There is an acute need to understand biogeochemical processes during winter in northern ecosystems. Root and microbial processes are surprisingly active in cold (0 - 5 oC) and even frozen soils and a significant portion (20 - 50%) of ecosystem C and N cycling, drainage losses and soil-atmosphere trace gas fluxes occurs during winter. Overwinter processes are highly variable, subject to perturbation, highly responsive to global climate change, and strongly influence nutrient losses during the subsequent growing season.
In previous NSF-funded research, we manipulated snow cover to induce soil freezing events. This research found that relatively mild freezing events (soil temperatures never decreasing below -4 oC) produced significant increases in soil inorganic nitrogen (N) levels, solute leaching, root mortality and nitrous oxide flux. These responses to mild freezing were surprising in view of laboratory studies by us and others that suggested significant biogeochemical effects would only occur with a freezing event sufficiently severe (< -10 oC) to directly cause root and microbial mortality. These results raised important questions about the mechanisms by which freezing affects soil biogeochemistry.
In this project we explored these intriguing effects of soil freezing on nutrient loss at the Hubbard Brook Experimental Forest (HBEF), a northern hardwood dominated forest in the White Mountains of New Hampshire. We hypothesized that physical disruption of soil by frost heaving and the formation of ice lenses represents an important disturbance to the ecosystem N cycle. We proposed four mechanisms by which mild soil freezing increases N loss; 1) rapid mineralization of dead roots killed by physical disruption, 2) reduced root uptake because of root mortality and/or injury, 3) physical disruption of soil aggregates and consequent release and mineralization of particulate organic matter (POM) and 4) physical effects of freezing on mineralization of fresh litter.
To test our hypotheses about these mechanisms we 1) established field plots to provide variation in soil freezing intensity, 2) quantified physical disruption of the soil ecosystem by freezing and 3) made detailed measurements of the perturbations associated with soil freezing. These measurements included, 1) quantification of root mortality and mineralization, 2) release of POM following freeze events and 3) decomposition and N release from 15N-labelled litter. These measurements were evaluated in the context of continuous measurements of N leaching (lysimeters), microbial biomass and activity, in situ net N mineralization and nitrification (intact cores) and soil:atmosphere fluxes of CO2, N2O and CH4 (field chambers).
To develop tools to extrapolate and evaluate results at larger scales, we developed distributed models of snow depth and soil processes and produced estimates of the nature and extent of soil freezing for the entire 3,000 ha HBEF valley. Ultimately, these models will be linked with biogeochemical models (PnET-CN, PnET-BGC) to assess variation in response to soil freezing events across the northern hardwood landscape at HBEF.