Modeling Forest Composition and Carbon Dynamics Under Projected Climate-Fire Interactions in the Sierra Nevada, California

Abstract:

The Sierra Nevada Mountains are occupied by a diversity of forest types that sort by elevation. The interaction of changing climate and altered disturbance regimes (e.g. fire) has the potential to drive changes in forest distribution as a function of species-specific response. Quantifying the effects of these drivers on species distributions and productivity under future climate-fire interactions is necessary for informing mitigation and adaptation efforts. In this study, we assimilated forest inventory and soil survey data and species life history traits into a landscape model, LANDIS-II, to quantify the response of forest dynamics to the interaction of climate change and large wildfire frequency in the Sierra Nevada. We ran 100-year simulations forced with historical climate and climate projections from three models (GFDL, CNRM and CCSM3) driven by the A2 emission scenario. We found that non-growing season NPP is greatly enhanced by 15%-150%, depending on the specific climate projection. The greatest increase occurs in subalpine forests. Species-specific response varied as a function of life history characteristics. The distribution of drought and fire-tolerant species, such as ponderosa pine, expanded by 7.3-9.6% from initial conditions, while drought and fire-intolerant species, such as white fir, showed little change in the absence of fire. Changes in wildfire size and frequency influence species distributions by altering the successional stage of burned patches. The range of responses to different climate models demonstrates the sensitivity of these forests to climate variability. The scale of climate projections relative to the scale of forest simulations presents a source of uncertainty, particularly at the ecotone between forest types and for identifying topographically mediated climate refugia. Improving simulations will likely require higher resolution climate projections.