Limnodrilus hoffmeisteri Claparède, 1862 is a common freshwater worm, often regarded as an indicator of organic pollution. The taxonomic status of this species is controversial due to great variation in morphological features. Numerous morphological forms of L. hoffmeisteri are recorded in the literature, especially from Europe and North America. Today, DNA-based species delimitation assumes that species boundaries can be more objectively and effectively estimated using genetic data rather than with morphological data alone. To investigate if L. hoffmeisteri is a single species, 295 worms identified as either L. hoffmeisteri or other similar (congeneric) morphospecies, using currently accepted morphological criteria, were collected from 82 locations in the northern hemisphere. The number of primary species hypotheses (PSHs) was first explored with cytochrome oxidase subunit I (COI), the proposed DNA barcode for animal species, and with data for all specimens. Both automatic barcoding gap discovery (ABGD) and the Bayesian general mixed Yule coalescent (bGMYC) model revealed the existence of ≥25 distinct PSHs (COI lineages) in our dataset. Then, smaller samples of individuals representing major COI lineages were used for exploration of a nuclear locus, the internal transcribed spacer (ITS) region. In the ITS gene tree (81 sequences), generated by BEAST, 16 well-supported terminal groups were found, but not all of these groups were congruent with the PSHs found in the COI tree. As results across these different analyses were inconsistent, we resorted to analyzing reciprocal monophyly between gene trees and used a minimum consensus of all evidence, suggesting that there are 13 separately evolving lineages (=13 species) within our sample. The smallest uncorrected COI p-distance between these species is 12.1%, and the largest intraspecific p-distance is 16.4%, illustrating the problem of species delimitation with a DNA-barcoding gap as the sole criterion. Ten of these species are morphologically identified as “L. hoffmeisteri,” the remaining three can be attributed to morphologically distinct congeneric species. An individual from the type locality in Switzerland was designated as a neotype of L. hoffmeisteri sensu stricto. This worm belongs to one of the ten species, and this lineage is widely distributed in Europe, Asia, and North America. The remaining nine species show a mixed distribution pattern; some appear to be endemic to a restricted area, others are Holarctic. Our results provide clues to the future revalidation of some of the nominal species today placed in synonymy with L. hoffmeisteri. A BEAST analysis, based on previously published and newly generated 16S data, suggested that this complex contains also other species than those studied by us. By integrating additional genetic data, it will be possible to identify these and additional specimens in future studies of Limnodrilus, and the neotype provides a baseline for further revisions of the taxonomy of the L. hoffmeisteri complex.
Read full abstract