# Ecological Impacts of Mining in the Amazon: Thematic Trends and Research Gaps

## TL;DR

This paper reviews 30 years of research on mining's ecological impacts in the Amazon, identifying key themes and significant gaps in scientific understanding.

## Contribution

The study introduces a systematic synthesis using LDA to map thematic trends and research gaps in mining-related ecological literature in the Amazon.

## Key findings

- Mercury bioaccumulation and human exposure are the most studied topics, reflecting toxicological concerns.
- Environmental monitoring and landscape changes receive limited attention, indicating research gaps.
- Semantic analysis reveals weak connections between monitoring indicators and biodiversity or toxicological studies.

## Abstract

Mining activities have expanded rapidly in the Amazon, generating ecological, social, and health concerns. Although the number of publications addressing mining impacts has grown, studies often lack integrative assessments. Here, we provide a systematic synthesis of how the ecological impacts of mining in the Amazon have been framed in the scientific literature over the last 30 years (1995–2025). We analysed 462 peer-reviewed articles and applied Latent Dirichlet Allocation (LDA) to identify the main themes discussed in the literature and the research gaps that structure this field. Twelve topics emerged and were grouped into five categories: Pollution and Toxicology, Land Use, Social–Ecological Systems, Biodiversity and Ecosystem Health, and Monitoring and Assessment. The most prevalent topics were mercury bioaccumulation, human mercury exposure, and post-mining restoration, highlighting persistent concerns with toxicological pathways and ecosystem recovery. In contrast, environmental monitoring & impact indicators, disease hotspots, and landscape changes due to resource extraction were the least represented, revealing limited attention to broader-scale ecological processes and early-warning indicators. Semantic similarity analysis showed close relationships among topics linked to contaminant pathways and human health, as well as between landscape alteration and stream biodiversity. The co-occurrence–based dissimilarity analysis revealed weak connections among several topic pairs and highlighted substantial gaps, particularly the infrequent association between monitoring indicators and either toxicological or biodiversity-focused studies. Together, these findings reveal structural imbalances in the scientific agenda. Bridging ecological, toxicological, and socio-political perspectives is essential to support evidence-based responses and safeguard both biodiversity and human well-being in the Amazon.