Range size patterns in European freshwater trematodes

While patterns in geographic range sizes in free-living species have received much attention, little is known on the corresponding patterns in parasites. For the first time, we report on patterns in geographic range sizes and dimensions of endoparasites, using published species lists of freshwater trematodes in 25 biogeographical regions of Europe. In general, the range sizes of trematodes showed a typical hollow curve frequency distribution, with most species having small ranges. Contrary to expectations, there were no differences in range sizes among trematodes using hosts with high (birds) and limited dispersal capacity (e.g. fish). This suggests that the well known importance of host dispersal capacity for parasite dispersal at small spatial scales is overridden by other factors on larger scales. Regression analyses and Rohde plots showed that the relationship between the latitudinal centre and trematode range size was hump-shaped in all host groups except for reptiles, for which it was linear. Most of the variation fell within the expectations given by null models, suggesting that the patterns mainly result from the geographic properties of the European continent and the biogeographical regions. Finally, trematode ranges tended to stretch more in east-west than in north-south directions, indicating dispersal barrier effects for parasite faunas, probably resulting from the geographical idiosyncrasies of the European continent. There is substantial variation in the size of the geographic ranges of species, spanning up to 12 orders of magnitude (Brown et al. 1996, Gaston 2003). Within any taxonomic group, this variation in the size of geographic ranges can be visualised with a frequency distribution of the species-specific range size, with must taxa showing a unimodal distribution of range sizes with a strong right skew, called the “hollow curve” (Willis 1922, MacDonald 2003). While the hollow curve distribution of geographic range sizes is a universal pattern, the mean size of geographic ranges differs among major phylogenetic lineages. Within the vertebrates, fish have the smallest range sizes, followed by, in increasing order, amphibians, reptiles, mammals and birds (Anderson 1977, 1984a, b, Anderson and Marcus 1992, Brooks et al. 2001). There are also geographic patterns in range size variation. For example, the range sizes of species tend to decrease from high to low latitudes (Rapoport's rule; Stevens 1989, Rhode 1996). Geographic ranges may not only vary in size but also in shape, i.e. ranges do not necessarily have the same extent in all dimensions, —major biogeographical barriers or large-scale climatic zones can restrict the contours of species' ranges (Brown and Maurer 1989). Finally, geographic patterns in range sizes may simply result from the geometric properties of the study area. The so-called mid-domain effect predicts that species richness is highest in the centre of a study area simply because if ranges were distributed at random, the overlap of ranges would be highest in the centre (Colwell and Lees 2000). Whereas patterns in geographic range sizes of free-living species have received much interest, little is known about the range sizes of parasites. Some studies have found a positive correlation between the range sizes of hosts and the numbers of parasites infecting them (Dritschilo et al. 1975, Gregory 1990, Brändle and Brandl 2001, Krasnov et al. 2004), but, to our knowledge, only two studies to date have investigated patterns in geographic range sizes of parasites themselves. The geographic ranges of ectoparasitic fleas (Siphonaptera) from small mammals show the same hollow curve distribution as seen in free-living species (Krasnov et al. 2005, 2008). The geographic range size of flea species was negatively correlated with their degree of host specificity, i.e. host specific flea species had smaller ranges than generalists that infect a wide range of host species (Krasnov et al. 2005). However, host specificity may not be the only determinant of geographic range sizes of parasites. As parasites depend on their hosts for dispersal, the dispersal capacity of their hosts should be a strong determinant of the range size of parasites. In parasites with complex life cycles, e.g. digenean trematodes, definitive hosts should be most relevant in this respect because, their intermediate hosts (first intermediate hosts: molluses; second intermediate hosts: invertebrates, amphibians or fish) have limited dispersal capacities (planktonic larval stages are not infected). Definitive hosts of trematodes are always vertebrates, which carry the adult parasites and then disperse the latter's eggs in their faeces. In general, the dispersal capacity of parasites in bird definitive hosts is considered to be higher than that of parasites of fish and other less vagile definitive hosts (Esch et al. 1988), an assumption supported by population genetics (Criscione and Blouin 2004). Hence, within a regional assemblage of trematodes, we expect parasites using freshwater fish as definitive hosts to have, on average, the smallest range sizes, and parasites utilising birds to have the largest range sizes. In this study, we investigate for the first time patterns in the size and shape of ranges of the entire European freshwater trematode fauna. Besides determining the frequency distributions of range sizes, we also investigate whether the range sizes are determined by the dispersal capacity of the definitive hosts. Based on the distributions of trematodes in 25 biogeographical regions of Europe, we ask the following questions: 1) do the range sizes of trematodes exhibit the typical hollow curve distribution known from free-living species? 2) Do the mean range sizes of trematodes differ depending on the type of definitive host they use? 3) Do the range sizes of trematodes increase with latitude (Rapoport's rule)? 4) Are the shapes of the ranges of trematodes in Europe constrained by major topographic features such as mountain ranges?