Drylands can have extremely rich floras (e.g. Cape floristic province); has aridity itself contributed to high speciation rates? If so, please provide references.
If aridity or harsh conditions fluctuated drastically in a particular region, you would have a mixed population of vegetation that need not necessarily adapt to the changes. On the other hand, if the location under consideration underwent a gradual change from fertile to arid, then the vegetation (native of that region) would adapt, leading to speciation.
Certainly, if aridity is linked with geographic (or climatic) isolation the number of endemic species can be high. We find a good example of this in several of the "Macaronesian" or Mid-Atlantic archipelagos, like the Canaries and the Cape Verdes. In some of the islands of this last archipelago (e.g. Sal, Boavista and Maio), many of the endemic plants are really "arid elements", mainly linked to NW Africa or the Sahel, and the same phenomenum is found in the Canary Islands, where there is a clear connection with the Moroccan and West Saharan flora.