Abstract
The widespread calanoid copepods Eudiaptomus graciloides and E. gracilis co-occur in a number of European lakes and are morphologically similar, but only E. graciloides is capable to produce diapausing eggs. Therefore, these two sister-species are ideal to investigate the genetic impact of a life-cycle with a diapause in a between-species comparison.
In addition to my interest in the population genetic structure of Eudiaptomus spp. in general, I hypothesized that diapausing eggs, as a consequence of their high dispersal capacity, lead to higher gene flow and, hence, to less genetic differentiation between populations. This was tested in samples of 16 E. gracilis and 17 E. graciloides populations at three different spatial scales (investigation of isolation-by-distance in a range of 100, 1000, and >1000 km). Length polymorphism of 10 microsatellite markers from previously developed, enriched DNA libraries was used for the population genetic analyses. Eudiaptomus species generally revealed complex and highly scale dependent patterns of genetic differentiation with geographic distance, i.e. isolation-by-distance. In E. graciloides, a positive correlation between genetic and geographic distances was detected both at the smallest (100 km) and at the largest spatial scale (>1000 km), but was absent at the intermediate range (1000 km). Genetic differentiation in E. gracilis populations never depended on geographic distances neither at a small nor at an intermediate spatial scale. Isolation-by-distance at the largest scale (>1000 km) could not be investigated with the data of this study, because E. gracilis was not found in the most easterly lakes of the sampling area. Above findings suggested that in Eudiaptomus species short- and long-distance dispersal were the result of different mechanisms, with diapausing eggs only being advantageous for short-distance dispersal. The results at a 1000 km range rather suggested either similar moderate gene flow in both Eudiaptomus species over large geographic distances or persistence of historic patterns after postglacial recolonization.
Secondly, I hypothesized that the occurrence of a diapausing egg bank in E. graciliodes results in higher effective population sizes (Ne) compared to E. gracilis, due to the maintenance of genetic diversity. Effective population sizes were estimated from four temporal samples taken from a lake in northern Germany. Allele frequencies were obtained by assessing the polymorphism displayed by altogether 10 microsatellite loci. Nes for both Eudiaptomus species were much smaller than one would expect for a zooplankton species that is passively transported by movements of the water. This might be due to a substructuring of populations, a skewed reproduction success, unequal sex-ratios, non-random mating or fluctuating population sizes. Despite the fact that a diapausing egg bank existed in E. graciloides only, no significant differences between Nes could be found between these species, suggesting that long-term diapausing eggs might not be as influential to the water column population in the lake under study than expected.
In order to obtain information on the genetic role of diapause in Eudiaptomus species over longer time scales, the variation of mitochondrial DNA sequences from populations from three lakes in northern Germany was studied in a joint project with A. J. Bohonak, M. D. Holland, B. Santer, C. M. Kearns, and N. G. Hairston (Jr.). Mitochondrial DNA variation was different in both Eudiaptomus species: Fewer alleles and a larger average divergence occurred in E. graciloides. Significant divergence among lakes in mitochondrial and nuclear markers (microsatellites) demonstrated that gene flow was low in Eudiaptomus species. Estimates of Neµ (µ = mutation rate) backwards in time derived from a Bayesian approach were not significantly different for most time points, but tended to be larger in E. gracilis compared with E. graciloides. Above findings could be explained by phylogeographic analyses and Bayesian skyline plots: these analyses gave evidence for historic range expansions that occured thousands of generations earlier in E. gracilis than in E. graciloides. Additionally, the populations of E. graciloides seem to have passed through a bottleneck. Thus, populations of E. graciloides are likely to have not yet reached equilibrium between gene flow and drift.