Modelling landscape-scale occurrences of common grassland species in a topographically complex mountainous environment

Résumé fourni par la source

Abstract Mountainous regions typically harbour high plant diversity but are also characterised by low sampling intensity. Coarse-scale species distribution models can provide insights into the distribution of poorly sampled species, but the required bioclimatic data are often limited in these landscapes. In comparison, several environmental factors that vary over relatively fine scales in mountain environments (e.g. measures of topography) can be quantified from remotely-sensed data, and can potentially provide direct and indirect measures of biologically-relevant habitat characteristics in mountains. Therefore, in this study, we combine field-sampled floristic data with environmental predictors derived from remotely-sensed data, to model the ecological niches of 19 montane plant species in the Maloti-Drakensberg mountains, South Africa. The resulting models varied considerably in their performance, and species showed generally inconsistent responses to environmental predictors, with altitude and distance to watershed being most frequently included in models. These results highlight the species-specificity of the forb species’ environmental tolerances and requirements, suggesting that environmental change may result in re-shuffling of community composition, instead of intact communities shifting along gradients. Furthermore, while the relatively high importance of altitude (a proxy for temperature) and topographic wetness index (a proxy for soil moisture) suggest that the flora of this region will be sensitive to shifts in temperature and rainfall patterns, several non-climatic environmental variables were also influential. Our findings indicate that local response to climate change in mountains might be especially constrained by soil type and topographic variables, supporting the important influence of non-climatic factors in microclimatic refugia dynamics.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.