A comparison of grass production and utilisation in sodic and crest patches on a semi-arid granitic savanna catena in the southern Kruger National Park, South Africa
Alard, Glynn Frank
Grass biomass productivity was measured in sodic and crest patches in the semi-arid (581.8 mm per annum) savanna of the southern Kruger National Park, straddling an herbivory exclosure located on the northern bank of the Sabie River. Sodic and crest patches are ubiquitous landscape elements within the undulating granitic terrain of the Lowveld region of north-eastern South Africa. Sodic patches consistently occupy footslope positions on the catena and are characterised by a sparse cover of woody shrubs, lower herbaceous standing crop and a higher concentration of large mammalian herbivores relative to adjacent crest patches. These characteristics imply that sodic patches possibly make a substantial contribution to the landscape in terms of high quantity or quality forage production. The long-term use of standing crop indices for grassland productivity assessment in the region are from areas subject to natural herbivory, and has created an impression of low grass biomass productivity on sodic patches. The crest and sodic patches were compared for grass biomass productivity, forage quality and utilisation, and were sampled in the growing season (November to May) from 2002 to 2005. Sequential monthly clippings were conducted in permanently located 1 m2 plots inside the exclosure to determine monthly productivity and outside the exclosure to determine utilisation. Clippings of grass standing biomass were simultaneously collected from 1 m2 once-off plots inside the exclosure to determine the effects of defoliation on productivity. Sodic productivity was 8.74 ± 2.37 g dry matter m-2 month-1 where mammalian herbivory was excluded and did not differ significantly from crest productivity both inside or outside the exclosure (p>0.05), but was significantly higher than sodic forage yields of 0.15 ± 2.82 g m-2 month-1 in the presence of sustained mammalian herbivory (p<0.05). This indicates that sodic forage is more heavily utilised than crest forage. Productivity was significantly higher in plots clipped once-off than in sequentially clipped plots (p<0.05). The patch utilisation differential was investigated by analysing dried productivity clippings for total nitrogen, sodium, carbon and phosphorus. Significantly higher (p<0.05) foliar nitrogen (>2% N) and sodium (>0.5% Na) in sodic forage both account for higher sodic utilisation, though relationships are not linear. Mixed and single species grass foliar samples from the study area were analysed for iii stable natural abundance nitrogen and carbon isotopes to investigate nitrogen availability and cycling, and water availability, respectively. Relatively higher δ15N ranges in sodic samples concur with higher sodic soil nitrogen availability and infer relatively higher rates of nitrogen cycling. Relatively higher δ13C ranges in crest samples concur with lower crest water availability and that crest grass species are more water use efficient. Compared to the sodic patch, crest patches are relatively water and nitrogen limited environments resulting in less fertile grass forage which is less utilised.