Abstract:
Understanding driving factors of spatial heterogeneity in plant species com-
position and functional structure is a key step towards wetland ecosystems restoration and
biodiversity conservation. We surveyed 60 randomly selected plots of 20 9 50 m each and
collected 12 explanatory variables to understand potential impacts of natural environ-
mental conditions, human disturbance and spatial position of plots variables on plant
functional groups (PFGs) and species composition within grasslands of the Kilombero
Valley Floodplain wetland, Tanzania. Ordination analyses were used to identify important
vegetation gradients and establish significant natural environmental conditions, human
disturbance and spatial position of plots correlates. Partial redundancy analysis (RDA) and
partial canonical correspondence analysis (CCA) were used to determine the individual
and shared effects of these three sets of explanatory variables on the PFGs and species
composition, respectively. In total, 115 plant species, including 22 weed and 3 invasive
species, and five PFGs were registered. Annual graminoid was the most abundant, whereas
perennial forb the least abundant PFG with 50 and 1 % relative covers, respectively.
Overall, spatial position, altitude, total organic carbon, cow dung, distance to the river and
distance to kraal (cattle enclosure near human settlement commonly made of wood
materials for animal protection) were important descriptors of both PFGs and overall
species composition. Separate CCA of only weed and invasive plant species showed that
some species, particularly the invasive Polygala paniculata were strongly associated with
cow dung, indicating that present increase in cattle numbers may result in future problems
associated with this species. Intensification of human activities and alteration of natural
environmental conditions associated with these factors should be discouraged to maintain
plant species composition and functional structure for wetland restoration and sustainable
biodiversity conservation.
