Other Research

The effects of savanna trees on grass: possible consequences for grazing herbivores in the Greater Makalali Conservancy, South Africa

The savanna biome is characterized by a dynamic balance between trees and grasses. This study was performed to give an insight into the effects of savanna trees on grass quality and quantity, and what the possible consequences are for grazing herbivores. Also a small study was performed to analyze the impact of elephants on trees. To determine the role of trees on quality and quantity of grass in subtropical savannas, grass and soil samples were taken inside and outside the canopy area of Acacia nigrescens (nitrogen-fixing) and Sclerocarya birrea (non-nitrogen-fixing) trees in three different tree densities in the Greater Makalali Conservancy, Limpopo Province, South Africa. Grass selection of blue wildebeest (Connochaetes taurinus), Burchell's zebra (Equus burchelli), warthog (Phacochoerus africanus) and impala (Aepyceros melampus) was recorded to determine whether the grass resource selection was related to the difference in grass quality. Finally, the impact of elephants (Loxodonta africana) on trees was observed to determine the indirect effects that elephants may have on the grass resource.

Soil nutrient contents, for both nitrogen and phosphorous, were higher in soil under the canopy compared to soil outside the canopy. Tree density did not significantly alter soil nutrient concentrations. The canopy seemed to increase soil nutrients in open, moderate and dense tree density areas. Trees with nitrogen fixing capabilities did not significantly change the concentration of soil nutrients. The above ground biomass was not significantly different inside the canopy compared to outside. In areas with high tree density the above ground dry grass biomass was about three times lower (103±9 g m -2 ) than in the open areas (304±35 g m -2) or in areas with moderate tree density (282±38 g m -2 ). Both the stem-to-leaf ratio and green-to-dead ratio were reduced with increasing tree density. The percentage alive grass was significant higher inside the canopy compared to outside. The tree canopy increased nitrogen and therefore the crude protein (CP) oncentrations in grass. Grass in areas with high tree density had a lower CP-content. CP-content in grass was higher around N-fixating trees. The phosphorous content in grass was reduced by increasing tree density. On the other hand, fibre content (%NDF) increased with tree density and was also lower under N-fixing trees. High nutritional grass species had a CP- and P-content of respectively 2.5 and 3.3 times higher than low nutritional grass species (range CP-content in grass species: 6.46-16.50%; P-content: 0.07-0.23%), and a lower %NDF (range: 62.9-73.9%).

The relative number of zebra observations was significantly higher in open woodland than expected. The relative grazing activity of both zebra and wildebeest seems to be higher in the open woodlands than in closed woodlands. This study showed that grass quality was higher in the areas with low tree density. This might indicate that resource selection of large grazing herbivores is influenced by grass quality, in particular by phosphorous content in grass. The impact of elephants on trees seems to be very low, 65 of the 347 recorded trees were utilized by elephants, and only 14 trees were utilized and dead. Tree utilization by elephants was mainly characterised by the breaking of main branches.

This study clearly shows that savanna trees have an effect on grass, and are therefore able to affect the resource of grazing herbivores. The most important factor is tree density, grass quality and quantity reduces with increasing tree density. This is possibly due to resource (nutrient, light and water) competition between trees and grasses which reduces the nutrient uptake and grass biomass production. The tree crown also plays a role by increasing the nitrogen availability in the soil which results in higher crude protein content in grass, and therefore a higher grass quality.