Predicting Diet Quality of White-Tailed Deer via NIRS Fecal Profiling

Abstract

Near infrared reflectance spectroscopy (NIRS) of feces for the prediction of diet quality in several species of livestock and wildlife has been reported. The technique has not been reported in deer. This study was conducted to determine the ability of fecal NIRS to determine dietary crude protein (CP), digestible organic matter (DOM), and phosphorus (P) in white-tailed deer (Odocoileus virginianus). Seventy-six diet reference chemistry:fecal spectrum (D:F) pairs were created ranging from 6.00 to 18.95% CP, 26.64 to 76.08% DOM, and 0.08 to 0.48% P. Calibration results (R2 and SE cross validation) were: 0.95 and 1.17, 0.88 and 3.62, 0.83 and 0.04 for CP, DOM, and P, respectively. These equations were used to predict a validation D:F set (n = 11). Results (R2 and SE prediction) were: 0.79 and 1.53, 0.49 and 5.46, 0.67 and 0.03 for CP, DOM, and P, respectively. These two D:F sets were combined and calibrations reformulated. Results (R2 and SE cross validation) were: 0.84 and 1.40, 0.89 and 3.55, 0.83 and 0.04 for CP, DOM, and P, respectively. These combined calibrations were used to predict diet quality characteristics using 11 fecal samples from wild deer. The diet quality characteristics were compared to NDVI greenness values for the study area in winter, spring and summer. High correlation (R2 > 0.7) between fecal NIRS predicted diet quality and NDVI greenness was observed with the exception of P in summer (R2 = 0.25). Fecal NIRS can be used to determine diet quality in white-tailed deer and thus become another tool to evaluate habitat suitability.