Keeping it cool: mammal responses to microclimatic variation in disturbed Indonesian lowland forest.

Abstract

The combined impacts of deforestation, forest fragmentation and climate change will push many mammal species towards extinction, unless appropriate long-term conservation measures are implemented. While most mammal conservation projects focus on protecting remaining primary habitat, it is also important to determine the long- term viability of populations in human-modified landscapes. Effective conservation requires a detailed understanding of species’ responses to fragmentation and climate change, yet physiological and behavioural responses of mammals to forest edge effects and climatic variation remain poorly understood. This thesis combines field data and observations with mechanistic models of microclimate and biophysical models of animal bioenergetics to determine the impacts of forest edges and microclimate variation on mammals at a site of disturbed forest in the Sikundur region of Sumatra, Indonesia. First, I identify mammal-habitat associations and the impacts of edge effects on terrestrial mammals using occurrence data from remote camera traps and fine-scale data of climate and habitat structure measured from the field. Next, I implement and test the performance of two freely available mechanistic models of microclimate, ‘NicheMapR’ and ‘microclimc’, and determine the effects of vegetation structure on their performance. Finally, I utilise an existing biophysical model ‘NicheMapR’ to predict the impact of forest edges on metabolism and water balance of a Critically Endangered mammal, the Sumatran orang-utan, Pongo abelii. Negative impacts of both forest edges and increased temperatures were observed. Mammal occurrence declined with increasing temperatures, and mammal diversity was much lower at the forest edge compared to the interior, while biophysical modelling suggested that orang-utans are already experiencing thermal stress at frequent intervals for short periods of time, with the frequency of these periods being higher at the forest edge. There are notable climate variations in both space and time which are not captured by coarse scale macroclimate data or yearly averages, and my study showed how useful available microclimate models can be to predict localised temperatures using readily available data. The results demonstrate that disturbed and edge forests are still of value to mammal conservation, but that this usefulness varies depending on the target species. This is one of only a few studies to date which has utilised mechanistic modelling approaches in tropical forest and for terrestrial mammals to fully incorporate microclimatic variation and its effects on tropical forest mammals. These approaches are promising tools which can be used to determine the underlying mechanisms of population responses to change, identify important local-scale threats to populations and appropriate mitigation strategies.