 
          
            International Journal of Molecular Evolution and Biodiversity 2016, Vol. 6, No. 2, 1-6
          
        
        
        
          3
        
        
          2.3.1 Relative density
        
        
          The recorded raw data were transcribed into the Ms Excel computer program, summarized and used to calculate the
        
        
          relative density for each tree species. We used Ms. Excel to plot histograms of trees' relative density against tree
        
        
          species. The aim was to compare descriptively relative density of old and regenerating tree populations.
        
        
          2.3.2 Trees species diversity
        
        
          Ms Excel computer program was used to calculate Shannon diversity indices and R statistics was used to analyze
        
        
          the diversity indices against tree stages. Response for this model was Shannon diversity index which is a continuous
        
        
          variable with trees stages as categorical variable. The predictor possessed of two categories that is tree species with
        
        
          Circumference at Breast Height (CBH) greater or equal to 20cm and trees with CBH less than 20cm. This response
        
        
          distribution and predictor type necessitated the use of linear model. This type of model requires linear relationship
        
        
          between variables, homoskedasticity, normal distribution of response and independence of observation. We used
        
        
          hist. command in R to test normality of response and noted that variables were fairly normal distributed. Furthermore
        
        
          we used box plot to visualize median and data spread (Zuur et al., 2010) and found that one observation fall outside
        
        
          the normal range of data hence an outlier. To curb the effect of the outlier we used median which is robust to outliers'
        
        
          effect. We plotted the median using box plot to report biological significance and ANOVA was used to determine
        
        
          statistical significance level.
        
        
          
            3 Results
          
        
        
          A total of 800 trees were measured and analyzed in our study. Out of these trees, 400 trees had Circumference at
        
        
          Breast Height less than 20m which represent "regenerating tree" population and 400 trees had Circumference at
        
        
          Breast Height greater or equal to 20cm representing "old tree" population.
        
        
          
            3.1 Relative density
          
        
        
          Results reveal that only regenerating tree population of
        
        
          
            Acacia zanzibarica
          
        
        
          and
        
        
          
            Terminalia spinosa
          
        
        
          have higher
        
        
          relative density than old tree populations of the same. Conversely, regenerating tree population of the remaining
        
        
          species have lower (or similar) relative density than their corresponding old populations (Figure 2).
        
        
          
            3.2 Trees species diversity
          
        
        
          Results reveal significant variation in species diversity between regenerating and the old tree populations (Figure
        
        
          3).
        
        
          
            4 Remarks
          
        
        
          Results show relative density of regenerating
        
        
          
            Acacia zanzibarica
          
        
        
          and
        
        
          
            Acacia mellifera
          
        
        
          is higher than corresponding
        
        
          old trees population. On the other hand, results show significant variation in species diversity between regenerating
        
        
          and the old tree populations as hypothesized.
        
        
          More than three quarter of the regenerating tree population is dominated by
        
        
          
            A. zanzibarica
          
        
        
          . Reduction of hot fires,
        
        
          grazing history and heavy loamy-clayey soils are said to have influenced establishment and spread of
        
        
          
            A. zanzibarica
          
        
        
          (Cochard and Edwards, 2011). Similarly, absence of medium-sized browsers like impala probably explains high
        
        
          recruitment rate and establishment of
        
        
          
            A. zanzibarica
          
        
        
          woodland (Prins and Van der Jeugd, 1993). Prins and Van der
        
        
          Jeugd (1993) urged that with high browsing pressure by impala and other small-medium browsers, establishment
        
        
          of
        
        
          
            Acacia
          
        
        
          seedling is rare and this impact declines with tree maturation as most foliage of a mature tree grows
        
        
          beyond reach of these browsers. Establishment of
        
        
          
            A. zanzibarica
          
        
        
          woodland to such extent in the study area
        
        
          significantly reduces grazing resources available to wildlife (Cochard and Edwards, 2011), provide a suitable habitat
        
        
          for tsetse flies (Trail et al. 1985) and block passage of large bodied animals thus restrict the use of the area. It has
        
        
          also been observed that dominance correlates negatively to measures of species diversity (Braun, 1950; Whittaker,
        
        
          1965; McNaughton, 1968; Berger and Parker, 1970). Therefore, the ultimate negative effect of dominance of
        
        
          
            A.
          
        
        
          
            zanzibarica
          
        
        
          in the study area is to cause a decline in plant and associated animals diversity.
        
        
          Our study reveals significant variation in species diversity between regenerating and the old tree populations.
        
        
          Results have shown that old tree population has higher diversity than regenerating population suggesting diversity