Sustainability & efficiency

Augmenting saved feed index with carbon dioxide breeding values

---

Abstract

In this study, we investigated the effect of augmenting a saved feed (SF) index with carbon dioxide (CO2) breeding values (BV). Incorporating an SF index into total merit selection has gained popularity in dairy cattle breeding. Recently, an alternative representation of the original SF index was suggested, consisting of BVs for dry matter intake (DMI), energy-corrected milk (ECM) and body weight gain (BWG). However, the accuracy of SF indices is often low due to limited feed intake data, and incorporating additional information would be beneficial. Efficient cows consume less feed and produce less heat per unit of product. The finding that heat and CO2 production are highly correlated, prompted studies on using CO2 measurements for developing feed efficiency metrics. The data used in this study were collected at Luke's Jokioinen research dairy farm exclusively for the purpose of feed efficiency research. The edited dataset included 20,665 weekly records from 537 first lactating Nordic Red dairy cows. Of these, the youngest 107 cows had 2,604 weekly CO2 observations derived from 16,395 daily measurements recorded using two GreenFeed units. Initially, multi-trait variance components were estimated for the traits DMI, ECM, BWG and CO2. The model fitted included the fixed effects of age, lactation stage and year×month, and the random effects permanent environment, additive genetic and residual. Estimated variance components were used to evaluate two SF indices: 1) a standard SF index SF=-BVDMI+0.456×BVECM +3.25×BVBWG, and 2) a CO2 BV augmented SF index aSF=-(b1×BVCO2)+0.456×BVECM+3.25×BVBWG, where b1 is the regression slope of DMI on CO2 production. To test the aSF index, different multi-trait models were applied on a test data set that included the observations of all 537 cows but for 92 cows with at least five CO2 records, all observations were either included, removed (pedigree index), or DMI observations removed but replaced with CO2 observations. Estimated heritabilities for DMI, ECM, BWG and CO2 were 0.30±0.08, 0.43±0.11, 0.29±0.10 and 0.31±0.29, respectively. The estimated genetic correlation between DMI and CO2 was 0.97±0.17. Based on the variance component estimates, heritabilities derived for SF and aSF were 0.11 and 0.20, respectively. The genetic correlation between SF and aSF was 0.97. The correlation between the 92 cows' SF indices based on the full data set with the corresponding pedigree index SF indices, the SF indices where CO2 observations were included as correlated trait or the aSF indices was 0.28, 0.79 or 0.82, respectively. The latter correlation agreed well with the high genetic correlation of 0.97, considering that reliabilities of the SF and aSF indices for the 92 cows were on average 0.47 and 0.56, respectively. Although the available dataset was limited, our findings indicate that CO2 observations could be used to increase the reliability of SF indices.

Keywords: 2026

How to Cite:

Bayat, A., Huhtanen, P., Kempe, R., Lidauer, M., Negussie, E., Salaudeen, A. & Stefanski, T., (2026) “Augmenting saved feed index with carbon dioxide breeding values”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2280530. doi: https://doi.org/10.31274/wcgalp.23444

Rights: 1