Paleoecological interpretations are based on our understanding of dietary and habitat preferences of fossil taxa. While morphology provides approximations of diets, stable isotope proxies provide insights into the realized diets of animals. We present a synthesis of the isotopic ecologies (δ¹³C from tooth enamel) of North American mammalian herbivores since ~7 Ma. We ask: 1) Do morphological interpretations of dietary behavior agree with stable isotope proxy data?; 2) Are grazing taxa specialists, or is grazing a means to broaden the dietary niche?; and, 3) How is dietary niche breadth attained in taxa at the local level? We demonstrate that while brachydont taxa are specialized as browsers, hypsodont taxa often have broader diets that included more browse consumption than previously anticipated. It has long been accepted that morphology imposes limits on diet; this synthesis supports prior work that herbivores with “grazing” adaptions, such as hypsodont teeth, have the ability to consume grass but are also able to eat other foods. Notably, localized dietary breadth of even generalist taxa can be narrow (~30 to 60% of a taxon’s overall breadth). This synthesis demonstrates that “grazing-adapted” taxa are varied in their diets across space and time, and this flexibility may reduce competition among ancient herbivores.

Isotopic data include all published stable isotope analyses of carbon from the carbonate portion of tooth enamel hydroxylapatite (δ¹³C) from herbivorous mammals (i.e., Perissodactyla, Artiodactyla, and Proboscidea) since the late Miocene (~7 million years ago) that occur in the contiguous United States below 37 °  latitude. Bulk data (1 sample taken parallel to a tooth’s growth axis, per individual, typically < 1 cm in length) and average values from serially sampled teeth (i.e., a series of samples taken perpendicular to a tooth's growth axis) were gathered via a Web of Science search using key words that included isotope, fossil, teeth, and other iterations of these words.

Newly acquired samples for this study (indicated in Dataset S1) were collected from the tooth enamel of late-erupting teeth using a low speed drill and carbide dental burrs. The powdered enamel was pre-treated with 30% hydrogen peroxide for at least 24 hours and 0.1N acetic acid for 18 hours, to remove organics and secondary carbonates, respectively. Samples (~1mg) were analyzed in the Department of Geosciences at the University of Florida. Analytical precision is ±0.1‰ based on replicate analyses of samples and NBS-19 standards. Stable isotope data were normalized to NBS-19 and are reported in conventional delta (δ) notation for carbon obtained from enamel (δ¹³C), where δ¹³C (parts per mil, ‰) = ((Rsample/Rstandard)-1)*1000, and R= ¹³C/¹²C. The standard is VPDB (Pee Dee Belemnite, Vienna Convention).

These data are correct and complete as of their data of upload. Users should be advised that the age and interpretation of fossil localities are occassionally revised.