Tree dominance and diversity in Minas Gerais, Brazil

Quantifying diversity is an old challenge for ecologists and is also a social demand given the increasing threats to natural areas. We sought to work on these issues by using data from 158 vegetation remnants (over 350,000 trees) in southeastern Brazil (Minas Gerais State-MG, nearly 600,000 km 2). Specifically, we sought to answer the following questions. (1) How many trees and tree species currently exist in MG vegetation remnants? (2) How much of such biodiversity is present in each vegetation domain (Atlantic Forest, Cerrado and Caatinga) and vegetation type (Seasonally Dry tropical forest "semideciduous'', Seasonally Dry tropical forest "deciduous'', Rain Forest, Swamp, Cerrado and Cerradao) in MG? (3) How much has been lost in regards to tree amount and tree species? We built a 0.1-degree cell grid to estimate the number of trees via spatial regression and used Fisher's alpha and Fisher's log series to provide estimates on how many tree species there are and were in MG. We found the number of trees in Minas Gerais to be approximately 24.5 x 10(9), and the number of tree species to be between 3592 and 3743. The most abundant species distribution among the vegetation domains and vegetation types followed inversely the environmental heterogeneity of the classes. Consequently, the most abundant species in MG belonged to the Cerrado domain, where there was less environmental heterogeneity. The numbers of trees and tree species lost were estimated at 68.54 and 4.03-8.42% of the original values, respectively. We discuss that due to the consequences of human impacts that go back in the region for over a thousand years, other processes not considered in this study, such as habitat degradation by isolation, alterations of food webs, unsustainable use, and climate change, might have caused local extinctions and potentially increased the number of species lost. We believe our results may guide conservation initiatives by providing a base for future environmental laws, parks planning, and the development of more appropriate vegetation management techniques in MG. In addition, our results may inspire future quantitative ecological studies in the tropics.