Piaroa shifting cultivation: temporal variability of soil characteristics and spatial distribution of crops in the Venezuelan Orinoco

The Piaroa is an indigenous community located in the Venezuelan and Columbian Orinoquia. This community has made shifting agriculture part of their agricultural production system. This activity has allowed them to maintain food sovereignty and has guaranteed their territorial sustainability. The globalization processes and demographic growth however, have caused indigenous communities in the Venezuelan Orinoquia, including the Piaroa, to experience significant changes over the last four decades. These changes are characterized by the application of external development models and the consequent productive dependency. These factors threaten the continuity of the existing system. Our current understanding and administration of Piaroan agroforestry is composed of information designed to provide technical tools for generating sustainable management strategies for the tropical forests. Several anthropological and ecological studies have described the Piaroa shifting agricultural system for the purpose of providing aid. These studies also propose possible microsites selected by Piaroa farmers; however, the selection has not yet been investigated concretely. This information gives important contributing guidelines for agroforestry management and the selection of species in the Amazon, while taking into consideration shifting agriculture understanding. For this reason the temporary variability of the soil characteristics and their relationship with the spatial distribution of the crops is important. In the Cuao River watershed, located in the Venezuelan Amazon, ten systems of Piaroa shifting agriculture were randomly selected from different developmental stages. Established within these systems were a total of ten circular areas, each with a four-meter radius. In each of these areas the following was performed: (i) soil sample from the horizontal Ap (15 cm depth), (ii) vegetation inventory and (iii) an estimated area of the soil covered by biodegrading woody biomass, leaf litter cover and total vegetation cover. The temporary variability was determined by applying statistical methods to the information taken from the sites, according to the soil characteristics and properties. In the same way, a linear regression model was applied to explain the distribution of the cultivated species based on the soil's properties and characteristics. The results indicate significant changes in Ca2+ and K+ throughout the sequence. Variations in the CO, N and P totals were also observed, however these attributes did not show significant differences. During the first part of the sequence (0-4 years), the applied models explained the cultivated species distribution, Manihot esculenta, Zea mays, Ananas comosus and Pouteria caimito, with R-2 > 0.5 and in a later stage (4-6 years) the model explained the distribution of fruit trees such as Theobroma grandiflorum and Pourouma cecropiifolia. Finally, in the most recent studies the model fails to explain the distribution of the species with noted exceptions such as Euterpe oleracea.