Search

Context: Habitat fragmentation per se (habitat subdivision independent of habitat loss) is a major driver of biodiversity change, potentially due to its impacts on climate. Habitat fragmentation may make landscapes hotter by increasing the amount of habitat edges, but can reduce landscape-level temperatures due to the "vegetation breeze" phenomenon. The plausibility of these two alternative hypotheses is unclear, as no study analyzed the effects of habitat fragmentation per se on temperature. Objectives: We quantify, for the first time, the impacts of habitat fragmentation on landscape-level temperature across the globe. Methods: We analyzed satellite data on forest cover and three climatic variables: mean daily temperature, albedo and evapotranspiration. The analyses were performed separately for tropical, temperate, and boreal regions. We compared the climatic variables between pairs of landscapes with similar amount of forest, but different levels of forest fragmentation (number of patches). Results: Habitat fragmentation reduced landscape-level temperature in all climatic regions. The magnitude of this cooling was stronger in the tropics and weaker in the boreal region due to different evapotranspiration rates. This landscape-scale cooling contradicts local-scale studies, which have indicated that edge effects rise local temperatures. However, habitat fragmentation may intensify vegetation breeze, resulting in final cooling at the landscape scale. Conclusions: Habitat fragmentation leads to colder landscapes. We propose a new conceptual model to unify local (edge-induced) and landscape-level effects of habitat fragmentation on temperature, advancing the understanding of the consequences of habitat fragmentation on climate globally. [ABSTRACT FROM AUTHOR]

Abstract: Tropical montane organisms are vulnerable to climate change because of elevational specialisation, but little is known of the variability in elevational specialisation across tropical insects. We assessed elevational specialisation across several insect taxa comprising four trophic groups 80–2263 m up an elevational transect in Brazilian Atlantic Rainforest, using community‐based and species‐based approaches. We sampled 697 species, of which 32% were found only in the top and 45% only in the bottom half of the transect. Considering only the 182 species with at least five individuals recorded, the percentage of species found exclusively in the top or bottom half drops to 16% and 24%. Across four vegetation belts (lower montane forest, montane forest, upper montane forest and Campos de Altitude) the Eumolpinae (herbivores) were more specialised than Scarabaeinae (saprotrophs), or Lampyridae (predators). This result was robust to the treatment of rare species, and the difference was most marked at higher elevations. Lampyridae lacked upper montane forest specialists. Using all species sampled, specialisation to the upper or lower half of the transect was greatest among Chrysomelidae, Curculionidae (both herbivores) and Lampyridae, and lowest among Tiphiidae (parasitoids) and Coccinellidae (predators). Considering only better sampled species, however, Lampyridae were the most specialised and Braconidae the least specialised. Trophic groups also varied consistent with these findings. Our findings suggest high elevational specialisation and concomitant extinction risk in Brazilian Atlantic Rainforest insects. Differences in elevational specialisation between taxonomic groups may alter the functioning of insect communities under climate change. [ABSTRACT FROM AUTHOR]