William Hoffmann

Understanding the factors that determine the current distribution of biomes is fundamental for projecting how vegetation will respond to future climates and disturbance regimes. Tropical savanna-forest boundaries mark the transition between the two most extensive tropical biomes, yet the main factors determining the location, structure, and dynamics of savanna-forest boundaries are poorly understood. Our understanding has been limited by strong positive feedbacks and other non-linear processes that result in complex dynamics and hysteresis (the dependence of a system not only on its current environment but also on its past environment). Hysteresis is particularly problematic because it can introduce large errors in model behavior if key processes are not represented realistically. We will combine field data and modeling to test for and quantify sources of hysteresis in savanna-forest dynamics at multiple scales. The primary objectives of this research are: (1) quantify processes that underlie switches in biome states between savanna and forest; (2) use this information to refine and parameterize the CLM(ED-SPITFIRE) model for simulating savanna-forest dynamics; and (3) perform simulations to understand environmental controls on the distribution of tropical savanna and forest ecosystems, with particular attention to understanding causes of hysteresis. Within the first objective, we have three specific aims designed to fill gaps in our understanding of vegetation-fire feedbacks that regulate the dynamics of savanna. These aims are: (1) understand the unique roles of savanna and forest tree species for mediating ecosystem dynamics during biome switches from savanna to forest; (2) understand the roles of savanna and forest tree species under the reverse process of forest retreat; (3) quantify flammability thresholds across savanna-forest gradients to understand controls over vegetation-fire feedbacks during biome switches.

Abstract 1. Objective: Management of military lands aims to ensure long-term persistence of threatened and endangered species while sustaining the military mission (DoD Instruction 4715.03). Evaluating local extinction risk of plants is complicated by the long lifespan and slow rate of dynamics of most species. To this end, we will combine new data with a rare-plant monitoring study that is putatively the largest, longest, and most comprehensive undertaken on military lands with the aim to 1) quantify the factors that influence local extinction, including population structure, management activities, and physical environment, 2) generalize these results across 3) provide resource managers with guidelines for assessing risk of local extinction and for remediating this risk. In addition to meeting these aims, we will re-evaluate a set of hypotheses that were originally conceived for this study and that were previously tested by Gray et al (2003) when only relatively short data series was available. 2. Technical Approach: For over 20 years, the Endangered Species Branch of Ft. Bragg has monitored 1396 populations of 45 rare plant species, resulting in the Monitoring and Assessment of Rare Species (MARS) database. The MARS database has over 32,000 population-years of data, thereby offering a rich resource for understanding the drivers of local extinction. We will combine this database with information on plant species traits, local site characteristics, management history to identify generalizable risk factors for local extirpation of rare plant populations. Identifying these risk factors will require detailed information regarding each population and species in the MARS database. Information on population size and distance to nearest conspecific population extracted from the MARS database. Information on management history and local conditions will be extracted by overlaying the geographic coordinates of each population on existing GIS layers. The influence of these factors on population persistence will be evaluated by using logistic regression and GLM methods to test for their influence on the probability that an existing population disappears with in the interval between successive censuses. 3. Benefits: The findings will provide DoD resource managers on multiple southeastern installations with guidance related to species management. The expected benefits of the proposed work are: ��������������� Deeper understanding of the metapopulation dynamics of the 45 rare plant species on Fort Bragg ��������������� Extension of the initial inferences to other species and installations in the southeastern United States through incorporation of functional traits ��������������� Provide a model framework for other ecosystems

The proposed research will examine the effects of fire on population dynamics and physiological ecology of common and rare species in inclusional wetlands

Our objective is to quantify the effect of warming on assimilation, transpiration, and growth of saplings of four dominant tree species within an existing warming experiment. We will combine field measurements and process modeling to deconvolve the temperature and moisture effects of warming on assimilation and transpiration. Finally, we will generate quantitative estimates of the uncertainty introduced by our incomplete knowledge of future changes in soil and atmospheric moisture under warming scenarios.

Habitat loss, fragmentation, and altered disturbance regimes are among the greatest threats to biodiversity and species conservation. These situations can be ameliorated with appropriate management, but only when there is sufficient information on which to base this management. Unfortunately, it is often necessary to implement management strategies with an incomplete understanding of the demography and natural history of the target species. We propose population and community studies to provide this information for five species (Amorpha georgiana var. georgiana, Astragalus michauxii, Lilium pyrophilum, Pyxidanthera brevifolia, Stylisma pickeringii) selected for intensive study at Ft. Bragg. For L. pyrophilum, we also propose a population genetics study to quantify genetic structure, outcrossing rates, and level of genetic differentiation in relation to its sister species.

Savannas and evergreen forests are the two most important tropical vegetation types in terms of area, biodiversity, total carbon stocks, and use by humans. These biomes originally comprised 82% of the tropical land area (Houghton and Skole 1990), and are home to approximately 40% the global human population and perhaps 50% of all species of organisms (Solbrig et al. 1996). The savanna-forest boundary represents the natural limit of distribution of tropical forest, and is therefore fundamental for understanding how the extent of tropical forest will respond to changing environments. The primary objective of this study is to understand how species traits and positive feedbacks govern ecosystem processes at the savanna-forest boundary.

Species traits determine plant responses to the environment as well as plant effects on the environment. This interrelationship between plants and the environment often results in feedbacks that override the direct effects of the environment on plant distributions. In particular, positive plant-environment feedbacks can result in abrupt and stable ecotones, even where environmental gradients are gradual or non-existent or where past environmental change otherwise would have caused a shift in the location of the ecotone. Such feedbacks decouple vegetation distributions from climatic, hydrological, and edaphic factors, thereby confounding attempts to understand and model vegetation response to the environment. We hypothesize that positive feedbacks have an overriding role in the dynamics of savanna-forest boundaries, and that these feedbacks are largely dependent upon the contrasting structural, physiological and ecological characteristics of savanna and forest species. Understanding and modeling the response of tropical savanna and forest to changes in land-use, climate, and disturbance regimes will require a thorough understanding of the feedbacks that reinforce the savanna-forest boundary as well as non-feedback processes that destabilize the savanna-forest boundary Intellectual merit. We will integrate ecophysiological methods, experimental manipulations, studies of vegetation dynamics, and ecological modeling to understand how the ecology and physiology of savanna and forest trees govern the dynamics, structure, and functioning of savanna-forest boundaries. These activities will utilize a comparative approach based on phylogenetically-independent contrasts to meet three specific objectives 1) understand the role of environment-modifying behavior and plant-environment feedbacks at the savanna-forest interface 2) understand the effect of resource availability and disturbance on the ecological dynamics of savanna-forest boundaries and consequently the distribution of the two most important tropical biomes, 3) quantify the role of tradeoffs and other causes of correlated evolution among physiological and ecological traits in determining species traits under contrasting environments. Broader impacts. Savannas and evergreen forests are the two most important tropical vegetation types in terms of area, biodiversity, total carbon stocks, and use by humans. Unfortunately, they are currently subjected to extremely high rates of burning, logging, and large-scale landuse change. The end result is a "savannization" of previously forested areas whose long-term fate will depend on the very processes we propose to investigate that occur naturally at savanna-forest boundaries. This proposal involves a diverse team of investigators united by an interest in elucidating the fundamental processes determining the distribution, function, and dynamics of tropical ecosystems. This team has a history of working together in successful collaborations that have exposed US and foreign students to an international research environment. All members are equally dedicated to strengthening scientific research in Latin America through capacity building of Latin American students. The proposed research will provide students from the University of Brasilia and probably from other Latin American institutions with an opportunity to interact and collaborate with the investigators of this grant as well as with students from the US. This proposal includes funding for one US graduate student and one US undergraduate student per summer to conduct research at the field site in Brazil. In doing so, we will stimulate interest and expertise in tropical ecology among several US students while fostering interactions that we hope will grow into long-term collaborations.

Fire plays a dominant ecological role over much of the earth?s land surface, influencing ecosystem productivity, biogeochemical cycling and biome distributions. Fire frequency and intensity are highly sensitive to meteorological conditions and will likely respond quickly and strongly to climate change, so it has been argued that fire is the mechanism most likely to cause rapid changes in vegetation distribution under changing climates. Although we have a solid understanding of how the behavior of individual fires responds to meteorological conditions, predicting changes in fire regimes under future climates is beset by problems. Foremost among these is that over large parts of the globe, humans are directly responsible for setting most vegetation fires. This makes future projections difficult because fire use by humans also responds to climate. We will use combine sensing products (MODIS active fire mask), ground meteorological data (WMO Database) and an existing fire behavior model (McArthur forest fire danger index) to quantify, at regional scales, how human activity influences how fire occurrence responds to fire. This information will be used to produce an improved fire model for climate change research. Inferring the human role will be possible by comparing regions with very different traditions of fire use as well as comparing seasonal versus inter-annual signals in the climate-fire relationship. This research will contribute to the NASA Strategic Roadmap of Earth Science and Applications from Space, the NASA Science Roadmap of Carbon Cycle and Ecosystems, and the NASA Applications Roadmap of Ecological Forcasting.

I propose to conduct metapopulation and plant community studies in support of research on appropriate management for five rare plant species that occur on Fort Bragg Military Reservation.

I propose to integrate population-level dynamics of rare species, including demographic and ecophysiological research, into a community and ecosystem-level approach that helps to understand both species-species and species-environment interactions. The research will be conducted on five rare plant species that occur on Ft. Bragg military reservation.