Optimizing multifunctional agroecosystems in irrigated dryland agriculture to restore soil carbon – experiments and model

Advances in dryland irrigated agriculture can develop into sustainable multifunctional agroecosystems if land use and soil management are optimized to enhance soil organic carbon (SOC). We hypothesized that combining high inputs of plant-derived carbon with no-tillage will increase long-term SOC stocks in this ecosystem. We evaluated data from two field experiments under a mango orchard and melon crops, respectively. Both were managed with two tillage systems, no-till (NT) and conventional till (CT), and combined with three types of intercropping. Plant mixtures used 75% legumes + 25% grass and oilseed species (PM1), 25% legumes + 75% grass and oilseed species (PM2), and spontaneous vegetation (SV). The Caatinga shrubland was cleared in 1972 for use of arable (AC) and palm plantation (PP). The carbon turnover model Roth C was used to discriminate the effects of the cropping and tillage systems on SOC and to predict the impacts on the long-term dynamic of SOC. Plant-derived carbon, including roots, exudates, and aboveground residues, add on average 5.23 Mg ha-1yr-1, increasing SOC stocks between 0.23 and 0.42 Mg ha-1yr-1. If the current designs of agroecosystems will be maintained, the value of soil carbon stock equal from dry forest steady-state will be reached in about 20 years for annual crop and 30 years for the perennial crop. Given that agriculture in dryland areas is recognized as key mediators of soil carbon storage decline, our results warrant further efforts to increase carbon storage in irrigated multifunctional agroecosystems designs that include plant mixtures and no-till in semiarid regions.