The full text of this item is not available at this time because the student has placed this item under an embargo for a period of time. The Libraries are not authorized to provide a copy of this work during the embargo period, even for Texas A&M users with NetID.
Application of Partial Least Squares Regression to Predict Dry Matter Intake and Feed Efficiency Based on Feeding Behavior Patterns in Beef Cattle
MetadataShow full item record
Feed expenses are the largest driver of input cost in the beef industry, thus, increasing the genetic merit of beef cattle efficiency is an effective strategy to improve the environmental and economic sustainability of beef production. Residual feed intake (RFI) is a measure of feed efficiency independent of average daily gain (ADG) and body weight (BW), whereby feed-efficient animals consume less feed than expected. Numerous studies have documented that cattle with divergent RFI phenotypes have distinctly different feeding behavior (FB) patterns, demonstrating their potential as bio-markers to predict feed efficiency. The nexus of this research lies in the development of animal behavior tracking systems and understanding the relationships between FB patterns and RFI. The first objective of this research was to validate a high-frequency RFID system to quantify frequency and duration of bunk visit (BV) events in beef cattle. The accuracy of the system to measure these traits was determined to be 81 and 90% accuracy, respectively. The second objective was to develop predictive equations for feed efficiency traits using FB traits as independent variables. Because FB traits are highly correlated, partial least squares (PLS) regression was used in this study as this method is better suited to deal with collinearity among independent variables. This study was conducted using data collected from composite Angus steers (N = 508; Initial BW 309 ± 56 kg) fed a high-grain diet in pens equipped with electronic feed bunks (GrowSafe® Systems). Individual dry matter intake (DMI), FB traits, and 14-d BW were measured for 70-d, and RFI calculated as the residual from the regression of DMI on ADG and BWsuperscript0.75. Cattle were ranked by RFI and assigned to 1 of 3 RFI classes based on ± 0.5 SD from the mean RFI. For each animal, 17 FB traits were evaluated: frequency and duration of bunk visit and meal events, head-down (HD) duration, time-to-bunk (TTB) interval, maximum non-feeding interval, and corresponding day-to-day variation (SD) of these traits. Additionally, 3 ratio traits were considered: BV frequency per meal event, HD duration per meal event and HD duration per BV event. Data analysis was conducted using a mixed-model (SAS 9.4) that included fixed effects of RFI class, trial and pen within trial. Feed-efficient steers consumed 16% less feed, while ADG and BW did not differ from high-RFI animals. Compared to high-RFI steers, low-RFI steers had 18% fewer and 24% shorter BV events and 11% fewer meal events that were 13% shorter (P < 0.01) in length. Feed efficient steers exhibited 10% less (P < 0.05) day-to-day variation in DMI, as well as 11 to 33% less (P < 0.05) day-to-day variance in frequency and duration of BV and meal events. Furthermore, low-RFI steers had 9% longer (P < 0.05) TTB, and 7% greater (P < 0.05) day-to-day variation in TTB compared to high-RFI steers. Partial least squares analysis identified 9 FB traits that explained 42% of the inter-animal variation in RFI. These results demonstrate that feed-efficient animals spend less time eating, visit the bunk less frequently for less total time per day and have more consistent day-to-day FB patterns compared to less-efficient animals. Further, these results indicate that FB traits may be useful as bio-marker to identify cattle that are more biologically efficient.
Partial least squares
Computing and Technology
Parsons, Ira Lloyd (2018). Application of Partial Least Squares Regression to Predict Dry Matter Intake and Feed Efficiency Based on Feeding Behavior Patterns in Beef Cattle. Master's thesis, Texas A & M University. Available electronically from