Partitioning in Trees and Soils (PiTS): A field research facility for testing dynamic carbon partitioning representations within global models

Background/Question/Methods

The PiTS research project was established to assess performance of carbon (C) partitioning routines in mechanistic ecosystem models. Our objective was to measure C flux through the plant-soil system following pulse-labeling with ¹³CO₂ and short-term experimental manipulations of gross primary production (GPP) and C availability, and to use results to test and potentially modify partitioning routines in large-scale ecosystem models. The first manipulation was a stand of young loblolly pine trees, which were labeled with a pulse of ¹³C-enriched CO₂ and then subjected to shade treatments. The progression of ¹³C was tracked from the atmosphere through foliage, phloem, roots, and respiratory soil CO₂ efflux. Concurrent measurements used for model parameterization included photosynthesis, sap flow, root and stem growth, soil water content and soil CO₂ efflux. A second study in dogwood trees used seasonal ¹³CO₂ labeling to address temporal patterns of C partitioning, and also included assessment of C flow into and through mycorrhizal fungi. A final study took advantage of the residual depleted δ¹³C signature of soils in previously CO₂-enriched plots at the historical ORNL free-air CO₂ enrichment (FACE) experiment to quantify root C storage pools, growth, turnover and impacts on soil C and nutrient cycling.

Results/Conclusions

In pine, late-season shading reduced C uptake, sap flow, stem growth and fine root biomass, and increased soil water content. Although there were reductions in new C flux belowground, shading did not reduce the rate of soil CO₂ efflux, highlighting its dependence on older C sources, and environmental conditions. The ¹³C label was immediately detected in foliage on label day, progressed through phloem by day 2, roots by day 2-4, and subsequently was evident as respiratory release from soil between days 3-6. Similar patterns were seen in the dogwood study. In addition ¹³C label results suggests much of the new C partitioned belowground was rapidly respired from arbuscular mycorrhizal hyphae, including Glomus sp. ¹³CO₂ efflux from fungal ingrowth chambers was similar in timing (~3 days) to control chambers and enrichment (0 to +20‰) was similar to control (0 to +75‰) and well above exclusion chambers (-15 to -10‰). Initial results from sweetgum girdling found a strong and multi-year dependence on C stored in woody tissue. Data from each study are being used to test a point version of the Community Land Model.  Results provide data of mechanistic processes not well-refined in the models and will lead to improvements in model representation of C partitioning.