Exploring microscale heterogeneity as a driver of biogeochemical transformations and gas transport in peat

Abstract

Peat pore network architecture is a key determinant of water retention and gas transport properties and has therefore been hypothesized to control redox conditions in and greenhouse gas emissions from peat soils. Yet, experimental approaches to directly visualize the spatial heterogeneity of biogeochemical reactions in pore networks remain scarce. Here, we report on a 13C pulse–chase assay developed to functionally explain and visualize the centimeter-scale heterogeneity in greenhouse gas emissions in peat cores. We injected a 13C-labeled substrate (13C2 acetate) at 2 to 8 cm depths and monitored its conversion into CO2 and CH4. We then measured the pore network architecture of the same cores by X-ray microtomographic imaging and constructed the air-filled pore networks using pore network modeling. We applied this approach to peat cores collected at a drained peatland forest in southern Finland in an experiment to study the effects of water hysteresis, i.e., differences between peat cores that reached a given water potential (−20 hPa) from drier or wetter conditions. We find large heterogeneity among the replicate cores and injections, indicating the effects of centimeter-scale heterogeneity on biochemical processes and gas transport. These treatments resulted in similar average air-filled porosity but distinct pore networks (higher coordination numbers and clustering coefficients in drying compared to wetting soils) and within-core water distribution. Substrate injection experiments revealed less (potential) microbial activity (less of the substrate emitted as CO2) at greater depth in both treatments. In peat cores from the drying treatment we also find a slower microbial response to label additions at greater depths (slower release of label-derived CO2), while the timing of emissions did not vary in wetting treatments. Air-filled porosity and pore network metrics could not explain the fraction of label converted to CO2, but greater porosity was associated with slower CO2 emissions, whereas higher clustering coefficients and betweenness centrality (two measures of pore network properties) were associated with faster emissions.