Projeções futuras de índice de vegetação e os efeitos das mudanças climáticas em regiões áridas e semiáridas

Abstract

Uncertainties with the effects of climate change has increased the need for more studies in recent years. Some regions in the world are more vulnerable, such as arid and semiarid ecosystems whose future projections indicate an increase in the frequency of extreme climate events. Vegetation cover has a fundamental role in maintaining these environments for biodiversity and population. Generalize about the response to climate change of endemic species from arid and semiarid environments to climate change is difficult to make due to information gaps. In the first part, we generated climatic niche models of the vertebrate species of these ecosystems using the results in meta-analysis statistics to evaluate if there was any tendency. The 166 species used totaled 1660 results due to the combination of climate models and methods. Of these results, 37% indicated loss of favorable climate areas, 40.92% were inconclusive and 22.08% indicated a gain in favorable climate areas. The heterogeneity of the results as well as the inconclusive results indicates the influence of different methods used in niche modeling. Even so, there is a tendency at global level to reduce favorable climate areas for endemic vertebrates. In the second part, we considered the Normalized Difference Vegetation Index (NDVI) a proxy of the vegetal cover. We used the ARIMA time series model to design this variable for the future based on NDVI data between 2001 and 2017. The NDVI data are available every 16 days at 250m resolution. We converted into monthly data and 0.1° degree resolution. We used the Brazilian semiarid as model to project NDVI to the present, validate the results and subsequently to projected for the future (2050 and 2070). We obtained good results with of 0.7552 between the estimated and the observed. Pearson correlation values also varied between months, with January being the highest value (0.8887). The root mean square deviation (RMSE) values were also low, indicating good performance of the models. The observed differences between the months, lowest values of correlation and errors can be related to the presence of climatic anomalies as well as the effects of land use changes. The good performance indicates a great potential of use of the NDVI as proxy of vegetation cover, thus contributing in climate change research. The third part, we evaluated niche models of endemic vertebrates of the Brazilian semi-arid with the use of NDVI integrated with climatic variables. The five species studied lost favorable climatic areas with or without the use of NDVI. However, the frequency of the number of areas changes with the use of NDVI can indicate refuges for some species. Vegetation cover should be considered in biodiversity research investigating climate change. There is great potential for research into drought periods. These studies can provide important information both for the effects of biodiversity and population that depends on these natural resources for survival.