|Daniel, John -|
Submitted to: Transactions of the ASABE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: October 1, 2010
Publication Date: November 15, 2010
Citation: Northup, B.K., Daniel, J.A., Phillips, W.A. 2010. Distribution of soil bulk density and organic matter along an elevation gradient in central Oklahoma. Transactions of the ASABE. 53(6):1749-1757. Interpretive Summary: There has been a great deal of recent interest in storing carbon (C) in rangeland and cropland soils of the Great Plains. Selling this stored C could represent a potential new cash flow for agricultural producers. However, while the number of C storage projects has been increasing, there has also been growing concern as to how valid they are. A primary concern has been a lack of proper monitoring for many projects, and whether accurate and cost-effective measures of the amount of C stored can be achieved. A major factor that affects both monitoring capacity and accuracy of measurements is the influence of variations in distribution of C within landscapes. We examined how C was distributed in three, 4 acre pastures (under different forms of management from 1978 to 2004) situated across a common slope in central Oklahoma, and how such distributions might influence the level of sampling required to develop accurate estimates. We found that both a wheat pasture used for grazing, and a native pasture that received no management, for 26 years had slightly smaller amounts of C in the upper 10 inches of soil (19±2 and 21±3 metric tons/acre, respectively) than a grazed native pasture (22±2 metric tons/acre). This C was unevenly distributed along the elevation gradient in all pastures, and regularly differed from the pasture-scale averages by ±2 to ±6 metric tons/acre, often within 5 to 10 ft of area. This uneven distribution resulted in large sampling requirements (23 to 37 samples/acre in relict native; 23 to 30/acre in grazed wheat; 14 to 21/acre in grazed native) to develop accurate estimates of soil C. Such numbers indicate the sampling costs to develop accurate measures of stored C in soils of central Oklahoma will likely be high.
Technical Abstract: The distribution of total soil carbon (TSC) within Oklahoma paddocks may affect sampling requirements to accurately monitor carbon (C) sequestration. This study examined how TSC was distributed in 3-1.6 ha paddocks [under different forms of long-term (1978-2004) management] situated across a common slope of an upland site in central Oklahoma, and how local variance in TSC might influence levels of required sampling. Included were grazed, conventionally tilled winter wheat, and 2 levels of grazing (high stocking rates over short periods, and no grazing) applied to tallgrass prairie. Soil cores (25 cm depth) were collected from each paddock at 1.5 m intervals along 150 m transects, and divided into 3 increments (0-5, 5-10, and 10-25 cm depths). Total soil C (Mg/ha) was determined from measures of bulk density and carbon concentration. Mean differences among management regimes were compared, and spatial variation in TSC was defined. The grazed wheat and relict native paddocks had different amounts of TSC than the grazed native paddock (48±5 and 52±6 Mg/ha, respectively vs. 54±4 Mg/ha). High levels of variability were recorded in TSC along the elevation gradient in all paddocks, and regularly differed from paddock-scale medians by ±5 to ±15 Mg ha-1. Results indicated variance in TSC related to slope position influenced the level of sampling required. Random sampling to achieve accurate TSC estimates (5% confidence intervals; 90% assurance) was high for the relict native and wheat paddocks (n=159 and 129, respectively), while optimum sample allocation based on slope position improved sampling efficiency (n=91 and 92, respectively). Monitoring C sequestration in Oklahoma will likely require both intensive sampling and blocking by topographic features to ensure accurate and efficient estimates.