The intermediate disturbance hypothesis explains arthropod beta-diversity responses to roads that cut through natural forests

Abstract

Forests are sensitive ecosystems with a distinctive micro-climate to which forest arthropods have adapted. In intact primary forests, sunlight only fully reaches forest floors temporarily through tree fall, or more permanently at forest edges. Roads that cut through pristine forests are however permanent features that fragment forests and increase forest edge. In accordance with the intermediate disturbance hypothesis (IDH), these edges may allow generalist species to enter forests and replace some specialists, often leading to little change, or even higher numbers of arthropod species at edges. Here, we determined how roads and hiking trails affect epigaeic forest arthropods in the largest Afrotemperate forest complex in South Africa. Arthropods were collected using pitfall traps along transects set up perpendicular to different types of road, including wide arterial roads, narrower secondary roads and hiking trails. As expected, hiking trails affected arthropod assemblages to a much smaller extent than the roads. Edge effects, as measured by arthropod alpha diversity, were evident up to only 5 m for hiking trails yet as much as 20 m for secondary roads and >50 m into the adjoining forest for arterial roads. Our results support the intermediate disturbance hypothesis, as changes in beta-diversity, especially next to arterial roads, was driven by the replacement of species in conjunction with changes in species richness. The magnitude of edge effect was however different for the selected feeding guilds. We recommend that where possible roads and even many hiking trails are removed from many areas of the deep forests. These ancient Afrotemperate forests are fragmented by roads and to a lesser degree by hiking trails, which could have cascading effects on overall forest integrity and long-term impacts on these forests.