The evolution of herbivory in mammals
So there’s this 2013 study that examines mammalian diversity in relation to the spread of angiosperms, and there are several interesting results. For instance, carnivorous/insectivorous species underwent a decline with the spread of angiosperms, while Argentoconodon is apparently a carnivore while Volaticotherium is an insectivore (both relevant to my flying volaticotheres post). Most interestingly, it paints a rather interesting picture on the development of mammalian herbivory.
As you can see above, through most of the Mesozoic mammals were predominantly animalivorous. By the Early Jurassic there was already a vast diversity of mammalian and quasi-mammalian species; insectivores such as Megazostrodon and Kuehneotherium branched into hard-shelled and soft-prey specialists respectively, while we see the appearance of relatively large sized carnivores like Sinoconodon and the aerial volaticotheres. This trend continues into the Late Jurassic and Early Cretaceous periods, which see a further diversification of mammals into aquatic, fossorial, arboreal and even larger sized carnivorous species. Essentially, we see a guild of insectivores and carnivores in just about any niche available.
However, through most of the Jurassic and Early Cretaceous only one lineage of mammals, the multituberculates, appear to have ventured into herbivorous niches. Even haramiyidans, traditionally considered herbivores or omnivores, range within the insectivore space (though in fairness only one genus is taken into account). And through most of the Mesozoic, multituberculates only do so tentatively, mostly staying within granivore or animalivorous niches. Only in the Late Cretaceous do they venture into fully herbivorous niches, alongside at least two other mammalian clades, the eutherian zhelestids and dryolestoid mesungulatoids (both unaccounted for in the graph but the latter alluded to in the paper).
So, in essence, through most of the age of the dinosaurs mammals were predominantly carnivorous, only occasionally touching granivorous niches until the very end, when fully herbivorous mammals explode in diversity. In spite of their diversity in locomotion methods and size, mammals remained more or less barred from predominantly plant-eating habits until late in the game.
This is in contrast with other groups, such as lepidosaurs and theropods, which did experiment with herbivory early on. Crocodylomorphs appear to follow a similar pattern, with most of the Mesozoic seeing a variety of carnivorous species but only witnessing the rise of herbivorous taxa in the Late Cretaceous; the same might also apply to pterosaurs, if the edentulous tapejarids were in fact significantly herbivorous. The consistent dietary range through time seems to imply that preservation bias is not influencing the results (beyond showing exactly when the “herbivorous turnover” occurred).
This is quite interesting for a variety of reasons. While the paper does warn against equating the evolution of mammalian herbivory with the spread of angiosperms, the fact that the first mammalian herbivores were seed-eaters might imply that mammals were unable to convert into conventional herbivory directly, having to go through a granivore stage first. This clearly applies to multituberculates, though it remains to be seen if it also applies to zhelestids and mesungulatoids.
This might give an insight to how herbivory developed in tetrapods. Tetrapods as a whole are ancestrally animalivorous, but explored herbivorous niches multiple times. It is possible that granivory could have bridged between insectivorous or carnivorous habits and full-fledged herbivory it at least some groups, drawing in through their metabolic rewards but offering a degree of structural complexity that needs to be dealt with. This is particularly interesting in groups such as dinosaurs and anomodonts, in which herbivorous representatives are often beaked or have “buck-teeth” and, like mammals, are endothermic, higher energetic needs that could imply a need for such a transition.
Another important insight is how Mesozoic trophic dynamics changed through time. Mammals were for the longest time barred from an important part of terrestrial ecologies and fulfilled mostly secondary and above consumer roles. This might explain the decline of carnivorous and insectivorous species in the medial Cretaceous, as the higher trophic levels would render them more vulnerable to sudden ecological turnovers. As pointed out in the paper, the more omnivorous therians and meridiolestidans managed to thrive and expanded into the niches left by non-multituberculate mammal groups.
More importantly, this might be part of a much larger turnover. Amidst Jurassic tetrapods only dinosaurs and tritylodontid synapsids appear to have specialised significantly towards herbivory, with a few crocolymorphs and sphenodonts probably veering towards omnivorous habits. Given that the Late Cretaceous sees a much higher diversity of herbivorous tetrapods, including notosuchians, sphenodonts, squamates, turtles and of course mammals, it might suggest that Mesozoic ecosystems couldn’t support many herbivorous guilds we now take for granted, and that floral turnovers such as the spread of angiosperms created new ecological niches that didn’t exist before.
Michael J. Benton,Mikhail A. Shishkin,David M. Unwin, The Age of Dinosaurs in Russia and Mongolia
JENNIFER BOTHA-BRINK and KENNETH D. ANGIELCZYK, Do extraordinarily high growth rates in Permo-Triassic dicynodonts (Therapsida, Anomodontia) explain their success before and after the end-Permian extinction?, Version of Record online: 26 JUL 2010 DOI: 10.1111/j.1096-3642.2009.00601.x