A dissertation submitted to the faculty of science, University of the Witwatersrand, in fulfillment of the requirements for the degree of Master of science,
Johannesburg, 2018 / Defining the spatial distributions of species with regards to habitat selection and landscape structure is
an important part of biogeography, ecology and conservation research. I investigated reptile
occurrence and community structure in two patches of dry forest in north western Madagascar using
detection/non-detection data collected on repeated transect surveys for four years. A Bayesian
hierarchical occupancy model and multispectral satellite imagery were used to assess the effects of
vegetation structure, proximity to human development and edge proximity on the site presence of 37
squamate species in the context of taxonomic family and Threat Status. Mean species richness was
highest at sites within a forest patch (23 (4, 30)). Sites with dense green vegetation promoted the
highest levels of reptile occupancy among the Chamaeleonidae and Gekkoniidae families (with
regression coefficient estimates up to 0.75 (0.12, 1.53)) and all species were more likely to occur at
sites closer to the forest patch periphery. The Boidae had the widest 95% CRI for the regression
coefficient estimates representing the effects of habitat variables on occupancy, indicating that they
are highly variable in their habitat use. The regression coefficient estimates of mean reptile occupancy
had 95% credible intervals (CRIs) including zero due to high variability between species. The
proximity to roads did not have a statistically significant effect on any of the species, and Uroplatus
guentheri was the only species which showed a significant preference for being further from a
settlement (0.81 (0.05, 1.77)). The findings demonstrated the importance of using caution when
assessing Threat Status at a regional level and suggest that the Red List’s measure of Population
Trend may be significantly influenced by imperfect detection / MT 2019
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:wits/oai:wiredspace.wits.ac.za:10539/27925 |
| Date | January 2018 |
| Creators | Evans, Aleandra |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Thesis |
| Format | Online resource (106 leaves), application/pdf |
Page generated in 0.0014 seconds