This thesis questions whether sympatric divergence, brought about by disruptive selection in a spatially heterogeneous environment, can occur under natural conditions and in the laboratory.
I. An investigation, to detect micro-differentiation, was made on a Plantago lanceolata population, comprising a gradation of phenotypes and occurring in a small area (lx 5m) of dune pasture, heterogeneous by virtue of different vegetation heights. The prostrate phenotype with shorter leaves and inflorescences was correlated with low vegetation} the erect phenotype was correlated with taller vegetation. Vegetative propagules in standard conditions of garden and greenhouse showed persistent genotype differences of growth habit, leaf length and inflorescence length. Growth habit and leaf length also correlated with the original environment, indicating adaptive micro-differentiation. Some phenotypic plasticity was apparent. Investigation of the field population revealed flowering time differences between the phenotypes and rapid turnover of individuals less than six months old, particularly in low vegetation where the Plantago population was the most dense.
Selection pressures appeared to be operating to maintain differentiation within this heterogeneous environment. Population control was evident, with mortality matching recruitment, but the chances of survival of an individual were independent of the season of establishment.
II. In a second series of experiments, a population of randomly- mating Drosophila melanogaster was maintained for 20 generations in small ‘population cages’, heterogeneous because they contained two types of food medium, viz. normal food and normal food plus peppermint essence. The founder population yielded 40£ more progeny on the normal food. There were three control populations feeding on (l) normal food only, (2) peppermint food only, (3) homogeneous half-strength peppermint only. The experimental population initially responded to the heterogeneous environment (with its choice of food media), by yielding numbers of progeny and biomass in excess of expectation, which was calculated from the controls. This was thought to be an environmental response. The difference between the observed yield and expectation increased steadily for 10 generations, implying adaptation to the heterogeneity, but, after 17 generations of selection, the yield was significantly less than expectation.
This persisted for a generation of lapsed selection on normal food, indicating a genetic response to some factor within the heterogeneous environment. Because females reared on peppermint showed a behavioural change and tended to choose this less palatable medium on which to lay their eggs, it is suggested that a genetic component of behavioural flexibility contributed to this result. There was also evidence of improved adaptation to normal food, possibly a genetic response to highly competitive conditions on this densely- populated medium. Although sympatric divergence was not conclusively demonstrated, a measure of habitat selection for egg-laying sites developed and the population became non-random. Peppermint retards the life-cycle of the flies living on it by approximately one day. Therefore, the heterogeneous population was experiencing conditions which might promote assortative mating.
It was concluded from the two experiments, that a heterogeneous environment may act disruptively on a small, randomly-breeding population within a small area. The Plantago population, in an environment where selection pressures were probably high, showed evidence of micro-differentiation, indicating that sympatric divergence had occurred, although phenotypic plasticity was also evident in some morphological characters. The Drosophila population, in a heterogeneous environment where selection pressures may have been relatively low, also became non-random and evolved habitat-choice. In both investigations, forces enhancing assortative mating, helping to maintain genetic variation by reducing gene flow, were apparent. Therefore, it is concluded that sympatric divergence may be brought about by disruptive selection in a heterogeneous environment.