Page Banner

United States Department of Agriculture

Agricultural Research Service

Research Project: GLOBAL CHANGE: RESPONSES AND MANAGEMENT STRATEGIES FOR SEMI-ARID RANGELANDS

Location: Rangeland Resources Research

Title: The Role of Water Relations in Driving Grassland Ecosystem Responses to Rising Atmospheric Co2

Authors
item Wylie, Bruce - EROS NATIONAL CENTER
item Gilmanov, Tagir - SOUTH DAKOTA STATE UNIVER
item Frank, Al - RETIRED ARS
item Morgan, Jack
item Haferkamp, Marshall
item Meyers, T - NATIONAL OCEANIC & ATMOSP
item Fosnight, E - US GEOLOGICAL SURVEY
item Zhang, L - US GEOLOGICAL SURVEY

Submitted to: USDA Greenhouse Gas Symposium
Publication Type: Abstract Only
Publication Acceptance Date: January 15, 2005
Publication Date: March 21, 2005
Citation: Wylie, B., Gilmanov, T., Frank, A., Morgan, J.A., Haferkamp, M.R., Meyers, T.P., Fosnight, E.A., Zhang, L. 2005. The role of water relations in driving grassland ecosystem responses to rising atmospheric co2. USDA Greenhouse Gas Symposium. CDROM.

Technical Abstract: Rangeland Carbon fluxes are highly variable in both space and time. In the Northern Great Plains, rainfall is an important determinant of whether this ecosystem will be a carbon sink or source. Given the large areas of rangelands and their significant soil organic matter stocks, understanding how they respond to climatic variation is important for making future predictions. Rangeland carbon fluxes associated with Net Ecosystem Exchange (NEE), gross primary productivity (P<SUB>g</SUB>), total ecosystem respiration (R<SUB>e</SUB>) were quantified from multiple year data sets from five flux tower locations in the Northern Great Plains. Light response curve analysis was used to partition net fluxes into P<SUB>g</SUB>, and R<SUB>e</SUB>. These flux tower measurements were then combined with 1 km<SUP>2</SUP> spatial data sets of photosynthetically active radiation (PAR), Normalized Difference Vegetation Index (NDVI), temperature, precipitation, start of growing season, and soil derived data sets. Regression tree models were developed by removing infrequently variables or variables that had limited impact on model prediction. Cross validation and jackknifing approaches quantified model prediction accuracies. Maps of 10-day carbon dynamics of NEE, P<SUB>g</SUB>, and R<SUB>e</SUB> were produced for each growing season from 1998 to 2001. Growing season carbon fluxes were combined with estimates of winter fluxes to estimate annual carbon budgets. Carbon sinks and sources were mapped and regional averages were calculated.

Last Modified: 9/23/2014
Footer Content Back to Top of Page