Skip to main content
ARS Home » Plains Area » El Reno, Oklahoma » Grazinglands Research Laboratory » Forage and Livestock Production Research » Research » Publications at this Location » Publication #169809


item Carson, Lesley
item Northup, Brian
item Daniel, John

Submitted to: Research Day Abstracts: Regional Universities Research Day
Publication Type: Abstract Only
Publication Acceptance Date: 9/10/2004
Publication Date: 10/1/2004
Citation: Carson, L.B., Northup, B.K., Daniel, J.A. 2004. Predicting soil characteristics with near infrared reflectance spectroscopy (NIRS)[abstract]. Research Day Abstracts: Regional Universities Research Day. Paper No. S-238.

Interpretive Summary: Abstract Only

Technical Abstract: Soil characteristics are important to pasture condition in central Oklahoma. Rapid, low cost, and accurate estimates would be useful in improving management. However, traditional laboratory techniques require significant time, equipment and manpower. This study examined whether predictive relationships could be developed between results of traditional laboratory analyses and near infrared reflectance spectroscopy (NIRS). In November 2003, samples (n=100) from three sections of the soil profile (0-5 cm, 5-10 cm, 10-25 cm) were collected from 1.6 ha experimental paddocks, after 27 years of sustained management (tallgrass prairie grazed at high stocking rates, unmanaged relict prairie, and tilled and grazed winter wheat). Samples were ground to 2.0 mm particle size, sieved to remove roots and litter, and scanned by NIRS. Laboratory-analyzed samples (n=124) were randomly chosen, and total nitrogen and carbon (high temperature induction), and soil organic matter (loss on ignition [400oC for 16 hours]) determined. Relationships between peaks of reflected wavelengths from NIRS (750-2500 nm range) and levels of characteristics defined by laboratory procedures (n=62) were developed by partial least squares regression and validated (n=62). Strong relationships were noted between results of laboratory analyses and NIRS for C and N (R2 >0.97; P<0.01). The predictive equation for organic matter produced a significant (R2=0.87; P<0.01) relationship, but was less precise. Results indicate NIRS could be used to quickly estimate (with relative accuracy) important soil characteristics. Such assessments could be used to quickly define resource levels and management response. Data from additional sites are required to fully test the ability of NIRS to estimate soil characteristics.