The approach rests on a deceptively simple premise: time is the principal constraint on animal survival. For diurnal species, the day is finite, and four activities — feeding, travelling, enforced rest and social grooming — account for more than 95 percent of a primate’s waking hours. Physiological processes are rate-limited, so if an animal cannot ingest sufficient nutrients within a set period, it starves. Whatever time remains after feeding, moving and resting is available for social bonding, which in primates means grooming. Because grooming time scales predictably with group size, the models can calculate the maximum ecologically tolerable group size, Nmax, that a species could maintain at any given location. If Nmax falls below the minimum group size needed to avoid predation, Nmin, the species simply cannot survive there.

These time budget models have already been built for baboons, chimpanzees, gorillas, colobus monkeys, guenons, gibbons, spider monkeys and even feral goats, and they predict the continental distributions of these genera more accurately than conventional climate envelope models. Crucially, they also explain why a species is absent from a habitat: whether the bottleneck lies in feeding time, travel time or rest time. Bettridge and Dunbar exploited this machinery in reverse. They took the equations for baboons — an ecological analogue sharing similar open environments — and for chimpanzees — a phylogenetic analogue by ancestry — and asked whether australopithecines could have survived at the 34 fossil sites where their remains are known, had they behaved like either genus.

The answer was a resounding no. Adjusted for the differences in metabolic body mass and stride length, the baboon and chimpanzee models predicted that early hominins could have maintained viable groups at only three of the 40 site-date combinations tested. At many sites, the models returned negative group sizes — a mathematical signal that the four components of the daily time budget added together exceeded 100 percent of the available daylight, leaving the animals hopelessly overtaxed. Neither referential model could explain the known australopithecine distribution, confirming long-standing criticisms of the analogical approach: these hominins occupied habitats that no contemporary primate touches, and their social organisation was a flexible response to conditions rather than a fixed species trait.

The reverse engineering proceeded in three steps. First, the researchers adjusted the slope parameters of individual equations. When that failed to improve the fit, they switched equations between the two models — pairing the chimpanzee feeding equation with the baboon travel equation and vice versa. Finally, they added correction factors for bipedalism. The winning combination proved to be a hybrid: a chimpanzee feeding equation, reflecting the apes’ frugivorous ancestry, combined with a baboon moving equation, reflecting adaptation to more open terrain, both corrected for australopithecine body mass of roughly 48.8 kilograms and for the energetic efficiency of walking on two legs. Bipedal australopithecines are estimated to have walked at about 1.0 metres per second, compared with 0.83 metres per second for quadrupedal chimpanzees, and standing upright reduced exposure to direct solar radiation, cutting rest time by around 12 percentage points — the so-called Wheeler discount.

This hybrid model predicted viable groups at every site where australopithecine fossils have been found, with a mean maximum group size of about 50 — remarkably close to the 41.3 individuals predicted independently from brain size via the social brain hypothesis. Without the Wheeler discount, three sites — Swartkrans in South Africa and Kanapoi in East Africa — would have become untenable. Sensitivity analyses confirmed the model’s robustness: perturbing individual parameters by up to 25 percent, or all parameters simultaneously by 10 percent, left the predictions essentially unchanged. Only simultaneous 25 percent errors in every parameter would have produced significantly different results, giving the researchers confidence that modest measurement error does not bias the model.

The social brain hypothesis provided a second, anatomically independent triangulation. In anthropoid primates, group size scales tightly with neocortex ratio, a relationship validated by cross-species comparisons and by roughly two dozen neuroimaging studies in humans and macaques. Because the model’s Nmax values were not consistently smaller than the social brain estimate, australopithecines could have foraged as a single cohesive community, just as baboons do — rather than fragmenting into small foraging parties as chimpanzees and gorillas do. That matters enormously in the open flood plains they favoured, where fossil evidence records the presence of lions, leopards and hyaenas, the three most dangerous predators of living African primates. Without trees offering refuges, foraging in small parties would have been suicidal; the evolution of bipedality, by freeing time otherwise consumed by travel and rest, made large cohesive groups possible.

The models also explain why the australopithecines went extinct. Plotting maximum tolerable group sizes through time reveals a broadly dome-shaped distribution: all three main genera — Australopithecus, Paranthropus and H. habilis — became increasingly well adapted after first appearance, then went into steep decline as the climate deteriorated after 2.4 million years ago. Examining the predicted activity budgets shows that resting time, which accounts for 25 to 35 percent of the day, exhibited by far the largest variation, far more than travel time at 16 to 18 percent or feeding time at 45 to 48 percent. Rising temperatures at the occupied sites — exacerbated by the subsiding Rift Valley floor, which dropped over 1000 metres in altitude, producing a 6.5 degree Celsius rise in mean temperature — would have made it very difficult for fully furred australopithecines to avoid overheating below 1000 metres above sea level. Over the two million years leading up to 1.0 million years ago, the combined rest, travel and feeding budget climbed from 92.4 percent to 102.1 percent of available daylight. At 3 million years ago there was ample spare capacity for grooming, sufficient for groups of roughly 26; by 1 million years ago, when they vanish from the fossil record, the australopithecines had run out of time for the social bonding that holds primate groups together.

Early Homo, by contrast, escaped the trap. Had H. habilis and its relatives been ecological australopithecines, their post-2.0 million-year group sizes would have collapsed to around 25, exposing them to severe predation risk — yet their brain sizes would have supported groups of roughly 75. The models imply that early Homo had already evolved novel adaptations: taller stature and improved skeletal design enabling more efficient striding locomotion, the loss of body fur combined with enhanced evaporative cooling, and possibly some degree of passive scavenging that reduced feeding time demands. Genetic convergence analysis at the melanocortin receptor locus suggests humans lost their ancestral primate body fur not long before 1.2 million years ago, and certainly not earlier than 2.5 million years ago — timing that dovetails neatly with the models’ predictions.

The study’s broader message is a caution against naïve analogical modelling. Baboons, chimpanzees and australopithecines are each constrained by different components of their time budgets — feeding time for baboons, travel time for great apes, rest time for australopithecines — and no single living species can serve as a template for an extinct one. The australopithecines’ unique niche as flood plain specialists, dependent year-round on C4 underground storage organs with a signature closer to grazing antelope than to fruit-eating cercopithecine monkeys, explains why they were confined to a narrow corridor along the Rift Valley rather than spreading as widely as baboons or chimpanzees. By reverse engineering principled equations from living primates, Bettridge and Dunbar have shown that quantitative theory can illuminate the behavioural biology of species with no direct analogues — offering a window into the forces that drove evolution, and extinction, within one of the most consequential lineages in human history.

Subject of Research: Reconstructing early hominin behavioural ecology: a modelling approach

Article Title: Reconstructing early hominin behavioural ecology: a modelling approach

Article References: Bettridge, C., & Dunbar, R. (2026). Reconstructing early hominin behavioural ecology: a modelling approach. Frontiers in Ecology and Evolution, 14, Article 1891539. https://doi.org/10.3389/fevo.2026.1891539

Image Credits: AI Generated

DOI: 10.3389/fevo.2026.1891539

Keywords: Reconstructing, early, hominin, behavioural, ecology, modelling, approach, scientific research