**2.8 Species richness, diversity, and abundance of the ground- and tree-feeding birds**

Data were tested for normality by using the Shapiro-Wilk (W) test and the Kolmogorov-Smirnov (KS) test. The data were found not to be normally distributed even after transformation. Therefore a nonparametric Mann-Whitney U-test was used to examine differences in independent samples [40]. Species richness estimates were obtained using the Species Diversity and Richness (SDR) computer program [41]. The first-order Jackknife estimator was chosen for species richness because it has been shown to perform well on bird communities distribution [42, 43].

Abundance was measured as the number of individuals found per unit time. Comparisons between pairs of habitats for an abundance of birds at similar grass heights were tested for normality using Kolmogorov-Smirnov (KS) test. In this case, the data were found to be normally distributed. Therefore, differences in the abundance of ground bird species at similar grass heights were compared using the parametric paired sample *t-*test from the computer software Statistix-10 [44].

The diversity of bird species was determined by the Shannon-Wiener diversity (H′) index denoted as:

$$H' = -\sum\_{i=1}^{k} p\_i \ln p\_i,\tag{1}$$

where *k* is the total number of species and *p*i is the proportion of individuals found in the *i*th species. The Shannon-Wiener diversity (H′) value obtained from each habitat type was tested for differences by using the randomization test [41, 45] to observe whether bird diversity of ground- or tree-feeding bird species differ in the *Commiphora* spp. habitat when compared with those in *V. tortilis* and *V. robusta* and when grassland compared with *V. drepanolobium* and *V. seyal* vegetations.
