Factors influencing species diversity
A GLM repeated measures ANOVA analysis was performed to determine the influence of spatial (6 study areas each with 3 study sites), trophic (5 bait types) and temporal factors (2×24 h sampling days) on total species richness, diversity, and abundance across the climatic gradient using Statistica, version 6.0 (Statsoft, 2008). Trap data on each sampling day were treated as repeated measures. Tukey‟s HSD post hoc tests were also conducted to determine which subjects are responsible for any significantly different variance that is detected.
The Kalahari basin dung beetle species were classified according to their behavioural patterns, comprising four principal functional groups defined by Doube (1990). The major functional groups comprise ball-rollers which roll a portion of dung away from the original dung pat before it is buried, tunnellers which provision tunnels made under the original dung pat with piece meal(dung balls), kleptocoprids which utilize dung buried by other dung beetles, and endocoprids which breed within the original dropping in situ (see Davis & Scholtz, 2001). In the present study, only the first 3 functional groups were recognized given that scarabaeinae endocoprids tend to be relatively uncommon compared to other functional groups (see Doube, 1990; Davis, 1996c; Davis & Scholtz, 2001). This allowed for an assessment of the proportional representation of ball rollers, tunnellers, and kleptocoprids across the climatic gradient. For each study area the proportion of species in each functional group was expressed as a percentage of all species present in the study area.
Species abundance patterns
Dung beetle assemblages from the six study areas were compared using non-metric multidimensional scaling (MDS). Analyses were conducted on combined data for carrion and dung assemblages and on dung assemblages only. Ordination plots were used to display dung beetle assemblage relationships among the six study areas to reveal the differences in dung beetle community structure. In this case assemblages in areas or samples with higher similarity are placed closer to one another and the less similar assemblages are placed further apart.
Cluster analysis was used to both define dung beetle community divisions in relation to study areas across the climatic gradient and to support the ordination analysis. The same data were arranged as a matrix of mean species abundance per site per bait type. These data were fourth-root transformed before analysis and the nonmetric Bray-Curtis similarity coefficient was used to compute a similarity matrix. This matrix was subjected to the agglomerative clustering technique, group average linking, using the multivariate analytical computer package PRIMER Version 5.0 (Clarke & Warwick, 2001). Multiple paired comparisons of community clusters were conducted using ANOSIM, a subroutine of this package.
CHAPTER 1 GENERAL INTRODUCTION
1.1. GEOLOGICAL HISTORY AND CURRENT SETTING
1.2. RECENT CLIMATIC HISTORY AND CURRENT PHYSICAL SETTING
1.3. MAMMALS AND RADIATION OF DUNG BEETLES
1.4. RECENT MAMMAL HISTORY RELATIVE TO VEGETATION AND CLIMATIC ZONES OF BOTSWANA
1.5. OBJECTIVES AND THESIS OUTLINE
CHAPTER 2 STUDY AREA, TRAPPING METHODS AND HYPOTHETICAL APPROACH
2.1 STUDY REGION
2.2. HYPOTHETICAL APPROACH
CHAPTER 3 BIOGEOGRAPHICAL DISTRIBUTION PATTERNS OF THE BOTSWANA DUNG BEETLE FAUNA IN SOUTHERN AFRICA AND THE LOCAL BIOGEOGRAPHICAL COMPOSITION OF ASSEMBLAGES ACROSS THE BOTSWANA KALAHARI
3.2. ANALYTICAL METHODS
CHAPTER 4 PATTERNS OF SPECIES DIVERSITY, TURNOVER, AND FUNCTIONAL DIVERSITY ACROSS THE BOTSWANA KALAHARI
4.2. ANALYTICAL METHODS
CHAPTER 5 REGIONAL AND LOCAL SPATIAL PATTERNS ACROSS THE BOTSWANA KALAHARI
5.2. ANALYTICAL METHODS
CHAPTER 6 REGIONAL AND LOCAL PATTERNS OF FOOD ASSOCIATION ACROSS THE BOTSWANA KALAHARI
6.2. ANALYTICAL METHODS
CHAPTER 7 DISCUSSION AND CONCLUSION
7.1. EFFECTS OF THE ENVIRONMENTAL GRADIENT ACROSS THE BOTSWANA KALAHARI
7.2. IMPLICATIONS OF THE RESULTS FOR REGIONAL CONSERVATION AND MANAGEMENT STRATEGIES