Degree Name

Doctor of Philosophy


Department of Biological Sciences


For marine invertebrates which consist of non-contiguous local populations of sessile or sedentary adults, several studies have shown that omission of a planktonic larval stage leads to reduced gene flow and genetic divergence among local populations. However, data on their o w n provide little indication of the extent to which the overall genetic composition of local populations has been affected by genetic drift, site-specific natural selection or a combination of both forces. This study used genetic, ecological and experimental techniques to investigate the roles of gene flow, effective population size, genetic drift and site-specific selection in the evolution of local populations of the direct developing intertidal gastropod Bedeva hanleyi (Muricidae: Prosobranchia) on the east coast of Australia. It was expected that if B. hanleyi had been influenced by genetic drift, then multiple unlinked loci should reveal similar patterns of differentiation among local populations. Likewise, it was expected that if site-specific selection had been important, then the philopatric life-history of B. hanleyi should have allowed a high degree of localised adaptation and/or intra-genomic coadaptation.

Confirmation that direct development results in low gene flow among local populations was provided by a comparison of the genetic structure of Bedeva hanleyi with that of another direct developing gastropod, the buccinid Cominella lineolata, and, most crucially, that of the planktonically developing muricid Morula marginabla. Six allozyme loci were assayed in each species, and in each instance there was no evidence of significant linkage disequilibrium. A s expected, along the same 180 km stretch of the south-east coast of Australia (latitudes 34° 38' - 36° 14' S), both B. hanleyi and C. lineolata exhibited substantially higher levels of differentiation among local populations, and significantly less variability within local populations relative to M. marginabla. These findings imply that, under direct development, local populations evolve more or less independently of one another, whilst under planktonic development there is regular long distance gene flow and mitigation of diversifying forces.

Regardless of larval type, proportions of single-locus genotypes in each collection were consistent with Hardy-Weinberg equilibrium. For Bedeva hanleyi and Cominella lineolata this implies that there has been no close inbreeding despite the fact that the life history of both of these species would appear to promote highly philopatric dispersal. For Morula marginabla, the observation of Hardy-Weinberg equilibrium conditions within local populations is consistent with recruitment from a planktonic gene pool which combines the products of discrete but undifferentiated adult breeding populations.

Two analyses revealed that drift is a major factor underlying the high levels of genetic subdivision in Bedeva hanleyi. Firstly, for several unlinked loci, the level of subdivision in the region north of 33° S, where the major nearshore current (the East Australian Current (EAC)) flows parallel to the shore, was less than that detected south of 33° S, where currents flows in a predominantly offshore direction. Differences in the level of subdivision between regions were interpreted as indicating that, in the northern region, the EAC provides greater opportunity for gene flow via rafted dispersal, and therefore less opportunity for genetic drift. The second finding implicating the action of drift was that, again for several unlinked loci, northern populations exhibited a strong pattern of isolation by distance (sensu Wright 1943), whilst amongst southern populations, there was no correlation whatsoever between divergence and distance of separation. The significance of this finding lies with two facts; (1) isolation by distance occurs as a result of equilibration between genetic drift and stepping-stone gene flow (sensu Kimura & Weiss 1964), and (2), the rate at which gene flow and drift approach equilibrium is inversely related to the strength of gene flow. That the stepping-stone model should apply to B. hanleyi was predicted a priori on the basis of its philopatric life history.

Direct estimates of effective population size (Ne) for two populations that were typical of those sampled for the allozyme survey (Ne Å 400 to 1200) imply that populations are currently experiencing only low to negligible genetic drift. If the conclusion that strong drift is the primary cause of the high level of subdivision seen in Bedeva hanleyi is correct, then this latter finding implies that current Ne must be substantially higher that evolutionary Ne (the estimate of Ne which takes in to account all historical fluctuations in population size). Historical phenomena which may have served to depress evolutionary Ne relative to current Ne in B. hanleyi include population crashes which leave only a few survivors, and/or local extinctions followed by founder events involving only a small number of individuals. Even short periods of small population size have a significant effect on evolutionary Ne because they result in large random fluctuations in allele frequencies, the consequences of which may persist for many generations.

Whilst the loci that were assayed in Bedeva hanleyi appeared to have been primarily influenced by genetic drift, reciprocal transplantation of adults and intra-capsular embryos implied that other loci had been affected by site-specific natural selection. At the conclusion of the reciprocal transplant experiment using adults it was found that in a majority of cases, the recapture rate of local adults was higher than that of foreigners, and, locals exhibited the greatest mean growth in shell length. Reciprocal transplantation of egg capsules revealed that local egg capsules typically exhibited the highest percentage hatching success, and the lowest incidence of egg wastage (eggs that neither developed nor were consumed as nurse eggs) at all transplant locations. Binomial probabilities indicated that it was highly unlikely that these outcomes could eventuated through chance alone.

The transplantation of intra-capsular embryos provide the strongest evidence that fitness differences are the result of localised adaptation. Adults had had direct experience of their native habitat before reciprocal transplantation, and as such it is possible that the superior performance of local adults is attributable to prior acclimation and not localised adaptation. In addition to providing strong evidence of local adaptation in Bedeva hanleyi, this study was the first marine study to simultaneously investigate the sitespecific fitness of two different life history stages of the same species.

The conclusion that Bedeva hanleyi exhibits localised adaptation is also consistent with the earlier conclusion that local populations are currently sufficiently large that random drift has little effect on allele frequencies. Within a population, the immediate effects of strong genetic drift would be to inhibit localised adaptation, firstly by eroding variability upon which selection could act, and secondly by preventing the effects of site-specific selection from accumulating in a deterministic manner.

Since the conditions that permit adaptation to site-specific external conditions also facilitate intra-genomic coadaptation within local populations, it was predicted that the fitness of F1 progeny arising from matings between Bedeva hanleyi from the same population should typically be higher than that of similar progeny resulting from matings between individuals from different populations. If localised coadaptation has occurred in a species, matings between individuals from genetically divergent local populations should typically result in outbreeding depression due to the disruption of coadapted gene complexes. It had been intended that a test of the intragenomic coadaptation hypothesis would be provided by a mating experiment using adults collected from the same local populations as those used in the investigation of localised adaptation. Unfortunately, this experiment failed because many individuals died during the holding period intended to allow females to divest themselves of store sperm.

A number of evolutionary theories predict that the particular combination of population structure and evolutionary history that appears to characterise Bedeva hanleyi (i.e. genetic subdivision combined with temporal changes in the relative importance of drift and selection) is one which should permit the most rapid range-wide adaptive advance.

The two best known examples of this type of theory are the shifting balance theory (Wright 1931) and the founder effect speciation theory (Mayr 1942). Both theories rely on the fact that random drift is the only means by which a population can undergo the transition from a genotype of low fitness to one of higher fitness under circumstances where intermediate genotypes are less fit than either the antecedent or the novel genotype. In the shifting balance theory, universally adaptive gene combinations emerge by chance in a small isolated population, and are then spread range-wide by selection, enhanced local productivity and increased rates of emigration. Firstly in the local population where the genotype emerged, and subsequently in all local populations that received the genotype via immigration. In the founder effect speciation model, a few individuals found a new population and the stochastic transition to a higher adaptive peak is accompanied by reproductive isolation from other populations. With regards to Bedeva hanleyi, this thesis argues that, with or without speciation, the spread of novel genotypes of potentially range-wide adaptive value is likely to be constrained by the necessity for novel immigrant genotypes to compete with large numbers of resident genotypes that are highly locally adapted. However, it is also argued that if B. hanleyi has experienced a moderately high rate of local extinction and recolonistation, adaptive synergy between drift and selection may have allowed many populations to shift to a higher locally adaptive peak than that which could have been attained under isolation, site-specific selection and consistently large population