We reconstructed the genetic structure of a planktonic crustacean Daphnia longispina living in high mountain lakes and ponds in the Pyrenees to investigate whether it was shaped by persistent founder effects originating shortly after the last glacial maximum or by ongoing dispersal and effective migration (gene flow). We found that the genetic structure can largely be explained by a single colonization event following gradual deglaciation of the Pyrenees ~10 000-15 000 years ago. Nuclear genetic diversity declined steeply from southeast to northwest, suggestive of serial colonization of available habitats with advancing deglaciation. The spatial genetic structure suggests that founder effects were major determinants of the present-day diversity, both at the catchment level and at the level of individual water bodies, further supporting extremely low effective migration rates. This study reveals a prime example of a founder effect that is both long lasting and maintained at small spatial scales. Our data suggest a process of isolation by colonization as a result of strong priority effects and monopolization. We found evidence for the spread of haplotypes with Pyrenean ancestry across the Palaearctic over distances up to 5500 km, although the local genetic structure after colonization was hardly influenced by contemporary dispersal. Finally, our data also suggest that mitochondrial mutation rates in the studied populations were seven times higher than typically assumed. Overall, we show that founder effects can persist for centuries even at small spatial scales at which the potential for dispersal is high.

Cladocerans and copepods are globally important freshwater zooplankton groups, differing in reproductive modes and dispersal abilities. We compared genetic variation of two common taxa of these crustaceans, the Daphnia longispina species complex (known to harbour multiple cryptic lineages) and Eucyclops serrulatus (morphologically and ecologically variable morphospecies), in lakes of ten Eastern European mountain ranges. We expected to discover cryptic lineages in both groups, and to observe different geographical patterns of diversity because of differences in life cycles. Within E. serrulatus, limited sampling through lowland habitats indeed showed the presence of eight highly divergent clades, probably cryptic species, but most of these were not found in the studied mountain lakes. Such a pattern was congruent with the diversity of the D. longispina complex. Regional coexistence of multiple clades within respective species complexes (two in Eucyclops and three in Daphnia) was observed only in the Tatra Mountains (on the Polish-Slovak border). In all other studied mountain ranges (in the Balkans), only single lineages of Daphnia and Eucyclops, respectively, were present, showing similar intraspecific patterns and no evidence for stronger dispersal limitation in Eucyclops than in Daphnia. Our results indicate that substantial cryptic variation may be expected in seemingly widespread copepod taxa. However, detection of cryptic lineages is not a general pattern in mountain lakes, although these habitats harbour substantial genetic diversity in crustacean zooplankton.

BackgroundThe population structure of cyclical parthenogens such as water fleas is strongly influenced by the frequency of alternations between sexual and asexual (parthenogenetic) reproduction, which may differ among populations and species. We studied genetic variation within six populations of two closely related species of water fleas of the genus Daphnia (Crustacea, Cladocera). D. galeata and D. longispina both occur in lakes in the Tatra Mountains (Central Europe), but their populations show distinct life history strategies in that region. In three studied lakes inhabited by D. galeata, daphnids overwinter under the ice as adult females. In contrast, in lakes inhabited by D. longispina, populations apparently disappear from the water column and overwinter as dormant eggs in lake sediments. We investigated to what extent these different strategies lead to differences in the clonal composition of late summer populations.ResultsAnalysis of genetic variation at nine microsatellite loci revealed that clonal richness (expressed as the proportion of different multilocus genotypes, MLGs, in the whole analysed sample) consistently differed between the two studied species. In the three D. longispina populations, very high clonal richness was found (MLG/N ranging from 0.97 to 1.00), whereas in D. galeata it was much lower (0.05 to 0.50). The dominant MLGs in all D. galeata populations were heterozygous at five or more loci, suggesting that such individuals all represented the same clonal lineages rather than insufficiently resolved groups of different clones.ConclusionsThe low clonal diversities and significant deviations from Hardy-Weinberg equilibrium in D. galeata populations were likely a consequence of strong clonal erosion over extended periods of time (several years or even decades) and the limited influence of sexual reproduction. Our data reveal that populations of closely related Daphnia species living in relatively similar habitats (permanent, oligotrophic mountain lakes) within the same region may show strikingly different genetic structures, which most likely depend on their reproductive strategy during unfavourable periods. We assume that similar impacts of life history on population structures are also relevant for other cyclical parthenogen groups. In extreme cases, prolonged clonal erosion may result in the dominance of a single clone within a population, which might limit its microevolutionary potential if selection pressures suddenly change.

Mountain lakes often harbour morphologically or genetically unique populations of zooplankton species, including cladocerans. Daphnia lacustris Sars, predominantly found in Fennoscandia but also known from two Central European lakes in the Tatra Mountains, is one of such taxa. This Daphnia species often forms morphotypes with extremely long tailspines. Historical literature from a century ago documented similar morphs from another lake in the Tatra mountain range, presently inhabited by the phenotypically very different D. galeata. Using a paleogenetic approach (partial sequencing of the mitochondrial gene for 12S rRNA from preserved ephippial eggs in the lake sediment), we tested the hypothesis that Daphnia species composition changed in the lake due to anthropogenic disturbances, and that long-spined morphs were actually another relict population of currently extinct D. lacustris. Ephippia with extremely long spines were successfully retrieved from sediment cores. Despite being morphologically very well preserved, intact eggs were found in less than 2% of analysed ephippia. Genetic analyses, benefiting in most cases from amplification of short 12S fragments using internal primers, proved that long-spined ephippia belonged to D. longispina, which apparently coexisted with D. galeata in the mid-twentieth century. Our results confirm that paleogenetic methods are useful for studying the recent population structures of zooplankton species forming dormant egg banks but lacking reliably identifiable remains in sediments, and show that the extreme development of tailspines in mountain-lake Daphnia is associated with as-yet unclear environmental factors rather than taxonomic status.