Simple Models for Predicting Agricultural Trafficability on Cohesive Soils Cultivated with Sugarcane in Brazil

Abstract

Preconsolidation stress (σ p) is an important physical indicator of the load-bearing capacity of soil, which is used for predicting agricultural trafficability. This study aimed to compare prediction models for agricultural trafficability using soil physical attributes as explanatory variables on soils of coastal lands cultivated with sugarcane in the Northeast of Brazil. The study was carried out on a cohesive soil area cultivated with sugarcane at which 42 points were sampled in order to collect undisturbed soil samples at 0.10-0.13 and at 0.30-0.33 m layers, summing up 84 samples. The following variables were analysed: water content (θ), bulk density (ρ), penetration resistance (PR) and σ p. Three models were fitted to predict σ p as a function of ρ, θ and PR. Bulk density and water content can be used as predictors of agricultural trafficability on cohesive soils, and have improved predictions when used together in regression models. However, despite the fact that water content and bulk density are known to be correlated, soil penetration resistance can be used as a practical soil trafficability criterion, by fitting a simple linear regression, for it is easily and quickly measurable.