Potential carbon storage at the landscape scale in the Pacific Northwest, United States.

摘要

Estimates of potential carbon (C) storage can be used to constrain predictions of future carbon sequestration and to understand the degree to which disturbances, both natural and anthropogenic, affect C storage. An upper bound on C storage in the Pacific Northwest (PNW) of the United States was estimated using field data from old-growth forests, which are near steady-state conditions and have been relatively undisturbed for long periods of time. The sites were located across a broad, biogeographical gradient in western Washington and Oregon, allowing comparison of potential carbon storage given a wide range of climate, soils, and vegetation conditions. Total ecosystem carbon (TEC) ranged from 195 Mg C ha−1 in eastern Oregon to 1127 Mg C ha −1 at the Oregon coast. A simple, area-weighted average of TEC to a soil depth of 1 m was 671 Mg C ha−1. Compared to estimates of current C storage, up to 338 Mg C ha−1 could be stored in addition to current stores in this region. A new model called MAXCARB was developed to predict potential carbon storage over a large area (approximately 105 ha), in part to better understand the role of disturbances on potential carbon storage. MAXCARB simulates the effects of climate, soils, or vegetation on potential carbon storage at steady state, for a range of natural and anthropogenic disturbance regimes. Initial results indicate that as the average interval between disturbance events increased, the steady-state C stores at the landscape scale increased. Predictions were well correlated to observed C stores in the PNW. Spatial interactions affect C flux processes at multiple levels of spatial interactions. Using another model, STANDCARB, the relative effect of edge-induced, tree mortality (mainly due to wind), and light limitations, on C dynamics were assessed for several artificial forest landscapes. Emergent behaviors resulting from the interaction of these processes were present at all levels of spatial interaction (stand and landscape). However, the magnitude of the emergent behaviors depended on the spatial structure of the landscape and the level of spatial interaction that was considered. When wind-mortality was high (8 times above natural mortality rates), the dynamics of C processes in fragmented landscapes was not captured using an additive approach. The spatial arrangement of patches on the landscape led to emergent behaviors for one case. However, in many cases, emergent behaviors were not significant or could be accounted for with traditional modeling methods.