COS 8-4 - South African mesic grasslands are resistant to drought but not warming

Abstract: Worldwide, increases in temperature and more frequent climatic extremes are predicted and have already been observed to some extent. In South Africa, these extremes are predicted to take the form of more frequent, severe, long-lasting droughts and temperature increases of up to 5°C. This is expected to reduce grassland productivity through changes in species composition, tuft morphology, and reductions in basal cover. As expected, the general pattern in species composition change is the loss of species with high water resource requirements. In addition, shifts in sward reproductive strategies are expected, however, in South Africa, limited work has been done to confirm these predictions.



In 2019 we initiated a site of the International Drought Experiment at Ukulinga Research Farm, Pietermaritzburg, South Africa. We established a drought treatment (50% rainfall exclusion), using rainout shelters as per the Drought-Net protocol, and a warming treatment (2°C increase), using hexagonal open-topped chambers. We found that drought affected grass species composition in 2020 only (when ambient rainfall was 40% below average), where cover of Themeda triandra was replaced by cover of, drought tolerant, Heteropogon contortus. This dominance shifted back to Themeda triandra in 2021 when the ambient rainfall increased. Forb species were unaffected. Warming affected species composition from 2020 onwards with increasing abundance of tropical zone grasses, such as Brachiaria serrata, and forbs replacing temperate zone grasses. Grass biomass, basal cover and number of tillers per tuft were unaffected by drought and warming, but the dominant forb species were shorter under drought. Grass reproductive strategies were unaffected by drought and warming with no differences observed in the time and duration of budding and flowering, bud bank density, seed bank and seed rain composition. Grass phenology appeared to be more strongly linked to the onset, rather than the volume of rain. However, forb phenology was affected, with dominant species extending their flowering stage by one week and producing significantly less flowers under drought conditions.



South African grasslands are well adapted to rainfall fluctuations and it appears unlikely that cover and grass species composition, and thus ecosystem services, will be negatively affected by moderate multiyear droughts. However, warming causes significant shifts in species composition with a tendency towards forbs and weakly tufted tropical grasses. Long term increases in these species is likely reduce basal cover and lead to grassland degradation.