Fig 1.
Geographic position of the studied sulfidic caves (F = Frasassi cave system, M = Movile Cave).
Shaded dots indicate the distributional range of Niphargus in Europe, after [61].
Fig 2.
Phylogenetic relationships among 44 species of Niphargus obtained by Bayesian inference.
The two studied species pairs from sulfidic caves are indicated in bold. Numbers on the nodes indicate Bayesian posterior probabilities; only nodes with support > 0.7 are indicated. Distinct yet undescribed species from the ‘longicaudatus’ group are labeled geographically. The tree was rooted using topological information from a genus-wide molecular phylogeny [28].
Table 1.
Difference in body size measured in Euclidean distances for pairs of co-occurring Niphargus species and their inferred ancestors.
Table 2.
Difference in ventral channel measured in Euclidean distances for pairs of co-occurring Niphargus species and their inferred ancestors.
Fig 3.
Schematic summary of the results.
The y-axis shows the normalized ecological difference between coexisting species as inferred from Euclidean distances, whereas the dashed lines indicate the theoretical minimal and maximal Euclidean distances. Normalized Euclidean distances were calculated as (actual—theoretical minimal Euclidean distance) / (theoretical maximum—theoretical minimal Euclidean distance). Theoretical Euclidean distances were calculated based on regional species pools and can be exceeded in ancestral species pairs. The x axis represents ancestral (open symbols) and extant (solid symbols) species pair. Arrows are oriented from ancestral to extant species pairs for each studied trait. Frasassi species are labeled with diamonds, Movile species with circles. Differences between ancestral species pairs were calculated using the mean, upper and lower estimates for a given trait. Note that the direction of evolutionary change in all but one case indicates competition-driven divergent evolution.