What controls methane emissions from restored wetlands?
My current research aims to better understand the controls of methane emissions from restored wetlands in the Sacramento-San Joaquin Delta of California. Quantifying these emissions is important to the state, who funds our research, because if methane emissions are large enough it may offset any greenhouse gas benefit of restoration. This has consequences to funding these projects through California's Cap-and-Trade Program. My research aims to identify what drives these potent methane emissions so wetlands can be managed to limit overall greenhouse gas emissions.
I use a large network of eddy covariance towers operated by the Baldocchi Biometeorology Lab to study these systems. My current/past projects here assess (1) how these systems respond to salinity disturbance, (2) what controls emission rates across space, and (3) how well our flux measurements perform across diverse land uses in the region. I use machine learning, causal inference, and time series decomposition techniques to better understand the underlying controls of observed flux patterns. These approaches allows us to gain a strong process-based understanding of our observations that cannot be achieved through traditional parametric approaches. More description of my core projects can be found below.
Soil Properties and Sediment Accretion Modulate Methane Fluxes
My work published in 2018 in Global Change Biology sought to understand how spatially variable soil properties from the pre-restoration landscape impact current methane emissions from restored wetlands. Using an eddy covariance tower network, data-driven analyses, and soil sampling, we found that spatial gradients in soil iron, a legacy of pre-drainage geomorphology, explain differences in wetland methane emissions, where lower fluxes were observed from wetlands restored on high iron soils. Using information theory and time series decomposition techniques, we demonstrate that methanefluxes from high iron wetlands were decoupled from plant processes at daily timescales, as expected when iron presence inhibits methane production. However these differences were transient, and three years post-restoration methane emissions converged across sites. Changing methane emissions with time could not be explained by a generalized additive model based on dominant biophysical drivers, suggesting that soil factors, not measured by our eddy covariance towers, were responsible for these changes. Our study suggests that soil properties, likely Fe content, are capable of inhibiting ecosystem‐scale wetland CH4 flux, but these effects appear to be transient without continued input of alluvium to wetland sediments. This work is entirely reproducible in R, the code and data can be found here.