The fossils of an ancient crocodile with mammal-like teeth discovered by a Wits scientist, Zubair Jinnah, in 2008 has been described in 5 August edition of Nature.

The fossils were discovered in the Rukwa Rift Basin of Tanzania, and the unusual creature is changing the picture of animal life at 100-million years ago in sub-Saharan Africa.
“I discovered the specimen which has an articulated skull, vertebrae and limb elements, whereas previously discovered material found by our research team of the same species  in previous years was of isolated or incomplete elements,” says Jinnah, a sedimentologist  and an Associate Lecturer in the Wits School of Geosciences, whose research focuses on fossil-bearing sedimentary rocks. “I also described the geology of the site where the most complete fossil came from in order to understand how it was preserved and why it was so complete. This specimen will now form the holotype (reference material) of the new species.”
He describes how the discovery was made: “On the day of the fossil find, we were prospecting a new area where we had never been before. Every time you look at a previously unprospected area in a fossil-bearing rock unit, there is a small chance of finding something exciting and that is what happened that day. At the time of the discovery, three or four small vertebrae (each one about a centimetre across) were sticking out of a cliff-face. At that stage, we had no idea how much more was embedded until we did the excavation a few days later.”
Dr Eric Roberts (former Wits scientist who led the discovery in 2007 and 2008) was the lead geologist on the Tanzanian project. “I joined him in 2007 and 2008 to assist with his geological work with the aim of reconstructing past environments by looking at sedimentary rocks and also determining how old the rocks were. This information helps palaeontologists to try to understand past ecosystems,” adds Jinnah.
But why is this discovery special? Lead author of the study, Patrick O’Connor, Associate Professor of Anatomy at the Ohio University College of Osteopathic Medicine, explains: “If you only looked at the teeth, you would not think this was a crocodile. You would wonder at whether it is a strange mammal or mammal-like reptile.”
The team describes the new species of notosuchian crocodyliform as a small animal whose head would fit in the palm of your hand. It was not as heavily armored as other crocodiles, except along the tail and its gracile limbs suggests that the creatures were quite mobile. Other aspects of its anatomy suggest that it was a land-dwelling creature (unlike water-dwelling crocodiles) that likely feasted on insects and other small animals to survive.
The new species is not a close relative of modern crocodilians, but is a member of a very successful side branch of the crocodyliform lineage that lived during the Mesozoic Era.
The international research team have now recovered portions of seven different individuals of the species in southwestern Tanzania. The tooth row with molar-like teeth initially puzzled many experts. Other ancient and living crocodiles typically boast relatively simple, conical teeth that serve to seize and tear prey, and they swallow flesh in large chunks.
The molar teeth of the new species, named Pakasuchus (Paka is the Ki-Swahili name for cat and souchos is Greek for crocodile), possessed shearing edges for processing food, similar in form to the teeth of some mammalian carnivores. Pakasuchus lived alongside large, plant-eating sauropod and predatory theropod dinosaurs, other types of crocodiles, turtles and various kinds of fishes.
While the specimens of the newly discovered animal and its close relatives are unusual, the study suggests that the creatures were abundant during the middle Cretaceous period, from around 110-million until 80-million years ago.
“Understanding the African fossil record from the Cretaceous (145 to 90 million years ago) is important for a number of reasons. Gondwana (South America, Africa, Australia, Antarctica, India) had begun breaking up at that time and the types of animals we find in Cretaceous ecosystems help us hypothesize about how the continents broke up. This is why it is important to look at all members of an ecosystem, even tiny crocodiles, and not just big, meat-eating dinosaurs.
"Understanding the palaeogeography from the Cretaceous period, that is, how rivers flowed and where the large rivers were, also helps us understand how African landscapes have evolved through time. Africa’s records of sedimentary rocks and fossils are relatively poorly understood and documented, providing opportunity for new work in this area,” explains Jinnah. “Earth sciences are very strong in South Africa, and as such, we are in a very good position to be answering these exciting questions.”