DETERMINATION OF CHITINASE ACTIVITY IN TALL FESCUE BY NEAR-INFRARED REFLECTANCE SPECTROSCOPY

New cultivars of pasture-type tall fescue (Festuca arundinacea Schrebe r) lack the fungal endophyte, Acremonium coenophialum Morgan;Tones and W. Gams. Although Acremonium-free cultivars are less toxic to livesto ck than Acremonium-infected cultivars, they are less disease tolerant. Acremonium-free cultivars may be improved for disease tolerance using a biochemical marker such as chitinase, a defense hydrolase associate d with disease resistance in many crops. The objective of this researc h was to measure chitinase activity in tall fescue seedlings by near i nfrared reflectance spectroscopy (NIRS). Ninety-nine seedling samples were freeze-dried, ground, and analyzed for total and specific chitina se activity using tritiated chitin as a substrate. Near infrared spect ra were recorded for each sample, and an NIRS equation was developed b y regressing radiochemical data against spectral data; regression proc edures included forward stepwise multiple regression and modified part ial least squares (MPLS). In optimum equations, standard errors of cal ibration and validation were near or below 10% of the mean, similar to errors observed in routine chemical analysis of chitinase. The optimu m equation used MPLS to predict specific activity, resulting in a coef ficient of determination of 0.90 and a mean and standard error of 88.8 +/- 7.2 disintegrations min(-1) mg(-1) protein. The NIRS-chitinase pr ocedure is accurate and efficient. Once a spectrophotometer is calibra ted, the NIRS procedure is at least 10 times faster than chemical proc edures, permitting analysis in 60 s per sample.