The effect of indigenous tenure on deforestation in the neotropics – results from a data-intensive assessment of Central America and the Amazon

Daniel Schoenig and Christian Messier, Département des Sciences Biologiques, Université du Québec à Montréal, Montréal, QC, Canada, Jérôme Dupras and Christian Messier, Département des Sciences Naturelles, Institut des Sciences de la Forêt Tempérée, Université du Québec en Outaouais, Ripon, QC, Canada

Background/Question/Methods
Over the last decade, the neotropics again experienced a net loss of forest area, despite increased reforestation efforts. Avoiding deforestation in the first place therefore remains crucial to achieve objectives related to biodiversity conservation, and the provision of forest ecosystem services. In this context, indigenous people are of particular importance, as numerous case studies have demonstrated a link between indigenous tenure and reduced deforestation. However, our knowledge is still limited concerning the strength and universality of this relationship at the continental scale.
To close this gap, we conducted a data-intensive analysis of forest loss between 2011 and 2019 for the Amazon (watershed and bioregion) and the Republic of Panama: Using publicly available sets, we sampled 10 million small forest patches for each region, and used generalized additive models combined with self-organizing maps to assess the effect of indigenous territories and protected areas on the probability of forest loss occurring in a given patch. For the Amazon, the available data also allowed us to compare officially recognized indigenous territories against not recognized territories. In all models, we controlled for autocorrelation and covariates that have been documented as proxies for accessibility and remoteness.
Results/Conclusions
We found that indigenous tenure is strongly related to reduced deforestation (relative to non-indigenous and non-protected areas): In the Amazon, the risk of forest loss is reduced by 59.4% under indigenous tenure (95% point-wise confidence interval: 59.0–59.8%), while in protected areas the risk is reduced by 51.7% (51.2–52.2%). For Panama, the reduction amounts to 20.9% (20.0–21.8%) within indigenous territories (comarcas), and 31.3% (30.6–31.9%) inside protected areas. In the Amazon, the deforestation risk is reduced more strongly in officially recognized indigenous territories (61.1%, 60.7–61.6%) than in not recognized territories (51.0%, 49.8–52.2%).
Our results further show that the effect of indigenous tenure on reducing deforestation is: (1) widespread and not confined to individual cases or smaller areas, (2) not confined to particular indigenous groups, (3) not the result of remoteness or inaccessibility of indigenous territories, (4) comparable in magnitude to effects achieved within protected areas. In addition, this effect is reinforced if tenure rights are secure.
We thus argue that indigenous people must have more substantial opportunities to fulfill an active part in, and directly benefit from efforts directed towards reducing deforestation in the neotropics.
We further discuss how our method can be applied in other contexts that require spatial evaluation of conservation targets, such as forest biomass or endangered species abundance.