Linking the potential activities of methanogens and methanotrophs to their communities under different fertilization regimes in coastal saline paddy soils

Methane (CH4) emissions are primarily dependent on the net balance between the activities of methanogens and methanotrophs. Straw returning (straw residue incorporation) and nitrogen (N) fertilizer applications are two of the most frequent agronomic practices applied to ensure rice yield, and they also profoundly affect the behavior of methanogens and methanotrophs. However, the effect of rice straw or its biochar combined with N fertilizer on the activities and compositions of methanogens and methanotrophs is still not well understood, especially in rice root zone soil in coastal saline paddy fields. Moreover, only few studies linking communities and functions of methanogens and methanotrophs are available, resulting in a lack of adequate indicators to predict methane production potential (MPP) or methane oxidation potential (MOP), for improving fertilizer strategies. Here, through quantitative PCR (qPCR), high-throughput sequencing, network analysis and linear regression model, we investigated the response of the abundances, compositions and activities of methanogens and methanotrophs to rice straw or its biochar combined with N fertilizer in rice root zone soil in coastal regions, and explored the relationships between MPP and methanogenic compositions as well as MOP between methanotrophic compositions, respectively. Therefore, a combination of pot and mesocosm experiment was conducted including seven fertilizer treatments: CK, without N fertilizer and rice straw return; SN100, 100 % N plus straw return; SN80, 80 % N plus straw return; SN60, 60 % N plus straw return; BN100, 100 % N plus rice biochar return; BN80, 80 % N plus biochar return; and BN60, 60 % N plus biochar return. Results showed that irrespective of MPP, MOP, or methanogenic and methanotrophic abundance, they all decreased significantly (P < 0.05), with the reduction of N application under the condition of combined straw or biochar with N fertilizer in coastal saline paddy soils. Moreover, compared with straw, biochar could significantly (P < 0.05) reduce MPP and methanogenic abun-dances, whereas improve MOP and methanotrophic abundances under the same N application rate. The models incorporating mcrA gene copy and soil dissolved organic carbon (DOC) could accurately predict MPP, while MOP was best predicted by a model incorporating pmoA gene copy and NO3-N. The family Methanobacteriaceae and Methanosarcinaceae were dominant in the methanogenic co-occurrence and keystone patterns. Correspondingly, the genera Methylocystis and Metilomonas were dominant in the methanotrophic co-occurrence pattern, and genus Methylocystis was also the keystone members in coastal saline root zone soil of rice. Soil DOC and salinity were the key factors in shaping the methanogenic co-occurrence and keystone patterns, whereas the methanotrophic co-occurrence and keystone patterns were mainly regulated by NO3-N. Furthermore, methanogenic keystone taxa were more suitable for predicting MPP than their co-occurrence pattern. Our findings provide insights into the possibility of key microbial taxa for predicting MPP or MOP for low-carbon production fertilizer application strategies in coastal saline paddy soils.