Evaluation Of Nutritional Composition Of Three Gramineous Grasses And Establishment Of NIRS Prediction Models

With the continuous development of animal husbandry in China,the demand for high-quality forage is also increasing.Among them,oat grass,festuca arundinacea,and Bromus inermis,as high-quality forage,have rich nutritional value and are the important populations of forage resources in China.Their economic value is high.Therefore,evaluating the nutritional value of forage has important practical significance for ruminant nutrition production.Due to its advantages of fast,sensitive,non-destructive,pollution-free,low cost,and good reproducibility,near-infrared spectroscopy(NIRS)is widely used in the study of animal nutrition.This study analyzed the nutritional composition of three grasses in the grass family,oat grass,festuca arundinacea,and Bromus inermis,and evaluated their nutritional value;And NIRS was used to construct predictive models for the nutritional composition of three species of forage,providing a basis for the scientific and precise application of forage in ruminant production.Experiment 1: The wet chemical analysis method was used to analyze the dry matter(DM),crude protein(CP),organic matter(OM),neutral detergent fiber(NDF),acidic detergent fiber(ADF),Ash,K,Ca,P,Mg,Fe,Al,and Mn contents of 108 oat grass,101 festuca arundinacea,and 101 Bromus inermis,and to evaluate their nutritional value.The results showed that by analyzing the nutritional composition of three grass species,including oat grass,festuca arundinacea,and Bromus inermis,it was found that there were significant differences in nutritional value among the three grass species.The RFV from high to low was Bromus inermis,festuca arundinacea,and oat grass,providing a data basis for measuring forage quality and rational utilization of forage resources in the future.Experiment 2: By using NIRS technology to combine spectral information and wet chemical analysis of nutritional composition results of oat grass,festuca arundinacea,and Bromus inermis,combined with improved partial least squares method(MPLS),a near-infrared prediction model was constructed for DM,CP,OM,NDF,ADF,Ash,K,Ca,P,Mg,and other nutritional composition contents of single and mixed varieties of three grasses.The results showed that the external validation relative analysis error(RPD)values of each nutrient composition prediction model in a single variety of oat grass were 3.18,4.00,3.63,2.75,3.15,1.48,3.63,2.85,2.11,1.97,2.17;The RPD values of festuca arundinacea are 3.14,3.28,1.51,2.48,2.21,2.56,1.88,3.67,2.45,1.30,and 1.56;The RPD values of Bromus inermis are 1.28,2.59,1.03,2.19,3.09,3.84,1.03,2.56,2.39,2.55,and 2.28;The RPD values of the mixed nutrient composition prediction model for festuca arundinacea and Bromus inermis are 1.54,3.13,1.34,2.86,2.37,3.76,1.34,2.81,2.38,2.84,and 2.17;The RPD values of the mixed nutrient composition prediction models for oat grass,festuca arundinacea,and Bromus inermis were 3.23,5.97,2.27,4.07,2.78,3.44,2.27,3.06,2.09,2.07,and 2.50.Indicating that in a single variety,except for EE and P in oat grass;OM,Ash,P,Mg in festuca arundinacea;The prediction model for the content of nutrients other than DM,OM,and Ash in Bromus inermis can be used for practical production measurements.In addition to the prediction models for DM,OM,Ash in the mixed varieties of festuca arundinacea and Bromus inermis,other prediction models for nutrient composition content can be used in actual production.Compared to a single variety,the accuracy of all indicator prediction models for mixed grasses has improved.Among them,all indicators of the three mixed prediction models,oat grass,festuca arundinacea,and Bromus inermis,can be used in production practice.