The use of coarser taxonomic resolution in studies of predation on marine sedimentary fauna

Abstract

Given the logistical difficulties, cost, and time involved in species-level identifications, several authors have proposed the use of coarser taxonomic resolution (e.g. family, order) in studies of pollution. The use of surrogates instead of species relies on their sufficiency to detect community responses to the pollution gradient without appreciable loss of information. No studies, however, have applied this approach to experimental studies such as community responses to predation disturbance and evaluated the performance of surrogates at the spatial scales typical of experiments. We addressed both problems by analyzing the results of three predation experiments carried out in Bonne Bay, Newfoundland. We pooled species data into coarser taxonomic categories (family to class) and determined whether effects of predation that were evident at the species level were also evident with the use of each coarser surrogate and increasing data transformation. Our results indicate that non-transformed data at the family level represent a reasonable surrogate of species; however, the ability to discriminate between ambient and (predator) manipulated sediments is gradually lost with data transformation and with the pooling of species into coarser taxonomic categories. Successive data transformation indicates that in this system predation plays a strong role on dominant but not necessarily rare species. Moreover, our results suggest that varying reliability of surrogates precludes the identification of a single general level of taxonomic sufficiency to be used in experimental studies. The use of surrogates, therefore, is suggested only after scrutiny and evaluation, and should be limited to preliminary studies where biodiversity has been well described.