Australopithecus afarensis skull

Australopithecus afarensis skull (© Iliya Mitskavets - stock.adobe.com)

In A Nutshell

  • Behavior led evolution: Early primates began eating grasses hundreds of thousands of years before their teeth adapted to handle them, supporting the theory of “behavioral drive.”
  • Theropithecus monkeys show a 900,000-year lag between adopting grass-heavy diets and developing longer molars to process them.
  • Early Homo shifted away from grasses around 2.3 million years ago, likely turning to underground tubers—a higher-calorie food source requiring more cognitive effort to harvest.
  • This dietary pivot coincided with rapid brain expansion and smaller molars, suggesting energy-rich foods and tool use helped fuel human evolution.

HANOVER, N.H. — Scientists have discovered something extraordinary about human ancestors: they were munching on tough, fibrous grasses for hundreds of thousands of years before their teeth evolved to handle the job properly. This discovery overturns a long-held assumption about evolution and shows that behavioral changes, not physical adaptations, may have been the primary driving force behind human development.

A new study published in Science shows that early human relatives and other primates began consuming graminoid plants (grasses and sedges) as far back as 4 million years ago. But in some species, notably grass-eating monkeys, it took nearly 900,000 years before their molars adapted to the dietary shift. This substantial delay supports a concept called “behavioral drive,” where behavioral innovation leads the way, and physical changes follow only much later.

For early Homo, the evolutionary story takes a different but equally compelling turn. Around 2.3 million years ago, our ancestors made a major dietary pivot away from grass-based foods and toward underground plant parts. It’s a shift that may have provided the energy needed to support bigger brains and more complex tools.

How Scientists Decoded Ancient Diets Using Fossil Chemistry

Lead researcher Luke Fannin and his team from Dartmouth College analyzed chemical signatures preserved in fossilized teeth from various primate species across Africa. These isotopic signatures act like ancient dietary fingerprints, revealing what these species were eating over the last five million years.

Carbon isotope values help distinguish between plants that use different photosynthetic pathways (C3 vs. C4), while oxygen isotope values provide clues about the types of water in the foods consumed, whether from evaporated above-ground leaves or underground sources like tubers.

In total, the team examined 457 fossil specimens, including 176 from early human ancestors. They focused on three groups that developed grass-eating habits independently: colobus monkeys (Cercopithecoides), baboon-like monkeys (Theropithecus), and early hominins, including australopiths and members of the Homo genus.

Eating grass poses serious challenges for primates. Unlike zebras or antelopes, primates lack specialized digestive tracts and continuously growing teeth. Grass leaves are not only low in energy but also packed with abrasive particles like silica, which can rapidly wear down tooth enamel.

The researchers report that hominin teeth, especially molars, became smaller and longer over millennia to accommodate a growing diet of tough grass-like plants known as graminoids and their underground storage organs.
The researchers report that hominin teeth, especially molars, became smaller and longer over millennia to accommodate a growing diet of tough grass-like plants known as graminoids and their underground storage organs. They found that the turn toward grasses began about 3.8 million years ago with the distant human relative Australopithecus afarensis (left). About 2.3 million years ago, the early human Homo rudolfensis (center) gained regular access to carbohydrate-rich underground plant organs such as tubers, bulbs, and corms. But this dietary shift toward outpaced tooth evolution until about 2 million years ago when species such as Homo ergaster (right) exhibited a spurt of change in tooth size and shape better suited to eating and breaking down cooked plant tissues to derive their nutrients. (Credit: L to R: Public domain; Don Hitchcock; Fernando Losada Rodríguez (rotated))

The 900,000-Year Gap Between Behavior and Biology

By plotting dietary preferences against molar tooth length over time, the researchers uncovered striking lags between behavior and physical adaptation. In Theropithecus monkeys, for instance, a clear preference for grasses emerged around 4.2 million years ago, but the lengthening of their molars didn’t begin in earnest until 3.3 million years ago. That’s a 900,000-year gap between adopting a grass-based diet and evolving the teeth to handle it.

Similar patterns were observed in other mammals. Ancient pigs and elephants both showed delayed dental changes following dietary shifts, reinforcing the broader principle that behavioral innovation often leads morphological evolution.

Among early hominins, dietary patterns became more complicated. While some species like Paranthropus boisei continued to consume large amounts of above-ground C4 plants, others in the Homo lineage began to shift course dramatically around 2.3 million years ago.

C3 vs C4 Plants: What’s the Difference?

C3 and C4 plants use different photosynthetic pathways, which leave distinct chemical fingerprints that scientists can detect millions of years later.

C3 Plants:

  • Most trees, shrubs, and cool-season grasses
  • Dominate in cooler, wetter environments
  • Include forests and woodland vegetation
  • Less efficient in hot, dry conditions

C4 Plants:

  • Tropical grasses and sedges
  • Dominate in hot, dry savanna environments
  • More efficient at photosynthesis in high temperatures
  • Include the tough grasses that early human ancestors began eating

Why This Matters for the Study: The two plant types incorporate carbon differently during photosynthesis, creating distinct carbon isotope ratios. When animals eat these plants, the chemical signatures get preserved in their tooth enamel. This allows scientists to determine whether ancient primates were eating forest plants (C3) or savanna grasses (C4) by analyzing fossilized teeth.

Underground Tubers: The Key to Human Brain Evolution

Using oxygen isotope data, the researchers detected a notable change in the types of water early Homo species were ingesting, moving from water evaporated from grass leaves to that found in underground storage organs like tubers, bulbs, and corms.

Unlike above-ground grasses, these underground plant parts offer more concentrated energy and are available year-round. They also require new foraging strategies: locating them, digging them up, and preparing them, all of which may have promoted tool use and cognitive development.

This dietary transition coincided with two major anatomical changes in the Homo lineage. Brain size began expanding rapidly around 2.1 million years ago, and molars started to shrink approximately 200,000 years later, a sign that early humans were eating foods that required less chewing. While some scholars attribute these changes to increased meat consumption, the study authors note there’s little zooarchaeological evidence of intensified carnivory during this period.

Instead, they argue that a switch to underground tubers better explains the combination of isotope data, brain expansion, and tooth reduction. As the authors note, “This distinction raises the possibility of technical disparities between Paranthropus and early Homo, and it is tempting to argue, as others have using different lines of evidence, that tubers fueled the emergence of Homo erectus.”

Why This Discovery Rewrites Human Evolution

Digging for tubers requires more than raw strength; it demands tools, memory, and ecological knowledge. That behavioral flexibility may have been a catalyst for many defining traits of our species: bigger brains, better tools, and greater adaptability.

Modern hunter-gatherer societies that rely on underground plant foods demonstrate the kind of skillsets that would have been advantageous for early Homo: recognizing plant cues, understanding seasonal cycles, and planning for future meals. These cognitive traits are hard to see in the fossil record—but they leave traces in behavior and diet.

Rather than viewing behavior as merely following physical evolution, the study shows that behavior often leads the way. “Dietary shifts and corresponding morphological changes can sometimes evolve in succession, not concurrently—an evolutionary process called behavioral drive,” the researchers write.

In the story of human evolution, grass-eating primates who eventually unearthed calorie-dense roots paved the way for everything that came next. Through risk-taking, experimentation, and flexibility, our ancestors set evolution in motion before their bodies had even caught up.


Paper Summary

Methodology

Researchers analyzed carbon and oxygen isotope ratios from 457 fossil teeth across Africa, representing multiple primate lineages over five million years. Carbon isotopes (δ¹³C) indicated reliance on C3 versus C4 plants, while oxygen isotopes (δ¹⁸O) revealed water sources tied to above-ground or underground plant tissues. Jacob’s D, a dietary selectivity index, was used to determine preference for C4 plants relative to availability. Molar length data were used to detect morphological shifts and timed against behavioral changes using regression models and phylogenetic rate analysis.

Results

The study identified a ~900,000-year gap between grass consumption and molar adaptation in Theropithecus. Among early Homo, researchers observed a dietary shift away from C4 plants around 2.3 million years ago, based on declining δ¹³C values and increased consumption of ¹⁸O-depleted water—pointing to underground tubers. This shift preceded molar reduction by about 700,000 years and coincided with brain expansion. Stem lineages showed 29% greater dietary variation than their descendants, supporting the “flexible stem hypothesis” for behavioral innovation.

Limitations

Isotope analyses cannot identify specific foods—only general plant types and water sources. Temporal resolution is limited by fossil availability, and inferred behavioral patterns are interpreted indirectly. Findings are geographically concentrated in East Africa, and sample sizes vary by time period and species. Evolutionary rate analyses depend on assumed phylogenies, which carry inherent uncertainties.

Funding and Disclosures

The study was funded by the David and Lucile Packard Foundation, the Institute of Human Origins, and the National Science Foundation’s Graduate Research Fellowship and Dissertation Grant programs. The authors declared no competing interests. Data and code are available via Zenodo.

Publication Details

Title: Behavior drives morphological change during human evolution
Authors: Luke D. Fannin, Chalachew M. Seyoum, Vivek V. Venkataraman, et al.
Journal: Science, Volume 488, July 31, 2025
DOI: 10.1126/science.ado2359

About StudyFinds Analysis

Called "brilliant," "fantastic," and "spot on" by scientists and researchers, our acclaimed StudyFinds Analysis articles are created using an exclusive AI-based model with complete human oversight by the StudyFinds Editorial Team. For these articles, we use an unparalleled LLM process across multiple systems to analyze entire journal papers, extract data, and create accurate, accessible content. Our writing and editing team proofreads and polishes each and every article before publishing. With recent studies showing that artificial intelligence can interpret scientific research as well as (or even better) than field experts and specialists, StudyFinds was among the earliest to adopt and test this technology before approving its widespread use on our site. We stand by our practice and continuously update our processes to ensure the very highest level of accuracy. Read our AI Policy (link below) for more information.

Our Editorial Process

StudyFinds publishes digestible, agenda-free, transparent research summaries that are intended to inform the reader as well as stir civil, educated debate. We do not agree nor disagree with any of the studies we post, rather, we encourage our readers to debate the veracity of the findings themselves. All articles published on StudyFinds are vetted by our editors prior to publication and include links back to the source or corresponding journal article, if possible.

Our Editorial Team

Steve Fink

Editor-in-Chief

John Anderer

Associate Editor

Leave a Reply