The objectives of this project are (1) to quantify main vs. interactive effects of experimental warming, added and reduced precipitation on ecosystem processes and community structure and (2) to integrate experimental results into models to improve our ability of predicting ecosystem responses to warming and altered precipitation.

We will quantify main vs. interactive effects of warming, increased and reduced precipitation on (1) carbon processes, (2) nitrogen processes, (3) water processes, and (4) phenology and plant community structure. We proposed three hypotheses for each of the four aspects of ecosystem attributes, two on main effects of temperature and precipitation respectively, and one on their interactive effects, totaling 12 hypotheses. Examples of the hypotheses are:

Warming stimulates decomposition and nitrogen mineralization, resulting in increased nitrogen availability and plant uptake. Although plant nitrogen content increases, nitrogen concentration in plant tissues decreases due to warming-stimulated plant growth, leading to decreased litter quality and negative feedback to the warming effects on nitrogen cycling over time.

Effects of warming and altered precipitation are additive for soil respiration and soil carbon content but are interactive for biomass production and root/shoot ratio.

Warming stimulates evapotranspiration, causes soil surface drying, and reduces runoff. The combination of reduced runoff and increased biomass growth leads to an increase in rain use efficiency under warming.

To test those hypotheses, we are conducting a new experiment in a tallgrass prairie in central Oklahoma to manipulate temperature and precipitation.

We have constructed a warming and precipitation facility in central Oklahoma that will warm the ecosystem by 1.5/3.0 ˚C day/night temperature, and add or reduce precipitation by 50% in the treatment plots from their respective ambient levels. We will measure aboveground biomass; belowground biomass; carbon and water fluxes at leaf, canopy, and ecosystem levels; nitrogen mineralization and availability; plant nitrogen uptake and use efficiency, runoff; evapotranspiraiton; phenology, and plant community structure to detect warming and precipitation effects. Advanced statistical, modeling, and inversion approaches will be employed to understand mechanisms underlying complex interactive effects of warming and altered precipitation on various ecosystem and community processes.

This project has the potential to fundamentally improve our understanding of main and interactive effects of climate warming and altered precipitation on ecosystem processes and to improve models for their capability of projecting ecosystem responses to climate change.

2. Modeling studies of forest responses to elevated CO2:

We have been conducting modeling studies for the Duke Forest Free-Air CO2 Enrichment (FACE) project since 1996. The overall objective of the FACE project is to quantify carbon sink in forest ecosystems and to examine ecosystem processes regulating forest responses to global change. Our modeling research component of the FACE project focuses on development of conceptual and quantitative frameworks that facilitate synthesis of experimental results and extrapolation of them over time and space to predict carbon sink in terrestrial ecosystems. The FACE experiments employ a "perturbation-response" approach in which ecosystems are perturbed with a step increase in CO2 and responses to the perturbation are measured. Our study indicates that observed responses in the step experiments are transient and may not adequately represent the responses of ecosystems to a gradually changing CO2 environment.  However, by decomposing observed responses in the FACE experiments into their constituent processes and deriving parameter values from experimental results, we are able to predict ecosystem responses to a gradual increase in atmospheric CO2 that is occurring in the natural world. The major components of this project include (1) development of deconvolution and inverse analysis methods; (2) scaling up of leaf- and soil core-level measurements to estimate ecosystem fluxes and pools; (3) data-model assimilation to improve model predictions and assess data collection strategy; and (4) uncertainty analysis on extrapolation of short-term measurements to predict long-term responses of ecosystems to global change. 

[Duke FACE website]

3. Data-model assimilation to quantify regional and continental carbon sequestration:

The objective of this proposal is to improve prediction of terrestrial carbon sequestration at ecosystem and regional/continental scales with data-model assimilation techniques to estimate and constrain the three sets of parameters from AmeriFlux, soil carbon, and isotope data. We recognize that individual data sets from AmeriFlux, soil carbon, and isotope measurements each provide partial information on terrestrial carbon cycling and thus constrain subsets of the model parameters.  It is essential to synthesize and assimilate data from different sources to effectively constrain carbon modeling.  We are (1) developing a common platform for data-model assimilation in terrestrial carbon research; (2) conducting inverse analysis at AmeriFlux sites to quantify temporal variations of photosynthesis and respiration parameters; and (3) analyzing regional and continental carbon sequestration using data from FACE, AmeriFlux, soil carbon, and isotope measurements.  

4. Long-term responses of grassland ecosystems of experimental warming:

It is an ecological certainty that global warming will affect almost all ecosystem processes.  However, we still lack knowledge on long-term impacts of warming on feedback processes that operate on time scales of years, decades, and even millennia.  We have been conducting a long-term experiment since 1999 to (1) evaluate consequences of phenological shifts on biomass production and timing of carbon and nitrogen processes; (2) long-term warming effects on soil respiration and ecosystem production; (3) changes in nitrogen transformation processes under warming; and (4) changes in the plant and microbial community structure under climate warming and clipping. 

5. Seasonal and inter-annual variability in net ecosystem exchange:

Seasonal and interannual variability (SIAV) in net ecosystem exchange (NEE) is a ubiquitous phenomenon observed at almost all eddy-flux sites over the world.  Our project is to identify mechanisms underlying and then to develop process-based models towards prediction of SIAV in carbon fluxes observed at AmeriFlux sites. We employ multiple approaches, including data processing, statistical analysis, process-based modeling, and inverse analysis to (1) characterize SIAV in climatic variables and C flux variables to facilitate site, year, and variable intercomparison; (2) improve the methods of NEE partitioning into photosynthesis and respiration so that we can quantify subtle differences of the two components over years; (3) decompose observed SIAV in NEE into direct effects of climatic variables and climate-induced indirect effects on ecosystem processes; and (4) develop a process-based model that enhances our ability to predict SIAV. 

6. NSF workshop: Data-model assimilation in ecology: Techniques and applications:

Ecology has been rapidly transformed from a data-poor to data-rich field in the past decades. With such rich data sources, it is urgent to develop our research capability of data-model assimilation to convert large volumes of data into knowledge of ecosystem functioning. Thus, the research community can be prepared for a data-rich, NEON-type era of ecological
 research. The workshop is to organize a community effort to develop and apply the data-model assimilation approaches to address ecological issues and device paths to prepare the community for a data-rich era of ecological research. 

[About the workshop]

7. Research projects in China.

We are currently working with two research groups in China.  One group at Fudan University is studying biogeochemical consequences of invasion by an exotic species. Spartina alterniflora was first introduced to China in 1970s from USA to enhance sedimentation and accelerate island formations in the month of Yangtze River. Since then Spartina has aggressively invaded most of the coastal ecosystems in China. We are conducting research projects to measure carbon and nutrient pool sizes and fluxes in three islands formed in the past 50 years. The other group at Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, is applying the data-model assimilation approaches to analysis of eddy-flux data from ChinaFlux network and regional carbon fluxes in China. We are using data sets from biometrical measurements, eddy flux towers, and satellite to estimate parameters of carbon processes and uncertainties in estimated parameters, and projected carbon sinks and sources.

[Fudan University]

[Institute of Geographical Sciences and Natural Resources Research, CAS]