Phylogeography of competing sexual and parthenogenetic forms of a freshwater flatworm: patterns and explanations

Abstract

Background: Models of the maintenance of sex predict that one reproductive strategy, sexual or parthenogenetic, should outcompete the other. Distribution patterns may reflect the outcome of this competition as well as the effect of chance and historical events. We review the distribution data of sexual and parthenogenetic biotypes of the planarian Schmidtea polychroa.

Results: S. polychroa lives in allopatry or sympatry across Europe except for Central and North-Western Europe, where sexual individuals have never been reported. A phylogenetic relationship between 36 populations based on a 385 bp fragment of the mitochondrial cytochrome oxidase I gene revealed that haplotypes were often similar over large geographic distances. In North Italian lakes, however, diversity was extreme, with sequence differences of up to 5% within the same lake in both sexuals and parthenogens. Mixed populations showed "endemic" parthenogenetic lineages that presumably originated from coexisting sexuals, and distantly related ones that probably result from colonization by parthenogens independent from sexuals.

Conclusions: Parthenogens originated repeatedly from sexuals, mainly in Italy, but the same may apply to other Mediterranean regions (Spain, Greece). The degree of divergence between populations suggests that S. polychroa survived the ice ages in separate ice-free areas in Central, Eastern and Southern Europe and re-colonised Europe after the retreat of the major glaciers. Combining these results with those based on nuclear markers, the data suggest that repeated hybridisation between sexuals and parthenogenetic lineages in mixed populations maintains high levels of genetic diversity in parthenogens. This can explain why parthenogens persist in populations that were originally sexual. Exclusive parthenogenesis in central and western populations suggests better colonisation capacity, possibly because of inbreeding costs as well as hybridisation of sexuals with parthenogens.

Figure 1

Overview of the geographic distribution of diploid sexual and polyploid parthenogenetic Schmidtea polychroa in Europe. Numbers indicate sample ID for COI phylogeny (see Table 2).

Figure 2

Bayesian phylogenetic analysis of haplotype sequences hp01-hp31 of S. polychroa. hlE03 (S. lugubris, biotype E) and hlF01 (S. nova, biotype F) were used as outgroups. Numbers adjacent to the nodes indicate the posterior probability for the Bayesian analysis. Locality names are followed from left to right by ploidy number (between brackets), locality name (between square brackets) and haplotype code. Haplotypes with equal number but followed by a different letter (e.g. hp03a, hp03b, hp03c, etc.), represent identical haplotypes (identical COI sequences) found in animals with different ploidy and/or from a different geographical locality. Grey boxes highlight all sexual (diploid) S. polychroa included in this study. Outgroup branch with dashed line has been shortened for aesthetics of the figure.

Figure 3

Minimum spanning network of all identified haplotypes. Haplotype codes, and the number of sexual (S) and partenogens (P) that have a particular haplotype are indicated inside the circles. Small filled circles separating haplotypes represent a single nucleotide substitution difference. Squares represent the ancestral haplotype of every particular network. Dashed lines indicate a possible joining place for the most divergent haplotypes, whose connection to other haplotypes could not be justified by the parsimony criterion.