Ecological Legacies of Pre-Columbian Settlements Evident in Palm Clusters of Neotropical Mountain Forests

TL;DR

This study uses deep learning and remote sensing to reveal how pre-Columbian human activity influenced palm distributions in Colombian forests, suggesting ancient settlements had a larger ecological footprint than previously documented.

Contribution

The paper introduces a scalable deep learning approach to detect ancient human impact on forests through palm distribution patterns using high-resolution satellite imagery.

Findings

Palms are more abundant near archaeological infrastructure.

Largest palm clusters suggest ancient human activity affected areas up to 100 times larger.

Palm distributions can help predict archaeological site locations.

Abstract

Ancient populations inhabited and transformed neotropical forests, yet the spatial extent of their ecological influence remains underexplored at high resolution. Here we present a deep learning and remote sensing based approach to estimate areas of pre-Columbian forest modification based on modern vegetation. We apply this method to high-resolution satellite imagery from the Sierra Nevada de Santa Marta, Colombia, as a demonstration of a scalable approach, to evaluate palm tree distributions in relation to archaeological infrastructure. Our findings document a non-random spatial association between archaeological infrastructure and contemporary palm concentrations. Palms were significantly more abundant near archaeological sites with large infrastructure investment. The extent of the largest palm cluster indicates that ancient human-managed areas linked to major infrastructure sites may be up to two orders of magnitude bigger than indicated by current archaeological evidence alone. These patterns are consistent with the hypothesis that past human activity may have influenced local palm abundance and potentially reduced the logistical costs of establishing infrastructure-heavy settlements in less accessible locations. More broadly, our results highlight the utility of palm landscape distributions as an interpretable signal within environmental and multispectral datasets for constraining predictive models of archaeological site locations.