Kaedan O'BRien

Analyses of dental microwear and mesowear shed light on dietary ecology in extinct species and contribute to paleoenvironmental reconstructions. Interpreting dietary signals in extinct taxa requires calibration with data from extant species. In taxa with high dietary plasticity, collecting baseline data is complicated by potential dietary variation across space. We assembled a dataset of over 100 springbok (Antidorcas marsupialis) samples to test whether dental microwear and mesowear vary intraspecifically across biomes. We compared dental wear in Pleistocene Antidorcas specimens to our modern dataset and simulated random sampling to test how dietary interpretations change with sampling strategy. We find that microwear and mesowear vary across and within biomes: fynbos microwear suggests more browse consumption, grassland microwear is consistent with more grass consumption, and savanna specimens show high variability in dental microwear, which is partially explained by ecoregion. Savanna specimens show the most wear in mesowear analysis. Dietary interpretations differ by biome, with some samples showing primarily a browser/frugivore signal and others a more ambiguous signal. Random subsampling produces similar results, with most subsamples consistent with browsers/frugivores, and many subsamples also consistent with browser-grazers/generalists. Fossil specimens share microwear patterns with the fynbos sample. Our results reveal that dietary signals in mixed feeders can vary intraspecifically within and between biomes, emphasizing the importance of diverse reference collections and careful subsampling strategies. Mixed-feeding species are valuable sources of palaeoecological data, providing insights into vegetation composition and environmental conditions via their realized diets, as inferred by dental microwear and mesowear analyses.

Modern African ungulates navigate seasonal variation in resource availability through diet‐switching (primarily mixed‐feeders) and/or migrating (primarily grass grazers). These ecological generalisations are well‐documented today, but the extent to which they apply to the non‐analog ecosystems of the Pleistocene are unclear. Drawing from serially‐sampled stable isotope measurements from 18 Kenyan large herbivore species from the Last Glacial Period (LGP), we evaluate how diet, diet‐switching, and migration compare to observations from present‐day settings. We find a higher grazing signal in most LGP species and a greater magnitude of diet‐switching than in the present. Additionally, we find that the relationships between grass intake, migration, diet‐switching, and body size during the LGP were unlike those observed today. This establishes a revised paleoecology of LGP herbivore communities and highlights that LGP herbivores were behaviourally non‐analog. Our results imply that ecological observations from present‐day settings offer an incomplete perspective of herbivore‐environment interactions.

Eastern Africa is home to the largest terrestrial migrations on Earth. Though these migratory systems have been well studied for decades, little is known of their antiquity and evolutionary history. Serially sampled strontium stable isotopes (⁸⁷Sr/⁸⁶Sr) from tooth enamel can be used to track migration in mammals. Here we analyse ⁸⁷Sr/⁸⁶Sr for 79 bovid and equid individuals representing 18 species from four localities in Kenya to characterize prehistoric migratory systems during the Last Glacial Period (115–11.7 ka). Of the species analysed, 16 lack definitive evidence for migration, including blue wildebeest (Connochaetes taurinus), Thomson’s gazelle (Eudorcas thomsonii) and plains zebra (Equus quagga), which are long-distance migrants today in the Greater Serengeti Ecosystem and historically in the Athi-Kapiti Plains. Only two species, the extinct wildebeests Rusingoryx atopocranion and Megalotragus sp., were migratory. These findings suggest a possible alternative narrative about ecosystem dynamics during the Last Glacial Period and shed light on the behaviour of both extant and extinct species at this time. In particular, these results indicate that migratory behaviour in extant species either emerged during the Holocene or was more spatiotemporally constrained in the past. Our results contribute to a growing body of evidence suggesting that the structure and function of geologically recent large mammal communities in eastern Africa differed considerably from those observed in the present day.

Although modern humans left Africa multiple times over 100,000 years ago, those broadly ancestral to non-Africans dispersed less than 100,000 years ago¹. Most models hold that these events occurred through green corridors created during humid periods because arid intervals constrained population movements². Here we report an archaeological site—Shinfa-Metema 1, in the lowlands of northwest Ethiopia, with Youngest Toba Tuff cryptotephra dated to around 74,000 years ago—that provides early and rare evidence of intensive riverine-based foraging aided by the likely adoption of the bow and arrow. The diet included a wide range of terrestrial and aquatic animals. Stable oxygen isotopes from fossil mammal teeth and ostrich eggshell show that the site was occupied during a period of high seasonal aridity. The unusual abundance of fish suggests that capture occurred in the ever smaller and shallower waterholes of a seasonal river during a long dry season, revealing flexible adaptations to challenging climatic conditions during the Middle Stone Age. Adaptive foraging along dry-season waterholes would have transformed seasonal rivers into ‘blue highway’ corridors, potentially facilitating an out-of-Africa dispersal and suggesting that the event was not restricted to times of humid climates. The behavioural flexibility required to survive seasonally arid conditions in general, and the apparent short-term effects of the Toba supereruption in particular were probably key to the most recent dispersal and subsequent worldwide expansion of modern humans.

Assemblages of mammal skeletal remains provide a powerful tool for censusing wildlife populations to establish zoological baselines required for evaluating biogeographic trends over varying timescales. Caves provide an ideal depositional setting to preserve these skeletal remains despite potential time averaging and taphonomic filtering. We describe a Holocene paleontological assemblage from Boomerang Cave in the Bear River Range of Cache County, northern Utah, United States, at an elevation of 2,231 m, and at the boundary between the Great Basin and Rocky Mountain biogeographic provinces. We analyzed 1,228 surface-collected specimens from six areas within the cave, and identified a minimum of 22 nonoverlapping mammalian taxa, comprising all size classes present in the region. Compared to museum records for mammals from the Bear River Range and individuals trapped or observed in the vicinity of the cave, specimen-based rarefaction demonstrates that our assemblage captures most of the mammalian diversity expected in the area. This is particularly apparent for carnivorans and soricids, which are particularly well-represented in the Boomerang Cave assemblage, with the former clade represented by at least nine taxa. This high level of diversity can be attributed to the relatively random nature of natural trap cave deposition, reducing accumulation biases due to size or diet. We also record the first occurrence of Merriam’s Shrew (Sorex merriami) from the Bear River Range. Our analysis does not indicate any mammalian changes between late Holocene and present-day communities, but these data establish a new zoological baseline for an alpine community at the interface between two key biogeographic provinces in western North America. Our work highlights the value of collecting skeletal remains from cave assemblages as a convenient and fast method for censusing terrestrial mammalian communities.

For many animals, migration is an important strategy for navigating seasonal bottlenecks in resource availability. In the savannas of eastern Africa, herds of grazing animals, including blue wildebeest (Connochaetes taurinus), Thomson's gazelle (Eudorcas thomsonii), and plains zebra (Equus quagga), travel hundreds of kilometers annually tracking suitable forage and water. However, we know nearly nothing about migration among the extinct species that often dominated Late Pleistocene communities. Using serially sampled ⁸⁷Sr/⁸⁶Sr and δ¹³C, we characterize the prehistoric movement and diet of the enigmatic wildebeest Rusingoryx atopocranion from two localities (Karungu and Rusinga Island) in the Lake Victoria Basin of western Kenya. We find clear evidence for migration in all four individuals studied, with three ⁸⁷Sr/⁸⁶Sr series demonstrating high-amplitude fluctuations and all falling outside the modeled isoscape ⁸⁷Sr/⁸⁶Sr ranges of the fossil localities from which they were recovered. This suggests that R. atopocranion exhibited migratory behavior comparable to that of its closest living relatives in the genus Connochaetes. Additionally, individuals show seasonally-variable δ¹³C, with a higher browse intake than modern and fossil eastern African alcelaphins indicating behavioral differences among extinct taxa otherwise unrecognized by comparison with extant related species. That this species was highly migratory aligns with its morphology matching that of an open grassland migrant: it had open-adapted postcranial morphology along with a unique cranial structure convergent with lambeosaurine dinosaurs for calling long distances. We further hypothesize that its migratory behavior may be linked to its extinction, as R. atopocranion disappears from the Lake Victoria Basin fossil sequence coincident with the refilling of Lake Victoria sometime after 36 ka, potentially impeding its past migratory routes. This study characterizes migration in an extinct eastern African species for the first time and shapes our ecological understanding of this unique bovid and the ecosystems in which Middle Stone Age humans lived.