Forest Resources Center Research

Partitioning in Trees and Soil (PiTS) Part 1 - Loblolly Pine

- A field research facility for developing dynamic carbon and nitrogen partitioning representations for global models and applications.

"Belowground observation units are placed in a large pit dug to a depth of 1 meter, adjacent to eight trees. Instruments are placed above and below ground to track carbon dynamics."

Members of the Oak Ridge National Laboratory (ORNL) research team conducting the experiment state that the objective of this experiment "is to improve the carbon partitioning routines in existing ecosystem models based on the concepts gathered from plant partitioning models and tested against field observations and manipulations." They propose "to use short-term, comprehensive field measurement of processes related to carbon partitioning from leaves to roots and roots to soil." The experiment focuses on "belowground carbon partitioning." They further state that "An improved understanding of the relative amount and fate of belowground partitioning will lead to improvements in model representation of carbon partitioning and the fate of carbon under elevated carbon dioxide and other climate perturbations".

Experiment Background

According to the team, the starting point for models of ecosystem carbon cycling, "is with a representation of photosynthetic assimilation of atmospheric carbon dioxide by vegetation." The next step is to model how a plant internally distributes carbon, a process called carbon allocation or partitioning.

In their description of the experiment, the team states how studies have shown that "an incomplete understanding of carbon partitioning currently limits the capacity to model forest ecosystem metabolism and accurately predict the effects of global change on carbon cycling." In particular, they cite the need for ecosystem model improvements "in their representation of partitioning to belowground structures and processes because plant-soil interactions are central to the biogeochemical cycles of carbon, nitrogen, and water, and they are important for modulating climate change impacts and regulating feedbacks of greenhouse gas emissions."

According to the team, "by coupling short term experimental manipulations with intensive measurements above- and belowground, we will advance our understanding of the biological and environmental influences on carbon partitioning, and their consequences for carbon-nitrogen interactions in plant and soil."

Experiment Construction

Eight loblolly pine (Pinus taeda) trees - less than 10 meters in height - have been chosen for the experiment. Belowground observation units are placed in a pit, dug to a depth of 1 meter, adjacent to the trees. Instruments are placed above and below ground to track carbon dynamics.

Total carbon uptake by the trees is calculated from measurements of Photosynthetic Active Radiation (PAR) by sensors above the tree canopy, transpiration by sapflow gauges on the stem, and leaf-level photosynthesis using gas exchange cuvettes, with access to the canopy provided by a hydraulic lift. In the belowground system, minirhizotron tubes track the timing and amount of root growth. Root observation windows allow access to newly produced roots for chemical analyses.

Also, carbon dioxide analyzers, soil gas and water samplers provide additional data on carbon and water dynamics in the soil.

The experiment is designed so that the carbon balance of the tree canopy can be altered. This provides data on how the carbon flux belowground varies with short-term changes in the canopy-integrated stomatal conductance and photosynthesis.

On September 1, 2010, the ORNL researchers enclosed the pines in a large plastic chamber and released carbon dioxide that was highly enriched in the non-radioactive, stable isotope 13C. The researchers are currently tracking the 13C signal in the leaves, wood, roots and soil in order to determine where carbon dioxide taken up during photosynthesis is partitioned within the tree and soil system.

During the course of the experiment, the team anticipates many other manipulations, including nitrate or ammonium applications to specific locations - at different depths - to measure root-specific nitrogen uptake rates as a function of soil depth. They also cite the value of the chosen facility's flexibility.

Model Interaction

According to the research team the "goal is to develop and test a dynamic carbon partitioning module that will be integrated into the Community Land Model (Thornton et al., 2007)*. This goal will need to be achieved without substantially increasing the computational burden of the model, which is designed for global-scale operations. Thus it is important to find out what processes are essential and how to represent them efficiently." They plan to "start from the dynamic, individual tree based carbon partitioning model and then progressively simplify its structure to reach a balance between computational demand and process representation." They describe the model as operating "at half-hourly time steps" and "driven by meteorological measurements at the site." "Sugar measurements at different positions of the tree and different times of the day and the season will be used to test model behavior. Allometric relationships obtained with destructive sampling at the end of the study will be used to test the long-term integration of the model."

Phase 1 of the experiment - the first of three phases - began in April 2010 with the digging of the pit and placement of the measuring instruments. Observations during the 2010 summer growing season have generated preliminary data for use in evaluating the second phase of the experiment and for input in the dynamic carbon partitioning model. Phase 2 is scheduled to begin in December 2010.

Members of the ORNL research team include:
ORNL Environmental Sciences Division
Richard J. Norby, Task leader
Colleen M. Iversen, Plant-soil interaction
Charles T. Garten, Jr., Stable isotope labeling
David J. Weston, Physiological/biochemical measurements
Jeffrey Warren, Sap flow and water relations
Peter E. Thornton, Ecosystem model
Lianhong Gu, Physiological model
Joanne Childs, Minirhizotron measurements
University of Tennessee Institute Of Agriculture
Jennifer Franklin, Associate Professor, Department of Forestry, Wildlife and Fisheries

*Thornton PE, Lamarque JF, Rosenbloom NA, Mahowald NM. 2007. Influence of carbon-nitrogen cycle coupling on land model response to CO2 fertilization and climate variability. Global Biogeochemical Cycles 21:15.

Empirical results from the initial FRREC PITS experiment in loblolly pine were published in: Warren JM, Iversen CM, Garten CT, Norby RJ, Childs J, Brice DJ, Evans RM, Gu L, Thornton PE, Weston DJ (2012). Timing and magnitude of carbon partitioning through a young loblolly pine (Pinus taeda L.) stand using 13C labeling and shade treatments. Tree Physiology, 32:799-813.