Abstract
Aim:
To predict the distribution of suitable habitats for Malayan gaur (Bos
gaurus) at a highly fragmented forest area in Peninsular Malaysia and to
identify the potential connectivity between suitable habitat patches.
Methodology: Maximum entropy (MaxEnt) approach was used to
predict the distribution of suitable habitats of the Malayan gaur. Gaur
presence-only data and six environmental variables were collated for the
habitat suitability modeling, and area under curve (AUC) value was used to
estimate the performance of the model. The resulting model was then used to
derive a potential connectivity map through least-cost analysis using
Corridor Designer toolbox in ArcGIS 10.4.
Results:
The AUC value of the habitat suitability model was 0.84. Distance from urban
areas indicated the highest relative contribution to the model (26.9%),
followed by distance from water body (24.2%) land use (18.0%) elevation
(14.3%), slope (14.0%) and lithology (2.6%). Predicted suitable habitats for
gaur were found mostly in lowland forest areas, especially in the vicinity of
rivers within forest reserves. A total of five wildland blocks were derived
from the habitat suitability model, and several potential corridor swaths
were identified connecting the wildland blocks.
Interpretation: The absence of gaur occurrence in
suitable habitats suggest that fragmented habitats greatly affected gaur
distribution and population. Road network and agricultural lands are the
major barriers of gaur movement as they are very sensitive towards
disturbances and conflict. Thus, this research proposes potential
connectivity at a regional scale for Malayan gaur for use in future planning
in conservation, management and development.
Key
words:
Forest fragmentation, Habitat connectivity, Malayan gaur; Maximum entropy
|