Species complexes within epiphytic diatoms and their relevance for the bioindication of trophic status.

The popularity of aquatic bioassessments has increased in Europe and worldwide, with a considerable number of methods being based on benthic diatoms. Recent evidence from molecular data and mating experiments has shown that some traditional diatom morphospecies represent species complexes, containing several to many cryptic species. This case study is based on epiphytic diatom and environmental data from shallow fishponds, investigating whether the recognition and use of fine taxonomic resolution (cryptic species) can improve assessment of community response to environmental drivers and increase sharpness of classification, compared to coarse taxonomic resolution (genus level and species level with unresolved species complexes). Secondly, trophy bioindication based on a species matrix divided into two compartments (species complexes and remaining species) was evaluated against the expectation that species complexes would be poor trophy indicators, due to their expected wide ecological amplitude. Finally, the response of species complexes and their members (cryptic species) to a trophic gradient (phosphorus) were compared. Multivariate analyses showed similar efficiency of all three taxonomic resolutions in depicting community patterns and their environmental correlates, suggesting that even genus level resolution is sufficient for routine bioassessment of shallow fishponds with a wide trophic range. However, after controlling for coarse taxonomic matrices, fine taxonomic resolution (with resolved cryptic species) still showed sufficient variance related to the environmental variable (habitat groups), and increased the sharpness of classification, number of indicator species for habitat categories, and gave better separation of habitat categories in the ordination space. Regression analysis of trophic bioindication and phosphorus concentration showed a weak relationship for species complexes but a close relationship for the remaining taxa. GLM models also showed that no species complex responded to phosphorus concentration. It follows that the studied species complexes have wide tolerances to, and no apparent optima for, phosphorus concentrations. In contrast, various responses (linear, unimodal, or no response) of cryptic species within species complexes were found to total phosphorus concentration. In some cases, fine taxonomic resolution to species level including cryptic species has the potential to improve data interpretation and extrapolation, supporting recent views of species surrogacy.