Soil Multifunctionality Predicted by Bacterial Network Complexity Explains Differences in Wheat Productivity Induced by Fertilization Management
Wenguang Li,Fan Shi,Shusheng Yi,Tianyu Feng,Caiyu Wang,Ziyan Li,Wei Zheng,Bingnian Zhai
DOI: https://doi.org/10.1016/j.eja.2023.127058
IF: 5.2
2024-01-01
European Journal of Agronomy
Abstract:Fertilization will cause changes in crop growth, soil microbial community and multiple ecosystem functions, but it is unclear whether and how the optimal productivity obtained by manure application and optimizing nitrogen is driven by microbial community and ecosystem multifunctionality. We explore this mechanism based on a field fertilization experiment using a split-plot design that began in 2014 with five N rates (N0, N75, N150, N225, N300) as main plots and two manure rates (M0, NPK group; M1, MNPK group) as subplots, respectively. In general, (1) grain yield was parabolically related to the N application rate. The N rates that obtained the highest yield were 150 kg ha-1 and 225 kg ha-1 for the MNPK and NPK groups, respectively. The average yield of the MNPK group was 10.5% higher than that of the NPK group. (2) Long-term fertilization management resulted in regular changes in soil multifunctionality (SMF) and the properties of various microbial communities. The richness, network complexity and soil multifunctionality of bacterial and fungal communities also increased initially and then decreased as the N rate. On average, the response ratio of soil multifunctionality to N75, N150, N225, N300 (control is N0) and M1 (control is M0) were 0.11, 0.41, 0.53, 0.44 and 1.19, respectively. The soil multifunctionality of MNPK was 260%, 722% and 101% higher than that of NPK at the tillering, jointing and harvest stages, respectively. (3) Soil multifunctionality not only always had a significant positive effect on aboveground biomass at all growth stages, but also had significant positive correlations with bacterial richness and network complexity. Based on the existing framework, we further demonstrated that high microbial richness supports multifunctionality by ensuring strong associative complexity among microorganisms. However, only bacterial network complexity always positively drove soil multifunctionality and indirectly influenced wheat growth at all growth stages in SEMs where multiple drivers were considered simultaneously (i.e., pH, SOC and biological factors). Random forest regression analysis showed that rare bacterial taxa had stronger predictive effects on soil multifunctionality than abundant taxa. Our results validate that M1N150 can ensure high yield in the study area while maintaining a high level of ecosystem function in the soil, which also contributes to the reduction of adverse environmental risks caused by fertilizer application. Importantly, the potential of microbial network complexity, especially bacterial network complexity, to influence crop growth by driving farmland ecosystem functions deserves to be explored under different crop types, cropping systems, and irrigation conditions, compared to previous microbial richness that has been simply quantified.