Landau S Y, Dvash L, Roudman M, Muklada H, Barkai D, Yehuda Y, Ungar E D
1Department of Natural Resources,Institute of Plant Sciences,Agricultural Research Organization,the Volcani Center,Bet Dagan 50250,Israel.
2Department of Natural Resources,Gilat Experimental Station,M.P. HaNegev 2,Israel.
Animal. 2016 Feb;10(2):192-202. doi: 10.1017/S175173111500169X. Epub 2015 Sep 1.
Rapid assessment of the nutritional quality of diets ingested by grazing animals is pivotal for successful cow-calf management in east Mediterranean rangelands, which receive unpredictable rainfall and are subject to hot-spells. Clipped vegetation samples are seldom representative of diets consumed, as cows locate and graze selectively. In contrast, faeces are easily sampled and their near-IR spectra contain information about nutrients and their utilization. However, a pre-requisite for successful faecal near-infrared reflectance spectroscopy (FNIRS) is that the calibration database encompass the spectral variability of samples to be analyzed. Using confined beef cows in Northern and Southern Israel, we calibrated prediction equations based on individual pairs of known dietary attributes and the NIR spectra of associated faeces (n=125). Diets were composed of fresh-cut green fodder of monocots (wheat and barley), dicots (safflower and garden pea) and natural pasture collected at various phenological states over 2 consecutive years, and, optionally, supplements of barley grain and dried poultry litter. A total of 48 additional pairs of faeces and diets sourced from cows fed six complete mixed rations covering a wide range of energy and CP concentrations. Precision (linearity of calibration, R2cal, and of cross-validation, R2cv) and accuracy (standard error of cross-validation, SEcv) were criteria for calibration quality. The calibrations for dietary ash, CP, NDF and in vitro dry matter digestibility yielded R2cal values >0.87, R2cv of 0.81 to 0.89 and SEcv values of 16, 13, 39 and 31 g/kg dry matter, respectively. Equations for nutrient intake were of low quality, with the exception of CP. Evaluation of FNIRS predictions was carried out with grazing animals supplemented or not with poultry litter, and implementation of the method in one herd over 2 years is presented. The potential usefulness of equations was also established by calculating the Mahalanobis (H) distance to the spectral centroid of a calibration population of 796 faecal samples collected throughout 2 years in four herds. Seasonal trends in pasture quality and responses to management practices were identified adequately and H<3.0 for 98% of faecal samples collected. We conclude that the development of FNIRS equations with confined animals is not only unexpensive and ethically acceptable, but their predictions are also sufficiently accurate to monitor dietary composition (but not intake) of beef cattle in east Mediterranean rangelands.
快速评估放牧动物所采食日粮的营养质量，对于地中海东部牧场成功进行母牛-犊牛管理至关重要，该地区降雨不可预测且常出现热浪。由于母牛会有选择性地定位和采食，剪下的植被样本很少能代表其所采食的日粮。相比之下，粪便易于采样，其近红外光谱包含有关营养成分及其利用情况的信息。然而，成功进行粪便近红外反射光谱分析（FNIRS）的一个先决条件是校准数据库要涵盖待分析样本的光谱变异性。我们利用以色列北部和南部的圈养肉牛，根据已知日粮属性与相关粪便（n = 125）的近红外光谱的个体配对，校准了预测方程。日粮由单子叶植物（小麦和大麦）、双子叶植物（红花和豌豆）的鲜切青饲料以及连续两年在不同物候期采集的天然牧草组成，并且可选择添加大麦籽粒和干家禽粪便。另外还有总共48对来自采食六种全混合日粮的母牛的粪便和日粮，这些日粮涵盖了广泛的能量和粗蛋白浓度范围。校准质量的标准是精密度（校准的线性度，R2cal，以及交叉验证的线性度，R2cv）和准确度（交叉验证的标准误差，SEcv）。日粮灰分、粗蛋白、中性洗涤纤维和体外干物质消化率的校准R2cal值>0.87，R2cv值为0.81至0.89，SEcv值分别为16、13、39和31 g/kg干物质。除了粗蛋白外，营养物质摄入量的方程质量较低。我们对补充或未补充家禽粪便的放牧动物进行了FNIRS预测评估，并展示了该方法在一个牛群中两年内的实施情况。通过计算与在四个牛群中两年内收集的796份粪便样本校准群体的光谱质心的马氏（H）距离，也确定了方程的潜在有用性。充分识别了牧草质量的季节性趋势以及对管理措施的反应，并且所采集的98%的粪便样本的H<3.0。我们得出结论，利用圈养动物开发FNIRS方程不仅成本低廉且符合伦理要求，而且其预测足够准确，能够监测地中海东部牧场肉牛的日粮组成（但不是摄入量）。
