4 Developing a Dry Matter Intake Prediction Equation for Grazing Animals based on Real-Time Enteric Emissions Measurements

• Published in: Journal of animal science 2022-10, Vol.100 (Supplement_4), p.1-1
• Main Authors: McFadden, Lily J, Menendez, Hector M, Olson, Ken, Brennan, Jameson R, Ehlert, Krista, Blair, Amanda
• Format: Article
• Language: English
• ISSN: 0021-8812; 1525-3163
• DOI: 10.1093/jas/skac313.000

Cattle dry matter intake (DMI) is an essential component of calculating cattle stocking rates, determining nutrient requirements, and evaluating grazing efficiency. Cattle DMI and digestion of forages impact enteric greenhouse gas (CO2e) emissions. Enteric emissions include methane (CH4) and carbon dioxide (CO2), that are eructated by ruminants. The amount of methane produced is affected by consumption, quality, and type of feedstuffs. Intake of grazing animals varies on environmental factors and physiological stage. Additionally, increased GHG levels indicate energy loss during the rumen fermentation process. However, there may be a silver lining to enteric GHG emissions to predict DMI of grazing animals since they are highly correlated with DMI and forage composition. There is limited data on the relationship of DMI and GHG on extensive rangeland systems. Obtaining data for beef cattle DMI and enteric emissions on forage-based diets similar to extensive rangelands is needed to develop an equation capable of predicting DMI for grazing cattle. Therefore, our objective was to determine the relationship between CH4, CO2, oxygen (O2), and hydrogen (H2) emissions and DMI of dry beef cows to develop a mathematical model that predicts grazing DMI from enteric emissions. The predictive equation or precision system model (PSM; Menendez et al., 2022) was developed using data from two feeding trials that were conducted using GreenFeed, SmartFeed Pro, and SmartScale (C-Lock Inc. Rapid City, SD). This study was conducted at the SDSU Cottonwood Field Station (Cottonwood, SD). The two feeding trials consisted of dry beef cows (n=10) receiving low (8% CP) or high (15% CP) quality grass hay using a 14-day adaptation period and a 14-day period of data collection. Regression, artificial neural network, and dynamic-mechanistic models were developed using these data and assessed to identify a model that accurately and precisely predicts forage DMI for dry beef cows on pasture. Model evaluation of the machine learning algorithms used a training, testing, and cross-validation scheme to determine model accuracy. Evaluation of mechanistic models used the Model Evaluation System (MES; Tedeschi, 2006) to measure accuracy (mean bias, Cb, RMSEP), precision (R2, MEF, CCC), and screening for systematic errors. This study successfully integrated three precision technologies which improve research capabilities on extensive rangeland systems through precision enhanced data collection. Dep