Prediction of clean mohair, fiber diameter, vegetable matter, and medullated fiber with near-infrared spectroscopy

Four experiments were conducted in three separate years to test the utility of near-infrared spectroscopy (NIRS) to predict the clean mohair content of Angora goat fleece. Mohair fleece samples were obtained each year from yearling billies at the conclusion of the Angora Goat Performance Test conducted at the Texas A&M University Research Station, Sonora. In Exp. 1 (n = 293) and Exp. 2 (n = 256), fleeces were scanned with a Pacific Scientific (Silver Spring, MD) near-infrared spectrometer fitted with a fiber-optic probe, and calibrations were developed for clean mohair content. In Exp. 3, 59 mohair fleeces collected at the Texas A&M Research Station in San Angelo were sampled four times each. Each sample was scanned with the same spectrometer in reflectance mode fitted with a transport mechanism. This mechanism allowed the instrument to scan a 15-cm(2) segment of the fleece sample. Conventional procedures to determine reference values for mohair yield, vegetable matter content, fiber diameter, and percentage of medullated and kemp fibers were conducted. Prediction equations were developed that related NIR spectra to reference values for yield and diameter parameters and were used to predict mohair characteristics for each fleece sample. The predicted and reference values were subjected to a simple analysis of variance to determine variation within and across samples. In Exp. 1, mohair base was related to NIR spectra with R(2) = .46 and standard error of calibration (SEC) = 2.84%. In Exp. 2, similar repeatability errors for mohair base could be obtained for both reference- and NIRS-derived values. Fiber diameter and medullated fibers were poorly related to NIR spectra. When samples were scanned using the transport mechanism (Exp. 3), R(2) and SEC were .82 and 1.19% for mohair base and .93 and .98 micrometers for fiber diameter, respectively. The CV for mohair base and diameter were 1.0 and 1.4%, whereas those for predicted mohair base and diameter were 1.4 and 3.4%, respectively. The increased variation within samples for predicted values represents sampling error and lack of fit between NIRS and the laboratory determined values. When the samples from Exp. 1 and 2 were rescanned with the NIRS transport (Exp. 4), R(2) and SEC were .79 and 2.03% for mohair base and .52 and 3.49 micrometers for fiber diameter. The fiber optic probe would facilitate real-time analysis on the shearing floor, but our data indicate that the spectral limitations so far are too severe. A large sample device such as the transport gave excellent results for predicting mohair base and fiber diameter.