Brazil nut journey under future climate change in Amazon

rainforest and other types of vegetation [22]. The Amazon biome has been altered by defores- tation and forest degradation, affecting 38% of the area [23]. Current and future climate variables datasets We used the bioclimatic variables of WorldClim 2.0 [24] (https://worldclim.org - accessed in 09/19/2023) as predictors with a spatial resolution of 0.0833˚ (~10 km). We selected the follow- ing variables: bio1—Average Annual Temperature, bio12—Annual cumulative precipitation; bio7—Annual temperature amplitude; and bio15—Seasonality of Precipitation. These predic- tors are often used for distribution models because of their relationship to species physiology and water/energy sources. Regarding the simulated climate for the future, the bioclimatic variables mentioned above were downloaded from WorldClim 2.0 from the most recent simulation generated by CMIP6 (Coupled Model Intercomparison Project Phase 6). We used four distinct time periods until the end of the century: 2021–2040; 2041–2060; 2061–2080 and 2081–2100. To minimize the effect of uncertainty associated with the different Global Circulation Models (GCMs) [25], we used an average ensemble of five GCMs (BCC-CSM2-MR, CNRM-CM6-1, CNRM-ESM2-1, MIROC6, MIROC-ES2L). Such GCMs were selected considering the values of the Equilibrium Climate Sensibility (ECS) and Transition Climate Response (TCR) parameters [26]. Anthropogenic Greenhouse Gases (GHG) emissions are mainly driven by population size, economic activity, lifestyle, energy use, land use patterns, technology, and climate policy. In

this sense, the Intergovernmental Panel for Climate Change (IPCC) conceptualized the SSPs— Shared Socioeconomic Pathways (SSPs) based on CMIP6 climate simulations [27]. We have selected four (04) simulated climate scenarios available on the WorldClim 2.0 (https:// worldclim.org/data/) [28]. Among the scenarios used in this study, the most optimistic is SSP1-2.6, which points to a sustainable future where there are low challenges to mitigating and adapting to climate change [29]. We also use a scenario called Middle of the Road (SSP2-4.5), where the challenges of adapting to and mitigating climate change can be considered intermediate. Finally, we use two pessimistic scenarios: SSP3-7.0 (Regional Rivalry) and SSP5-8.5 (Fossil-fueled Development). The difference between them concerns the great challenges for adapting to climate change, where SSP3-7.0 points to high challenges in promoting adaptation. In other hand, SSP5-8.5, despite the economy and use of natural resources being based on the intensive use of fossil fuels, there would be more conditions for the generation and sharing of new adaptive technologies. Occurrence database To prospect the species distribution of Bertholletia excelsa Bonpl., in the Amazon biome, we used geographical occurrences from the Global Biodiversity Information Facility (GBIF) (https://www.gbif.org/species/3083180; accessed in 16/05/2024), which is an international net- work and research infrastructure funded by the world’s governments and aimed at providing open access to data about much life on Earth. To download the occurrences, we used the

’dismo’ package [30]. After downloading the occurrences, we performed coordinate accuracy testing and duplicate exclusion with the ’biogeo’ package [31]. We compiled 255 unique occur- rence records for Bertholletia excelsa, covering the entire area of study. Distribution model for B. excelsa To build the species distribution model for B. excelsa, we used the package ’biomod2’ [32] implemented in the R software [33]. We built a consensus model based on an ensemble of PLOS ONE Effects of climate change on Brazil nut PLOS ONE | https://doi.org/10.1371/journal.pone.0312308 November 13, 2024 3 / 13