Welcome to the machine: A Pan-Continental overview machine learning applications in ecology and conservation
DOI:
https://doi.org/10.17161/bi.v19i.23622Abstract
Machine-learning emerged as an excellent alternative to understanding ecological patterns and processes at different spatiotemporal scales. Our study aimed to offer a pictorial overview of the status quo on the use of machine-learning in ecology and conservation globally. Using keywords in the Scopus engine, we indexed all publications in ecology and conservation using machine-learning. We employed descriptive statistics and regressions models to provide an overview and predict geopolitical patterns. The majority of manuscripts were condensed in economically affluent countries, such as the United States (USA) and China (CHN) which together amount to 91 (36.8%) studies. There is a spatial aggregation in the authors’ affiliations, once 182 (73.7%) studies derived from both Nearctic and Palearctic teams, whereas Tropical teams published 65 (26.3%) manuscripts and the most-cited papers also are concentrated in northern regions. In ecology and conservation, machine-learning first appear in the literature in 2003. Yet, increased exponentially since the 2010s. In 2010, this overview indicated nine manuscripts, whereas 10-yrs later reached 120 publications. Most studies (N = 173; 70.1%) are focused on landscape and vertebrate ecology. The primary aims of the publications were widely variable but strongly adherent to providing the best-information on both landscape-scale classifications and species distribution modeling. The manuscripts encompass different methods, from maximum entropy to boosted regression trees and random forest, sometimes using a gamma of deep-learning architectures. Finally, the predictive variables (i.e., mammal diversity and per capita GDP) do not exert significant influences on the number of studies published.
Metrics
Downloads
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Juliano Bogoni, Derick Souza Campos, Claumir Muniz, Manoel Santos-Filho, Jéssica Poletto

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Copyright for articles published in this journal is retained by the authors, with first publication rights granted to the journal. All articles are licensed under a Creative Commons Attribution Non-Commercial license.
Competing Interests: The authors have declared that no competing interests exist.