Hydrological Controls on Ecosystem CO2 and CH4 Exchange in a MIXED Tundra and a FEN within an Arctic Landscape UNDER Current and Future Climates

Wednesday, 17 December 2014
Robert F Grant, University of Alberta, Edmonton, AB, Canada, Elyn Humphreys, Carleton University, Ottawa, ON, Canada and Peter Lafleur, Trent University, Peterborough, ON, Canada
Variation in CO2 and CH4 exchange in years with contrasting weather is strongly affected by hydrology in landscapes underlain by permafrost. Hypotheses for this variation were incorporated into the ecosystem model ecosys which simulated CO2 and CH4 fluxes along a topographic gradient within an arctic landscape at Daring Lake, NWT, Canada. Fluxes modelled at mixed tundra and fen sites within the gradient were compared with CO2 fluxes measured at eddy covariance towers from 2006 to 2009, and with CH4 fluxes measured with surface chambers in 2008. Slopes and correlation coefficients from regressions of modelled vs. measured CO2 fluxes were 1.0 ± 0.1 and 0.7 – 0.8 for both sites in all years. At the mixed tundra site, rises in net CO2 uptake in warmer years with earlier snowmelt were constrained by midafternoon declines in CO2 influxes when vapor pressure deficits (D) exceeded 1.5 kPa, and by rises in CO2 effluxes with greater active layer depth (ALD). Consequently annual net CO2 uptake at this site rose little with warming. At the fen site, CO2 influxes declined less with D and CO2 effluxes rose less with warming, so that rises in net CO2 uptake in warmer years were greater than those at the mixed tundra site. The greater declines in CO2 influxes with warming at the mixed tundra site were modelled from greater soil-plant-atmosphere water potential gradients that developed in drier soil, and the smaller rises in CO2 effluxes with warming at the fen site were modelled from O2 constraints to heterotrophic and below-ground autotrophic respiration that limited their responses to greater ALD. Modelled and measured CH4 exchange during July and August indicated very small influxes at the mixed tundra site, and larger emissions at the fen site. Emissions of CH4 modelled during soil freezing in October – November contributed about one-third of the annual total, and so should be included in estimates of annual emissions. These contrasting responses to warming under current climate modelled and measured at the mixed tundra and fen sites were apparent in their contrasting responses modelled under long-term climate change.