Using partial volumes to estimate available browse biomass in Southern African semi-arid savannas

Abstract

Aims With the increase of game farming in woody areas, as well as the pressing need to assess the spatial distribution of carbon, the quantification of available browse biomass has become an important subject to refine. We aimed to (1) calculate tree canopy volume (CVol) for varying tree canopy shapes and for strata within those shapes to cater for different herbivore browse heights, (2) develop a relation to estimate available browse of deciduous (micro- and mesophyllous) and evergreen (macrophyllous) semi-arid savanna tree species using our calculated CVols, and (3) demonstrate the application of the use of partial volumes and the resultant regression models in the determination of available browse biomass for browsing game species. Location Four key savanna tree species were sampled in the open wooded grassland areas of three game reserves in northeastern KwaZulu-Natal, South Africa: Bonamanzi Game Park, Mun-Ya-Wana Private Game Reserve and Pongola Game Reserve. Methods Tree CVol determination involved calculation of partial volumes, using the integration of a series of circles, allowing a wider range of tree shapes to be included. Non-linear regression was used to investigate the relation between available browse biomass of deciduous microphyllous Acacia nilotica, evergreen macrophyllous Euclea divinorum and Gymnosporia senegalensis and deciduous mesophyllous Spirostachys africana and their respective calculated partial CVols. Results Exponential non-linear regression of the natural logarithm of CVol data provided an accurate and precise description of the tree CVol–available browse biomass relation. Significant relations were obtained over a 2-yr period (2003–2004) for species–site models and general species models. The incorporation of tree foliage densities in available browse biomass estimation further improved model accuracy and precision. Conclusion The general models provided an objective means of determining the browsing capacity of a tree population. In addition, estimation of carbon sequestration could be related to the model output, particularly when changes over time are recorded. Regular measurement of available browse comprises a useful monitoring tool to evaluate impact of animals on trees and shrubs.