Results

All calibrated models performed well with an average TSS value above 0.8 (Supporting Information Fig.S2). The results reported below were consistent across all GCMs explored (Supporting Information Fig. S5). The number of grid cells projected suitable for the 1,527 cultivated alien species in Southern Africa under current conditions varied from 0 to 9,244, or approximately 51% of Southern Africa’s area. The number of species projected to encounter climatically suitable conditions in at least one grid cell under the current conditions also varied greatly, ranging from 0 to 313 species per grid cell, or approximately 20% of the modeled cultivated alien flora of Southern Africa (Fig. 1a). Under climate warming scenarios (SPP1 and SSP5), these numbers were projected to decrease (Fig. 1b-e). Projected changes in Maximum Temperature of the Warmest Month resulted in the highest future contraction of potential ranges of alien cultivated plants in Southern Africa (Supporting Information Fig. S3).
Of Southern Africa, 10.0% were defined as invasion hotspots that were considered climatically suitable under current climatic conditions for at least 128 from the pool of 1,527 modeled cultivated alien species (Fig. 2a). Until the end of the century, the size of the invasion hotspot is predicted to decrease slightly under SSP1 climatic scenario (to 7.1%; Fig. 2b), but substantially under the worst-case climatic scenario SSP5 (to 2.0%; Fig. 2c).
We compared the number of alien species predicted to find suitable conditions in the various biomes in Southern Africa under current and future climatic conditions and found slight differences among them (Fig. 3a, b). Based on both future scenarios, tropical biomes are likely to experience less of a reduction in species richness, while other biomes may either not differ significantly from random expectations or even lose more species richness than anticipated by chance. Furthermore, under the worst-case scenario SSP5, the average potential species richness was significantly less than zero in all biomes (indicating a significant decrease in cultivated alien species richness) (Fig. 3a, b).
Under current climatic conditions, naturalized plants were predicted to have a larger potential range than non-naturalized cultivated plants in Southern Africa (GLM: Z = 9.64, P ≤ 0.001; Fig. 3a, Supporting information Table S1). However, as climate becomes warmer, the size of potential ranges is projected to decline for all species (Fig. 3b&c). While contraction is projected to be more significant for non-naturalized species than for naturalized ones under the moderate future climatic scenario SSP1 (GLM: Z = 2.15, P = 0.031; Fig. 3b, Supporting information Table S1), under the worst-case scenario SSP5, there will be no significant difference between the two groups (GLM: Z = 1.87, P = 0.060; Fig. 3c, Supporting information Table S1). While the average cultivated plant in Southern Africa will experience a reduction of its potential range, an increase is projected for ~13% and 9% of species under the scenarios SSP1 and SSP5, respectively.
Our results indicated that under the SSP1 climate-change scenario, species native to TDWG continents mainly located in the Southern Hemisphere (i.e., Pacific Islands, Australasia, Tropical Asia, and Southern America) would likely lose less climatically suitable areas than the average cultivated species. Species native to Pacific Islands and Australasia have an average change in range size significantly greater than zero, indicating a significant increase in areas with suitable climates. In contrast, species native to continents mainly located in the Northern Hemisphere (i.e., Europe, other parts of Africa, and Asia Temperate) will likely lose climatically suitable range in Southern Africa at an above-average rate (Fig. 4a). However, for SSP5 climate-change scenario all species are predicted to have an average change in range size significantly less than zero, indicating a significant contraction in areas with suitable climates (Fig. 4b).