Soil microbial biomass P status affecting P runoff loss from paddy fields under different agricultural practices

Phosphorus (P) runoff loss in agroecosystems is known to be soil-dependent. While the physicochemical processes of P runoff have been extensively studied, there is a lack of research on its microbial effects. This study investigated variations in P runoff loss, soil P fractions, and microbial biomass (MB) across 11 different agricultural treatments in a long-term positioning experiment of double-cropping rice cultivation. Results revealed a P-deficient state in the paddy soil, with an average total soil P content (TSP) of 0.55 g kg^-1 and Olsen-P of 9.17 mg kg^-1 across all treatments. Among the treatments, application of pig manure equivalent to 50 % conventional nitrogen resulted in the highest soil MB, soil P fractions, and flow-weighted dissolved-P (DP_wc) and total-P (TP_wc) concentrations in surface runoff water. Conversely, the treatment without P fertilizer application showed the lowest values. While long-term agricultural practices profoundly affected the characteristics of soil MB, soil P status, and P runoff loss, a common pattern emerged across all treatments. Specifically, all treatments exhibited wider ranges of soil microbial carbon (C):P ratio (MBC:P) (48.49-175.95) and soil microbial nitrogen (N):P ratio (MBN:P) (3.83-12.56) compared to soil microbial C:N ratio (MBC:N) (9.55-20.12). Additionally, soil P fractions decreased in the order of TSP > Citrate-P > Olsen-P > Enzyme-P, and the average DP_wc (0.12 mg L^-1) accounted for approximately one-fourth of the average TP_wc (0.48 mg L^-1), suggesting a critical and similar mechanism for paddy soil P runoff loss. The addition of exogenous C and P created favorable conditions for microbial growth, leading to increased MBP and subsequently elevated soil P contents, particularly Olsen-P. However, the structural equation model (SEM) analysis revealed that the mediation effect of MBP and MBN:P weakened the relationships between agronomic practices and P runoff losses, with the path coefficient decreasing from 0.34 to 0.06. Furthermore, the relative contribution of microbial biomass-induced effects to P runoff loss was about one-fifth of soil physicochemical effects in the double paddy rice ecosystem. Overall, our study quantified the effects of microbial biomass P status on P runoff loss in paddy fields under long-term agronomic measures, and these results will help us understand and manage nutrient cycling in paddy field ecosystems, thereby reducing environmental pollution, improving agricultural productivity and achieving sustainable agriculture.