Effects of rising temperatures on forest composition and biogeochemical cycles.
This research is to determine the effects of plant and soil warming on tree demography, soil plant available N and P dynamics, and plant-mycorrhizal interactions in a mixed deciduous forest in the southern Piedmont region of Georgia. Demographic aspects of this experiment will be linked to two separately-funded (DOE grant to Melillo, Mohan and Clark for juvenile tree demographic work) climate warming experiments in southern New England (Harvard Forest) and Mid-Atlantic (Duke Forest) settings, facilitating synthetic analyses of demographic responses to warming across the eastern deciduous forest biome. We will use natural variations in temperature and soil moisture across the sites to examine climate impacts on demography of major eastern tree species – originating from the same genetic sources – and the importance of genetic variation to infer how future forests will look and how future species’ ranges will shift as a consequence of climate change. Mechanistic explorations of soil N, P and mycorrhizal dynamics relative to tree responses to warming will increase our capability to predict future eastern forest range shifts and compositional changes.
Five questions organize our research:
(1) Will tree species at the northern “cool” limit of their range increase in abundance in response to projected warming? (2) Will temperate tree species at the southern “warm” limit of their range decline in abundance during the coming century due to projected warming? (3) How important are underlying functional traits (shade tolerance, growth form, photosynthetic capacity) for determining demographic responses to warming? (4) Will low-nutrient soils of the Southeast show less of a response to warming than northern systems by mineralizing less N and P over short (weeks, months) and/or longer (years) time scales, and how will this impact tree responses? (5) How important are responses of mycorrhizal fungi to warming for determining plant demographic responses?
The 280 ha Whitehall Forest is a warm temperate forest managed by UGA’s Warnell School of Forestry and Natural Resources, located ~3 miles from campus in the Piedmont of Georgia (33o54'N, 83o22'W). Our research site includes 3600 m2 of even-aged, mixed deciduous forest dominated by Quercus alba, Q. rubra, and A. rubrum, and is representative of deciduous forests of the Georgia Piedmont (Forkner and Hunter 2000). Soils are of the Cecil/Pacolet association (fine, kaolinitic, thermic Typic Kanhapludults) with low organic matter content and medium to slow permeability. Mean January temperature is 9.1oC and mean July, 28oC. Precipitation is evenly distributed throughout the year and growing season droughts are relatively common due to high evapotranspiration. The long-term precipitation mean is 125 cm yr-1 (1945-2003); recent mean is 113 cm yr-1 (2003-2007). Our site at Whitehall Forest complements the upland Q. rubra, Q. alba, A. rubrum forests examined in separately-funded experiments at Duke and Harvard Forests. All three sites are naturally-recruited, unmanaged stands originating in the first half of the 20th Century following agricultural land abandonment. The proposed site is in a long-established research area and has easy road access, line power and is near major laboratory facilities. We will leverage existing infrastructure for our experimental manipulations.
We have established six experimental blocks; three blocks in closed canopy conditions, and three blocks in canopy gap conditions. The 1500-m2 gap was manually cleared to minimize site disturbance (summer 2008). Each block will contain three open-top chambers – one maintained at ambient temperature, one with soils at +3oC above ambient, and one with soils and plants at +3oC above ambient using infrared (IR) heat lamps to warm the trees. July 2009 mid-day light levels in the nine gap chambers averaged 92±2% full sunlight, and the nine chambers in the forest understory averaged 2.1±0.2%. Each block will have an additional 18.3-m2 plot marked off as an unchambered control (N=3 in both Gap and Forest understory sites). This plot is used as the temperature reference control for the other chambers in the block, and will have the same tree seeds planted and monitored as the chambered plots.
The experimental design incorporates soil-warming cables in chambers. Previous work shows that heating air without heating soils creates anomalous conditions because the target soils are surrounded by an ‘ocean’ of unheated soils, and heating with just IR lamps only warms the soil surface by ~1oC and provides no warming at depths >1 cm (Harte and Shaw 1995, Melillo 2002).
Species will include those near the southern (Pinus virginiana, Quercus rubra, Q. prinus, Q. velutina) and northern (Magnolia virginiana, Q .nigra, Q. virginiana, Pinus palustris) limits of their current ranges, as well as those with more extensive distributions. For species currently growing in the Georgia Piedmont, seeds will be obtained from and near Whitehall Forest. In addition, we will do a cross-site, “common garden” experiment with species occurring at Whitehall, Duke and Harvard Forestswhere we will plant at each site seeds collected locally, as well as seeds collected at the other sites, to assess the importance of geographical genetic variation in response to global change (Mohan et al. 2004). By planting seeds from the same species collected across the eastern biome, this research will inform questions of the importance of genetic variability in responses to changing climate. Further, demographic data from our Georgia seedlings will be used in Bayesian hierarchical modeling incorporating results from all three sites to forecast how eastern temperate tree species will differ in abundance and range response to climate change. We will also plant remote species from the Coastal Plain to evaluate recruitment and competitive dynamics of potential immigrants to the Piedmont region with climate warming. Potential immigrants are selected based on their southern proximity to Whitehall Forest, and on predictions of future distributions from CE models (Iverson and Prasad 2002). For example, longleaf and slash pines, sweetbay magnolia and live oak are common components of sandy Coastal Plain forests, and in a warmer climate, these species are predicted to establish in Piedmont forests, yet it is unclear how well they will survive and compete in the clay soils of the Piedmont. Southern magnolia is another Coastal Plain native, but has become naturalized following anthropogenic planting in the Piedmont. This species may become more competitive in a warmer Piedmont climate.