Response of soil organic carbon fractions to cover cropping: A meta-analysis of agroecosystems

Agroecosystems have a large potential for soil carbon (C) sequestration, and cover cropping is one management practice that can enhance soil C while providing other ecological and agronomic benefits. Still, we lack an understanding of the distribution of this added C across soil organic C (SOC) fractions. Two major operational pools make up SOC: easily decomposed, faster-turnover particulate organic C (POC) and slowly decomposed, slowerturnover mineral-associated organic C (MAOC). Though POC is generally more sensitive to environmental change in the short term, recent projections hold that MAOC stocks could face considerable loss in croplands where it is the dominant form of SOC. We used a systematic meta-analysis to determine responses of POC and MAOC concentration and stock to cover cropping and how these responses are influenced by soil, agronomic, and climatic variables. Across 49 studies concentrated in the Americas and Europe, total POC increased by 14.94% and total MAOC increased by 5.56% on average in cover-cropped soils vs. non-cover cropped or fallow control soils, which would effectively reverse or offset projected losses of SOC with climate warming. The effect of cover cropping on POC was greater in soils sampled from 0 to 10 cm vs. 10-30 cm depths and under higher mean annual temperature, but the effect of cover cropping on MAOC was not influenced by either of these factors. The response of POC to cover cropping was significantly positive for certain cover crop growing seasons, cover crop types, and tillage practices in which the MAOC response was not significant, indicating that changes in POC rather than MAOC are responsible for driving the positive response of total SOC to cover cropping. One commonality between POC and MAOC responses to cover cropping was that both dramatically increased with the amount of aboveground C produced by cover crops. Still, POC and MAOC responses to cover cropping were not strongly altered by nitrogen fertilization rate, cover crop diversity, experimental duration, annual precipitation, or baseline soil properties (texture, total organic C, and pH). Our results show that POC is more responsive to cover cropping than MAOC in an average of five years. However, more experimental data are needed to determine how these SOC fractions respond in deeper soil as well as in different geographic regions, climates, and soil orders. Based on our results, we discuss drivers of total and fractional SOC dynamics in agroecosystems and lessons learned from our synthesis of available fractional SOC data.