Fig. 4

a A comparison of carbon (C) transfers out of the ecosystem, either through emissions (combustion or decay) or transfers to harvested wood products (HWP) in the year of disturbance, for different disturbance matrices. The main pathway for removal of C in the year of harvest is transfer to HWP and for wildfire it is combustion of dead organic matter (DOM). Removals of C from the ecosystem in the year initiating surface mining are almost double that of harvest or wildfire because merchantable wood is removed to HWP and the remaining biomass and DOM are piled and burned to clear the land. The amount of C removed in the year of disturbance from aboveground pipelines and corehole pads far from a mill are similar to surface mining but all are emissions to the atmosphere, with no transfer of C to HWP. Emissions of C from seismic lines (regardless of vintage) are similar to one another (low and from enhanced decomposition) and much lower than for the other disturbance types shown here. Results were obtained from application of the disturbance matrices to a theoretical 1 ha of forest lands. PDM, primary disturbance matrix (see Table 1). “DOM dist. Emission” are emissions directly attributed to the disturbance, whereas “DOM ann. Emission” are emissions from the annual process of decay. b The relationship between the ecosystem indicators net primary productivity (NPP), heterotrophic respiration (Rh); and net ecosystem productivity (NEP) in the years after disturbance from wildfire, harvest, and old seismic lines (established between 1960–1999) simulated by the Generic Carbon Budget Model (GCBM). When NPP recovers immediately after disturbance, the time to recover to a net sink is influenced by the type of disturbance because it determines the amounts and types of biomass that transfers to DOM pools that subsequently release C as they decay (Rh). Recovery is fastest after wildfire (10 years), followed by harvest (12 years) and then seismic lines (16 years). If recovery of NPP is delayed (e.g., difficulties in initiating reclamation success on seismic lines) for 20 years (in this example) the effect is to further delay recovery of the system to a net sink (34 years vs. 16 years) and total net emissions over the simulation period are higher when reclamation success is delayed