The Eastern Cape contains the majority of the remaining forested areas in South Africa (95.8 percent Coastal forests and 47.6 percent Afromontane forests), however these occur in fragments. Due to the rapid rise in the human population and its needs, conversion of forests to agricultural land and the extraction of timber increases, natural vegetation are lost and this therefore leads to forests becoming fragmented into small forest patches. One of the main consequences of forest fragmentation is loss of contiguous habitat, which is the dominant threat to species globally as it negatively affects both species richness and genetic diversity. This research investigates the effects that forest fragmentation has on small mammal diversity and abundance. The study sampled small mammals in Southern Mistbelt Afromontane forest patches in the Eastern Cape. The first aim of this study was to identify which environmental variable, or combinations of variables, affect the diversity and abundance of small mammals in fragmented forest patches in the Eastern Cape during the austral summer. The second aim was to compare seasonal changes in small mammal diversity and abundance at a forest edge compared to the forest core at Beggars Bush, an Afromontane forest in the Eastern Cape, South Africa. To identify the environmental impacts that different forests have on small mammals, nine different forests were chosen in the Eastern Cape and sampled during the summer in 2013. These nine forests were: Fort Fordyce Forest, Thaba Ndoda Forest, Hogsback Forest, Dassie Kraans Forest, Langeni Forest, Burchell’s Reserve Forest, Maiden Dam Forest and Kagaberg Forest. At each site, three transects of 30 Sherman Traps each were used. Trapping periods within each forest patch lasted between three to five consecutive nights and traps were inspected once a day in the morning. Nine environmental variables were initially identified and after being tested for normality and colinearity, five variables were chosen. These variables were Altitude, Mean Annual Temperature, Gradient, Patch Size and Mean annual potential evaporation. The number of individuals captures and species richness were then incorporated as dependent variables for best subset multiple regression model selection using the Akaike Information Criterion (AIC). For the second aim, the same trapping methods were used when comparing seasonal changes to diversity and abundance within the forest core. However at the forest edge, due to the small size of the forest, only 20 Sherman traps were used in each transect with five traps (25 percent) placed in the grassland and 15 (75 percent) placed in the forest along each transect. This study was one of few to survey small mammal diversity and abundance in fragmented forests in the Eastern Cape. Gradient, patch size and mean annual potential evaporation were the variables that best predicted the individual number of small mammal captures while only gradient best explained species richness. It does however, need to be noted that capture rates and species richness were very low and this would therefore effect the analysis of environmental variables. Future studies should have a larger sample size of forest patches and include more microhabitat environmental variables to determine their effects on small mammal diversity and abundance. However, it should be noted that through climatic extinction filtering, forest mammals are resilient generalists that can tolerate fragmentation effects. Furthermore, it was found that forest edges appear to play a significant role in small mammal diversity and abundance in the Beggars Bush Afromontane forest. Some species were habitat specialists such as Rhabdomys pumilio preferring the grassland habitat and Graphiurus murinus and Aethomys namaquensis the forest habitat, while Myosorex varius was the only species that was not habitat dependant. It was found that there was a greater diversity and abundance at the forest edge compared to within the forest core throughout most of the seasons. One possible flaw was that the sampling methods were different at the edge and within the forest and therefore future studies should ensure that the method stays uniform throughout the study.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:rhodes/vital:4270 |
| Date | January 2015 |
| Creators | Junkuhn, Kyle Peter |
| Publisher | Rhodes University, Faculty of Science, Botany |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Thesis, Masters, MSc |
| Format | 77 p., pdf |
| Rights | Junkuhn, Kyle Peter |
Page generated in 0.0017 seconds