Modelagem de nicho ecológico para a conservação de espécies florestais diante dos cenários de mudança climática na Amazônia

Abstract

The Amazon phytogeographic domain is considered the largest tropical forest in the world with recognized importance for climate regulation. Despite its importance, the Amazon forest has been the target of intense deforestation involving natural ecosystems and resulting in the loss of biodiversity. Forest species and those considered commercially important timber are targets of illegal deforestation activities on the Amazon plus, they are vulnerable to the effects of climate change. Studying their behavior in the face of climate change will help to understand the species' ability to perpetuate. The aim of this research was to develop a consensus model to simulate the potential distribution of six Amazonian species Cupiúba (Goupia glabra Aubl), Mandioqueira (Qualea paraenses Ducke), Cardeiro (Scleronema micranthum (Ducke) Ducke), Louro-vermelho (Ocotea rubra Mez.), Faieira (Roupala montana Aubl.) e Cedroarana (Cedrelinga cateniformis (Ducke) Ducke), to design and map their potential area of environmental suitability through ecological niche modeling based on environmental variables aiming at the conservation and use of their genetic resources, in addition to applying the stable isotope technique for the purpose of traceability of the wood of the tree species G. glabra. The georeferenced occurrence points were obtained from GBIF, SpeciesLink and BIEN databases. The bioclimate environmental variables were obtained from Worldclim and the edaphic variables were obtained from the Harmonized World Soil Database. Data from future scenarios were projected for the periods 2021 – 2040 and 2041-2060 based on the sixth IPPCC report. Thirteen algorithms available in the ENMTML package were used: BIO, MAH, DOM, ENF, GLM, GAM, BRT, RDF, SVM, MXS; MXD, MLK and GAU, and analyzed in the R software. Three metrics AUC, TSS and Sorensen were used to evaluate the performance of each algorithm. The composition of the consensus model was obtained for the five models that presented performance values > 0.7. The evaluation of statistical metrics showed that the RDF, SVM, BRT, GAU and MXD algorithms obtained the best performances. The estimated niche in the Brazilian Amazon for the six species predicts that climate change will cause a large loss of area of climatic suitability for all of them in the future scenarios analyzed. Notorious losses were observed for the species S. micranthum, while R. montana showed smaller losses of areas with environmental suitability, under the evaluated conditions. Temperature will have a great influence on the suitability area. The study suggests that public policies be aimed at conserving timber forest species and proposing measures to mitigate climate impacts. In order to build the isotope database, 53 G. glabra trees were sampled in six different locations in the Amazon. Geographic coordination data were obtained at each collection. Analysis of the de 13C and 15N 9 isotopes were carried out at the Stable Isotopes Forensic Laboratory of the Federal Police of Amazonas. The data was organized in Microsoft Excel spreadsheets. Graphs and statistical analysis were performed using R software (version 4.4.2). The study showed variability in the de 13C and 15N isotopic ratio of G. glabra samples between trees and locations. The δ13C values ranged from -29.30‰ to -25.07‰ and from -1.52‰ to 6.56 for δ15N, considering all locations. The use of stable isotopes in addition to other environmental variables will further improve the identification of the origin of wood species. This method is promising for combating the national and international illegal trade in Amazonian species.