![Sample our Environment & Agriculture journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5934/TOC-Subject-Sample-2021-Environment-Agriculture.jpg"})
Abstract
Habitat suitability and species distribution models have both become essential tools in biodiversity conservation and management. However, very few of these studies exist from Arctic habitats and hardly any on Arctic species diversity modelling. The basic goal of this study was to develop a statistical model based on vascular plant species’ spatial distribution data on the Svalbard archipelago and their dependence on a set of available environmental variables. The obtained model was then implemented into GIS, enabling us to calculate plant diversity indices for the Svalbard archipelago. Svalbard is easily accessible for research and contains well-known flora with plentiful ancillary data layers available. This location thus constitutes a suitable study area for analysing and modelling biodiversity. Georeferenced data on vascular plant species diversity were gathered from 184 study sites widely distributed on the archipelago. Thirteen environmental raster layers were generated based on a digital elevation model, a geological map, as well as climatic and remote sensing data. Environmental data were extracted from the raster layers at each of the 184 field study plots. Both field study plots and raster layers were studied at 1 km2 resolution. Analysis using forward stepwise multiple regression revealed that growth season temperature sum (GDD), mean July precipitation (PREC) and the vegetation indices ‘normalised deviation vegetation index’ (NDVI) are the best predictors of Svalbard's vascular plant biodiversity. Despite a 48% precision of the statistical model in predicting Shannon diversity index (SDI), the output map seems to reflect well the expected distribution based on knowledge of the influence of the environmental variables considered. All variables in the model, and most other data tested in the model, are easily available and with global coverage.