Impact Of A Rainfall Gradient On Carbon Storage In The Dry Woodland Ecosystem Of Kavango And Zambezi Regions Of Namibia

Abstract

Namibia is currently a signatory on the United Nations Framework Convention on Climate Change (UNFCCC) as a non-Annex I party. As a non-Annex I party, Namibia is required to ‘provide information on greenhouse gas inventories’, but they currently lack sufficient baseline data. The study aims to answer whether increased rainfall increases the amount of carbon stored above and below ground for two important timber species (Burkea africana and Pterocarpus angolensis) found along the natural rainfall gradient in North-eastern Namibia. From this aim, three research objectives were set. The first is to compare whether rainfall results in statistically different non-destructive measurements in the tree species investigated. The second is to develop and compare different allometric models that link non-destructive measurements of trees to above and below ground carbon values, to determine which produces the highest accuracy (about 30% uncertainty level). The third is to determine whether rainfall influences woody biomass for either species.

Four study sites along this rainfall gradient (Long term average rainfall-LTAR) were chosen (Nkurenkuru where the LTAR amount falls between 550-600 mm, Mashare where the LTAR falls between 500-550 mm, Divundu where the LTAR amounts falls between 550-600 mm and Katima Mulilo where the LTAR is greater than 600 mm). Non-destructive sampling was conducted at all sites, at these sites multiple tree parameters (height, dbh, etc.) were measured in 5 dbh classes from 5cm - >45cm. In addition, at one site, destructive sampling was conducted where the above and below ground woody components were weighed in the field. Between 3 -5 sample discs and one root sample were taken from each destructively harvested tree and dried in the laboratory. From this site, allometric biomass models were derived and extrapolated using the measured non-destructive parameters to the remaining 3 sites.

For B. africana, all tree parameters differ significantly across the sites and higher rainfall causes higher woody biomass in this species. For P. angolensis, all tree parameters except bark thickness differ significantly across the sites, but an increase in rainfall does not cause a higher woody biomass in this species. For both species, allometric models were developed to model the above and below ground carbon levels, but the models which included total biomass did not meet the 30% uncertainty level. When looking at only above ground biomass, models were fitted for both of the species which met the 30% uncertainty level. The mean biomass and carbon values followed the rainfall gradient as expected with Mashare having the lowest values with the lowest LTAR and Katima Mulilo having the highest values with the highest LTAR.

Keywords: Pterocarpus angolensis (kiaat), Burkea africana, carbon storage, rainfall gradient