Did vegetation change drive the extinction of Paranthropus boisei?

Early evaluations of the masticatory morphology of Paranthropus (e.g., Robinson, 1954) pointed to it being an ecological specialist. This prompted a decades-long series of evolutionary narratives suggesting that a narrow dietary niche prevented Paranthropus from successfully responding to environmental change in the Pleistocene (reviewed in Wood and Strait, 2004). In a recent example, Quinn and Lepre (2021) draw from the eastern African carbon isotope record of pedogenic carbonates (i.e., paleosol isotopes) to hypothesize that an increase in mean fraction woody cover (f_wc) during the Mid-Pleistocene Transition (MPT; ∼1.3–0.7 Ma) influenced the extinction of Paranthropus boisei. Enamel stable isotope data from P. boisei are consistent with its heavy reliance on C₄ resources (Cerling et al., 2011a). Quinn and Lepre (2021) propose that this made P. boisei vulnerable to an abrupt—on geological time scales—increase in mean f_wc and decrease in C₄ biomass, which would have in turn increased competition for C₄ resources, decreased P. boisei population size, and ultimately resulted in its extinction.

Quinn and Lepre (2021) variably characterize the mean f_wc increase as pertaining to the whole Eastern African Rift System (EARS; e.g., their section header, “Causes of the EARS C₃ excursion”; Quinn and Lepre, 2021: 3) as well as to individual depositional basins within the EARS (e.g., “The Awash (Ethiopia) and Turkana (Kenya) basins have the highest δ¹³C_PC sampling resolutions across the MPT … and both record significant reductions in C₄ vegetation… Olduvai (Oldupai) Gorge and Tugen Hills, yielding small samples from the MPT interval… show persistent grassy vegetation structures”; Quinn and Lepre, 2021: 2). With respect to the larger scale of the EARS, we show here that the ‘C₃ excursion’ presented by Quinn and Lepre (2021) is an artifact of an analysis that was overly influenced by data from two localities: Dikika in the Awash and Nachukui in Turkana. Spatially uneven pedogenic carbonate sampling resulted in a data set and an analysis that does not reflect paleovegetation patterns across the currently known geographic range of P. boisei. Concerning their basin-scale characterization, we discuss why vegetation change at this scale is insufficient to cause the extinction of a geographically widespread species. At most, a change at this scale might cause a species' extirpation, but extinction requires vegetation changes across most, if not all, of the localities occupied by the species. Lastly, even if we accept that the C₃ excursion presented by Quinn and Lepre (2021) reflects what was happening across all of the EARS, we show—via an analysis of the distribution of mean f_wc tolerated by modern mammals in African ecosystems—that the observed mean f_wc increase during the MPT was unlikely to have had a significant influence on the extinction of P. boisei.