Conhecimento indígena, atributos florísticos, estruturais e espectrais como subsídio para inventariar diferentes tipos de Florestas de Campinarana no rio Içana, Alto Rio Negro.

Abstract

Baniwa indigenous knowledge and perception; vegetation composition and structure; and canopy spectral reflectance in the visible and near infrared bands of the Landsat 5 satellite sensor were all used to identify different white sand forests (campinaranas). These three kinds of information were used to distinguish different types of campinarana near the three Baniwa indigenous communities of Aracu, Jandu and Juvitera, along 50 km of the Içana River in the upper Negro River Basin. Ten people were interviewed and 14 campinarana types (Hamáliani in Baniwa) detected. Fifteen quantitative inventories covering several types were conducted, each in an area of 0.2 hectares. Parts of each inventory were spread over an area of about 200 x 200m to facilitate collecting average spectral reflectance from the Landsat image. These 15 inventories covered replicates of six campinarana types, including two subtypes (“high” and “low” canopy stature), making eight distinct types for quantitative analyses and comparisons. In Baniwa, these were: Anerima madoape kaawa (low) e Anerima dzenonipe (high), Waittirima madoape kaawa (low), Waittirima dzenonipe kaawa (high), Maarolina, Heridzolima, Añholima, and Waapalima. An ethno-environmental classification allowed this knowledge to be recorded and organized. Baniwa classes of forest were compared with ordination (MDS, PCA) and cluster analyses of the 15 inventories. The ordinations were conducted separately for structural, floristic and spectral attributes to allow comparison between gradients patterns derived from each of these attribute types. Floristic information was based exclusively on Baniwa identifications in the field. Using quantitative similarity indices, there was a strong floristic gradient that structured the Campinarana forests of the Içana. This was seen in two single-axis MDS ordinations that used density and basal area as the species-importance measures. These ordinations explained, respectively 77% and 87% of the floristic dissimilarity between plots. This composition gradient was highly congruent with a PCA ordination of the same inventory plots using three spectral bands, and was strongly correlated with two structural attributes: average canopy height and total basal area. A floristic ordination based on presence/absence data, however, was very different, grouping the inventories by the three study locations along the river. This was due either to differences between the knowledge of the three informants used to identify the trees at each location, or to high turnover in the rare (but not the common) species between locations. An MDS ordination of all types of campinarana in the Baniwa classification system was also performed based on the number of times a species was reported present in each type by ten different informants. This also resulted in a gradient, but with two distinct groups of vegetation types, corresponding to high and low forests. A description of indigenous knowledge systems that record uses and perceptions of the environment allows one to identify those areas used for subsistence activities such as hunting, fishing, agriculture, extractivism and crafts. Studies of this kind provide a basic knowledge of environmental resources and are an instrument for improved planning, zoning and management, as well as a link with other segments of society. They also contribute to a better understanding of habitat diversity in Amazonia and are an efficient method to characterize poorly known environments.