When Humans Began Writing the Forest: Neanderthals, Mesolithic Hunters and the Making of European Landscapes

The forest that was never truly “wild”

A persistent image in ecological thought and public imagination casts the Pleistocene and the early Holocene as a pristine Europe: continuous dense forests, untouched soils, and roaming megafauna setting the tempo of nature. A recent multinational study, coordinated in part by researchers at Aarhus University, challenges this narrative. By combining continental-scale pollen records with coupled climate and ecological simulations and an AI-assisted optimization algorithm, the team concludes that both Neanderthals and Mesolithic hunter-gatherers shaped vegetation structure long before the advent of agriculture.

Fire, spears and simulations: a method reshaping the narrative

The paper’s methodological contribution is twofold. First, it assembles an extensive palynological dataset covering two warm intervals: the Last Interglacial (roughly 125,000–116,000 years BP) and the Early Holocene (approximately 12,000–8,000 years BP). Second, it deploys spatially explicit models that integrate climatic variables, the dynamics of large herbivores, natural fire regimes and, crucially, parameterizations for human activities such as deliberate burning and subsistence hunting or disturbance.

Using an AI algorithm to run and compare thousands of scenarios, the team identified combinations of drivers that best reproduce continental pollen signals. Models that exclude human influence fail to capture specific patterns of vegetation opening and shifts in dominant plant types. Incorporating scenarios with controlled fires and localized reductions in megafaunal populations substantially improves fit with empirical data, indicating a measurable anthropogenic imprint on landscape structure well before farming.

Invisible cascades: how hunting reshapes the forest

A central insight of the study is the significance of indirect effects arising from hunting. Reductions in the density of megafauna—elephants, rhinoceroses, aurochs and bison—alter grazing and browsing pressure, which in turn influence understory development and canopy closure. In ecological terms, these changes affect landscape “openness” and the spatial distribution of plant communities.

The published figures are notable: Mesolithic activities could explain up to 47% of the observed distribution of plant types, while the Neanderthal signal, smaller but detectable, accounts for roughly 6% of plant distribution variation and about 14% of changes in landscape openness. These percentages denote persistent, spatially coherent phenomena across millennia and extensive regions, not isolated local disturbances.

“The study paints a new picture of the past,” says Jens-Christian Svenning. “The Neanderthals did not hold back from hunting and killing even giant elephants.”

Neanderthals versus Homo sapiens: two tempos of transformation

The contrast between Neanderthals and Mesolithic Homo sapiens is not merely quantitative but also temporal and spatial. Neanderthal populations tended to be smaller and more dispersed, producing pronounced but geographically limited effects. Mesolithic hunter-gatherers, with different demographic densities and social strategies, produced broader and more enduring landscape modifications. The authors argue that in some regions Homo sapiens progressively degraded the ecological role of megafauna—not necessarily through instantaneous extermination, but through sustained population erosion that precipitated cascading vegetation changes.

This reframing has implications beyond paleontology: it challenges the notion of pre-agricultural humans as passive consumers and repositions them as active ecological agents who used fire, created transient pathways and applied hunting pressure in ways that altered other habitat components. The label of an “unaltered nature” therefore becomes epistemologically and politically fraught.

What remains uncertain—and why it matters

Interpreting these simulations requires caution. Large-scale ecological modeling depends on assumptions regarding human population densities, the frequency and spatial patterning of anthropogenic burning, life-history parameters of megafauna, and the temporal resolution of pollen archives. Pollen signals are mediated by processes of transport, deposition and preservation; taphonomic biases can favour certain taxa, and time-averaging may obscure short-lived but intense events.

Other drivers are incompletely quantified: disease outbreaks affecting megafauna, abrupt climatic perturbations, or intrinsic changes in animal behaviour unrelated to human predation could produce similar outcomes. The AI-driven framework produces probabilistic inferences, not definitive proofs; results hinge on the initial variable set and the weighting of interactions within the model.

Nevertheless, even acknowledging these constraints, the conclusion that humans acted as “co-creators” of European landscapes constitutes a conceptual shift. If prehistoric peoples were continuous ecological agents, the benchmark of a “primordial nature” used in conservation and restoration must be reconsidered.

From science to policy: practical consequences

This reinterpretation has direct ramifications for rewilding, conservation and restoration initiatives. Policymakers and practitioners must decide what kind of forest or landscape they aim to re-establish: one shaped periodically by human presence and intervention or an imagined, people-free state that may never have existed at large scales. The findings can be marshalled to support active management—prescribed burns, targeted herbivore control or strategic reintroductions—but they also complicate narratives that frame anthropization as solely a modern aberration.

There is a real risk of politicization. Economic actors might invoke deep-time anthropogenic signatures to legitimate ongoing exploitation or to downplay contemporary biodiversity losses. Clear, nuanced science communication is therefore paramount: empirical history should inform policy design, not provide rhetorical cover for unsustainable practices.

A call for regional nuance

The authors emphasize that continental-scale models offer broad maps, not local prescriptions. Europe’s heterogeneity—its islands, peninsulas, plains and mountains—reflects widely varying human demographies and faunal communities. The logical next step is the integration of high-resolution archaeology, paleogenomics and geoarchaeological fieldwork to map not just where humans altered landscapes but how those processes operated in particular ecological and cultural settings.

The Warhial Perspective

This study effects a durable conceptual rupture: the notion of an untouched European wilderness is increasingly untenable. The point is not to indict or absolve, but to recognize an empirical reality—people were ecological forces long before agriculture. From a journalistic and policy-oriented perspective, this recalibration ought to be an instrument of governance rather than an excuse for regression. My forecast: over the next five to fifteen years we will see this conceptual transfer reflected in conservation practice—more integrated active management (prescribed burning, herbivore controls, strategic reintroductions) coupled with intensified ethical debate over who decides how landscapes are “restored”.

Two caveats remain. First, science does not license the normalization of every human intervention: not all transformations are sustainable or ethically acceptable. Second, there is a danger of oversimplification in public discourse: the slogan “humans made everything” can be misused to dilute contemporary responsibilities for biodiversity stewardship. The Warhial view anticipates a future in which conservation must weave paleoecological knowledge with the rights and knowledge of local communities and an adaptive ecological ethic: not a quest to return to an imagined primordial state, but the design of viable, resilient and scientifically justified landscape futures.