Here are some snippets from the original Nature article with some telling words/statements bolded, compliments of Creation Safaris' editor:
During most of the Cenozoic era, from the Cretaceous–Tertiary boundary 65 million years ago [sic] until roughly 24 million years ago [sic], Afro-Arabia was [sic] an island continent drifting steadily [sic] northwards towards Eurasia. Fossil mammals documenting this period are scarce and belong almost exclusively to endemic forms restricted to Afro-Arabia, such as proboscideans, hyraxes and elephant-shrews. But by around 24 million years ago [sic], a permanent land bridge had formed between the two landmasses. A burst of faunal interchange followed: many Eurasian mammals, such as rhinos and ruminants, dispersed into Africa, and some Afro-Arabian mammals, such as elephants, migrated in the opposite direction.
Among the proboscideans recorded are primitive [sic] forms such as Palaeomastodon and Phiomia (also known from older deposits in Egypt). But there are also representatives of modern families, for example taxa such as Gomphotherium, the earliest proboscidean on the branch leading to extant elephants. Another surprise is the oldest occurrence of deinotheres, peculiar proboscideans with downward-curved lower tusks, which were previously recorded only from rocks younger than 24 million years old [sic]. The new species of deinothere displays molars that are more ‘bunodont’ in form (that is, made of several distinct cusplets arranged in transverse crests) than its descendant, whose molars display plain transverse crests. This discovery seems to rule out the possibility that deinotheres are derived from an ancestor bearing plain, transverse-crested molars, as was formerly supposed, and provides new evidence about proboscidean evolution [sic].
Finally, the discoveries of Kappelman et al. highlight two other palaeobiological issues. First, on northern continents glaciation caused a significant cooling around 33 million years ago [sic], which resulted in numerous extinctions [sic] among mammalian communities. From these new data, however, it seems that large Afro-Arabian herbivores were not affected, either at that time or later, implying that the climatic changes were less severe on southern continents. Second, the fossil record of the Afro-Arabian continent is not only scanty but also largely concentrated on the northern edge. This has led to the proposal that other groups of mammals existed in Afro-Arabia during its period of isolation, but that they were restricted to more southern latitudes. However, the Chilga mammal community is rather like that found at Fayum in Egypt, which is some five million years older [sic], providing hints that there was little provinciality among Afro-Arabian mammals at that time. As yet, though, we have unveiled [sic] only a few of the secrets of mammal evolution on the Afro-Arabian continent. Many more surprising discoveries are to be expected.