An interlaboratory comparison was designed to test the precision and relative accuracy of a method using near infrared spectroscopy for the determination of nitrogen, lignin, and cellulose concentrations in green and senescent foliage. A total of five laboratories were involved. Two methods of nitrogen analysis (Kjeldahl and CHN analysis) and two methods of carbon-fraction (lignin and cellulose) analysis were used. Equations for converting near infrared reflectance spectra into estimates of nitrogen and carbon-fraction concentrations are presented, along with estimates of relative bias between laboratories and methods. Analytical errors (variation among repeated measurements of the same sample) associated with each of the different methods are also presented. Results show that the near infrared reflectance method gives values that lie within the range of values generated by wet chemistry methods and so introduces no additional bias. Analytical errors are one-third to one-fifth those of wet chemistry methods for nitrogen concentration and one-fifth to one-eighth those of wet chemistry methods for lignin and cellulose concentration.
