Robust predictive performance of indicator species despite different co-occurrence patterns of birds in natural and managed boreal forests

Abstract Indicator species are widely used biodiversity surrogates that allow the assessment of biodiversity without the expensive and time-consuming construction of species inventories. The selection of indicator species often relies on species co-occurrence patterns, which may be altered by anthropogenic disturbance such as forest harvesting, imposing a unique challenge to their application in managed forest. Here, we studied boreal bird communities in natural forests originating from wildfire and managed forests originating from clearcutting. We aimed to (1) compare species co-occurrence patterns in natural and clearcut forest stands, (2) select indicator species based on species co-occurrence patterns to predict avian diversity, and (3) evaluate the predictive performance of indicator species under both natural and clearcutting disturbance regimes using the same training data set and an independent testing data set. We found that species co-occurrence patterns differ substantially between natural and clearcut stands, suggesting that forest harvesting alters species-environment relationships and/or interspecific interactions. Consequently, we selected different sets of species as indicators of avian diversity based on data from natural or clearcut stands. However, according to internal and external evaluation, selecting indicator species using data from both natural and clearcut stands produced surrogates that predicted avian diversity accurately and precisely in both types of forests. Our results suggest that, despite forest harvesting altering species co-occurrence patterns, a comprehensive understanding of species co-occurrence patterns across natural and managed forests can be used to develop robust biodiversity surrogates. Our study shows that small sets of indicator species can represent the biodiversity of a wide range of species in ecosystems undergoing anthropogenic disturbance, which has important implications for the application of biodiversity surrogates for conservation.