Simulations of snag dynamics in an industrial Douglas-fir forest

Abstract

Industrial silviculture is known to reduce snag density, but snag dynamics in industrial forests are poorly understood. I developed a simulation model that integrated a snag model and a well-known forest growth model, the forest vegetation simulator (FVS). A new snag model was developed by averaging the outputs of four independently created snag models. The four models were for Douglas-fir snags in forests west of the Cascade Crest in Oregon, Washington, and British Columbia. Forest growth and snag dynamics were simulated under a typical silvicultural regime and current occupational safety and environmental regulations. The results indicate that management practices like those simulated yield: (1) small and medium diameter snags at moderate densities (20 snags per hectare (sph)) for short periods of time (5–10 years); (2) a snag population with high temporal variation fluctuating between 4.2 and 22.5 sph; (3) mean densities of small, medium, and large snags equal to approximately 3.9, 6.2, and 0.1 sph/decade; and (4) a soft snag density of 0.1 sph/decade. Snag recruitment curves generated through simulations showed that to increase mean snag density per decade by 1 sph, the number of snags retained must be increased by about 1.4 sph. The mean density of snags per decade produced under the typical silvicultural regime was projected to be about 20% that found in unmanaged stands. The density of large snags was projected to be less than 1% that found in unmanaged stands. Sensitivity analysis showed that simulator output was slightly sensitive to error in snag model parameter estimates. Considering the lack of long-term empirical studies on snag dynamics, and the dearth of information on compliance with safety and environmental regulations, models like this one represent the best available scientific information upon which to base forest management decisions affecting snag resources.