A 100-Year Retrospective Landscape-Level Carbon Budget for the Sooke Lake Watershed, British Columbia: Constraining Estimates of Terrestrial to Aquatic DOC Transfers.

Abstract:

To address how natural disturbance, forest harvest, and deforestation from reservoir creation affect landscape-level carbon (C) budgets, a retrospective C budget for the 8500 ha Sooke watershed from 1911 - 2012 was developed using historic spatial inventory and disturbance data. Data was input to a spatially-explicit version of the Carbon Budget Model-Canadian Forest Sector (CBM-CFS3), an inventory-based C budget model used to simulate forest C dynamics at multiple scales. In 1911 the watershed was dominated by mature/old Douglas-fir forests with aboveground biomass C (ABC) of 262 Mg C/ha and net ecosystem production (NEP) of 0.63 Mg C/ha/yr. Land was cleared around Sooke Lake, a dam built and lake expanded from 370 to 450 ha in 1915, 610 ha in 1970, 670 ha in 1980 and 810 ha in 2002. Along with deforestation, fires and localized harvest occurred from 1920 - 1940, reducing ABC to 189 Mg C/ha, with NEP varying from -1.63 to 0.13 Mg C/ha/yr. Distributed harvest occurred 1954 - 1998, with a minimum ABC of 148 Mg C/ha in 1991. By 2012 ABC (177 Mg C/ha) and NEP (2.29 Mg C/ha/yr) had increased. Over 100 years, 2430 ha forest was cut and replanted and 640 ha deforested. CBM-CFS3 includes transfers of dissolved organic C (DOC) to aquatic systems, however data has not been available to parameterize DOC flux. DOC fluxes are modelled as a fraction of decay loss from humified soil C with a default of 100% of losses to CO₂ and 0% to DOC. Stream flow and [DOC] data from 1996 - 2012 for 3 watershed catchments, Rithet, Judge and Council were used to estimate annual DOC fluxes. Rithet, Judge and Council differed both in area % disturbed (logging or fire) over 100 years (39%, 93%, 91%) and in area % mature/old forest (>80yrs in 2012) (67%, 56%, 21%). DOC flux for Rithet and Judge ranged from 0.037 - 0.057 Mg C/ha/yr, Council averaged 0.017 Mg C/ha/yr. Low DOC fluxes were likely due to influences of a small lake in the catchment. Constraining CBM-CFS3 to observed DOC fluxes, required parameterization of humified C losses of 5.5 - 6.5% to DOC for Rithet and Judge, 2.5% to DOC for Council. All other catchments were assigned a fraction loss to DOC based on physiographic similarities to the 3 gauged catchments and the watershed C budget recalculated. Assuming none of the exported terrestrial DOC was respired to CO₂, over 100 years up to 30640 Mg C may have been sequestered in sediment.