Vegetation dynamics over 18 years in response to seasonal climatic conditions and land use in the Kamiesberg mountains of the succulent Karoo, South Africa

Projections of vegetation change in response to future global change conditions are challenged by the lack of long-term, empirical data on how vegetation responds to the variability in climatic conditions. This is particularly true for biodiverse systems that are threatened by climate change such as the winter-rainfall Succulent Karoo biome of southern Africa. In this study we analyse long-term data (2002-2019) of two biodiversity observatories from the Kamiesberg Mountains in the Succulent Karoo. We employed the R package GLMMtmb to analyse counts of abundance and estimated cover values of vegetation as response variables and seasonal Standardised Precipitation-Evapotranspiration Index (SPEI), annual livestock density, topography and soil surface structure as predictors for the interannual change of life forms under low and high land use intensity. We compared the life form composition at the two contrasting land use types and found the cover and abundance of leaf succulents to be higher under low grazing intensity and that of stem succulents to be higher under high grazing intensity. The cover and abundance of herbaceous life forms varied over the years depending on the seasonal weather conditions. Our GLMM results reveal counts of abundance to be a sensitive predictor for population dynamics in various herbaceous and shrub life form types. Estimated cover values showed less of a response to interannual differences in the environmental drivers. All life forms responded strongly to high humid and cool conditions in the second half of the winter rainfall period, when the temperatures increase again. The abundance of leaf-succulent and non-succulent shrubs also responded positively to moderate conditions towards the end of the dry season. The extensive and hot dry summer season in turn had a significant negative effect on the survival of particularly young plants with low cover. Cover values showed no response to the conditions of the late dry season and were generally less sensitive to the variance in seasonal climatic conditions. Our results suggest that future changes in interannual patterns in climatic conditions will affect life form composition of the vegetation.