A study of the effect of the interaction between site-specific conditions, residue cover and weed control on water storage during fallow

In the semiarid central region of Argentina the probability that rainfall meets crop requirements during growing season is less than 10%, therefore fallowing has been the most important practice to assure water availability during the growing season. Various site-specific and management factors have been identified as crucial for defining fallow efficiency (FE) and final available water contents (AW). The objective of the present study was to improve our knowledge about the interactions between residue cover, weed control, soil profile depth and water storage capacity (WSC) on FE. In 10 sites covering the environments of calcareous plains and sandy plains of the semiarid central region of Argentina and with different WSC, experiments with 3 different levels of residue cover (H, M, L) and with and without weed control (C and W respectively) during fallow were set up. A completely randomized block design with four repetitions and splits plots to consider weed control was used. Soil texture and organic matter were determined in samples of the A horizon (0.20 m). Bulk density, field capacity, permanent wilting point and soil water contents (monthly frequency) were measured at depth intervals of 0.20 m to the depth of the calcite layer or to 2.00 m depth. Soil temperature was taken in weekly intervals at 0.05 m depth and weed plants, separated by species, were counted at the end of fallow in 4 repetitions of 0.25 m2 in each treatment. An empirical model was developed to predict final AW under these experimental conditions. Model parameters were: Residue level, weed control, WSC, profile depth, and rainfall during fallow. Site-specific conditions (WSC and profile depth) affected water storage during fallow; soils with highest values for both parameters showed highest final AW. Weed density was the most important factor that controlled AW, with on average 35 mm less AW in W than in C treatments. Residue level had a positive effect on final AW in both C and W treatments, with a difference of 18.5 mm between H and L. An interaction between residue level and weed density was observed, indicating weed suppression in H treatments. This was also confirmed by correspondence analysis between residue level and weed species which revealed that different species were related to each level. High residue levels also decreased soil temperature, thus affecting germination of post-fallow crops. The empirical model had an overall average prediction error of 13.7% and the regression between measured and predicted values showed a determination coefficient of 0.77.