Population structure, species composition and distribution of riparian woody plant communities in the Okavango Delta, Botswana

Abstract

There is limited information on the species composition and distribution, population structure and dynamics and the underlying environmental controls of riparian woodland communities in the Okavango Delta. This study aimed to investigate these parameters, regeneration, and environmental controls of riparian woodland in the Okavango Delta. Tree species were identified in 101 (20 × 50m) plots. Sampling was done in 7 sites with different hydrological characteristics: Boro, Jao, Upper Panhandle, Seronga, Vumbura, Moremi and Santawani. Vegetation sampling was done from February 2012 to November 2013. The sampling plots were placed along a groundwater gradient from the main surface water body to the drier fringe of the riparian zone. Percentage canopy cover was estimated for each tree species from each plot using a Braun- Blanquet cover abundance scale. In each plot the circumference (>15cm) was measured from trees at 1.3m above ground level. The effect of land use on regeneration structure was studied by comparing the regeneration status of tree species between Moremi Game Reserve (Protected area) and Seronga (Communal, open-access area). Height of trees was estimated and individuals were assigned to 0-0.5m, 0.5-1m, 1-2m, 2-4m and >4m height classes. The influence of proximity to surface water was sought through subdividing each 20 × 50m plot into five 20m × 10m sub-plots, to give distance classes of 0-10m (closest to surface water), 10-20m, 20-30m, 30- 40m and 40-50m (furthest from surface water). Tree height, basal area, species richness, cover and diversity were determined for each distance class. Vegetation communities were determined across sites with different flooding frequency regimes based on agglomerative hierarchical cluster analysis and indicator species analysis. Plant species and diversity were determined for each vegetation community. Multi-Response Permutation Procedures (MRPP) were used to evaluate the statistical validity of the separation between vegetation clusters. The Kruskal-Wallis test was used to compare species diversity and richness between vegetation communities. Independent sample Student‟s t-test was used to determine statistical difference in seedling density/ha and sapling density/ha between different sites. Indicator species analysis was used to determine the dominant species at each distance class and site. One-way ANOVA with Tukey test post hoc analysis were used to compare species diversity, richness, mean tree height, cover and basal area between distance classes and site. The four major vegetation communities identified were: Croton megalobotrys-Hyphaene petersiana; Acacia erioloba-Diospyros lycioides; Syzygium cordatum-Phoenix reclinata; and Garcinia livingtonei-Acacia nigrescens. Species richness was highest in the Garcinia livingstonei-Senegalia nigrescens community and lowest in the Syzygium cordatum-Phoenix reclinata community. MRPP showed that there was significant (p<0.05) difference between these vegetation communities. Species diversity and richness were significantly (p<0.05) higher in the Garcinia livingstonei-Acacia nigrescens community than in all others. Distance from surface water was positively correlated with all vegetation parameters except for mean species richness. Woodland community composition, density, cover and diversity varied significantly (p<0.05) between different sites. The population and regeneration structures of different riparian tree species varied with sites in the Delta. This study provided an inventory of riparian woodland plant communities which can be used as a baseline for future monitoring of vegetation change in the Okavango Delta. The results of this study imply that seed dispersal, germination and recruitment processes vary spatially. There is a potential to use woodland plant community composition and distribution and population structure to reconstruct the historical hydrological patterns of a given area where such information is not available.