No abstract

Mussels were collected at high and low shore locations from two Mytilus edulis/Mytilus galloprovincialis populations, Croyde Bay and Whitsand Bay, in S.W. England. Genotype-dependent length-at-age values were determined. At high and low shore locations at both sites, M. edulis-like mussels had significantly smaller length-at-age values than M. galloprovincialis-like and putative F1 hybrid individuals. The putative F1 hybrids exhibited length-at-age values between those of the parental types, but much closer to those of M. galloprovincialis-like rather than M. edulis-like individuals. Genotype frequencies as a function of age were determined and relative viability coefficients estimated from comparisons of genotype frequencies of young versus old mussels. At high and low shore locations at both sites, the relative viability coefficient of M. galloprovincialis-like individuals was greater than that of M. edulis-like mussels. Putative F1 hybrids at both sites had relative viability coefficients intermediate between those of the parental types. These data indicate that the length-dependent variation in allozyme frequencies that characterizes sympatric populations can be attributed to a small but significant genotype-dependent difference in length-at-age values, but mostly to large and highly significant differences in viability.

Biological (predation) and physical (wave actlon) factors influencing mortality of MyUus edulis and M , gallopromcialis mussel types within hybrid and pure populations were investigated with the aim of identifying possible selective factors responsible for observed length-and age-dependent genotypic variation In laboratory tests, shore crabs Carcinus maenas did not significantly prefer any type when choosing between mussels from a hybrid population. When presented with edulis and galloprovinciaLis from allopatric populations the crabs expressed no significant preference for either mussel type. Neither crab sex nor crab size significantly affected mussel type chosen. Dogwhelks also did not express a significant preference for mussel type when feeding upon mussels of a hybrid population, but did exhibit significant preference for edulis over galloprovincialis from pure populat i o n~. Mussel strength of attachment (SOA) to the substrate was tested in 2 hybrid mussel populations. Croyde and Whitsand in SW England. Shell length and genotype, but not population, explained significant variation in SOA. At all lengths, edulis-like mussels had a significantly lower SOA than galloprovincialis-like mussels. Overall, these results suggest that neither predator is likely to be responsible for the higher frequency of galloprovincialis in larger and older mussels of mixed population~. However, the SOA data indicate that this physical factor, which is related to site-specific wave action, is correlated with genotype-dependent mortality. It is concluded that the strong negative correlation between eduhs frequency and length (and also age) in many hybrid populations results partly from a slight galloprovincialis growth advantage (measured by in situ growth experiments), but mainly from selective mortality of edulis individuals.

A population of Mytilus edulis is described whlch is numerically dominated by young ( < 2 yr old) individuals. Growth within this population is indeterminate, i.e. no maximum shell length is attained throughout the realized life span. Predation pressure, principally by Carcinus maenas and Asterias vulgaris, results in a population mortality rate of 42 % in a 6 mo period and also probably prevents growth from being determinate. The partitioning of organic matter between shell and body growth is discussed in connection with the cycling of energy within the ecosystem of the rocky intertidal zone.

A wild and a cultured greenshell mussel (Perna canaliculus) population were compared for biochemical genetic variation at seven polymorphic and four monomorphic allozyme loci. Significant heterozygote deficiencies were observed for all polymorphic loci Lap,2, Pgi, and Pgm) for both populations (except the Pgi locus of the wild mussel population). Genotypic disequilibrium was calculated for both populations: genotypic frequencies were significantly nonrandom at three pairs of loci among the wild mussels, and significantly non-random at three different pairs of loci among the cultured mussels. All six pairs of loci which exhibited significant genotypic disequilibrium involved aminopeptidases, suggesting that these loci form a linkage group, and that neither the Pgi nor the Pgm loci are associated with this group. Exact tests for population differentiation based upon populationspecific allele distributions indicated that four of the polymorphic loci were significantly heterogeneous among the two populations, whereas the remaining three polymorphic loci were not. Based upon the private allele system, the number of migrants (N m ) between the populations was estimated to be 2.009, which, according to the private allele system, represents a high level of gene flow. These findings are discussed with regard to the population biology and genetic structure of this species. M96028