Ten thousand years ago, the Americas teemed with mastodons, giant ground sloths, and saber-toothed cats. Within a few millennia, nearly all of them were gone. A study published in April 2025 in the Proceedings of the National Academy of Sciences now shows that those losses did not just empty the landscape. They permanently rewired the predator-prey networks that hold mammal communities together, and the damage is still visible in the food webs of two continents.
Simpler webs, smaller animals
The PNAS paper, led by ecologists Laura Beaudrot and Evan Fricke along with colleagues including Henry Hsieh, Daijiang Li, and Daniel Gorczynski, reconstructed mammal food webs at 389 tropical and subtropical sites across the Americas, Africa, and Asia. The team drew on trait data, geographic ranges, documented interactions, and machine-learning models covering more than 440 species.
Their central finding is stark: modern American mammal food webs contain fewer predator-prey connections and are dominated by smaller-bodied species compared with those in Africa and Asia, where Late Quaternary megafauna losses were far less severe. Africa, which retained elephants, hippos, lions, and a deep bench of large herbivores, still supports multi-layered networks with many potential interaction pathways. The Americas, stripped of their largest players, do not.
The analytical approach itself was validated in an earlier Science paper by an overlapping team. That study showed trait-based machine learning could reliably reconstruct terrestrial mammal food webs across deep time and documented a broad collapse of those networks since the Late Pleistocene. The PNAS work extends that foundation by quantifying how much structural complexity each continent lost and retained.
Beyond who eats whom
Predator-prey links are only part of the story. A 2013 analysis in Nature Geoscience by Christopher Doughty and colleagues estimated that the extinction of large mammals in Amazonia triggered a greater than 98 percent reduction in the lateral transport of phosphorus across the region. Phosphorus is a limiting nutrient for plant growth in many tropical soils. Animals that once carried it in their guts across hundreds of kilometers effectively fertilized distant forests. When those animals vanished, so did the nutrient pipeline.
That finding, now more than a decade old, remains one of the most concrete illustrations of how ancient extinctions reshaped modern ecosystems through biogeochemical pathways rather than trophic ones. Large mammals acted as mobile nutrient pumps. Where they survived, as in parts of sub-Saharan Africa, those functions persist. Where they disappeared, soil chemistry still reflects the absence.
Why the giants vanished is still debated
The causes of Late Quaternary megafauna extinctions remain actively contested. A synthesis published in Cambridge Prisms: Extinction frames the debate between two main drivers: human hunting pressure and rapid climate shifts at the end of the last ice age. In some regions, extinction pulses align closely with the arrival of human populations. In others, they track abrupt warming or cooling events. Most researchers now suspect some combination of both forces, but the relative weight of each factor varies by continent and time period, and no consensus has emerged.
The timing itself resists tidy summary. Globally, the extinction window stretches from roughly 50,000 to 10,000 years ago. Australia experienced severe losses around 46,000 years ago, well before the end-Pleistocene pulse that hit the Americas. Africa lost comparatively few species. Referring to these extinctions as happening “10,000 years ago” captures the approximate endpoint in the Western Hemisphere but flattens a drawn-out, geographically uneven process.
A 2015 modeling study in Proceedings of the Royal Society B by Mathias Pires and colleagues added a structural dimension to the debate. Using paleontological and ecological data combined with extinction simulations, the team found that megafaunal interaction networks grew more vulnerable after human arrival, suggesting that even modest additional pressures could push already stressed systems past tipping points. The precise mechanisms, whether overhunting of keystone prey, disruption of predator guilds, habitat modification, or cascading indirect effects, remain difficult to disentangle from concurrent climate changes. The models offer strong theoretical support for human-driven destabilization but depend on assumptions about ancient species interactions that the fossil record cannot directly confirm.
How fragile are the networks that remain?
One question the PNAS analysis raises but does not fully answer is how close today’s simplified American food webs are to functional thresholds. The study documents the structural impoverishment clearly. What it cannot yet tell us is how much additional loss these networks can absorb before critical ecological functions break down.
The Americas still harbor tapirs, jaguars, pumas, bears, and peccaries, species that anchor what remains of large-bodied interaction networks. But those species face mounting pressure from habitat loss, hunting, and climate change. Because the redundancy once provided by multiple overlapping large-bodied species is gone, losing even one or two of these remaining players could disproportionately erode network connectivity.
Some ecological functions may be partially compensated by other organisms. Mid-sized mammals and birds have taken over certain seed-dispersal roles once filled by megafauna, though research quantifying the extent of that compensation is still in early stages. The phosphorus-transport deficit documented in Amazonia, by contrast, appears to have no obvious substitute: no living animal in the region moves nutrients at anything close to the scale that giant herbivores once did.
What this means for protecting large mammals now
For conservation planners, the practical implication is pointed. Modern threats are not acting on pristine ecosystems. They are acting on systems already stripped of structural complexity by extinctions that occurred thousands of years before the first European colonists arrived in the Americas. Policies that treat current species assemblages as a natural baseline risk underestimating how much resilience has already been lost.
This framing also sharpens the case for protecting and, where feasible, restoring populations of the largest surviving mammals. The documented roles of megafauna in both food-web architecture and nutrient cycling suggest that large-bodied species deliver ecological functions that smaller animals cannot replicate at the same scale. Rewilding proposals, from reintroducing European bison to experimental projects like Pleistocene Park in Siberia, draw part of their rationale from exactly this logic.
But the uncertainties flagged by the Cambridge Prisms synthesis and the Royal Society B modeling counsel realism. Ancient communities cannot simply be reassembled. Introducing large animals into modern landscapes that have spent millennia adapting to their absence carries ecological risks alongside potential benefits. The strongest takeaway from this body of research is not a prescription for any single intervention. It is a recognition that the food webs we see today in the Americas are not the ones evolution built over millions of years. They are what remained after the giants disappeared, and they are more fragile than they look.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
