State-space approach to evaluate effects of land levelling on the spatial relationships of soil properties of a lowland area

•The study shows the potential of the state-space to evaluate the land levelling.•The state-space approach can be applied for data sets collected in a sampling grid.•Levelling affected the local spatial relationships among soil properties.•Levelling affected the state-space model to estimate field capacity.•Levelling affected the state-space model to estimate the permanent wilting point.

Land levelling is an agricultural practice to correct soil surface irregularities turning the area more efficient for management of agronomic inputs. The technique has been adopted mainly in lowland areas used for flood irrigation of rice. As a result sub-surface soil layers may be exposed with possible impact on crop production by changing soil properties. This study aimed to evaluate the effects of land levelling on the spatial relationships of soil properties. Relations among them were quantified using a state-space approach in different scenarios constructed from data sets in a 1 ha lowland area sampled as a grid. The grid consisted of 100 sampling points (10 m × 10 m), with samples collected from the 0–0.20 m soil layer, before and after levelling, totalizing 200 samples. Soil water contents at field capacity (θFC) and permanent wilting point (θPWP) were taken as response variables due to their importance on rice water management. Sand, silt and clay contents, soil microporosity (Micro) and bulk density (BD), cation exchange capacity (CEC), organic carbon content (Corg), and the depth of the top of the B horizon in relation to soil surface (DTB) were used as co-variables through an evaluation of their spatial auto- and cross-correlation behaviors with θFC and θPWP. Eight data-array scenarios were tested. Results showed that levelling induced negative effects on soil quality since this procedure decreased θFC and θPWP, Corg, and CEC as well as increased BD. Using the state-space approach, we concluded that its performance in estimating θFC and θPWP was affected by the scenarios after levelling, the best performance being for the vertical scenarios. DTB and CEC contributed to the estimation of both soil water contents for all scenarios after levelling. The main problem associated to the levelling is the cutting of shallow soils which decreases their capacity to store water and to exchange cations in the effective arable depth which will be explored by the crop root system. As DTB is directly related to the spatial distribution of taxonomic soil profile properties, the state-space approach, which considers sampling location coordinates, may be a potential on-site-specific tool aiming at the recuperation of degraded soils through amendments since it opens the possibility for farmers to manage a crop field based on local environmental properties, with their spatial association and localized variation being a function of the distance between their measurements.