Dillon 1984

Dillon RT. 1984. Geographic distance, environmental difference, and divergence between isolated populations. Systematic Zoology 33: 69-82.

This author examines the relative contributions to population divergence of selection and gene flow (or the lack thereof) using 25 populations of freshwater snails occurring in extremely stable drainages in the south-eastern USA. The system used here has clear advantages for a study that attempts to disentangle these frequently-confounded variables.

Divergence between populations can be correlated by distance in two non-mutually-exclusive ways. A reduction in gene flow that may be associated with longer dispersal distances means that novel mutations take increasingly long times to reach further populations. Environmental differences tend to be spatially autocorrelated such that distant populations are likely to have more different environments and selection will therefore be different. However, if divergence/distance and divergence/environmental difference can be separated, then gene flow and selection can be examined independently.

The drainages of the southern Appalachians appear to have been highly stable since the Cretaceous. The snail Goniobasis proxima appears to be incapable of dispersal overland, though very rare cases of transport by birds or mammals may be responsible for establishing some populations. It is restricted to smaller streams of intermediate flow rates, many of which are distributed on both sides of the southern Appalachians and on the Piedmont (low plateau of small foothills) east of the mountains. Most of the populations examined in this study are completely isolated from each other, such that snails would have to either pass through the marine environment or over the (often very short) land barriers between populations. Development is direct, with egg masses attached to solid substrates producing crawling juveniles. This author notes that at any time, the majority of individuals are crawling upstream against the current, which apparently allows populations to stay approximately in the same place despite the occasional individual that must lose its grip and be swept downstream.

The analysis of population divergence here included comparisons among eight 25 x 25 symmetric matrices, constructed using a comprehensive range of variables including morphological features (shell height, aperture width, etc.), allozyme alleles for seven loci, and 15 environmental variables (11 components of water chemistry, plus temperature, flow rate, stream gradient, and parasite infections by trematodes) and a further independent assessment of environmental similarity derived from an examination of the diatom species diversity in the diets of each population.

Varying levels of genetic divergence were found throughout the study system, but most differences were relatively high compared to similar studies of other organisms. Allozyme alleles fell primarily into two categories: either they were present in all four study drainages, or they occurred in only a single population or small group of neighbouring populations. This suggests that all alleles arose either during a period when rapid spread across drainages was possible, or during a later period when dispersal was more difficult. The geological evidence strongly indicates extreme drainage stability, indicating that something about either the environment and / or the dispersal capabilities of G. proxima was different, perhaps during the Tertiary, than today.

No cline in morphology or allozymes was observed, which may be the result of a lack of gene flow preventing the spread of beneficial alleles. In other words, while nearby populations (overland) may experience very similar environments, adaptations in one population cannot spread to the other.

This author summarizes with a statement that both selection and gene flow restriction seem to be equally important in promoting morphological divergence in isolated populations. However, time since divergence may be the more important diversifying factor, and may underlie both selection and gene flow in this system. Furthermore, measures of divergence using allozymes indicate that time since isolation or gene flow may be more important than selection in structuring differences between populations.