Phenotypic modelling of residual feed intake using physical activity and methane production as energy sinks.

Feed efficiency is usually expressed as the amount of milk produced per unit of feed intake. However, this definition is a ratio trait, which is challenging to incorporate in selection indices. Residual feed intake (RFI) is an alternative expression of feed efficiency. It is defined as the difference between observed and predicted dry matter intake (DMI). Predicted DMI is obtained through multiple regression of DMI on energy sinks, such as energy corrected milk (ECM) yield, metabolic body weight (MBW) and body condition score (BCS). For similar levels of energy sinks, cows eating less than the herd average have negative RFI values and are deemed more efficient, whereas cows eating more than the average have positive RFI values and are judged less efficient. The residual term, RFI, is an accumulation of true feed efficiency, unknown sinks, and trait-recording and model-fitting errors. This residual error part can contribute to a large fraction (41 to 47%) of residual energy intake. Nutrition studies suggest that methane emission originates from this residual intake. It is a significant waste (2.7 to 9.8% of gross energy intake of dairy cattle). Similarly, physical activity requires energy and thus may also be considered an energy sink. The importance of methane emission and physical activity as energy sinks or unwanted sinks requires investigation. Therefore, Dr D.W. Olijhoek and co-workers include methane production and physical activity as additional energy sinks in their model to evaluate their relation with DMI. Along with this, they also examined the correlations of RFI among the first 36 weeks of lactation (WOL) and evaluated whether lactation numbers and breeds are similar in these aspects. The authors published their results in the Journal of Dairy Science, Volume 103 of 2020, page 6967 to 6981, the title being: Phenotypic modelling of residual feed intake using physical activity and methane production as energy sinks. 

The authors modelled RFI of first- and second-lactation Holstein and Jersey dairy cows within nine lactation segments (consecutive segments of four weeks each), covering the first 36 weeks of lactation. The intention was also to evaluate physical activity and daily methane production as additional energy sinks in the estimation of RFI, to examine the correlations of RFI among the first 36 weeks of lactation (WOL), and to evaluate whether lactation number and breeds show similar results. Records of 449 first-lactation Holstein, 298 second-lactation Holstein, 195 first-lactation Jersey and 146 second-lactation Jersey cows were used. Model 1 included the following energy sinks: ECM yield, MBW, BCS, daily changes in BW (ΔBW) and BCS (ΔBCS), and physical activity. Model 2 was based on a subset of the data and only for Holstein cows, and included the same energy sinks as Model 1, plus daily methane production. 

The segment specific partial regression coefficients (PRC) of DMI on activity were similar across lactation but differed slightly between breeds. For daily methane production, the PRC decreased over lactation segments for first- and second-lactation Holstein cows. The PRC of DMI on ECM yield, MBW, BCS, and ΔBW were generally similar across lactations, except for ΔBCS. Activity accounted for on average 7.3, 6.8, 7.2, and 6.4% of DMI for first-lactation Holsteins, second-lactation Holsteins, first-lactation Jerseys and second-lactation Jerseys, respectively. Methane losses accounted for 8.7% and 8.5% of DMI for first- and second-lactation Holstein cows, respectively. Repeatability estimates for RFI over 36 WOL for Model 1 were 0.63 for first-lactation Holsteins, 0.65 for second-lactation Holsteins, 0.76 for first-lactation Jerseys and 0.80 for second-lactation Jerseys. For Model 2, the estimates were 0.59 and 0.61 for first- and second-lactation Holstein cows, respectively. Correlations of RFI between WOL varied in strength, with weak correlations for the first two to three WOL.  

In conclusion: Physical activity and daily methane production were shown to be energy sinks for Holstein and Jersey dairy cows. These traits accounted for part of DMI and improved the model, although, when including these sinks in the model, their effect on RFI was small. Residual feed intake of dairy cattle is not the same trait throughout lactation, which may reflect the strong effects of energy mobilization during early lactation or inaccuracies in modelling the metabolic status of cows over lactation. This warrants the modelling of traditional energy sinks and RFI within lactation segment.