Nature Conservation Research: Заповедная наука (Aug 2020)
Brown bear (Ursus arctos) ecological niche and habitat suitability modeling in the southern taiga subzone using the method of GNESFA
Abstract
The article presents the results of ecological niche and habitat suitability modelling of Ursus arctos, performed in the Central Forest State Nature Reserve and its buffer zone (Tver region, Russia), using the GNESFA factor analysis method. Briefly discussed are the basic rules for performing such a study, approaches and modelling techniques. The ecological niche was modelled using FANTER, ENFA and MADIFA. The habitat suitability was modelled using Mahalanobis distances analysis. As ecogeographical variables, vegetation indices, morphometric characteristics of the relief, proximity rasters and types of landscape cover were used. The presence points of the species were recorded on permanent routes using a GPS navigator during 2008–2018. As a final result, six models with a different combination of input data (presence points and environmental variables) were selected. As a set of independent test data, we used U. arctos' presence points recorded by the forest rangers and research staff. The scenarios of the influence of utilisation weights on presence points and multicollinearity on variables are separately modelled. The global marginality of the best model was 1.17, while the tolerance was 0.69. The distance to anthropogenic food sources, the NDVI and the projective cover of grasslands made the highest contribution along the marginality axis. In the first three axes of specialisation, coniferous forests, Sphagnum pine (Pinus sylvestris) forests, and the distance to the rivers made the largest contribution. AUC values ranged from 0.66 to 0.73. Values of the continuous Boyce index were from 0.85 to 0.93. The area of the partially violated buffer zone turned out to be more suitable for U. arctos than the inner area of the Central Forest State Nature Reserve. Utilisation weights and the number of ecogeographical variables had a significant impact on the results of all analyses.
Keywords
