Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Pays Bas → 2015 → Modelling the distribution of the cheetah (Acinonyx Jubatus) in Namibia.

International Institute for Geo-Information Science and Earth Observation (ITC) 2015

Modelling the distribution of the cheetah (Acinonyx Jubatus) in Namibia.

Mwendera, N.Y.(Nyasha Yvonne)

Titre : Modelling the distribution of the cheetah (Acinonyx Jubatus) in Namibia.

Auteur : Mwendera, N.Y.(Nyasha Yvonne)

Etablissement de soutenance : University of Twente International Institute for Geo-Information Science and Earth Observation (ITC)

Grade : Master of Science in Geo-Information Science and Earth Observation 2015

Cheetah numbers and overall areas of occupancy have rapidly declined, relegating their conservation status to vulnerable . Factors contributing to the decline of the cheetah occupied range include the depletion of their wild prey base, conflict with humans, attacks by larger predators ; and loss and disintegration of their preferred habitat . Sufficient information on the cheetah distribution and the factors affecting it is needed in order to achieve operative conservation strategies . The aim of this study was to understand the distribution of cheetah in Namibia in terms of biophysical and anthropogenic variables for evidence based species conservation . One objective was to identify the environmental variables important in explaining the cheetah spatial distribution. The environmental variables tested were elevation, slope, rainfall, temperature, vegetation, prey, large carnivores , management and land tenure. The study also to aimed to establish whether the bushland and desert areas of Namibia c an further aid in determining the environmental variables which can be used to predict the presence of cheetah. The third objective of this study was to show the change in time of the cheetah occupied range. Species Distribution Models were used to establish the important environmental variables pertaining to predicting cheetah presence. Forward stepwise Maxent modelling was done to compute S DMs and the highest performing SDMs were taken to be representative. Variables for consideration in the modelling process were chosen based on correlation tests, chi- squared test, VIF analysis and jackknife of predictors. Models were evaluated using the AU C of the ROC plots, True Skills Statistic (TSS) and Kappa statistics. Species Distribution Modelling was done using Maxent software. The change in time of the occupied range was calculated using the kernel density estimations in ArcGIS and isopleth tools in Geospatial Modelling Environment (GME). The environmental variables important for explaining the cheetah spatial distribution were Elevation, Kudu, Land Tenure, Leopard, Lion and Vegetation. An SDM computed from these environmental variables performed significantly well and proved to be robust (AUC = 0.821). Elevations above 1500m were determined to be associated with a high probability of presence of cheetah. Cheetah presence probability increased with an increase in the number of Kudu per head per square kilometre. Cheetah presence increased with an increase in the number of larger carnivores but reduced as the numbers became significantly high. The Cheetah presence was found to be high associated with land tenure. Vegetation also has an impact on the cheetah presence. SDMs modelled in the desert areas proved to perform better than those modelled over the whole study area or the bush land alone. Results showed that the occupied range of the cheetah had decreased by approximately 52% over the years from 1 982 to 2014. The findings of this study contribute to the baseline knowledge needed for effective cheetah conservation. The results can be used to establish which areas are useful for further conservation efforts, relocations and establish whether some conservation efforts already in place have a significant positive effect on cheetah range .

Version intégrale (ITC )

Page publiée le 5 avril 2018