Carbon input and allocation by rice into paddy soils: A review

Knowledge of belowground C input by rice plants and its fate is essential for managing C cycling and sequestration in paddy soils. Previous reviews have summarized C input and the pathways of root-derived C in upland soils by labeling with C-14 or C-13(C-13(/)14), while rice rhizodeposition and C input in paddy soils have not been comprehensively evaluated. Here, we analyzed the results of C-13(/)14 pulse and continuous labeling studies using 112 datasets from 13 articles on the allocation and pathways of photosynthesized C by rice plants to assess C input, budget, and amount stabilized in paddy soils. Overall, C-13(/)14 partitioning estimated by continuous labeling was 72% to the shoots, 17% to the roots, 10% to the soil, and 1.3% was recovered in microbial biomass. Pulse-labeling studies showed a similar C partitioning: 79%, 13%, 5.5%, and 2.1%, respectively. The total belowground C input estimated based on continuous labeling was 1.6 Mg ha(-1) after one rice season, of which rhizodeposition accounted for 0.4 MgC ha(-1). Carbon input assessed by pulse labeling was slightly lower (total belowground C input, 1.4 Mg ha(-1); rhizodeposition, 0.3 MgC ha(-1); 14 days after labeling). Rice C input after one cropping season was lower than that by upland plants (cereals and grasses, 1.5-2.2 Mg ha(-1)). In contrast to upland crops, most paddy systems are located in the subtropics and tropics and have two or three cropping seasons per year. We conclude that (1) pulse labeling underestimates the total belowground C input by 15%, compared with that by continuous labeling, and (2) rhizodeposition of rice accounts for approximately 26% of the total belowground C input, regardless of the labeling method used. Based on allocation ratios, we suggest a simple and practical approach for assessment of the gross C input by rice into the soil, for partitioning among pools and for long-term C stabilization in paddies.