New insight into soil carbon fixation rate: The intensive co-occurrence network of autotrophic bacteria increases the carbon fixation rate in depositional sites

Abstract

Whilst it has been shown that different microbial populations can significantly affect ecosystem functions, little is known about the potential effect of soil erosion on autotrophic bacterial co-occurrence networks and if the networks can influence carbon fixation potential. Here, the random matrix theory (RMT)-based molecular ecological network (MEN) approach with a stable isotope labeling experiment were performed to discern the relationship between autotrophic bacterial networks with the carbon fixing ability and quantify the 13CO2 (carbon dioxide) fixation rate in different soil erosion regions (the Chinese Loess Plateau, the T groups, and the hilly red soil erosion region in Southern China, the H groups). The rates of carbon fixation (RS) were ranked in the following order: H group depositional sites (HD; 30.86 Mg C km−2 yr−1) > H group erosion sites (HE; 18.04 Mg C km−2 yr−1) > T group depositional sites (TD; 5.00 Mg C km−2 yr−1) > T group erosion sites (TE; 1.03 Mg C km−2 yr−1). The results of the random forest modeling and variation partitioning analysis indicated that the bacterial co-occurrence networks were significantly associated with the carbon fixation rate (P < 0.05), and can explain 85.9% variations of carbon fixation rate in different erosion and deposition sites. With the results of multiple stepwise linear regression analysis, pH and C: N were considered to be the key abiotic factors that influenced the carbon fixing rate, which resulted in the carbon fixation rate of H groups were higher than what in T groups. Further, the ZP-plot suggested that the Proteobacteria were the keystones connecting and promoting other species to perform the C-fixing function. The statistical evidence of co-occurrence networks, which are positively associated with the autotrophic bacterial carbon fixation rate, advances the understanding of microbial networks in the function of carbon fixation.