The earliest mammals are generally believed to date back at least as far as the Cretaceous period in the Mesozoic era, which ended roughly 66 million years ago with the extinction of most of the dinosaurs. Juramaia, the earliest known placental mammal, has been reliably dated back 160 million years. Although the haramiyids go back even further than that, there is controversy on whether they were true mammals.
Introduction to haramiyids
The extinct suborder known as Haramiyida is usually thought to be the earliest known terrestrial herbivores among the root ancestors of mammals. They lived in the Jurassic and possibly Triassic period of the Mesozoic era, before the dinosaurs ruled and the supercontinent Pangaea broke up. This could potentially make them as much as 212 million years old.
The haramiyids were rat-like animals which are known mostly through fossils of their teeth. All haramiyid teeth are similar to those of their multituberculate and rodent cousins.
Now, two new haramiyid fossils have now been found in China and Mongolia, which are much more complete than anything in the suborder which had been found before. These new fossils, which have been named Megaconus mammaliaformis and Arboroharamiya jenkinsi, show conflicting evidence about how far back mammalian traits go. This raises questions about how many mammalian traits an animal can have before it is classified as a mammal.
Similarities between haramiyids and mammals
Like mammals, M. mammaliaformis had fur on their abdomen. Fur is one of the defining characteristics of mammals.
While M. mammaliaformis was a ground-based herbivore, A. jenkinsi was an omnivorous tree dweller. The hands and feet of A. jenkinsi were small, with unusually large digits. Its tail may have been prehensile, like those of many tree-dwelling mammals today.
Mammalian movement is based on different leg, ankle and hip anatomy from the movement of other species. With its two fused lower leg bones, the shambling gait of M. mammaliaformis was probably similar to that of the armadillo or rock hyrax.
M. mammaliaformis also had a keratinous and possibly poisonous spur on its heels, similar to the modern platypus. This may be a male-only trait, as it is in the platypus.
The way the teeth fit together in A. jenkinsi also has an open space where mammalian ear bones could fit. These small bones do not preserve well and are not usually found with fossil skeletons.
Reptilian jawbones are made of multiple parts. This is the feature which allows a snake to swallow prey which is bigger across than it is. The jawbone of A. jenkinsi is composed of a single bone on each side, which has only been seen before in mammals.
Are these haramiyids really haramiyids?
This jawbone is a serious problem for identifying the relationship of this species in the mammalian family tree. Apart from Haramiyavia, all previously known haramiyids did not have a solid jawbone. In fact, before 1997, very little was known at all about haramiyids except their jawbones and two parallel rows of cusps on molariform teeth. This was a primary reason for classifying them separately from mammals.
However, the rodent-like teeth of A. jenkinsi are also very similar to those of multituberculates, which are a major side-branch of crown mammals. While multituberculates look superficially like true mammals, with a one-piece jaw and mammalian middle ear, their tooth structure is functionally completely different. They would even have to chew differently from a true mammal.
Thus, this may be a case of parallel evolution. Although multituberculates developed some features which were similar to cousin mammals, the differences probably mean that multituberculates were a side branch outside the mammals.
As a result, it is likely that A. jenkinsi and possibly M. mammaliaformis are really multituberculates, and not haramiyids at all. However, much more of the fossil record will need to be discovered before this can be known for sure.