On the determination of residual feed intake and associations of infrared thermography with efficiency and ultrasound traits in beef bulls

Abstract To determine the relationship of infrared thermography (IR) and ultrasound measures (US) with the variation in feed efficiency and daily dry matter intake (DMI), alternative models for calculating residual feed intake (RFI) were tested using DMI, average daily gain (ADG), mid-trial body weight (BW), body surface temperature measured using IR, and mid-trial US (ribeye area, backfat thickness and marbling score) from 154 crossbred bulls. The original model (Koch's model) for RFI (RFI koch ), based on a regression of DMI on BW and ADG had the lowest coefficient of determination ( R 2 ; 0.58) and was included as the base model in all the other alternate extended models tested. The residual or error term from the model represents the calculated RFI. RFI calculated including ribeye area (RFI rbea ) had the highest R 2 (0.62) of all US at mid-trial. RFI including all US traits (RFI usdt ) had the same R 2 (0.63) but higher, less desirable, Bayesian information criterion (BIC; 456.2 vs. 334.9) than the model where only feet temperature was added to Koch's model. A model combining mid-trial US and feet IR (RFI usir ) had the greatest R 2 (0.67) and BIC (327.7). RFI koch was correlated with eye, cheek and feet temperature (0.24 to 0.43) but not with ribs, rear area or scrotum temperature. Bulls were also sorted into RFI koch groups (high, medium, low). Eye, cheek and feet had lower temperature in low-RFI bulls compared to high-RFI bulls (less efficient). The feed to gain ratio (F:G) and DMI was higher for high-RFI than for low-RFI bulls. Further analysis demonstrated that 28% of the RFI koch variation was explained by US (4%) and IR (24%), where feet and cheek were the major IR traits, representing 74% of the IR contribution to the explained variation in RFI koch . Our results provide evidence that: (1) Koch's model might be improved by adding US traits; (2) US and IR traits are useful to explain the DMI variation; (3) IR of different body locations have distinct relationships with RFI (RFI kock and RFI usdt ), and with other efficiency traits and US; and (4) heat production (IR) explains a greater proportion of the variation in RFI than body composition (US). This study demonstrates the potential for improving RFI determination and also the application of IR in the assessment of feed efficiency. Further studies to develop standard procedures for performing infrared imaging might increase the predictability of this technology for assessing feed efficiency, especially residual feed intake, in cattle.