Robert Scheller

Temperate forests are responsible for about 14% of total carbon (C) storage and 26% of net C uptake globally. In North America, forests are the largest land sink, annually absorbing 7-24% of US fossil fuel emissions and storing large quantities of C in their wood and soils. However, much of that results from recovery from agricultural abandonment, fire suppression, and reduced logging, and the future of this C sink is highly uncertain. Higher rates of individual tree mortality, more large-scale disturbances (particularly fire), and higher decomposition rates due to higher temperatures may increase the variability and/or magnitude of sink strength in U.S. forests over long-time spans. Coupled with these emerging dynamics are the spatial legacies of previous forest management practices, which may render some sinks more vulnerable or resilient to climate change or disturbances due to different stand ages and compositions. Newly implemented but untested management practices (young growth harvesting and maximizing C for sale on carbon markets) add further ecological and economic complexity. Yet our knowledge of how management choices will influence the magnitude and spatial patterning of C cycling and long-term C storage at broad scales is relatively poor. Our work will address four fundamental questions: How has historical management influenced the spatial distribution of C stocks and fluxes relative to natural disturbances? Does that spatial pattern promote stability or vulnerability to climate change and natural disturbance? To what degree and where are current management practices constrained by historical patterns of management? What will these shifts in management practices and their historical constraints mean for future C distributions and their value on C markets under climate change? We will use a combination of data integration, remote sensing, and modeling to answer these questions.

Recent studies have shown that forest restoration via mechanical thinning and prescribed fire can have beneficial effects on drought mortality even with future changes in climate. Additional work has shown that forest restoration can be effective in reducing fire severity under climate change thus reducing the tree mortality from fire in the future. But little is known about the interacting effects of fire and drought on tree mortality and the effect of forest restoration on these two processes. TNC will develop statistical models of drought mortality for 3 5 conifer species in the Four Forests Restoration Initiative landscape based on Forest Inventory and Analysis (FIA) data and climate data. NCSU will incorporate those relationships into the LANDIS-II model. NCSU, with help from TNC, will update the Four Forest Restoration Initiative (4FRI) landscape LANDIS-II input data, including updating initial communities and creating input data for the wildfire SCRPPLE extension. NCSU, with assistance from TNC, will run all LANDIS-II simulations for the 4FRI landscape to evaluate drought mortality, mortality from fire, and the effect of restoration. NCSU will take a lead role in writing a manuscript on drought mortality, fire mortality, and climate change in the 4FRI landscape, TNC collaborators will be co-authors. TNC will take a lead role in drafting a manuscript on the interacting effects of drought and fire and the effect that forest restoration can have on both in a changing climate, NCSU collaborators will be co-authors.

Significant changes to the historical disturbance complex have altered ecological function in many Southern Appalachian forested ecosystems. To maintain oak and hickory and perpetuate the forest types that have been ecologically and economically important to the region, it is necessary to seek alternative management approaches that will restore species, structural, and functional complexity to the Appalachian region. We are proposing to evaluate oak regeneration under traditional silvicultural systems and use these results to guide the design of an alternative expanding-gap approach; to initiate baseline sampling imperative in the long-term evaluation of the expanding-gap approach; and use stand- and landscape-scale simulations to test the degree to which a gap-based, silvicultural approach will increase: 1) oak regeneration, 2) structural complexity and species diversity; and 3) carbon sequestration and storage. Specifically we will evaluate the capacity for alternative hardwood management practices to increase the regeneration of oak and hickory within the Southern Appalachian mixed oak forest. We will assess the interactions among forest structure, composition, regeneration and ecosystem processes and integrate our empirical research into a spatially-explicit landscape model to simulate multiple scenarios of management, disturbance, and climate interactions. With strong support from local and regional forestry professionals and non-government organizations, our team of University and Forest Service scientists will ensure that the results will reach managers and resource professionals. We specifically address AFRI Program Area D, Priority 1 with the goals of advancing our understanding of processes and interactions and assessing and developing new management practices to improve ecosystem services.

Many species of forest birds have experienced population declines in recent decades and are at risk of further declines due to climate change. Active management is necessary to maintain habitat for these species, but shifts in the structure of forests caused by climate change or disturbance may necessitate changes in management strategies. Because silviculture or other management strategies directly affect forest structure, models of bird habitat which are sensitive to forest structure are needed, but these are not widely available at relevant spatial scales. We will quantify direct and indirect threats to bird habitat and assess climate-adaptive management strategies across the southern Appalachians.

Isle Royale is internationally recognized for its populations of wolves and moose. Despite wolf predation on the moose population, both the short and longer-term impacts of moose browsing can be seen across the island. In response to a decade-long decline in the wolf population of Isle Royale National Park, ultimately resulting in just two wolves remaining, the National Park Service (NPS) began reintroducing wolves to the island over the past winter. It is not, however, possible to disentangle all of the factors that might influence patterns of vegetation change over large landscapes with traditional experimental approaches or by conducting observational studies on their own. We will combine landscape-scale simulation models and data from experiments and observational studies to foster a more integrated understanding of vegetation, herbivore, carnivore dynamics. By using a simulation model to evaluate hypothetical scenarios, we will isolate the possible effects of one or more factors while holding others constant. Climate change will also interact with wolf predation and moose browsing to alter the future forests of Isle Royale and will be considered in our forecasting.

There are two major objectives of this project. The first is to reparametrize and re-run LANDIS-II for the Tahoe Central Sierra Initiative landscape. The output of these model runs will then be used as input to a hydrologic model. The second objective is to couple the output of a FORSYS model run, which will provide an optimized portfolio of management activities, to LANDIS-II, and to run LANDIS-II over a 40 year time horizon in order to verify that the management portfolio provides adequate levels of ecosystem services. A third objective is to lead or assist with at least one peer reviewed publication or report.

We will build on prior efforts to develop a spatiotemporal statistical model for the 2011 Texas drought that relates forest conditions measured from FIA data to the SPEI or a similar drought measure. The model will account for trends in drought and mortality over time and space, as well as variations in drought effects based on forest species composition, drought tolerance of tree species, soil moisture, and other climatic, biophysical, and environmental correlates.

While managers continue to set goals for increasing prescribed fire in the southern Appalachians, population density is also increasing and WUI communities are expanding. Locations of existing communities and future WUI development strongly influence locations and timing for implementing prescribed burns. Little work has been published that addresses cumulative smoke exposure from prescribed fire in southern Appalachian communities, with specific consideration for existing community vulnerabilities. The proposed research responds to the issues of smoke exposure, community vulnerability, and expanding WUI development using a long-term modeling approach. Landscape change, including dynamic fuels and fire emissions, is linked to VSMOKE to model multi-decadal smoke dispersal from priority burn sites in western North Carolina. Modeled emissions and smoke dispersal will be spatially overlain with social vulnerability indices and hypothetical WUI development, such that managers can more readily analyze and understand community characteristics adjacent to prescribed fire locations.

Emerging plant disease and pest outbreaks reduce food security, national security, human health, and the environment, with serious economic implications for North Carolina growers. These outbreaks may accelerate in coming decades due to shifts in the geographic distributions of pests, pathogens and vectors in response to climate change and commerce. Data-driven agbioscience tools can help growers solve pest and disease problems in the field more quickly but there is an urgent need to harness game-changing technologies. Computing devices are now embedded in our personal lives with sensors, wireless technology, and connectivity in the “Internet of Things” (IoT) but these technologies have yet to be scaled to agriculture. Our interdisciplinary team will build transformative sensor technology to identify plant pathogens, link local pathogen data and weather data, bioinformatics tools (pathogen genotypes), and use data driven analytics to map outbreaks, estimate pest and pathogen risk and economic damage, in order to coordinate response to emerging diseases, and contain threats. Sensor-supported early and accurate detection of pathogens before an outbreak becomes wide-spread in growing crops will significantly reduce pesticide use and increase crop yields.

Successful outreach to forest managers and landowners requires improved visualization of landscape change model outputs. This grant will provide funding to add/improve features for the LandViz visualization tool and other changes necessary for the LANDIS model.