Biodiversity thresholds in invertebrate communities: The responses of dung beetle subgroups to forest loss

DOI
10.1371/journal.pone.0201368
Publication Year
2018
Publication Site
PLoS ONE
Journal Volume
13
Family
Scarabaeidae
General topic
Ecology
Biodiversity/Biogeography
Specific topic
community structure
habitat disturbance
Author

Pinto Leite, Clarissa Machado; Mariano-Neto, Eduardo; Bernardo da Rocha, Pedro Luís

Abstract Note

Extinction thresholds have been predicted to be critical values of habitat loss in which an abrupt reduction in populations occurs through the interaction between reduced habitat and increased isolation in the landscape. In communities, extinction thresholds are referred to as ‘biodiversity thresholds'. The biodiversity threshold values documented so far occur between 30% and 50% of habitat cover in landscapes. However, the assessment of biodiversity thresholds has mainly focused on vertebrate and plant communities. Here, we evaluated the occurrence of biodiversity thresholds in dung beetle communities by sampling ten 3,600 ha Atlantic Forest landscapes with forest cover ranging from 5% to 55%. We analysed the response patterns (abundance, gamma and mean alpha diversity) of community subgroups with different levels of forest dependency (forest species, generalist species, and open-area species) using model selection, comparing null, linear, bell-shaped and logistic models. The response of the community of forest species equally fits both linear and logistic models predicting a biodiversity threshold at 25% forest cover. Generalist species showed peak abundance at 20% forest cover although this result reflects a very poor generalist assembly. Open-area specialists did not respond to the amount of forest. The two most plausible models for forest species suggest two different biodiversity management options. Since the biodiversity threshold model represents a more dramatic scenario for the loss of biodiversity in Atlantic forest landscapes, we suggest, based on precautionary principle, that our results should strength guidelines that consider minimum values of forest cover in management strategies to avoid abrupt biodiversity loss and impacts on ecosystem services.