Jb. Reeves et Js. Van Kessel, Near-infrared spectroscopic determination of carbon, total nitrogen, and ammonium-N in dairy manures, J DAIRY SCI, 83(8), 2000, pp. 1829-1836
The objective of this study was to investigate the feasibility of using nea
r-infrared reflectance spectroscopy (NIRS) to determine nutrient concentrat
ions in dairy manures. We assayed diverse dairy manures (n = 107), collecte
d from dairy farms in the northeastern United States (CT, MD, NY, PA, and V
A) by conventional means and NIRS for total C, total N, NH3-N, moisture, P,
K, and pH. Samples were scanned from 400 to 2498 nm in polyethylene bags o
n a FOSS-NIR-Systems Model 6500 scanning monochromator equipped with a samp
le transport device. We developed calibrations using a one-out cross valida
tion procedure under partial least-squares regression. Preliminary results
showed that eight samples were outliers either because of inaccurate conven
tional analysis or because they were uncharacteristic (i.e., two samples ha
d moisture content below 72%, while all others were above 78%). These outli
ers were removed from further consideration. Final calibration results with
the remaining 99 samples demonstrated that NIPS can accurately determine t
he moisture (r(2) = 0.945, root mean squared deviation or RMSD = 1.0%), tot
al C (r(2) = 0.950, RMSD = 0.40%), total N (r(2) = 0.956, RMSD = 0.030%), a
nd NH3-N (r(2) = 0.967, RMSD = 0.013%) concentrations, but not P or K conce
ntrations in dairy manures. In conclusion, NIPS was shown to be a viable al
ternative to conventional analysis procedures for determining moisture, tot
al C, total N, and NH3-N in a very diverse set of dairy manures.