Unraveling hidden biodiversity: An example of integrative taxonomy applied to the hybridizing Brachionus calyciflorus species complex

Abstract

B. calyciflorus, undoubtedly the most commonly studied freshwater rotifer taxon, was suggested to be a species complex already more than a decade ago yet it has only recently been formally described. This description was based on a comprehensive phylogenetic and morphometric analysis and suggested the existence of four species. Nevertheless, molecular analyses have revealed widespread mitonuclear discordance between each pair of species whereas successful hybridization and backcrossing between species have also been observed in the laboratory. The latter facts suggest that there is a need for additional evidence according to other criteria to support the definition of these species. The aim of this thesis was to provide additional complementary evidence following an integrative taxonomy approach, using data about life history strategy, reproductive isolation and phylogeography. In Chapter 2, we compared life history traits, population growth rates and population demographic structure between the four species. According to these results, B. fernandoi differed strongly from the other species and was characterized by significantly lower investment in sexual reproduction and longer egg development times. Demographic costs associated with long egg development times seemed compensated by reduced costs associated with sexual investment as B. fernandoi was able to realize population growth rates equal to that of B. calyciflorus s.s., and higher than those of B. dorcas and B. elevatus. In Chapter 3 and 4, we investigated some of the mechanisms that may potentially result in reproductive isolation between the two phylogenetically most closely related species, B. calyciflorus s.s. and B. elevatus. We found evidence for pre- and postzygotic barriers that are very likely to hamper cross-fertilization. Prezygotic barriers were revealed by higher conspecific than interspecific fertilization rates. Postzygotic barriers were shown to result from hybrid zygote mortality which manifested itself as a morphological abnormality of the dormant propagule although we found no differences in hatching rates between healthy-looking hybrid and non-hybrid dormant propagules. Despite the reproductive barriers, we were still able to develop viable F1 hybrids. Although F1 hybrid genotypes showed a high performance in clonal growth rate most of them showed a lower tendency to reproduce sexually. Clonal growth rate and sexual reproductive output were found to be strongly determined by parental genotype combination. In Chapter 5, we explored the phylogeographic structure of the four species throughout the Eurasian continent. Phylogenetic analysis based on the mitochondrial marker COI and the nuclear marker ITS1 revealed deep intraspecific phylogeographic subdivisions between and within subcontinents. This result indicates that despite the high potential for long-distance dispersal, the effective gene flow at the continental level is very limited. Furthermore, we detected extensive mitonuclear discordance across the Eurasian continent for all species combinations. We observed large-scale patterns of mitochondrial displacement in three of the four species (i.e. B. calyciflorus s.s., B. elevatus and B. dorcas). Although such displacements could have been caused by a variety of causes, we propose that they most likely originated during postglacial expansions. Persistent patterns of mitonuclear displacement also suggest that hybridization is currently not affecting the genetic structure of Dutch populations any more.