Suitability of presence vs. absence indicator species to characterize stress gradients: Lessons from zooplankton species of boreal lakes

Published in Ecological Indicators, 2013

Recommended citation: Anas, M.U.M., K.A. Scott and B. Wissel (2013). "Suitability of presence vs. absence indicator species to characterize stress gradients: Lessons from zooplankton species of boreal lakes." Ecological Indicators. 30:90-99. https://doi.org/10.1016/j.ecolind.2013.01.038

Indicator species are species that readily reflect some measure of habitat characteristics and have become an increasingly appealing tool in environmental monitoring. Traditionally, habitat conditions were derived based on the presence of specific indicator species, but the absence of indicator species may be just as informative. To evaluate the importance of presence vs. absence of indicator species for characterizing habitats across environmental gradients, we evaluated the interactions of zooplankton and acid-stress in 244 boreal lakes. We adopted the statistical methods proposed by Dufrene and Legendre (1997) to identify presence and absence indicator species to characterize high, intermediate and low acid-stress lake categories. Presence indicator species (identified by the statistical analysis) for highly stressed lakes were not entirely appropriate because further evaluation identified them as ubiquitous generalists. In contrast, absence indicators for highly stressed lakes were more appropriate as these habitat specialists were specifically absent from this category of lakes. On the other hand, presence indicators for the low acid-stress category were largely habitat specialists and therefore appropriate indicators. However there were no presence or absence indicators for lakes at intermediate acid-stress level. Thus the combined use of both presence and absence indicators is recommended to characterize habitats across a stress gradient. To evaluate if the successful application of this combined approach is dependent on a stress gradient, we applied the same analyses to a sub-set of uninfluenced (non-sensitive) lakes representing three different environmental conditions. This approach identified statistically significant presence and absence indicator species for all three different reference conditions. Yet, the absence-based approach was not essential under unstressed conditions, as presence indicator species were habitat specialists for all reference categories. Finally, this study also emphasizes the importance of meaningful ecological traits of species in order to ensure the appropriateness of statistically selected indicator species.