The demise of the giant ape *Gigantopithecus blacki*

Gigantopithecus blacki is known from Early to Middle Pleistocene cave deposits in southern China and is the largest primate known, with a fossil record largely restricted to mandibles and isolated teeth. Previous chronologies placed its presence between about 2.2 Ma and 420–330 ka, with indications of morphological and potential dietary changes over time. However, the timing and causes of its extinction remained uncertain due to limited regional dating and a lack of integrated environmental and behavioural evidence. This study aims to resolve the timing of extinction and explore environmental and behavioural drivers by applying a regional, multiproxy approach across 22 caves in southern China.

Prior work established G. blacki as a key member of Early to Middle Pleistocene faunal zones in (sub)tropical Asia and documented its large body size, unique dental morphology, and potential dietary specialization. The fossil record has been limited to four mandibles and nearly 2,000 isolated teeth, with few sites dated using more than one radiometric method. Earlier timelines centered on specific sites (for example, Baikong and Hejiang) and suggested a presence to 420–330 ka. Reconstructions based on dental anatomy suggested a specialized herbivore adapted to abrasive, fibrous, and fruit-rich diets. Regional vegetation studies indicated diverse forests capable of supporting multiple primate communities earlier in the record, with later range contraction of G. blacki to Guangxi. Despite debates, the lack of a regional, multiproxy approach integrating behaviour and environment left the extinction cause unresolved.

The study targeted 22 caves in southern China (Chongzuo and Bubing Basin), including 11 Gigantopithecus-bearing and 11 non-Gigantopithecus-bearing sites, combining material from earlier excavations (1999–2016) and newly discovered caves (2017–2020). A total of 157 radiometric ages were obtained using six independent techniques applied to sediments and fossils: PIR-IRSL, OSL, ESR on quartz, U-series on speleothem (sediments), and U-series on teeth plus coupled US-ESR (fossils). Bayesian modelling was used to establish age ranges per site and derive a regional extinction window (EW). Palaeoenvironmental reconstruction employed pollen, charcoal, palaeontological, stable isotope, and microstratigraphic analyses. Behavioural and dietary shifts were assessed by trace element mapping (Sr/Ca, Ba/Ca, Pb), stable isotope analyses (δ13C, δ18O), and dental microwear texture analysis (DMTA) on G. blacki and Pongo weidenreichi teeth. Analyses focused on temporal bins: pre-EW (2,300–700 ka), transitional phase (700–295 ka), EW (295–215 ka), and post-EW (215 ka to present).

• Chronology: Fossil evidence from the 22 caves spans 2,300–49 ka. The presence of G. blacki is extended to 2.3 Ma and as late as 255 ka. A regional extinction window is constrained to 295–215 ka.
• Palaeoenvironment: Pollen data show pre-EW dominance of arboreal taxa (Pinaceae, Fagaceae, Betulaceae) with grassland patches. During the transitional phase, forest communities changed with an increase in disturbance taxa and more open forests. Post-EW (~200 ka), there was a substantial decrease in arboreal cover, increases in ferns and grassland (e.g., Poaceae), and more charcoal evidence.
• Faunal shifts: Pre-EW assemblages include G. blacki (relatively numerous), Ailuropoda microta, Procynocephalus, Sinomastodon, Stegodon, Hesperotherium, Hippopotamodon. Before the EW, G. blacki occurrences decrease and assemblages shift to Ailuropoda baconi, Stegodon, and Elephas. Post-EW, G. blacki is absent.
• Microstratigraphy: Pre-EW microfacies characterized by fine grains, higher clays/oxides, bioturbation, and guano-induced phosphatization. At the EW, grain sizes increased with lower oxides, bioturbation, and bone/tooth alteration, improving fossil preservation; post-EW conditions reverted to pre-EW features.
• Stable isotopes: For G. blacki, pre-EW δ13C and δ18O range from −16.2 to −13.8‰ and 9.7 to −7.0‰, respectively; during the EW, values shift slightly to −15.3 to −10.3‰ and −9.3 to −6.3‰. For P. weidenreichi, pre-EW δ13C and δ18O range from −14.7 to −13.7‰ and −7.1 to −6.3‰; during the EW, δ13C extends to −14.7 to −13.3‰ and δ18O changes to −4.9 to −4.4‰.
• Trace elements: Pre-EW G. blacki teeth show distinct, synchronous Sr/Ca and Ba/Ca banding in enamel and dentine, transitioning to less visible, diffuse banding approaching the EW. Distinct lead banding evident pre-EW becomes less distinct during the EW, indicating physiological stress and environmental change.
• DMTA: No statistically significant dietary differences between G. blacki- and P. weidenreichi-bearing sites overall, but significant dietary differences are observed within four G. blacki-bearing sites between pre-EW and just before the EW. G. blacki shows greater fluctuations in mean anisotropy and complexity over time, whereas P. weidenreichi remains more stable, especially in anisotropy. Sample sizes: G. blacki n=16; P. weidenreichi n=22.
• Demography and range: By ~300 ka, the number of G. blacki caves and teeth decreases, indicating dwindling populations and a dramatic range reduction to Guangxi prior to extinction.

The multiproxy regional dataset indicates that enhanced environmental variability and increasingly open, disturbed forest structures prior to and during 295–215 ka undermined the ecological niche of G. blacki. As a dietary specialist relying on fruits with lower-nutrition fallback foods, with a large, less mobile and terrestrial lifestyle, G. blacki was more vulnerable to forest structural changes and seasonality increases. Trace element banding changes indicate chronic stress, and microwear trends reflect shifts from preferred dietary behaviour. In contrast, P. weidenreichi was more arboreal, mobile, and behaviourally flexible, potentially moving in smaller groups and adjusting to open habitats, exhibiting less stress and stable isotopic/microwear signals. Body-size trends (G. blacki increasing, P. weidenreichi decreasing) imply differing adaptive capacities, with G. blacki likely having slower growth and lower reproduction rates. There is no evidence that archaic hominins played a role in this extinction event in southern China. The findings provide a robust, species-specific framework linking timing, environmental change, and behavioural responses to explain G. blacki’s extinction.

This study presents the most comprehensive, robustly dated regional record for G. blacki, constraining its extinction to 295–215 ka, extending its presence to 2.3 Ma and as late as 255 ka, and linking its demise to enhanced environmental variability and forest opening that a dietary and ecological specialist could not overcome. In contrast, generalist primates, including Homo and Pongo weidenreichi, displayed behavioural and dietary flexibility in mosaic environments. The multiproxy, genus- and species-specific approach provides strong regional insights into primate resilience and megafaunal fates in Southeast Asia and contextualizes the demise of the largest primate known.

• Fossil record constraints: G. blacki remains are limited to four mandibles and nearly 2,000 isolated teeth, with no postcranial fossils, potentially restricting behavioural and ecological inferences.
• Dating uncertainties: While 157 radiometric ages from six techniques were modelled, the paper notes that detailed limitations and uncertainties are discussed in the Supplementary Information.
• Prior chronological gaps: Historically, few sites were dated with multiple techniques, and earlier timelines were site-specific; this study addresses but cannot entirely eliminate such biases.
• Proxy sample sizes: DMTA analyses involved limited samples (G. blacki n=16; P. weidenreichi n=22), which may affect the statistical power and generalizability of microwear-based dietary inferences.
• Cave-based record: Reliance on cave deposits may introduce taphonomic and depositional biases that could affect regional representativeness.