Ecological assessment of highly heterogeneous systems: The importance of taxonomic sufficiency

We assess the effects of taxonomic resolution (genus-family levels) on the ecological assessment of 39 highly heterogeneous ponds located in north-western Spain.Non-metric multidimensional scaling (MDS) and one-way analysis of similarities (ANOSIM) were used to investigate the effects of taxonomic resolution on the macroinvertebrate assemblage structure. The Mann Whitney U-test and discrimination efficiency were used to assess the ability of nine diversity measures (total richness, rarefied richness samples of 25, 50 and 100 individuals, Margalef's index, Pielou's evenness, Shannon-Weaver's index, Simpson's index and percent dominant taxon) to discriminate between acceptable (best available and good conditions) and unacceptable (moderate, poor and very poor) conditions using three levels of taxonomic resolution: (i) family, (ii) family plus subfamilies of benthic non-biting midges and (iii) genus level.Based on non-metric MDS, the macroinvertebrate assemblages of ponds of acceptable (A) and unacceptable (N) conditions were statistically undistinguishable, both at genus and family levels. On the other hand, based on several community metrics (total richness, Margalef index, etc.) the two sets of samples were statistically different, although only when the genus or the subfamily level was used and after Bonferroni correction. These results suggest that the structure of macroinvertebrate assemblages by itself is more sensitive than the specific composition in distinguishing the fauna living in acceptable and unacceptable conditions. Moreover, dealing with families including many taxa generally showing different tolerance to disturbance may lead to misclassification of ponds. We agree, however, that the two approaches, i.e. assemblage composition and diversity measures, are conceptually different and hence they should be used in combination for a better understanding of the response of single metrics.