SciELO - Scientific Electronic Library Online

vol.51 issue9Mapping soil carbon, particle-size fractions, and water retention in tropical dry forest in BrazilSoil class prediction by data mining in an area of the sedimentary São Francisco basin author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Pesquisa Agropecuária Brasileira

Print version ISSN 0100-204XOn-line version ISSN 1678-3921


ANDRADE, Sandra Fernandes de  and  MENDONCA-SANTOS, Maria de Lourdes. Prediction of soil fertility of the agricultural hub of the state of Rio de Janeiro using soil x landscape modeling. Pesq. agropec. bras. [online]. 2016, vol.51, n.9, pp.1386-1395. ISSN 0100-204X.

The objective of this work was to predict soil fertility in the agricultural hub of the state of Rio de Janeiro, Brazil, using soil x landscape modeling. The studied area comprised the most productive regions of the state of Rio de Janeiro: northern, northwestern, and Serrana (highlands). Soil chemical traits - pH H2O and cation exchange capacity (CEC) - and environmental variables - elevation, plane of curvature, curvature profile, moisture content index, aspect and slope of the terrain, in addition to soil types, normalized difference vegetation index (NDVI), Landsat 7 images, and lithology - were used as prediction variables. Exploratory data analysis identified outliers, which were purged, in the preparation for analysis by multiple linear regression (MLR). The kriging results of the regression residuals were added to the results of MLR, using a soil digital mapping (SDM) technique known as regression-kriging. In the Serrana region, the environmental variables explained the chemical variables. The NDVI variable was significant in all three regions, showing the importance of vegetation cover for the prediction of soil fertility. Generally, the tested soils showed low pH. The CEC values in the studied regions are within the range considered good for soil fertility.

Keywords : digital soil mapping; pedometrics; regression-kriging; spatial variability.

        · abstract in Portuguese     · text in Portuguese     · Portuguese ( pdf )