussion strains (r = 0.0399, P > 0.5). It appears that the 2 main Lasaea strains are not segregated in McNeill Bay on this The absence of putative intermediate protein pheno spatial scale and show a high degree ofoverlap. types and the consequent deviations from random mat ing expectations in spatially overlapping, sympatric field populations and in the progeny of pair mating experi 21 ments suggest that (1) several to many strains of Lasaea . coexist and are widespread along the west coast of North

. 2 I America, and (2) if mating occurs between the various 0. aE strains, it must be very rare. Population genetic structure ‘¿I.)2( of our study populations resemble those described by

S Crisp el al. ( 1 983) and Crisp and Standen (1988) for Eu 0. ropean Lasaea but are markedly different from the ran 0 @ C domly mating L. australis which has a planktotrophic 0 -@ 0. . larval development (O Foighil, 1988).

. S Formation of non-hybridizing, genetic strains can re

0. . •¿ sultfrom a number of reproductivemodes, including LU4 0. •¿ •¿ •¿ •¿ •¿â€¢¿ apomixis, autosegregation, pseudogamy, and self-fertil 6 @ Z@ •¿ : •¿ •¿ •¿ •¿ ization (Bell, 1982). Identifying the reproductive mode @ o a 12 ‘¿16 20 employedbyLasaeathatlackdispersivelarvaeisprov No. PEP-GL3 Phenotypes Per Sample ing to be problematical, due in part to the difficulty of Figure 9. Relative numbers of PEP-GL 1 and PEP-GL 3 pheno- identifying sperm in the ovotestis. In populations exam types in 25 100cm2 samples ofbarnacle cover from McNeill Bay, Vic- ined, a minute quantity of testis (averaging approxi toria, B. C. Large circle = double score. Correlation coefficient (r) mately 5% ofgonadal volume) is produced in the poste = —¿0.0399, P> 0.5. no-ventral lobe of the ovotestis; the sperm heads vary NON-HYBRIDIZING, SYMPATRIC, LASAEA STRAINS 225 in shape and exhibit incompletenuclear condensation cluded that L. rubra reproduces by a combination of (Pelseneer, 1903; Oldfield, 196 1; 0 Foighil, 1985a, 1987; pseudogamy and apomixis, citing 0 Foighil's (1987) de McGrath and 0 Foighil, 1986; Beauchamp, 1986). It is scription for sperm penetration in L. subviridis in sup now apparent that both British and Kerguelen Lasaea, port of this view. Although this remains a possibility, a respectively, described by Crisp et al. (1983) and Thiriot number of important points need to be clarified before Quiévreuxet al. (1988) as female, apomictic clones, are Crisp and Standen's (1988) interpretation can be ac actually simultaneous hermaphrodites which appear cepted. 0 Foighil (1987) rejected apomixis in L. subviri identical to northeastern Pacific Lasaea in their gonadal dis, not because of sperm penetration, but because the and sperm morphologies (Oldfield, 1961; 0 Foighil, egg apparently undergoes a meiotic division before first 1985a; McGrath and 0 Foighil, 1986; 0 Foighil, un cleavage, producing two polar bodies. It is quite possible pub.). Despite their unusual morphology, these sperm that European L. rubra eggs are likewise meiotic. How appear to be functionally mature and each is capable of ever, this remains to be established by observation of cy swimming and undergoing an acrosomal reaction, bind tological events that occur before first cleavage. An alter ing to, and penetrating an egg (0 Foighil, 1987). nate interpretation ofthe protein phenotype patterns ob Important data on post-spawning cytological events in served by Crisp and Standen (1988) is that they represent isolated northeastern Pacific individuals are available (0 multiple homozygous loci. C. Thiriot-Quiévreux(pers. Foighil, 1987), as reviewed in the introduction. These comm.) found very high chromosome numbers in L. ru data, especially the evidence for production oftwo polar bra, and these represent potential sources of such loci. bodies before first cleavage, indicate that individuals Finally, pseudogamy is employed by all-female clones from populations we studied reproduce not by apomixis (Moore et al., 1956; Kallman, 1962; Uzzell, 1964; but by automixis. Extrapolation to all northeastern Pa Schultz, 197 1, 1977; Kiester et al., 198 1; Stenseth et al., cific strains assumes that results from these isolated mdi 1985) or hermaphrodites (e.g., the enchytraeid oligo viduals ofunknown electrophoretic phenotype represent chaete Lumricillus, see Christensen and O'Connor, the norm. Additionally, since sperm were seen to pene 1958; Christensen, 1980) that are sexual parasites of trate eggs (0 Foighil, 1987), these data suggest that the closely related cross-fertilizing species and are incapable particular automitic mode must be either self-fertiliza of reproducing in isolation. L. rubra can reproduce in tion or meiotic parthenogenesis with some form of isolation to at least an F2 generation (Crisp et al., 1983). pseudogamy. In practice, self-fertilizing organisms and The combination of pseudogamy with apomixis pro meiotic parthenogens are distinguished because the for posed by Crisp and Standen (1988) would indeed be mer produce both sperm and eggs, giving differentiation noteworthy, iffurther substantiated, because we are pres ofgender at the gametic level (Bell, 1982). ently unaware of any species where individuals are Ideally, electrophoretic examination ofthe progeny of known to use their own sperm to initiate parthenogenic isolated heterozygous (Aa) adults could provide strong development. evidence for self-fertilization, or for some types of au Thiriot-Quiévreuxet al. (1988) concluded that Ker tomictic parthenogenesis that have similar consequences guelen Lasaea were apomictic based on the absence of to selfing (White, 1973; Suomalainen et al., 1987), if the sperm and meiotic stages. It is now known that Ker progeny do not deviate significantly from an expected guelen Lasaea are simultaneous hermaphrodites with 1AA:2Aa: laa ratio. However, lack of segregation of the greatly reduced sperm production (0 Foighil, unpub.). protein phenotypes in sampled populations makes posi Because of the relatively tiny testis, it is possible that tive identification of heterozygous loci difficult, because sperm were overlooked by Thiriot-Quiévreux et al. multiple bands on a gel that resemble typical heterozy (1988) in their chromosome preparations. We consider gous monomers or dimers could also be explained as the that a self-fertilizing mode of reproduction has yet to be results ofmultiple loci. Similarly, ifall progeny resemble ruled out for European and Kerguelen Lasaea popula their parent at a particular locus, as they did in our study tions. As is the case for northeastern Pacific congeners, a and as reported by Crisp et al. (1983), this is not conclu detailed cytological study is needed to firmly establish sive evidence for apomixis. Only a detailed study of male the reproductive mode ofthese populations. and female pronuclear interaction is likely to resolve the Our present working hypothesis is that northeastern reproductive mode in these cases. Pacific Lasaea reproduce by automixis, probably by self Crisp and Standen (1988) used a different buffer sys fertilization. This interpretation is more parsimonious tem from the one used by Crisp et al. (1983) and found than that ofpseudogamy combined with automictic par non-segregating protein phenotypes they interpreted as thenogenesis and, as discussed above, pseudogamy is suggestive evidence for fixed-heterozygosity at PGM and only known to be employed by sexual parasites incapa GPI loci of European Lasaea (L. rubra). They then con ble of reproducing in isolation. If self-fertilization is in 226 D. O FOIGHIL AND D. J. EERNISSE deed the predominate norm in northeastern Pacific La combinations to selection and eventual elimination. saea populations, heterozygous loci should be extremely Thus prolonged inbreeding dilutes the genetic penalty of rare within each strain, and each band in the complex inbreeding depression caused by self-fertilization. If a multi-banded PEP-GL, PEP-LGG and LAP electro self-fertile individual happens to arise in an already in morph patterns recorded in this study should represent bred population, and it additionally produces relatively a separate homozygous locus. Each “¿B―strain typically few sperm, it would then be at a reproductive advantage, yielded a greater number of protein phenotypes (i.e., because of the lowered cost of spermatogenesis and at a bands) than did each “¿A―strain in our study. One inter genetic advantage due to the reduced “¿costof meiosis― esting possibility is that the strains differ in ploidy and (Williams, 1975; Maynard Smith, 1978; Bell, 1982). In perhaps one or more ploidy duplication events has been deed, the genetic advantage of a reduced “¿costof meio responsible, in part, for the reproductive isolation(s) of sis―implies that once developed, the evolution of corn Lasaea strains (C. Moritz, pers. comm.). The only pletely self-fertilizing lineages may be an irreversible step published work to date on Lasaea karyology is that of (Bull andCharnov, 1985). Strathmann et al. (1984) were Thiriot-Quiévreux et al. (1988) who found a record concerned primarily with explaining hermaphroditism (though variable) number of chromosomes for the class in groups that are normally gonochoric (e.g., echino in Kerguelen Lasaea (100 2n 120). A sim derms, anemones, chitons, sipunculans, etc.), but they ilarly high chromosome complement may be present in also point out that the basic model should apply to ex northeastern Pacific Lasaea (especially the “¿B―strains) planing exceptional self-fertilizers in hermaphroditic and may represent a source ofmultiple homozygous loci. taxa that normally have effective blocks to self-fertiliza Species that reproduce by self-fertilization still, in the tion (e.g., tunicates, nudibranchs, certain bivalve taxa in ory, retain the potential for cross-fertilization, and exclu cluding Lasaea). The lack of a plankionic dispersive sively self-fertilizing populations [e.g., Rivulus marmora stage, simultaneous hermaphroditism, and apparent tus (Kallman and Harrington 1964; Vrijenhoek, 1985)] high level of self-fertilization of northeastern Pacific La are thought to be extremely rare in nature (Bell, 1982). saea are consistent with the proposal (Strathmann et al., A mixed self- and cross-fertilizing mating pattern has the 1984; Eernisse, 1988) that having crawl-away offspring potential of creating a wide variety of new genotypes, as can lead to departure from cross-fertilization. occasional outcrossing between inbred lineages will re Cross-fertilizing marine invertebrate species that are sult in highly heterozygous progeny (Antonovics, 1968; typically sedentary as adults and lack a planktonic larval Bucklin et aL, 1984). Cross-fertilizations in northeastern stage show significant interpopulational genetic diver Pacific Lasaea populations might be exceptionally rare gence on a relatively small geographic scale (Berger, or absent, judging from the lack of intermediate protein 1973,1977;Snyderand Gooch,1973;Campbell,1978; phenotypes encountered in this study and the apparent Ward and Warwick, 1980; Bulnheim and Scholl, 1981; absence of sperm transfer and sperm storage mecha Burton, 1983; Janson and Ward, 1984; Palmer, 1984; nisms that enable other suprabranchial brooding bi Grant and Utter, 1988). Interpopulational genetic drift valves to cross-fertilize (Coe, 193 1;Townsley et al., 1965; in these cases may originate from a founder effect (Hol Turner and Yakovlev, 1983; 0 Foighil, 1985b, c). Still, gate, 1966; Nei et al., 1975) during an initial coloniza even rare cross-fertilization could provide an alternative tion, or a later genetic bottleneck event, that may be mechanism, besides ploidy duplication, for the creation maintained and enhanced by infrequent genetic ex ofnew strains. change with other populations over time. In contrast, our The majority of marine invertebrates including bi results for northeastern Pacific Lasaea show that some valves cross-fertilize, so it is interesting to consider how strains (1, 4) are present in geographically distant sites. A predominate self-fertilization might arise. Various ge number of potential factors may have contributed to netic models of the evolution of self-fertilization predict this, including alternative dispersal techniques such as that a history of inbreeding predisposes originally cross drifting (Lane et al., 1985) and rafting (High fertilizing populations to the development of automixis smith, 1985) (short and long distances respectively), low (self-fertilization) by removing recessive deleterious al rates of mutation, and a predominantly self-compatible leles (Charlesworth and Charlesworth, 198 1; Lande and reproductive mode. Self-fertilization is genetically con Schemske, 1985; Uyenoyama, 1986). This forms the servative because new alleles formed by mutation are premise for Strathmann et al.'s (1984) hypothesis to ac rapidly expressed in homozygous combinations and are count for the association ofsimultaneous hermaphrodit thus directly exposed to selection (Bell, 1982). Newly ism and brooding ofyoung to a crawl-awayjuvenile stage formed populations will not experience a founder effect in many marine invertebrate taxa. Strathmann et al. sug if the initial colonizers previously existed as reproduc gest that reduced dispersal promotes inbreeding which tively isolated strains at the source site. lowers heterozygosity, exposing deleterious homozygous Taxonomic interpretations based on a small number NON-HYBRIDIZING, SYMPATRIC LASAEASTRAINS 227

ofloci can lead to a potentially serious bias (Nei, 1972). plies best to randomly mating organisms, would seem Even though we examined relatively few loci, it is clear inappropriate for these populations. Ponder (197 1) con that northeastern Pacific Lasaea populations are com cluded that most nominal species of Lasaea are merely posed of a variety of reproductively isolated strains as regional subspecies, or ecotypes ofthe type species L. ru Crisp et al. (1983) found for British populations. These bra. Alternatively, one could conclude that every distinct strains are readily distinguishable by electrophoretic electrophoretic strain is a distinct historical entity or, per analysis. haps, species. Efforts to determine the level at which the Diagnostic separation of strains from Victoria, B. C. category “¿species―best applies would best wait for repro populations on the basis of shell phenotype appears un ductive investigations of other Lasaea populations and reliable, except for some apparent color differences be a better understanding ofLasaea historical relationships, tween the strains. We can predict with a high degree of which are undoubtedly complex among Lasaea that lack @ confidence that white-shelled Lasaea from Victoria will dispersive larval stage (Ponder, 197 1; Crisp et a!., 1983; express particular protein electromorphs. However, 0 Foighil, l986a; Thiriot-Quiévreux et al., 1988). Reso specimens with the same protein phenotypes may also lution ofthese relationships will require a multidisciplin have mixed shell coloration (red/white) or totally red ary approach applied to a variety of populations of this shells. Shell pigmentation is known to be light-induced near-cosmopolitan genus. in juvenile mussels (Trevelyan and Chang, 1987), and, in hinemoa, the extent ofred coloration in Acknowledgments the valves is thought to be related to the degree of expo sure to sunlight (Ponder, 197 1).This may also be the case Funding was provided by a University of Victoria for at least strain 3 of northeastern Pacific Lasaea be graduate student fellowship and a FridayHarbor Labo cause specimens recovered from deep crevices had the ratories post-doctorate fellowship to D. 0 Foighil; and whitest shells. There may be a partial habitat difference NSF grant OCE-84 15258 to R. R. Strathmann and D. J. between the two main strains in Victoria, with strain 3 Eernisse. We are indebted to D. J. Crisp for sending us occurring in both deep and shallow crevices and strain 1 an unpublished manuscript and to C. Thiriot-Quiévreux found only in shallow crevices. Data from the laboratory for letting us examine her histological material. A. R. pair mating experiments hint at physiological or spawn Palmer, R. Grosberg, C. Moritz, F. Sly, R. R. Strath ing differences between the two main strains: 17 of the mann, and anonymous reviewers gave valuable com first individuals to spawn from the 18 heterogeneous ments on earlier drafts. We thank A. 0. D. Willows, di pairs expressed a PEP-GL 1phenotype. rector ofFHL, for the use of facilities. Criteria used by Keen (1938) to separate two nominal Lasaea species in Californian populations are made Literature Cited quate when applied to Victoria Lasaea ofknown protein Antonovics, J. 1968. Evolution in closely adjacent plant populations. phenotype. Keen (1938) distinguished L. cistula from L. V. Evolutionofself-fertility.Heredity23:219—238. subviridis by its smaller size, less oblique outline, darker Ayala, F. J., J. R. Powell, M. L Tracey, C. A. Murao, and S. Perez color, higher umbones and by its more abrupt slope from Salas. 1972. Enzyme variability in the Drosophila willistoni group. IV. Genic variation in natural populations of Drosophila the umbone to the anterior margin. However, Ponder willistoni. Genetics70: 113—139. (197 1) and Beauchamp (1985) could not distinguish on Barnard, K. H. 1964. Contributions to the knowledge ofSouth Afri morphological grounds two different forms ofLasaea in can marine . Pt. 5. Lamellibranchiata. Ann. S. Afr. Mus. Californian populations. As in this present study, Ponder 47:361—593. (197 1) found Keen's ( 1938) distinguishing morphologi Beauchamp, K. A. 1985. The reproductiveecology ofa brooding, her maphroditic clam, Lasaea subviridis. M. Sc. Thesis, University of cal characteristics to be highly variable among individual California, Santa Cruz. 56 pp. shells. For the present, it would appear that electropho Beauchamp, K. A. 1986. Reproductive ecology ofthe brooding, her retic analysis is necessary for the reliable identification maphroditic clam, Lasaeasubviridis. Mar. Biol. 93: 225—235. of northeastern Pacific Lasaea strains. Whether the two Bell, G. 1982. The Masterpiece ofNature. Croom Helm, London. main strains detected in this study correspond to L. sub 635 pp. Berger, E. M. 1973. Gene-enzyme variation in three sympatric spe viridis and/or L. cistula will require careful comparisons cies ofLittorina. Biol. Bull. 145:83—90. of any diagnostic morphological distinctions, should Berger, E. M. 1977. Gene-enzyme variation in three sympatric spe they be found, with existing type material for northeast cies ofLittorina. II. The Roscoffpopulation with a note on the ori ern Pacific Lasaea. gin ofNorth American L. littorea. BioL Bull. 153: 255—264. There is as yet no evidence for cross-fertilization be Booth, J. D. 1979. Common bivalve larvae from New Zealand: Lep tonacea. N. Z. J. Mar. FreshwaterRes. 13: 24 1—254. tween the strains found in British (Crisp et al., 1983) and Bucklin,A., D. Hedgecock,andC. Hand. 1984. Geneticevidenceof northeastern Pacific Lasaea populations. Therefore the self-fertilization in the sea anemone Epiactis prolifera. Mar. BioL biological species concept (Mayr, 1957, 1963), which ap 84: 175—182. 228 D. O FOIGHIL AND D. J. EERNISSE

Bull, J. J@and E. L. Charnov. 1985. On irreversible evolution. Evolu zyme and shell characters in Littorina saxatiis Olivi (Prosobran tion39:1149—1155. chia: Littorinidae). Biol. J. Linn. Soc. 22: 289—307. Bulnbeim, H. P., and A. Scholl. 1981. Genetic variation between geo Kallman, K. D. 1962. Population genetics ofthe gynogenetic teleost graphic populations ofthe amphipods Gammarus zaddachi and G. Molliensaformosa (Girard). Evolution 16:497—504. salinus. Mar. Biol. 64: 105—115. Kallman, K. D., and R. W. Harrington. 1964. Evidence for the exis Burton, R. S. 1983. Protein polymorphisms and genetic differentia tence of homozygous clones in the self-fertilizing hermaphrodite tion of marine invertebrate populations. Mar. Biol. Lett. 4: 193— teleost Rivulus marmoratus(Poey). Biol. Bull. 126: 101—114. 206. Kay, E. A. 1979. Hawaiian Marine Shells. Bernice P. Bishop Mu Cam@ll, C. A. 1978. Genetic divergence between populations of seum Press, Honolulu. 620 pp. Thais lamellosa (Gmelin). Pp. 157—170in Marine Organisms: Ge Keen, A. M. 1938. New pelecypod species ofthe genera Lasaea and Crassinella. Proc. Malacol. Soc. Lond. 23: 18—32. netics, Ecology and Evolution, B. Battaglia and J. A. Beardmore, Kiester, A. R., T. Nagylaki, and B. Sheller. 1981. Population dynam eds. Plenum Press, New York. ics ofspecies with gynogenetic sibling species. Theor. Pop. Biol. 19: Charlesworth, D., and B. Charlesworth. 1981. Allocation of resources 358—369. to male and female functions in hermaphrodites. Biol. I Linn. Soc. Lande, R., and D. W. Schemske. 1985. The evolution of self-fertiliza 15:57—74. tion and inbreeding depression in plants. I. Genetic models. Evolu Chavan, A. 1969. Leptonacea Gray 1847. Pp. 518—537in Treatise tion39:24—40. On Invertebrate Paleotology. Pt. N. Mollusca (2), R. C. Moore, ed. Lane, D. J. W., A. R. Beaumont, and J. R. Hunter. 1985. Byssus drift Geological Society of America and University of Kansas Press, ing and the drifting threads ofthe young post-larval mussel Mytilus Lawrence. edulis. Mar. Biol. 84: 30 1—308. Christensen, B., and F. B. O'Connor. 1958. Pseudofertilization in the Maynard Smith, J. 1978. The Evolution ofSex. Cambridge Univer genus Lumbricillus(Enchytraeidae). Nature 181: 1085—1086. sity Press, Cambridge, U. K. 222 pp. Christensen, B. 1980. Constant differential distribution of genetic Mayr, E. 1957. Species concepts and definitions. Pp. 1—22in The variants in polyploid parthenogenetic forms ofLumbricillus linea Species Problem, J. Huxley, A. C. Hardy and E. B. Ford, eds. Allen tus(Enchytraeidae, Oligochaeta). Hereditas 92: 193—198. and Unwin Press, London. Coe, W. R. 1931. Spermatogenesis in the California oyster, Ostrea Mayr, E. 1963. Species and Evolution. Harvard University lurida. Biol. Bull. 61: 309—315. Press, Cambridge, Massachusetts@ Crisp, D. J., A. Burfltt, K. Rodriguez, and M. D. Budd. 1983. Lasaea McGrath, D., and D. 0 Foighil. 1986. Population dynamics and re rubra: an apomictic bivalve. Mar. Biol. Lett. 4: 127—136. production ofhermaphroditic Lasaea rubra (Bivalvia: Galeomma tacea). Ophelia 25: 209—219. Crisp, D. J., and A. Standen. 1988. Lasaea rubra (Montagu) (Bival Moore, B. P., G. E. Woodroffe, and A. R. Sanderson. 1956. via: Erycinacea), an apomictic crevice-living bivalve with clones Polymorphism and parthenogenesis in a ptinid beetle. Nature 177: separated by tidal level preference. J. Exp. Mar. Biol. Ecol. 117: 27— 847—848. 45. Morton, J. E. 1960. The responses and orientation ofthe bivalve La Dell, W. H. 1899. Synopsis ofthe recent and tertiary Leptonacea of saea rubra (Montagu). J. Mar. Biol. Assoc. U. K. 39: 5—26. North America and the West Indies. U. S. NatI. Mus. Proc. 23: 18- Nei, M. 1972. Genetic distance between populations. Am. Nat. 106: 32. 283—292. Dell, R. K. 1964. Antartic and subantartic Mollusca: Amphineura, Nei, M., T. Maruyama and R. ChakkabOrty. 1975. The bottleneck Scaphopoda and Bivalvia. Discovery Rep. 33: 93—250. effect and genetic variability in populations. Evolution 29: 1—10. Eernisse, D. J. 1988. Reproductive patterns in six species of Lepi 0 Foighil,D. 1985a. FinestructureofLasaeasubviridisandMysella dochitona (Mollusca: Polyplacophora) from the Pacific coast of tumida sperm (Bivalvia: Galeommatacea). Zoomorphology 105: North America. Biol. Bull. 174: 287—302. 125—135. Fischer, E. A. 1981. Sexualallocationin a simultaneouslyhermaph oFoighil,D.1985b.Spermtransferandstorageinthebroodingbi roditic coral reeffish. Am. Nat. I 17: 64—82. - valve Mysella tumida. BioL Bull. 169:602—614. Glyna, P. W. 1965. Community composition, structure, and interre 0 Foighil,D. 1985c. Form, function and origin oftemporary dwarf lationships in the marine intertidal Endocladia muricata-Balanus males in Pseudopythina rugifera (Bivalvia: Galeommatacea). Veil glandula association in Monterey Bay, California. Beaufortia 148: ger 27: 72—80. 1—198. oFoighil,D.1986a.Reproductivemodesandtheirconsequencesin Grant, W. S., and F. M. Utter. 1988. Genetic heterogeneity on three galeommatacean clams: Mysella tumida, Pseudopythina rugi different geographic scales in Nucella lamellosa (Prosobranchia; ftra and Lasaea subviridis(Mollusca: Bivalvia). Ph.D. Thesis, Urn Thadidae). Malacologia 28: 275—287. versity ofVictoria, 251 pp. 0 Foighil,D. 1986b. Prodissoconchmorphologyisenvironmentally Haderlie, E. C., and D. P. Abbott. 1980. Bivalvia: the clams and their allies.Pp.355—411inIntertidalInvertebratesofCalzfornia.R.H. modified in the brooding bivalve Lasaea subviridis. Mar. Biol. 92: 517—524. Morris, D. P. Abbott and E. C. Haderlie, eds. Stanford University Press. oFoighil,D.1987.Cytologicalevidenceforself-fertilizationinthe brooding bivalve Lasaea subviridis. mt. j. Jnvertebr. Reprod. Dev. Heath, D. J. 1979. Brooding and the evolution of hermaphroditism. 12:83—90. J.Theor.Biol.81:151—155. oFoighil,D.1988.Randommatingandplanktotrophiclarvaldevel Highsmith, R. C. 1985. Floating and algal rafting as potential dis opment in the brooding hermaphroditic clam Lasaea persal mechanisms in brooding invertebrates. Mar. Ecol. Prog. Ser. australis(Lamarck, 1818). Veliger3l: (in press). 25: 169—179. Oldfield, E. 1961. The functional morphology of Kellia suborbicu Holgate, P. 1966. A mathematical study of the founder principle of laris (Montagu), Montacutafrrruginosa (Montagu) and M. substri evolutionary genetics. J. App!. Probab. 3: 115—128. ata (Montagu), (Mollusca; Lamellibranchiata). Proc. Malacol. Soc. Janson, K., and R. D. Ward. 1984. Microgeographic variation in allo Land.32:255—295. NON-HYBRIDIZING, SYMPATRIC, LASAEA STRAINS 229 Oldfield, E. 1964. The reproduction and development ofsome mem Suomalainen, E., A. Saura, and J. Lokki. 1987. Cytology and Evolu bersofthe Erycinidaeand Montacutidae(Mollusca, Eulamellibran tion ofParthenogenesis. CRC Press, Boca Raton, Florida. 216 pp. chiata). Proc. Malac. Soc. Land. 36: 79—120. Townsley, P. M., R. A. Rickey, and P. C. Trussel. 1965. The labora Palmer, A. R. 1984. Species cohesiveness and genetic control of shell tory rearing of the shipworm Bankia setacea (Tyron). Proc. Nail. color and form in Thais emarginata (Prosobranchia, Muricacea): Shellfish. Assoc. 56: 49—52. preliminary results. Malacologia 25: 477—491. Thiriot-Quiévreux,C., J. Soyer, F. de Bovéeand P. Albert. 1988. Pelseneer, P. 1903. Mollusques (amphineures, gastropodes et lamel Unusual chromosome complement in the brooding bivalve Lasaea libranches). in Résultatsdu voyage du S. Y. Belgica. Buschmann, Lasaea consanguinea. Genetica 76: 143—151. Anvers. Tracey, M. L., K. Nelson, D. Hedgecock, R. A. Schleser and M. L Ponder, W. F. 1971. Some New Zealand and Subantartic bivalves of Pressick. 1975. Biochemical geneticsoflobsters: genetic variation the Cyamiacea and Leptonacea with descriptions ofnew taxa. Rec. and the structure of the American lobster (Homarus americanus) Dom. Mus.Wellington7:119—141. populations. @J.Fish. Res. Board Canada, 32: 2091—2101. Roberts, D. 1984. A comparative study ofLasaea australis, Vulsella Trevelyan, G. A. and E. S. Chang. 1987. Light-induced shell pigmen spongarium, Pinna bicolorand Donacilla cuneata (Mollusca: Bival tation in post-larval Mytilus edulis and its use as a biological tag. via) from Princess Royal Harbor, Western Australia. J. MolL Stud. Mar. Ecol. Prog. Ser. 39: 137—144. 50: 129—136. Turner, R. D., and Y. M. Yakovlev. 1983. Dwarfmales in the Tere Rosewater, J. 1975. An annotated list ofthe marine mollusks of As dinidae (Bivalvia; Pholadacea). Science 219: 1077—1078. cension Island, South Atlantic Ocean. Smithson. Contrib. Zool. 189:41pp. Uyenoyama, M. K. 1986. Inbreeding and the cost ofmeiosis: the evo Schultz, R. J. 1971. Special adaptive problems associated with uni lution of selfing in populations practicing biparental inbreeding. sexualfishes.Am. Zoo!.11:351-360. Evolution 40: 388—404. Schultz, R. J. 1977. Evolution and ecology of unisexual fishes. Pp. Uzzell, T. M. 1964. Relations ofthediploid and triploid species of the 277—331in M. K. Hecht, W. C. Steere and B. Wallace(eds.). Evolu Ambystomajeffersonium complex (Amphibia, Caudata). Copeia 2: tionary Biology. Plenum Press, New York. 257—300. Snyder, T. P., and J. L. Gooch. 1973. Genetic differentiation in Lit Vrijenhoek, R. C. 1985. Homozygosity and interstrain variation in torina saxatilis(Gastropoda). Mar. Biol. 22: 177—182. the self-fertilizing hermaphroditic fish, Rivulus marmoratus. J. Soot-Ryen, T. 1960. Pelecypods from Tristan da Cunha. Res. Nor. Hered.76:82—84. Sci.Exped.TristandaCunha49:1-47. Ward, R. D., and T. Warwick. 1980. Genetic differentiation in the Stenseth, N. C., L R. Kirkendall, and N. Morgan. 1985. On the evo molluscan species Littorina rudis and Littorina arcana (Prosobran lution ofpseudogamy. Evolution 39: 294—307. chia: Littorinidae). Biol. J. Linn. Soc. 14: 417—428. Strathmann, R. R., M. F. Strathmann, and R. H. Empson. 1984. Does White, M. J. D. 1973. Animal Cytology and Evolution. Cambridge limited brood capacity link adult size, brooding and simultaneous University Press. 961 pp. hermaphroditism? A test with Asterina phylactica. Am. Nat. 123: Williams, G. C. 1975. SexandEvolution. Princeton University Press, 796—818. Princeton, N. J. 255 pp.