Regional vegetation and climate changes are inconsistent with global trends, but what now?

Global greening is a trend attributed to rising atmospheric CO 2 , warming and land-use change. However, certain regions may diverge from this pattern. South Africa, which occupies 0.82% of the global terrestrial surface area with ca. 10% of global Plantae species, is considered especially vulnerable to climate change. We analysed enhanced vegetation indices (EVI) from 1984 to 2021 across protected areas, correlating this with climate data from ground stations and reanalysis models (TerraClimate, ERA5, CFSv2). The data were aggregated for ten vegetation biomes and anomaly values calculated (i.e., departure of annual value from site mean for the entire period). Results showed significant declines in EVI anomalies over time in seven biomes (Desert, Forest, Fynbos, Grassland, Indian Ocean Coastal Belt, Savanna and Succulent Karoo) out of ten. Temporal EVI variability did not increase, and there was little evidence for changes in precipitation or aridity. Contrary to global patterns, regional ground station average temperatures were relatively stable between 1980 and 2021 whereas maximum temperature increased, and minimum temperatures decreased, resulting in a sharp increase in temperature ranges (0.455°C decade -1 ). A shift in average temperature trends, however, occurred in 2007, with a subsequent linear rise of 0.36°C per decade. While reanalysis data agree with ground station data on the lack of precipitation change, it diverged on temperature trends, showing low correlation with ground observations raising concerns about its accuracy at the regional level. We conclude that South Africa’s vegetation patterns diverge from the global greening trend. It is uncertain how the observed increases in temperature ranges, recent increases in average annual temperature, and varying disturbances may drive future vegetation change. • South Africa’s vegetation patterns diverge from global greening trends • Negative EVI trend in 7 of 10 biomes over 4 decades • Changed temperature ranges and disturbances may be drivers • Recent shifts in climate trends evident • Poor correlation between reanalysis and ground station data compared to global values • Regional, long-term observational analyses vital to guide management and policy