Net ecosystem productivity of boreal aspen forests under drought and climate change: Mathematical modelling with Ecosys

摘要

The net ecosystem productivity (NEP) of boreal aspen is strongly affected by comparative rates of annual potential evapotranspiration (E sub(a)) and precipitation (P sub(a)). Changes in E sub(a) versus P sub(a) during future climate change will likely determine changes in aspen NEP and consequently the magnitude of the carbon sink/source of a significant part of the boreal forest. We hypothesize that the effects of E sub(a) versus P sub(a) on aspen NEP can be modelled with a soil-root-canopy hydraulic resistance scheme coupled to a canopy energy balance closure scheme that determines canopy water status and thereby CO sub(2) uptake. As part of the ecosystem model ecosys, these schemes were used to model diurnal declines in CO sub(2) and latent heat (LE) exchange during a 3-year drought (2001-2003) at the Fluxnet-Canada Research Network (FCRN) southern old aspen site (SOA). These declines were consistent with those measured by eddy covariance (EC) at SOA, except that ecosystem CO sub(2) effluxes modelled during most nights were larger that those measured by EC or gap-filled from other EC measurements. Soil CO sub(2) effluxes in the model were close to, but sometimes smaller than, those measured by automated surface chambers at SOA. Diurnal declines in CO sub(2) exchange during the drought caused declines in annual NEP in the model, and in gap-filled EC measurements (model versus EC in g C m super(-2): 275 versus 367 +/- 110 in 2001, 82 versus 144 +/- 43 in 2002 and 23 versus 104 +/- 31 in 2003). Lower modelled NEP was attributed to the larger modelled CO sub(2) effluxes. Ecosys was then used to predict changes in aspen net biome productivity (NBP = NEP - C lost from disturbance) caused by 6-year versus 3-year recurring droughts during 100- year fire cycles under current climate versus climate change projected under the IPCC SRES A1B scenario. Although NBP was adversely affected during recurring 6- year droughts under current climate, it recovered quickly during non-drought years so that long-term NBP was maintained at 4 g C m super(-2) year super(-1). NBP rose by 10, 108 and 126 g C m super(-2) year super(-1) during the first, second and third centuries under climate change with recurring 3-year droughts, indicating a gradual rise in sink activity by boreal aspen. However recurring 6-year droughts during climate change caused recurring negative NBP (C losses), gradually depleting aspen C reserves and eventually causing dieback of the aspen overstory during the third century of climate change. This dieback was followed by a large decline in NBP. We conclude that NBP of boreal aspen will rise gradually under current projections of climate change, except under prolonged (e.g. 6 years) recurring droughts, which would eventually cause aspen to die back and substantial amounts of C to be lost.