Biomass in different types of land use in the Brazil's 'arc of deforestation'

This study aims to investigate the relationship of biomass with density and richness and soil characteristics. In particular, we focus on the different types of land use found in areas of family farming in the southern and southeast regions of Para State, Brazil. The study was conducted in three communities located in the Nova Ipixuna, Parauapebas and Pacaja municipalities, Para State, Brazil. In each of the three areas, nine small family farms were chosen; in each lot five collection sites were established, where the structure of the vegetation cover was inventoried. Nine different types of land use were considered: conserved, exploited and burned forests; old and young secondary forests; clean and invaded pastures; crops areas (with cassava and cocoa). At each of the 135 sampling sites, vegetation cover was inventoried in upper, middle and lower strata; soil samples were also collected for physico-chemical characterization. From these data, we calculated the biomass of the different strata and individuals density, species richness and physico-chemical of soil to investigate their relationships with biomass. Total biomass varied greatly in all types of land use and in all strata studied. Conserved and exploited forests were different from those of burned forests. Burned forests were similar to old secondary forests. Young secondary forests were similar to invaded pastures and cocoa plantations. In principal component analyses (PCA) of biomass three groups of land use were distinguished: one group was explained by important biomass from the upper stratum; a second group was explained by the absence of biomass in the upper stratum; and a third group was explained by biomass found in the middle stratum. This result was compared with the results of the PCAs of density and richness matrix per stratum and physico-chemical properties of soil matrix through a co-inertia analysis. The first co-inertia analysis showed that the two data matrices (density and richness structure and biomass) share a common structure, with 77% of the variability being explained. The second co-inertia analysis also showed that the two data matrices (soil and biomass) do share a common structure. Thus, biomass was strongly correlated with density and richness and with the edaphic characteristics analyzed. The different types of land use in areas of family farms can be categorized into three simple groups according to biomass; this categorization can serve as a reasonable indicator for the rapid estimation of stored biomass. (C) 2012 Elsevier I3.V. All rights reserved.