By David R. Foster
In pre-European times forest patterns in the New England landscape were broadly controlled by such factors as climate and physiography and were locally determined by soils, aboriginal impacts, and the effects of fire, windstorms, and pathogens. This landscape was transformed by European settlement, with its accompanying broad-scale deforestation, farming, and subsequent abandonment of agriculture and widespread reforestation. In 200 years forest cover dropped to 25 to 40% and then recovered to 65 to 90%. Associated with this landscape transformation were subtle impacts:
- Decimation of native animal populations
- Widespread alteration of aquatic ecosystems and natural drainage
- Introduction of pathogens
- Changes in atmospheric chemistry
- "Perforation" of the forest landscape by suburbanization, road construction, and forestry
The reforestation of the New England landscape has been heralded as an environmental cause celebre in recent book and media coverage but major ecological questions abound. How does the composition, structure and function of the modern forest landscape compare with that of pre-European times? Does forest response to novel versus natural disturbance and stress differ? What are the legacies of historical land-use in the modern landscape? These questions address major issues in policy debates and in the understanding of ecosystem process and pattern.
With a focus on New England terrestrial and aquatic ecosystems, they form the heart of the Harvard Forest LTER Program in Central Massachusetts.
In order to address these questions, the Harvard Forest LTER program has developed an integrated research approach that involves scientists from the biological, physical, and social sciences. Including projects funded by DOE (NIGEC), the Mellon Foundation, NASA, and NSF REU and CRUI programs, this research effort annually involves more than 80 senior scientists and 30 students representing more than 20 institutions.
The integrated approach that has been taken in the Harvard Forest LTER (see design below) 1) seeks to provide a long-term perspective on the development of the forest landscape; to investigate ongoing processes and relationships through long-term measurement; to study the response to critical stresses and disturbances through experimental manipulations and controlled environment studies; to integrate results across studies, disciplines, and scales; and to apply the resultant understanding to fundamental ecological questions and societally relevant issues. Investigation across spatial scales is an important part of this program because it allows us to place intensive, site-based studies within a broad context, to understand processes that operate across landscapes and physiographic regions, and to identify and address regionally important issues.
The value of this approach is apparent in recent abstracts from the Seventh Annual Harvard Forest Ecology Symposium concerning two very different processes: the effect and legacies of 18-19th C agricultural land-use and the impact of the extreme mid-summer drought in 1995. From paleoecological and historical data Janice Fuller, Emily Russell, and David Foster reported that European settlement resulted in the most rapid rates of vegetation change in the last 2000 years, the development of novel assemblages of tree species, and a decrease in the spatial variation of forest composition across the region, with little indication that the forests were returning to their pre-settlement condition despite more than 100 years of reforestation. Glenn Motzkin and Jason McLachlan used paleoecology, dendroecology, and forest surveys to highlight the long-term impact of land-use on modern stand structure and composition and Kathleen Donohue investigated the demography, architecture and reproductive biology of clonal species to explain the resulting species:land-use patterns.
Land-use legacies also persist in terms of ecosystem process and response to disturbance. Studies reported by Jana Compton indicated that the type of 19th Century land-use was important to modern soil characteristics: formerly plowed and pastured sites have higher nitrification potentials, higher counts of autotrophic bacteria responsible for conversion of nitrate into nitrate, and lower C:N ratios than permanently forested sites. As a consequence of extensive prior land-use the forests are a net sink for carbon, as highlighted by Bill Munger and Steve Wofsy from their eddy flux measurements; Eric Davidson is evaluating the relative contributions of soils and plant biomass to this uptake.
One ultimate policy implication of these studies was addressed by John Aber, who suggested that current patterns of nitrate leaching to streams in the northeastern United States may depend more on historical patterns of land-use and disturbance than on current rates of N deposition.
A major value of the LTER program is the ability to detect and evaluate the consequences of unusual events and environmental conditions. In Central New England the summer of 1995 was marked by an extreme mid- to late-summer drought that resulted in the wilting and premature leaf fall of many understory trees and herbs, extreme soil moisture deficits, and some unexpected consequences on forest net ecosystem exchange detected at our Environmental Measurement Station tower. Bill Munger and Steve Wofsy reported that the net uptake of carbon was similar or slightly greater than in previous years, as depressed photosynthesis rates were counteracted by greater declines in ecosystem respiration. Kathy Newkirk, Jerry Melillo, and Eric Davidson documented that soil respiration rates fell precipitously as the drought intensified. Interestingly and atypically, poorly drained sites and trenched sites without roots had the greatest respiration rates, presumably due to higher soil moisture. Oaks studied through the drought by Jeannine Cavender-Bares and Fakhri Bazzaz varied in their response by size class: seedlings had a much greater depression of photosynthetic activity than overstory trees, which were more deeply rooted and able to tap deeper stores of moisture.
Apparent species differences in the ability to tolerate the drought suggests that overall species composition, as controlled by land-use and successional status, may have a large influence on overall forest response.
Scaling these observations up Dave Fitzjarrald indicated that positive feedbacks may exist between the vegetation and climate during such a drought period; associated with drought-induced water stress are an increase in bulk canopy resistance to water vapor, decreased evapotranspiration, and a decrease in water vapor in the atmospheric boundary layer. These conditions restrict cloud formation, resulting in increased afternoon temperatures, increased plant stress and a lifting of the condensation layer in the atmosphere and further decrease in cloudiness. As a result of our long-term series of measurements, this change in cloud formation, as well as plant to ecosystem responses to this short-term climatic event, were readily detected.
The annual symposium, along with monthly science meetings enable the Harvard Forest LTER program to synthesize and summarize major findings and to highlight the connections between long-term studies and short-term responses. We invite other scientists and students to join us in these collaborative efforts.
For more information: David Foster, 508/724-3302, dFoster@LTERnet.edu