Pasture conversion is a promising alternative to cut down the carbon footprint of deforestation for oil palm expansion.

Increasing soil organic carbon (SOC) in agroecosystems is necessary to mitigate climate change and soil degradation. Management practices designed to reach this goal call for a deeper understanding of the processes and drivers of soil carbon input stabilization. We identified main drivers of SOC stabilization in oil palm plantations using the well‐defined spatial patterns of nutrients and litter application resulting from the usual management scheme. The stabilization of oil palm‐derived SOC (OP‐SOC) was quantified by δ13C from a shift of C4 (savanna) to C3 (oil palm) vegetations. Soil organic carbon stocks under frond piles were 20% and 22% higher compared with harvest paths and interzones, respectively. Fertilization and frond stacking did not influence the decomposition of savanna‐derived SOC. Depending on management zones, net OP‐SOC stabilization equalled 16–27% of the fine root biomass accumulated for 9 years. This fraction was similar between frond piles and litter‐free interzones, where mineral NPK fertilization is identical, indicating that carbon inputs from dead fronds did not stabilize in SOC. A path analysis confirmed that the OP‐SOC distribution was largely explained by the distribution of oil palm fine roots, which itself depended on management practices. SOC mineralization was proportional to SOC content and was independent on phosphorus availability. We conclude that SOC stabilization was driven by C inputs from fine roots and was independent of alteration of SOC mineralization due to management. Practices favouring root growth of oil palms would increase carbon sequestration in soils without necessarily relying on the limited supply of organic residues.

Severe constraints on grasslands productivity, ecosystem functions, goods and services are expected to result from projected warming and drought scenarios under climate change. Negative effects on vegetation can be mediated via soil fertility and water holding capacity, though specific mechanisms are fairly complex to generalise. In field drought experiments, it can be difficult to disentangle a drought effect per se from potential confounding effects related to vegetation or soil type, both varying along with climate. Furthermore, there is the need to distinguish the long-term responses of vegetation and soil to gradual climate shift from responses to extreme and stochastic climatic events. Here we address these limitations by means of a factorial experiment using a single dominant grassland species (the perennial ryegrass Lolium perenne L.) grown as a phytometer on two soils types with contrasted physicochemical characteristics, placed at two elevation sites along a climatic gradient, and exposed to early or late-season drought during the plant growing season. Warmer site conditions and reduced precipitation along the elevational gradient affected biogeochemistry and plant productivity more than the drought treatments alone, despite the similar magnitude in volumetric soil moisture reduction. Soil type, as defined here by its organic matter content (SOM), modulated the drought response in relation to local site climatic conditions and, through changes in microbial biomass and activity, determined the seasonal above and belowground productivity of L. perenne. More specifically, our combined uni-and multivariate analyses demonstrate that microbes in a loamy soil with low SOM are strongly responsive to change in climate, as indicated by a simultaneous increase in their C,N,P pools at high elevation with cooler temperatures and wetter soils. Contrastingly, microbes in a clay-loam soil with high SOM are mainly sensitive to temperature, as indicated by a strong increase in microbial biomass under warmer temperatures at low elevation and a