Near infrared spectroscopy (NIRS) was applied to determine faecal nitrogen and phosphorus using a temporo-spatially diverse dataset derived from multiple ruminant herbivore species (i.e. cattle, bison, deer, elk, goats, and sheep). Single-species NIRS calibrations have previously been developed to predict faecal constituents. Multi-species NIRS calibrations have previously been developed for herbivore faecal nitrogen but not for faecal phosphorus. Faecal samples representing a herd or flock composite were analysed via NIRS (400–2498 nm). Calibration sets for faecal nitrogen and phosphorus were developed from: (1) all samples from all six species, (2) all cattle samples only, (3) all samples except those from bison, (4) all samples except those from deer, (5) all samples except those from elk, (6) all samples except those from goats, and (7) all samples except those from sheep. Validation sample sets included: (1) each of the individual species (predicted with a cattle only-derived calibration), and (2) each of the individual species (other than cattle) predicted with a multi-species calibration constructed from all cattle samples plus those samples from the remaining four species (i.e. ‘leave-one-out’). All multiple coefficient of determination (R²) values for faecal nitrogen calibrations were ≥0.97. Corresponding standard error of cross validation (SECV) values were ≤0.13. Validation simple coefficient of determination (r²) and standard error of prediction (SEP) of each alternate species using the cattle-derived calibration ranged from 0.76 to 0.84, and 0.28 to 0.5 respectively. Similar values for the sequential species leave-one-out validation for faecal nitrogen were 0.67 to 0.89, and 0.17 to 0.47 respectively. All R² values for faecal phosphorus calibrations were ≥0.79; corresponding SECV were ≤0.14. Validation r² and SEP of each alternate species using the cattle-derived phosphorus calibration were ≤0.63 and ≥0.13 respectively. Similar values for the sequential species leave-one-out validation were ≤0.66 and ≥0.22 respectively for faecal phosphorus. Multi-species faecal NIRS calibrations can be developed for monitoring applications in which determination of faecal nitrogen is appropriate, e.g. free-ranging herbivore nutrition, nitrogen deposition from animal faeces on rangelands with declining forage quality, or runoff from confined animal feeding operations. Similar calibrations for faecal phosphorus require additional research to ascertain their applicability.

使用从多种反刍动物食草动物（即牛，野牛，鹿，麋鹿，山羊和绵羊）获得的时空分布数据集，应用近红外光谱法（NIRS）测定粪便中的氮和磷。以前已经开发了单种NIRS校准品来预测粪便成分。以前针对草食动物粪便氮而非粪便磷开发了多种NIRS标定。通过NIRS（400-2498 nm）分析了代表畜群或羊群复合材料的粪便样品。粪便氮和磷的校准集来自：（1）所有六个物种的所有样品，（2）仅所有牛的样品，（3）除野牛以外的所有样品，（4）除鹿外的所有样品，（ 5）除麋鹿以外的所有样品，（6）除山羊外的所有样品，（7）除绵羊以外的所有样本。验证样本集包括：（1）每个单独的物种（通过牛只标定预测），以及（2）每个单独的物种（非牛）通过从所有牛样本构建的多物种标定预测加上其余四个物种的样本（即“留一出”）。所有确定系数的倍数（粪便氮校准的R ²）值≥0.97。相应的交叉验证标准误差（SECV）值≤0.13。使用牛衍生的校准方法对每个替代物种的确认简单确定系数（r ²）和预测标准误（SEP）分别在0.76至0.84和0.28至0.5的范围内。粪便氮的连续物种留一法验证的相似值分别为0.67至0.89和0.17至0.47。粪便磷校准的所有R ²值均≥0.79；相应的SECV≤0.14。验证r ²使用牛衍生的磷校正的每个替代物种的SEP和SEP分别≤0.63和≥0.13。粪便磷的连续物种留一法验证的相似值分别为≤0.66和≥0.22。可以开发多种类的粪便NIRS校准品，以监测适合测定粪便氮的应用，例如自由放牧的草食动物营养，草料质量下降的牧场上动物粪便中的氮沉积或有限的动物饲养操作产生的径流。粪便磷的类似校准需要进一步研究以确定其适用性。