Skip to main content
ARS Home » Research » Publications at this Location » Publication #329449
Title: Hyperspectral canopy reflectance as a predictor for root concentrations of nitrogen and carbon in native and non native grass species

Author
item TREY, SCOTT - Redlands Community College
item BREKKE, PETERSON - Orise Fellow
item Starks, Patrick

Submitted to: Grazinglands Research Laboratory Miscellaneous Publication
Publication Type: Proceedings
Publication Acceptance Date: 5/29/2016
Publication Date: 6/14/2016
Citation: Trey, S., Brekke, P., Starks, P.J. 2016. Hyperspectral canopy reflectance as a predictor for root concentrations of nitrogen and carbon in native and non native grass species. Pp. 12-16. In: R.W. Todd and A. Campbell (Eds). Proceedings-Great Plains Grazing Field Research Symposium, 14 June 2016, Oklahoma State University, Stillwater, OK. Available at: https://drive.google.com/file/d/0B5YS3Y9RTDyiQV9IUURWY2NNNW8/view?pref=2&pli=1.

Interpretive Summary: Land managers, scientists, and crop professionals need real-time, inexpensive, and labor-saving methods to determine below-ground biomass and potential carbon (C) and nitrogen (N) inputs of that biomass. Remote sensing is a non-destructive tool that monitors vigor of vegetation and has been used to assess forage quality. A study was conducted at the USDA-ARS Grazinglands Research Laboratory, El Reno, OK in non-native, Old World Bluestem and native, tallgrass prairie to: 1) determine N and C concentrations of roots of non-native and native pasture, 2) determine if canopy hyperspectral reflectance data can produce a usable equation for non-destructive determination of total root C and N. Hyperspectral canopy reflectance was measured bi-weekly using an ASD FieldSpec FR radiometer. Destructive canopy and root samples were acquired immediately after the hyperspectral data were collected. Sampling occurred at toe-, mid- and upper-slope positions along four parallel and widely-spaced transects. Canopy and roots were separated, oven-dried at 65oC for 48 hr, ground and total C and N concentrations determined. Canopy reflectance and root concentrations of C and N were statistically analyzed using partial least square. Initial results indicate that prediction of root C and N from hyperspectral canopy reflectance is moderate. However, further study is needed to determine if this is an appropriate non-destructive method. Implications of this research could lead to quicker determination of below ground inputs to soil carbon and nitrogen cycles and provide a better understanding of perennial ecosystem services.

Technical Abstract: Land managers, scientists, and crop professionals need real-time, inexpensive, and labor-saving methods to determine below-ground biomass and potential carbon (C) and nitrogen (N) inputs of that biomass. Remote sensing is a non-destructive tool that monitors vigor of vegetation and has been used to assess forage quality. A study was conducted at the USDA-ARS Grazinglands Research Laboratory, El Reno, OK in non-native, Old World Bluestem (Bothriochloa sp.) and native, tallgrass prairie to: 1) determine N and C concentrations of roots of non-native and native pasture, 2) determine if canopy hyperspectral reflectance data can produce a usable equation for non-destructive determination of total root C and N. Hyperspectral canopy reflectance was measured bi-weekly using an ASD FieldSpec FR radiometer. Destructive canopy and root samples were acquired immediately after the hyperspectral data were collected. Sampling occurred at toe-, mid- and upper-slope positions along four parallel and widely-spaced transects. Canopy and roots were separated, oven-dried at 65 C for 48 hr, ground and total C and N concentrations determined. Canopy reflectance and root concentrations of C and N were statistically analyzed using partial least square. Initial results indicate that prediction of root C and N from hyperspectral canopy reflectance is moderate (R2=0.65). However, further study is needed to determine if this is an appropriate non-destructive method. Implications of this research could lead to quicker determination of below ground inputs to soil carbon and nitrogen cycles and provide a better understanding of perennial ecosystem services.