Thu, Aug 05, 2021:On Demand
Background/Question/Methods
Highly dynamic ecosystems and their capacities to respond to environmental change are incredibly difficult to understand in natural settings, in part due to the necessity of timeseries data to adequately characterize them. In turn, poor temporal resolution has led to coarse representation of fluctuating ecosystems in hydrobiogeochemical models, and scaling spatially and temporally resolved information from the molecular to watershed scale remains challenging. We use a timeseries of molecular measurements in a model ecosystem (dam-driven hyporheic zone) to develop a temporal understanding of microbial and biogeochemical processes in a variable ecosystem, their connectivity through time, and their dependency on prior environmental conditions. We hypothesize that (1) biogeochemical responses to surface water intrusion depend on recent history and (2) this is underlain by diverse microbiome activities with distinct core and auxiliary metabolic processes.
Results/Conclusions Using ultrahigh resolution environmental metabolomics, we show that metabolomic changes coincided with surface water elevation and generally preceded corresponding rates of aerobic metabolism by approximately one day. Actively expressed genes for every step in denitrification and DNRA were present in the hyporheic zone, as well as microbial activities related to the oxidation of lignin and chemically complex hydrocarbons, underscoring diverse metabolic activities. Additionally, we found auxiliary metabolomic processes consistent with lignin and hydrocarbon decomposition only present during periods of elevated metabolism. Multi ‘omic co-association networks also highlighted the importance of lignin oxidation, hydrocarbon degradation, and nitrate reduction as central metabolisms under hyporheic fluctuation in addition to phenotypic plasticity. Collectively, our results point to a conceptual model of highly dynamic ecosystems in which microorganisms have the capacity to upregulate auxiliary metabolic pathways during periods of optimal growth, but biogeochemical process rates lag these metabolic changes. Yet, process-based models continue to assume that contemporary factors are the primary determinant of microbially-mediated biogeochemistry, with instantaneous responses to change. Our results challenge the assumptions underlying model-based predictions of biogeochemistry and point to asynchronous processes at the molecular-to-ecosystem scale as a source of uncertainty in process-based models, even over relatively short timescales.
Results/Conclusions Using ultrahigh resolution environmental metabolomics, we show that metabolomic changes coincided with surface water elevation and generally preceded corresponding rates of aerobic metabolism by approximately one day. Actively expressed genes for every step in denitrification and DNRA were present in the hyporheic zone, as well as microbial activities related to the oxidation of lignin and chemically complex hydrocarbons, underscoring diverse metabolic activities. Additionally, we found auxiliary metabolomic processes consistent with lignin and hydrocarbon decomposition only present during periods of elevated metabolism. Multi ‘omic co-association networks also highlighted the importance of lignin oxidation, hydrocarbon degradation, and nitrate reduction as central metabolisms under hyporheic fluctuation in addition to phenotypic plasticity. Collectively, our results point to a conceptual model of highly dynamic ecosystems in which microorganisms have the capacity to upregulate auxiliary metabolic pathways during periods of optimal growth, but biogeochemical process rates lag these metabolic changes. Yet, process-based models continue to assume that contemporary factors are the primary determinant of microbially-mediated biogeochemistry, with instantaneous responses to change. Our results challenge the assumptions underlying model-based predictions of biogeochemistry and point to asynchronous processes at the molecular-to-ecosystem scale as a source of uncertainty in process-based models, even over relatively short timescales.