Dinâmica de carbono nas diferentes fitofisionomias de Cerrado em Minas Gerais

Abstract

Carbon storage is considered one of the most important ecosystem services provided by vegetation. Studies that evaluate different methods of carbon storage in the Cerrado domain can be useful for understanding the potential for carbon storage in the different vegetation type of the domain. Namely, the Cerrado is considered an important carbon sink, given its high potential for storing carbon dioxide (CO2). Thus, this study aimed to: (I) quantify the carbon storage present in the Cerradão, Campo Cerrado and Cerrado sensu stricto vegetation type, in the state of Minas Gerais, in the years 2005 and 2010; (II) estimate the difference in carbon storage between measurements (carbon delta); (III) compare the different carbon estimation equations; (IV) model the carbon stock in the Cerrado domain according to climate variables and vegetation dynamics. The data used for this analysis come from the Forest Inventory of Minas Gerais, and correspond to 25 Cerrado fragments, which were inventoried in 2005 and 2010. In the inventoried plots, all individuals with DBH ≥ 5 cm and the total height of all individuals that met the inclusion criteria were measured. The carbon stock estimate was performed per plot, based on four equations found in the literature, adjusted for each vegetation type of the domain. Based on the carbon estimates for each year of measurement, the carbon delta was calculated and, after the estimates, the Wilcox test and the Kolmogorov-Smirnov (K-S) normality test were performed. Among the areas studied, it is worth noting that the Cerrado vegetation type presented the greatest potential for carbon storage when compared to the other vegetation type, presenting an average gain of 4.31 Mg/ha. Of the equations used, the equation proposed by Scolforo et al. (2008) was considered the most accurate for estimating carbon in the Cerrado of the study region considered. After estimating the carbon delta, its modeling was performed according to climatic variables and vegetation dynamics rates using the Random Forest (RF). The model adjusted by the RF presented good performance, highlighting the variables that most influenced the variation of carbon stock in the Cerrado, these being the following variables: precipitation, rate of net change in basal area, turnover in basal area, rate of net change in number of individuals, and minimum temperature. These results demonstrate the importance of the Cerrado in storing atmospheric carbon and its relevance in mitigating climate change.