Genetic characterization of the hybridization betweenMytilus edulis andM. galloprovincialis on the Atlantic coast of France

A genetic study carried out on nine natural mussel populations on the French Atlantic coast from 1989–1990 revealed interdigitation between typicallyMytilus edulis and typicallyM. galloprovincialis populations and intermediate populations. The allele components of the populations followed aM. edulis/M. galloprovincialis gradient which does not correspond to a geographical gradient. Strong hybridization was evident in samples with intermediate allele frequencies. The respective importance of gene flow and selection is discussed in the light of the results and the evironmental features of the sampling zone.     

Morphological variation of Mytilus edulis and Mytilus trossulus in eastern Newfoundland

Allopatric populations of Mytilus species show distinct shell morphology which may be due to genetic and/or environmental effects. Sympatric populations of Mytilus species show similar shell morphology which may be due to hybridization eroding morphological differences and/or the influence of common environmental conditions. The present study examined shell morphology and shell shape from 16 sites in eastern Newfoundland where M. edulis L. and M. trossulus Gould coexist in common environments with limited hybridization. Shell morphology was based on measurements of eight characters, and shell shape was quantified by elliptic Fourier analysis of shell outlines. Significant differences were observed between species for both shell morphology and shell shape across 16 sites sampled. The relatively small differences in morphology and shape between the species were probably due to exposure to common environments rather than hybridization. Shell shape for M. edulis was more eccentric compared to M. trossulus which was more elongated. Shell shape analysis of a range of size classes at one site showed a change from an eccentric to an elongated shape going from the smaller to the larger size classes. Both species showed a similar trend, with the larger M. edulis more eccentric and the larger M. trossulus more elongated. Received: 17 July 1998 / Accepted: 6 January 1999     

Spatio-temporal patterns in the genetic structure of recently settled blue mussels (<Emphasis Type="Italic">Mytilus</Emphasis> spp.) across a hybrid zone

Previous studies of a hybrid zone between the mussels Mytilus edulis Linnaeus and M. galloprovincialis Lamarck have not resolved the relative importance of the genetic composition of settling larval cohorts versus post-settlement selection in determining the distribution of the parental species and their hybrids. In the present study, recently settled mussels (spat) were collected from 20 sites in southwest England throughout the summer and fall (May–October) in 1998 and 1999. This study investigated the spatio-temporal patterns of settlement and genetics of mussel spat by genetically identifying M. edulis, M. galloprovincialis and their hybrids using the diagnostic PCR marker Glu-5. Settlement was observed earlier in populations of M. edulis than in populations of M. galloprovincialis. Settlement occurred in hybrid populations at times intermediate to and overlapping with both of the parental populations. Temporal genetic variation within years was rare at most sites, while there was some variation between the two years. Spatial genetic variation, however, was common among spat settling within the hybrid populations and matched that observed in small, sub-adults at the same sites. No consistent directional changes in allele frequency were observed over the course of several weeks after settlement. These data suggest that the observed spatial variation in the adult populations is the result of spatial variation in settling larval cohorts and not of either temporal genetic variation or of selection soon after settlement.Communicated by J.P. Grassle, New Brunswick     

Assessment of natural selection in a hybrid population of mussels: evaluation of exogenous vs endogenous selection models

We examined natural selection within a population of marine mussels, sampled in southwestern England in June 1991, containing a high frequency of hybrids between Mytilus edulis L. and M. galloprovincialis Lmk. This system is particularly tractable for the assessment of natural selection because hybridization is common and individual mussels can be aged, allowing changes in the frequency of hybrid genotypes among age classes to be determined. We show that strong viability selection occurs among hybrid genotypes which results in the virtual elimination of M. edulis–like genotypes from the population over a period of 3 years. Recombinant hybrid genotypes are intermediate in fitness, with M. edulis–like genotypes having a lower survival rate and M. galloprovincialis–like genotypes having a higher survival rate than genotypes of mixed ancestry. Since intermediate fitness for hybrid genotypes is inconsistent with endogenous selection models we conclude that the structure and position of this hybrid zone is probably generated by exogenous selection. This pattern of selection is a recurring feature of this hybrid population and likely occurs elsewhere in the hybrid zone. Selection against M. edulis–like genotypes appears to be offset by extensive immigration of larvae dispersed from pure populations of M. edulis. Received: 14 July 1997 / Accepted: 24 February 1998     

Genotype-dependent fecundity and temporal variation of spawning in hybrid mussel (Mytilus) populations

Mussel samples were collected at 4 to 6 wk intervals throughout 1987 from two hybridMytilus edulis/M. galloprovincialis populations, at Croyde Bay and Whitsand Bay, in southwest England. These were analyzed at two polymorphic loci which are diagnostic for allozyme differences which typifyM. edulis andM. galloprovincialis. Dry mantle weight as a function of shell length was determined for all individuals of each sample. Size-frequency data for the two populations was obtained in September 1987 and March 1988. For all genotypes at both sites, fecundity was a function of shell length, and in both populations the frequency ofM. galloprovincialis alleles was positively correlated with shell length. At both sites, allozyme genotype explained a significant amount of variation in mantle weight either when assessed as a main effect or when assessed as an interaction with shell length or time of collection. At Croyde,M. galloprovincialis mussels had greater estimated fecundity per unit length than theM. edulis mussels. Differences in the timing of spawning activity between theM. edulis and theM. galloprovincialis mussels were inferred, and these differences might act to reduce the amount of interbreeding at Croyde. At Whitsand, a reduced level of variability in the timing of spawning activity and fecundity between the genotypes was observed and explained by a higher degree of genetic mixing. Because theM. galloprovincialis mussels had (1) a greater estimated fecundity at any length, and (2) a greater mean length than theM. edulis mussels, the mean genotypic annual fecundity perM. galloprovincialis mussel was 2.8 times greater than an individualM. edulis mussel at Croyde, and 2.2 times greater than an individualM. edulis mussel at Whitsand. This evidence thatM. galloprovincialis mussels have an advantage in fecundity, and thus perhaps in fertility, taken together with the evidence thatM. galloprovincialis also has a higher viability, indicates directional selection in favour of theM. galloprovincialis phenotype. Because of the observed temporal stability of the population it seems likely that this selection is counterbalanced by a massive imigration ofM. edulis spat from neighbouring populations.     

Mitochondrial DNA heteroplasmy in European populations of the mussel Mytilus trossulus

Mitochondrial DNA (mtDNA) of Mytilus trossulus from the Gulf of Gdansk (southern Baltic) and M. edulis from Swansea Bay, South Wales, UK, collected in 1991, was studied by restriction-enzyme analysis. These two species were more similar to each other in haplotype frequencies than either was to M. galloprovincialis from Britain. M. trossulus resembles M. edulis in having a high frequency of heteroplasmy restricted to males. However, in contrast to M. edulis where restriction-site heteroplasmy predominates, in M. trossulus heteroplasmic individuals possess two genomes which differ in length by up to 3 kilobases.     

Introgression and mitochondrial DNA heteroplasmy in the Baltic populations of mussels Mytilus trossulus and M. edulis

A strong clinal change in salinity occurs between the Baltic Sea and the North Sea, Atlantic Ocean, in the Danish Straits, where hybridization zone between mussels Mytilus edulis and M. trossulus has been reported. Eleven samples of mussels were studied from the Danish Straits and the inner Baltic Sea. Extensive introgression of M. edulis alleles from the North Sea into populations throughout the Baltic was ascertained for mitochondrial DNA (mtDNA) and two nuclear markers (ME15–16 and ITS). In the opposite direction, introgression of M. trossulus alleles into the M. edulis background was observed at the EFbis nuclear marker in populations from Kattegat (Danish Straits). While only M. edulis F (female) mtDNA was present in the Baltic, there were still strong differences in frequencies in the control region length variants between the Danish Straits and the inner Baltic samples, and weaker variation in coding region ND2–COIII haplotype frequencies. In the assays of the two mtDNA regions, various patterns of heteroplasmy were detected in 32% of all the studied individual mussels; this includes the presence of distinct, independently inherited M and F mitochondria in males, as well as the presence of two different distinguishable F genomes. The male-inherited M mtDNA genomes are quite common in the mussels from the Danish Straits, but very rare in males from the inner Baltic. Instead, a recombined control region variant (1r), which seems to have taken over the role of the M genome, was present in a number of specimens in the Baltic. Observations of heteroplasmy for two F genomes in some females and males confirm disruptions of the doubly uniparental inheritance mechanism in the hybrid Baltic Mytilus.     

The Mytilus edulis species complex in southwest England: effects of hybridization and introgression upon interlocus associations and morphometric variation

Loglinear analysis of electrophoretic data from two hybrid Mytilus edulis x galloprovincialis populations in southwest England revealed non-significant associations between genotypes at four allozyme loci, each of which is partially diagnostic for differences between the two taxa. Significant non-random genotypic associations within the context of the non-significant model involved all four assayed loci equally, consistent with their occurrence in a relatively tight linkage group. Multivariate analyses were used to examine electrophoretic variation from the two hybrid populations, and morphometric variation in the hybrid populations and in four allopatric (two M. edulis and two M. galloprovincialis) populations from western Europe While the number of hybrid mussels is high at both sites (22% at Croyde, 53% at Whitsand) the two taxa have largely maintained the genetic differences which exist between them in allopatry. However, morphological differences between the taxa have been eroded for mussels within the hybrid zone, whereas these differences are quite pronounced for mussels from allopatric populations. It is proposed that each taxon within the genus maintains its genetic identity, despite high dispersal potential, widespread hybridization, and high levels of introgression, as a result of adaptation to different environments. The worlwide occurrence of all four Mytilus hybrid zones at ecotones between recognized biogeographical provinces which are characterized by differences in temperature and salinity is consistent with such an interpretation.     

Genetic and nutritional control of the gametogenic cycle in Mytilus edulis

Individuals of Mytilus edulis were collected from two populations in easterm Long Island Sound, USA, at 6 wk intervals from January 1982 to February 1983, and the hypothesis that nutrient availability controls the timing of gametogenic events in this mussel was tested using genetic variants at the Lap locus. In eastern Long Island Sound, Lap genotypes differ in the net rate of nitrogen accumulation. A highly significant difference in reproductive cycle among genotypes occurred in a population of M. edulis where Lap genotypes differ in nitrogen budget. Genotypes with lower rates of nutrient accumulation delayed the reproductive cycle by approximately six weeks relative to genotypes with high rates of nutrient accumulation. No significant difference in reproductive cycle occurred at a locality where Lap genotypes do not differ in nitrogen budget. These data indicate that nutritional status can exert significant control over the timing of gametogenesis in M. edulis. Differences in reproductive cycle among Lap genotypes provide a mechanism of non-random mating among genotypes which may result in heterozygote deficiencies at this locus.     

A molecular approach to the ecology of a mussel (Mytilus edulis – M. trossulus) hybrid zone on the east coast of Newfoundland, Canada

Mytilus trossulus Gould and M. edulis L. coexist in mixed populations in Atlantic Canadian waters. Although there is evidence that the two species hybridize in natural populations and that hybrids produce progeny through backcrosses, no study of the microgeographic distribution of the two forms and their hybrids has been made. Here we examine subtidal samples of mussels taken in July 1997 from two locations in eastern Newfoundland (Canada) and from wave-exposed and protected environments within each location. Shell lengths ranged from 15 to about 90 mm. Mussels were classified as pure forms or hybrids (F₁, F₂ and from backcrosses) based on four diagnostic markers, two allozyme loci (Mpi and Est-D) and two nuclear PCR-based DNA markers (ITS and Glu-5). In addition, a PCR-based mtDNA marker (COIII) was used to characterize the distribution of mtDNA mitotypes among pure and hybrid individuals. There were differences in the proportions of pure M. edulis and M. trossulus between sites and between environments at one location. M. trossulus was the predominant species at one of the two exposed sites. In all four samples, M. trossulus was also the predominant form among small individuals. The frequency of hybrids was 26% overall and did not differ among samples. Hybrids consisted mostly of backcrosses that were M. trossulus-biased among small mussels and M. edulis-biased among large ones. We conclude that both intrinsic genetic factors and extrinsic environmental factors influence the relative frequency of M. edulis, M. trossulus and their hybrids. Received: 29 June 1998 / Accepted: 4 November 1998     

Geographic variation,niche breadth and genetic differentiation at different geographic scales in the mussels Mytilus californianus and M. edulis

Genetic differentiation was investigated in the marine mussels Mytilus californianus Conrad and M. edulis Linn. from the west coast of North America. In allopatry with M. californianus, M. edulis occurs throughout the intertidal zone; however, in microgeographic sympatry its ecological range is restricted to above the M. californianus mussel bed and to patches of substratum opened by natural disturbances within the bed. Over the same geographic scale, the broader-niched M. edulis shows greater among-locality genetic difference and greater levels of polymorphism than M. californianus at two enzyme loci. Genetic differentiation on a geographic basis was investigated in M. californianus at a single rock (on a scale of meters), on an island (on a scale of kilometers), throughout a strait (on a scale of 10² kilometers), and along the west coast of North America (on a scale of 10³ km). Differentiation was minimal over the west coast, and could be explained by microhabitat differences in a local area. The minimal differentiation in west coast mussels relative to strong geographic differentiation of M. edulis on the east coast of North America may be related to the steeper latitudinal thermal gradient on the east coast. Local ecologically related microgeographic variation can result in biased and misleading estimates of genetic heterogeneity but microgeographic variation at enzyme loci may not be due to selection at the loci investigated or may even be due to the breeding structure of the mussel populations, as suggested by Tracey et al. (1975).     

Genetic variation of mitochondrial DNA in mussel (Mytilus edulis and M. galloprovincialis) populations from South West England and South Wales

Mitochondrial DNA (mtDNA) and allozyme variation were analysed in samples of mussels collected in 1984 and 1985 from four localities in South West England and one locality in South Wales, a region of Britain where the common mussel (Mytilus edulis) occurs sympatrically and hybridises with the Mediterranean mussel (M. galloprovincialis). Significant differences in mtDNA genotype frequencies for three restriction enzymes (BstEII, XbaI, and EcoRI) were observed between mussels from M. galloprovincialis populations (Padstow and Bude) and those from an M. edulis population (Swansea). Some mtDNA genotypes at high-frequency in M. galloprovincialis were not observed in M. edulis, although there was no indication that mtDNA variation provides greater overall diagnostic power than allozyme variation in distinguishing between the two forms of mussel. Construction of a phylogenetic tree of multiple mtDNA genotypes revealed small mutational distances between the genotypes characterising M. edulis and M. galloprovincialis. The results were consistent with predominant mtDNA flow from M. edulis to M. galloprovincialis. This can be explained by the dispersal of larvae to South West England from M. edulis regions to the north and east, but little dispersal in the opposite directions. Samples from two hybrid populations (Whitsand and Croyde) were analysed. mtDNA genotype frequencies at Croyde were in line with predictions made on the basis of two partially diagnostic allozyme loci (Est-D and Odh), mtDNA frequencies at Whitsand were not. Frequencies of some mtDNA genotypes at Whitsand were characteristic of M. edulis, others of M. galloprovincialis. Differential selective mortality or flow of different mtDNA genotypes and allozyme variation are proposed as possible causes of these results.     

Delayed growth of mussel (Mytilus edulis) and scallop (Pecten maximus) veligers at low temperatures

Veliger larvae of Mytilus edulis (L.) from Menai Straits, North Wales, were maintained for up to 2 mo during 1981 at 5°C and then grown on to metamorphosis at 17°C. Larvae so treated showed similar low mortality and equivalent spat production to control larvae. Growth rate at 17°C was less in treated larvae than in controls, but treated larvae grew a little during the period at low temperature. Larvae of Pecten maximus (L.) from the Irish Sea suffered high mortality at low temperature but larvae surviving 2 wk at 8°C could be grown on to spat at 17°C. The longevity of M. edulis larvae is discussed in relation to the genetic homogeneity of adult populations around the UK.     

Reproductive isolation and reproductive output in two sympatric mussel species (Mytilus edulis, M. trossulus) and their hybrids from Newfoundland

The mussels Mytilus edulis L. and M. trossulus Gould are found sympatrically in most areas of Newfoundland, with a low frequency of hybrids. To assess the potential for reproductive isolation, we sampled mussels from three sites in an eastern Newfoundland Bay from May–October 1996 to determine if there were differences in the reproductive cycles of the two species and their natural hybrids. In mussels with sheil lengths of 38–42 mm, males and females with mature gametes were dominant in June for M. edulis and hybrids, while M. trossulus showed a lower frequency of individuals with mature gametes. M. trossulus and hybrids spawned over a prolonged period (from late spring to early autumn) compared with most M. edulis individuals that spawned over a period of 2–3 weeks in July. This asynchrony in spawning activity between the two species may partially explain the low frequency of hybrids found in previous studies of these mussel populations. Female and male hybrids between M. edulis and M. trossulus showed normal gonad development, ripening and spawning, providing an opportunity for the introgression of genes between the two species. M. trossulus had a higher reproductive output than M. edulis of similar shell length, while hybrids showed intermediate values of reproductive output. M. trossulus females produced smaller eggs than either M. edulis or hybrids. Differences in reproductive traits may partially explain the maintenance of the mussel hybrid zone in Newfoundland. Published online: 13 August 2002     

Implications of movement behavior on mussel dislodgement: exogenous selection in a<Emphasis Type="Italic"> Mytilus</Emphasis> spp. hybrid zone

Rocky intertidal habitats often exhibit high levels of environmental heterogeneity, and the ability of organisms to move between microhabitats is likely to have a profound influence on their rates of mortality and overall fitness. Mussels within the Mytilus edulis complex are morphologically very similar, yet at sites where these species hybridize in southwest England, populations repeatedly show evidence of selection against individuals with alleles specific to M. edulis Linnaeus, in favor of those with alleles specific to M. galloprovincialis Lamarck. Differential movement rates of these two species were examined within simulated mussel beds (gravel substrate) in the winter (February) and summer (July) of 2001. M. edulis-like mussels moved more frequently and more quickly to the exterior of gravel beds than did M. galloprovincialis-like mussels. Coupled with measurements of attachment strength in the field conducted in July 2001, we used a wave force model to examine the probability of dislodgement for each species under a range of water velocities. Results suggest that by preferentially moving to the exterior of beds, M. edulis experiences higher dislodgement rates due to exposure to large hydrodynamic forces than do M. galloprovincialis. As a consequence of lower attachment strengths, M. edulis is also predicted to have higher mortality rates than M. galloprovincialis in interior portions of the bed. Thus, differential movement behavior may contribute to the differential genotype-specific mortality rates observed in the Mytilus spp. hybrid zone in southwest England, and is an example of behavior potentially modifying rates of exogenous selection in an intertidal hybrid zone.Communicated by J.P. Grassle, New Brunswick     

Heterozygosity and growth in transplanted mussels

Growth comparisons were made involving mussels (Mytilus spp.) collected from five different localities in Britain in 1980–1981. Two of the localities, Mumbles, South Wales, and Bude, Southwest England, have pure populations of M. edulis and M. galloprovincialis, respectively. The other three, Whitsand Bay, Southwest England, Croyde Bay, Southwest England and Robin Hood's Bay, Northeast England, have hybrid populations with both M. edulis and M. galloprovincialis ancestry. To make growth comparisons, mussels from different populations were mixed in oyster nets and transplanted to three different localities. After periods of transplantation varying between several months and one year, growth was assessed by measuring increase in shell length or dry body weight. Starch-gel electrophoresis was used to assay variation in the transplanted mussels at three allozyme loci partially diagnostic for M. edulis and M. galloprovincialis. The results provide evidence of small growth differences between populations and between allozyme genotypes within populations. These differences accounted for no more than a few percent of the total variation in growth between mussels. Statistically significant results were obtained, but were frequently found not to be reproducible. There is no clear evidence of a growth difference between M. edulis and M. galloprovincialis. Allozyme heterozygotes appear to have growth rates intermediate between allozyme homozygotes; this study thus fails to provide evidence for overdominance with respect to growth rate.     

Molecular population genetics of male and female mitochondrial genomes in European mussels <Emphasis Type="Italic">Mytilus</Emphasis>

The doubly uniparental system of mitochondrial inheritance (DUI) is best known in marine mussels Mytilus. Under DUI there are two types of mitochondrial DNA (mtDNA). The female type (F) is transmitted to offspring of both genders and the male type (M) exclusively to sons; consequently two distinct mtDNA lineages exist. The M lineage evolves under more relaxed selection than the F lineage resulting in higher polymorphism within the M lineage. Though this polymorphism is expected to make inferences on fine population structure easier using M instead of F data, no comprehensive comparative data exist to support this claim. We sequenced a 1,205 bp fragment of M and F mtDNA comprising parts of the COIII and ND2 genes, and analysed 204 individuals representing three Mytilus species: M. edulis, M. galloprovincialis and M. trossulus from 13 European sampling sites. A clear distinction between Mediterranean and Atlantic populations was found with both M and F data, but much better geographic differentiation was found within the Atlantic using F rather than M data. In particular, Atlantic M. galloprovincialis can be differentiated from Atlantic M. edulis, and further subdivision of Atlantic M. edulis is possible using the F data but not the M data. Multiple tests of selection were carried out to attempt to explain this paradox. We concluded that the overall pattern of polymorphism is consistent with strong purifying selection; not only is this selection relaxed in the M lineage in comparison with the F lineage, but it is also more frequently interrupted by periodic selective sweeps within the M lineage.     

Paternity and condition affect cannibalistic behavior in nest-tending bluegill sunfish

Theory of parental care evolution predicts that a parent should invest more in a brood when its fitness value is greater than alternative investments such as the parent's own survivorship or future broods. In fish, filial cannibalism (eating one's own offspring) is widespread and represents a challenge to parental care evolution. In this study, I investigated filial cannibalism in bluegill sunfish (Lepomis macrochirus). Bluegill are characterized by alternative mating tactics referred to as "parentals" and "cuckolders". Parentals delay maturation, construct nests, court females and provide sole parental care for the developing offspring. Cuckolders mature precociously and parasitize parentals using two tactics called "sneakers" and "satellites". I found that parentals that obtained fewer eggs during spawning appeared more likely to completely cannibalize their brood (total filial cannibalism: P=0.07), regardless of their condition. Among parentals that provided care, partial cannibalism was greater during the egg phase as compared to the fry phase of care, but it was unrelated to brood size. Throughout the care period, parentals in better condition were less likely to partially cannibalize their brood, indicating that parentals use cannibalism to replenish energy reserves. Independent of condition, parentals that were cuckolded more were more likely to eat part of their brood. This relationship was evident only after the eggs had hatched, which is consistent with data showing that parentals can use olfactory cues produced by fry but not eggs to assess their paternity. This latter result proposes that parentals may be selectively culling cuckolder offspring from their nest. These data provide empirical support for parental care theory, and the first evidence for the importance of paternity on cannibalistic behavior.Communicated by M. Abrahams     

Genetic structure of local populations and divergence between growth forms in a clonal invertebrate,the Caribbean octocoral Briareum asbestinum

Although the genetic structure of many populations of marine organisms show little deviation from panmixia, in those marine species with limited larval dispersal, patterns of microgeographic genetic differentiation may be common. The octocoral Briareum asbestinum should show local population differentiation because colonies reproduce asexually by fragmentation, most matings occur between colonies in very close proximity, and the sexually produced larvae and sperm appear to disperse only short distances. Variability in secondary chemistry of individual B. asbestinum colonies from different populations in close proximity also suggests local population differentiation. We determined the genetic composition of local populations by surveying allozyme variation of three shallow and two deep populations within a 300 m² area at San Salvador Island, Bahamas and at a site 161 km away on Little San Salvador, Bahamas in July 1990. As B. asbestinum occurs as either an erect branching form or an encrusting mat often at the same sites, we sampled both morphs to examine the extent of genetic exchange between them. Five of 21 loci were polymorphic and most populations showed a deficit of heterozygotes. Allele frequencies differed significantly between morphs at each site where they occurred together. The mean genetic distance (D=0.065) between morphs is consistent with the interpretation that the two morphs are genetically isolated. Despite the close spatial proximity of the San Salvador populations, both the branching and encrusting morphs showed significant genetic heterogeneity among neighboring populations. Similarly, pooled allelic frequencies for samples collected from the islands of San Salvador and Little San Salvador differed significantly at 1 locus for the branching morph and at 3 out of 5 loci for the encrusting morph.     

Genetic structure of crown-of-thorns starfish (Acanthaster planci) in Australia

Significant genetic structuring is reported for crown-of-thorns starfish Acanthaster planci (L.) populations within Australia, sampled in 1987 and 1989, with Eastern Australian and Western Australian populations forming different groups. Eastern Australian populations sampled included the three most southerly, on isolated reefs of Elizabeth and Middleton Reefs, and Lord Howe Island. The Lord Howe Island population formed a discrete outlier from all other Australian populations, with average genetic distances (Nei's unbiased D) from other populations of 0.05 compared with 0.01 among all other populations. This is greater than the levels of genetic differentiation described for A. planci from sites at the extremes of the Pacific. The genetic distinction of Western Australia and Lord Howe Island is considered to reflect restricted dispersal as a result of current patterns. The Western Australian populations also showed less genetic diversity than the Great Barrier Reef populations, and this might reflect a founder effect or recent bottleneck in the Western Australian populations.Contribution No. 554 from the Australian Institute of Marine Science