LAUSR

Methane emissions from aquatic ecosystems are increasingly recognized as substantial, yet variable, contributions to global greenhouse gas emissions. This is in part due to the challenge of modeling biologic parameters that affect methane emissions from a wide range of sediments. For example, the impacts of fish bioturbation on methane emissions in the literature have been shown to result in a gradient of reduced to enhanced emissions from sediments. However, it is likely that variation in experimental fish density, and consequently the frequency of bioturbation by fish, impacts this outcome. To explore how the frequency of disturbance impacts the levels of methane emissions in our previous work we quantified greenhouse gas emissions in sediment microcosms treated with various frequencies of mechanical disturbance, analogous to different levels of activity in benthic feeding fish. Greenhouse gas emissions were largely driven by methane ebullition and were highest for the intermediate disturbance frequency (disturbance every 7 days). The lowest emissions were for the highest frequency treatment (3 days). This work investigated the corresponding impacts of disturbance treatments on the microbial communities associated with producing methane. In terms of total microbial community structure, no statistical difference was observed in the total community structure of any disturbance treatment (0, 3, 7, and 14 days) or sediment depth (1 and 3 cm) measured. Looking specifically at methanogenic Archaea however, a shift toward greater relative abundance of a putatively oxygen-tolerant methanogenic phylotype (ca. Methanothrix paradoxum) was observed for the highest frequency treatments and at depths impacted by disturbance (1 cm). Notably, quantitative analysis of ca. Methanothrix paradoxum demonstrated no change in abundance, suggesting disturbance negatively and preferentially impacted other methanogen populations, likely through oxygen exposure. This was further supported by a linear decrease in quantitative abundance of methanogens (assessed by qPCR of the mcrA gene), with increased disturbance frequency in bioturbated sediments (1 cm) as opposed to those below the zone of bioturbation (3 cm). However, total methane emissions were not simply a function of methanogen populations and were likely impacted by the residence time of methane in the lower frequency disturbance treatments. Low frequency mechanical disruption results in lower methane ebullition compared to higher frequency treatments, which in turn resulted in reduced overall methane release, likely through enhanced methanotrophic activities, though this could not be identified in this work. Overall, this work contributes to understanding how animal behavior may impact variation in greenhouse gas emissions and provides insight into how frequency of disturbance may impact emissions.