Biotic and abiotic controls on whole-ecosystem carbon balance in forest ecosystems


Forested ecosystems, which represent the largest component of the terrestrial sink for atmospheric CO2, play a critical role in the global carbon cycle. Better understanding of the mechanisms underlying long-term trends and annual-to-decadal-scale variability in forest carbon sink strength (J103, J101) is therefore a key to reducing uncertainties about future levels of atmospheric greenhouse gases, and hence global temperature. From an applied perspective, research in this area has far-reaching social and economic policy implications.


We are actively involved in the “flux tower” measurements conducted at the Bartlett Experimental Forest (New Hampshire) (J058, J031) and Howland Forest (Maine) (J033) AmeriFlux sites. At these long-term research stations we use the eddy covariance technique to measure surface-atmosphere exchanges (“fluxes”) of CO2, water, and energy. These high-frequency (30 minute resolution), continuous (24 hours a day, 365 days a year) measurements offer profound insights into relationships between environmental drivers and ecosystem function, and how these vary over time (e.g., diurnal, seasonal, and annual cycles; J098, J056, J033) and across sites (e.g., between evergreen and deciduous forest ecosystems; J092, J081, J062, J047, J045). Of primary importance is the ability to use such data to study ecosystem carbon cycling. For example, integrating CO2 flux measurements made over the course of a whole year yields an estimate of the total carbon balance (i.e., net carbon sequestration) of the forest ecosystem—is it a sink or source for atmospheric CO2 (J103)? At a general level, we strive to answer the general question: Is global change increasing or reducing the sink strength? We can also ask more specific questions such as: What is the impact of a hot, dry summer on the forest carbon balance (J068, J033, J018)? Or, how much less carbon is sequestered in a year with cool, damp spring (J075)? Related to this, we have recently begun new measurements of forest-atmosphere fluxes of methane to investigate the role of CH4 in the overall ecosystem-scale greenhouse gas budget at Howland.


Increasingly, we are applying cutting-edge model-data fusion techniques (J064, J055, J036, J033) to provide an integrated framework for the analysis and interpretation of eddy covariance flux data. This research is highly analytical and quantitative, and is currently at the forefront of carbon cycle model-data fusion research. In a recent study (J076, see figure below), we combined tower-based measurements of whole-ecosystem CO2 exchange together with plot-level measurements of forest composition and biomass, leaf area index (J085), litterfall, and soil respiration (J059, J046), to constrain the parameters of process-based forest carbon cycling model. The main objective of this work was to investigate relationships between abiotic factors (e.g. weather), internal system dynamics, and ecosystem processes. A key secondary goal of this analysis was to examine how combinations of different types of data reduced model uncertainties and improved model predictions. Recent work within the lab (J109, J103, J090) has built on this foundation.


Both DALEC (shown below, J076), and the more recently developed FöBAAR (J103), have been used extensively in our lab to study carbon cycling processes in New England forests.