Edge effects constraint endemic but not introduced arthropod species in a pristine forest on Terceira (Azores, Portugal)

Abstract

Pristine Azorean forests have been deeply fragmented since human colonization. Fragmentation increases the length of edges and it, therefore, promotes edge habitats. Studying the impact of edge habitat on species assemblages is crucial to highlight the importance of forest connectivity and guide management strategies. This study explores the impact of forest edges on arthropod assemblages, and particularly investigates the differences of arthropod communities between three habitats, along a distance gradient from the forest edge near a pasture matrix to the core forest. We also compare patterns of arthropod communities with different biogeographic status (endemic, native non-endemic, and introduced species). We sampled in a pristine forest on Terceira island bordered by semi-natural pastures, using flight interception traps. Traps were established along the ecological gradient of three habitats: the forest edge, an intermediate habitat at 100 m from the border, and the core forest at 300 m deep from the border. We took a multi-taxon approach and used both univariate and multivariate techniques to understand the variation in arthropod species abundance, richness and composition between the three habitats in the native forest. Overall, endemic species dominated arthropod abundances whereas species richness and diversity were similar between the three biogeographic categories. We found evidence of a strong edge effect on arthropod assemblages, adjusted both by biogeographic categories and seasonality. Indigenous (endemic and native non-endemic) species abundances were higher in the forest interior than at the edges or intermediate habitats, suggesting that indigenous arthropod assemblages were sensitive to the distance from the edge, whereas introduced species abundances did not show an edge effect. Species diversity and richness did not differ between the three habitats either, regardless of the biogeographic categories. The composition of arthropods between the three habitats differed significantly when we considered all species or endemic species only, but not with native non-endemic or introduced species. However, the difference got obscured when seasonality was included in the analyses, suggesting that even though edges impact species composition, this impact varies seasonally and endemic species are particularly affected in early summer. Our results indicate that forest edges impact arthropod assemblages but endemic species are more likely to be constrained by the increase of edges than introduced species. Since most of these endemic species are of conservation concern, we urge to avoid forest management strategies that increase fragmentation and call for action to increase the size of native forest habitat.