Dental mesowear patterns challenge the hypothesis of a 1.7 Ma transition to open grasslands in South Africa's Cradle of Humankind

The assessment of palaeoenvironmental conditions in South Africa's Cradle of Humankind has traditionally relied on two main lines of evidence: the identification of fossil herbivore taxa and the stable carbon isotopic signatures preserved in their dental enamel. However, interpretations drawn from taxonomic composition often depend on taxonomic analogy – that is, the assumption that extinct species had similar diets and habitat preferences as their modern relatives. This assumption can be problematic as ecological roles can shift significantly over evolutionary time. Furthermore, many of these studies were conducted within broad chronological frameworks, which may obscure important temporal and spatial variability in vegetation structure. The traditional model is that the Cradle experienced a transition from woodlands to open grasslands around 1.7 million years ago. While foundational, this model rests on assumptions about ecological continuity and inexact chronological resolution that can mask finer-scale environmental differences. Here, we assess the hypothesis of an abrupt environmental shift at 1.7 Ma by comparing dental mesowear profiles across seven of the most important Cradle sites spanning the time period 3.2–1.3 Ma. These uranium‑lead dated sites are: Cooper's Cave, Drimolen, Haasgat, Hoogland, Malapa, Sterkfontein, and Swartkrans. We analysed mesowear signatures from fossil bovids (n = 623) across seven bovid tribes. All deposits exhibit similar mesowear scores, consistently reflecting a stronger grazing than browsing signal. Findings do not support a transition from woodlands to grasslands ca. 1.7 Ma. Instead, they are consistent with a heterogeneous landscape that remained unchanged from 3.2 to 1.3 Ma. We also identify variation in dietary ecology between members of the same tribe (e.g. Alcelaphini) in the same deposit (e.g. Swartkrans Member 2), indicating generalist tendencies among some tribes. Moreover, certain tribes (Tragelaphini, Reduncini, Antilopini) differ substantially in feeding preference compared to their modern counterparts. Our study highlights the importance of a regional approach to reconstructions of the South Africa Plio-Pleistocene, and the value of conducting such studies within an accurate, local and direct chronological framework.