The responses of leaf litter invertebrates to environmental gradients along road edges in subtropical island forests.

Abstract

Knowledge of how roads affect forest biodiversity can be improved by measuring the responses of indicator species to complex environmental gradients caused by these infrastructures. We studied litter invertebrate species responses to road edges in laurel and pine forests in Tenerife, Canary Islands. We sampled invertebrates from litter and assessed the environmental variation related to road proximity. We also assessed the effect of relevant environmental predictors on a diverse array of potential indicator species. We applied canonical ordination and non-parametric regression (Lowess) to classify invertebrate species responses to roads and their associated gradients. Three types of responses to road edge proximity were defined for the most common invertebrate taxa: edge-preferring or edge specialists, interior-preferring or edge-avoiders, and edge-indifferent or neutral species. Those species appearing most frequently and with higher population density between 1 and 20 m from the edge (commonly peaking at 10 m from the road) were categorized as edge-preferring. We classified taxa attaining peak population densities at or beyond 60 m from the edge (and most commonly 100 m) as interior species. Edge-neutral species were those without an evident pattern of stabilization in abundance along the gradient and with peaks in abundance at varying distance intervals. These edge litter communities contain a high native and endemic diversity but also a significant density of alien fauna. The specific patterns of penetration of road edge effects on invertebrate species should be seen as having a pervasive and cumulative impact considering the exceptionally large number of roads in these forests and the high population densities of alien invertebrates. Future management plans for forest conservation on the Canary Islands should include the highly altered but valuable litter communities along road edges.