Key Microbial Processes Of Soil Carbon Dynamics And Their Responses To Extreme Drought In Evengreen Broad-leaved Forests: Data-model Fusion

Soil microbes are important decomposers,which play a key role in regulating soil carbon(C)processes and their responses to climate change in terrestrial ecosystems.Recently,more and more microbial modules have been incorporated into ecological process-based models to improve the simulation of microbial process and soil C dynamics.As an important global C sink,evergreen broad-leaved forests are facing the threat of extreme drought with increased frequency.Extreme drought affects soil C storage via altering microbial processes.However,the key mechanisms by which microbes regulate soil C processes in response to extreme drought,and how the addition of microbial modules affect the terrestrial ecosystem model to simulate soil C dynamics under drought conditions are unclear.Therefore,based on the extreme drought experimental platform of Zhejiang Tiantong Forest Ecosystem National Observation and Research Station,multiple terrestrial ecosystem models(including 3 traditional models: 3?5-pool model;5 microbial-enzyme models:3-pool+enzyme model,4 or 5-pool+1 or 2-enzyme model)were selected by data-model fusion method.The effects of microbes on soil C processes in evergreen broad-leaved forests under drought conditions were simulated.The main results were as follows:(1)Comparison and screening of traditional and microbial-enzyme model structures: the addition of microbial processes(enzyme modules)improved the accuracy of terrestrial ecosystem C cycle models in simulating soil C processes in evergreen broad-leaved forests.Specifically,microbial-enzyme models were more accurate(4.8%)than the traditional models in simulating the C process except for the4-pool+enzyme model.,The accuracy of soil C process simulation in microbial-enzyme model(5-pool+enzyme pool and 4-pool+2-enzyme model)were11.60% and 6.49% higher than that in traditional model(4-pool model)under drought condition.Considering the complexity of the model,the 4-pool+2 enzyme model could be used as the most optimal model to simulate the C process in evergreen broad-leaved forest.Moreover,our results indicate that the addition of enzyme activity data could constrain more parameters related soil C process,reduce the uncertainty of parameters,and improve the accuracy of model simulation.(2)The effect of extreme drought on soil C dynamics simulation based on microbial-enzyme models: the soil C process was simulated under drought condition by using the screened microbial-enzyme model(5-pool+enzyme pool and4-pool+2-enzyme model).The results showed significantly reduction in total soil C pool and microbial C pool under drought,mainly due to the decline in net primary productivity of ecosystems and increased turnover rate of microbes under drought condition.The difference between microbial-enzyme model and traditional model was manifested the effect of drought on slow organic C and structural litter in the short-term.Relative to 4-pool model,the relative contribution of 4-pool+2-enzyme model and 5-pool+enzyme model in simulating the effect of drought on slow organic C pool decreased by 54.64% and 78.68%,respectively.The relative contribution to the effect of structural litter pool increased by 394% and 803%,respectively,showing obviously weakened utilization of slow organic C and enhanced utilization of structural litter by microbes under short-term drought.This might be associated with enhanced physical protection of soil aggregates and the change of microbial preference for substrate utilization owing to the shifts in microbial community under drought.In summary,the addition of microbial process could improve the simulation of terrestrial ecosystem C model to simulate evergreen broad-leaved forest soil C process and its response to drought,it also could reveal the regulatory role of microbes in soil organic C response to drought.The results of this study provide a reference for a better understanding of soil organic Cresponse to extreme drought and its microbial regulation mechanism,and provide a scientific basis for reducing the uncertainty of model simulation.