HR: 17:15h
AN: B12G-06    [PDF]
TI: Accounting for age Structure in Ponderosa Pine Ecosystem Analyses: Integrating Management, Disturbance Histories and Observations with the BIOME-BGC Model
AU: * Hibbard, K A
EM: kathy.hibbard@oregonstate.edu
AF: Oregon State University, Department of Forest Science, Corvallis, OR 97331 United States
AU: Law, B
EM: bev.law@oregonstate.edu
AF: Oregon State University, Department of Forest Science, Corvallis, OR 97331 United States
AU: Thornton, P
EM: thornton@ucar.edu
AF: National Center for Atmospheric Research, CGD Division 1850 Table Mesa Drive, Boulder, CO 80303 United States
AB: Disturbance and management regimes in forested ecosystems have been recently highlighted as important factors contributing to quantification of carbon stocks and fluxes. Disturbance events, such as stand-replacing fires and current management regimes that emphasize understory and tree thinning are primary suspects influencing ecosystem processes, including net ecosystem productivity (NEP) in forests of the Pacific Northwest. Several recent analyses have compared simulated to measured component stocks and fluxes of carbon in Ponderosa Pine (Pinus ponderosa var. Laws) at 12 sites ranging from 9 to 300 years in central Oregon (Law et al. 2001, Law et al. 2003) using the BIOME-BGC model. Major emphases on ecosystem model developments include improving allocation logic, integrating ecosystem processes with disturbance such as fire and including nitrogen in biogeochemical cycling. In Law et al. (2001, 2003), field observations prompted BIOME-BGC improvements including dynamic allocation of carbon to fine root mass through the life of a stand. A sequence of simulations was also designed to represent both management and disturbance histories for each site, however, current age structure of each sites wasn't addressed. Age structure, or cohort management has largely been ignored by ecosystem models, however, some studies have sought to incorporate stand age with disturbance and management (e.g. Hibbard et al. 2003). In this analyses, we regressed tree ages against height (R2 = 0.67) to develop a proportional distribution of age structure for each site. To preserve the integrity of the comparison between Law et al. (2003) and this study, we maintained the same timing of harvest, however, based on the distribution of age structures, we manipulated the amount of removal. Harvest by Law et al. (2003) was set at stand-replacement (99%) levels to simulate clear-cutting and reflecting the average top 10% of the age in each plot. For the young sites, we set removal at 73%, 51% and 61% for sites averaging 9,16 and 23 years, respectively. It was assumed that changes in long-term pools (e.g. soil C) were negligible within these timeframes. In Law et al. (2003), the model performed well for old and mature sites, however, model simulations of the younger sites (9-50Y) were weak compared to NEP estimates from observations. Error for the young plots in Law et al. (2003) ranged from 150 - $>$400% of observed NEP. By accounting for the observed age structure through harvest removal, model error from this study ranged from 20-90% in young plots. This study is one of a few that have sought to account for age structure in simulating ecosystem dynamics and processes.
DE: 1615 Biogeochemical processes (4805)
DE: 1851 Plant ecology
SC: Biogeosciences [B]
MN: 2003 Fall Meeting