Using a Global Vegetation Model to Plan Local Natural Resource Management

Like politics, all ecology is local. Vegetation structure, composition, and production are strongly affected by local soils, topography, climate, and management. Local effects can be particularly strong in vegetation transition zones, areas that are often especially sensitive to climate variability, and in national parks or reserves, where management often differs substantially from surrounding areas. Natural resource management planning for future climate conditions in the latter is complicated by government-mandated or publically expected management priorities, such as maintaining a viable population of individual species of interest or the character of a landscape. Wind Cave National Park (WCNP), a 13,000-ha natural area in the Black Hills of South Dakota, lies on a prairie-forest transition where grass production controls the population viability of important wildlife and fire strongly influences the extent and character of the forest within the park. Both of these processes are quite climate-sensitive and park managers have been looking for ways to prepare for the challenges of climate change. The dynamic global vegetation model MC1 incorporates ecosystem science (C, water and N cycling; wildfire; CO2 effects), climate (temperature, precipitation, humidity), and natural resource management practices (fire suppression, prescribed fire, grazing) to simulate vegetation dynamics, thereby providing a means for natural resource managers to anticipate the effects of climate change, their management actions, and the interactions of the two on critical resources at the park scale. We parameterized MC1 to approximate the historical balance between forest and grasslands at WCNP, then ran 100-year-long simulations into the future using three fire and grazing scenarios and statistically downscaled climate projections from three general circulation models (GCMs). Under all fire/climate scenario combinations some forest remains in the park, but with lower biomass due to increased fire frequency. The simulated forest extent and biomass differ across scenarios, from an area similar to what is observed for current conditions but with lower biomass in a "natural" fire scenario to very low forest biomass over a larger area in a fire suppression scenario. The character of the landscape is thus projected to change substantially, with fire playing an important role in the outcome. Simulations suggest that fire management will become more difficult as increasing temperatures (all climate scenarios) and decreasing humidity (2 of 3 scenarios) increase the frequency of high fire danger days from approximately 10 per year in the 20th century to 20, 60, or 100 days per year by 2100. Simulated grass production in two of the climate scenarios remains similar to current levels well into the 21st century when grazing is moderate, but in the hottest, driest scenario grass production decreases to approximately half its current level by 2100. Thus, wildlife management could require alternate food sources or reduced population size. Despite the limitations of applying a vegetation model designed for the global scale to a small, local area, the results of this work have provided tangible, quantitative projections of future natural resource management challenges to park management staff, thereby stimulating the development of new management strategies.