Relationship between feed efficiency and slaughter traits of French Charolais bulls

Development and validation of a model for early prediction of residual feed intake in beef cattle using plasma biomarkers

Identification of plasma biomarkers for feed efficiency in growing beef cattle offers a promising opportunity for developing prediction models to improve precision feeding strategies. However, these models must accurately predict feed efficiency at early stages of fattening. Our study aimed to evaluate the reliability of candidate biomarkers previously identified in late-fattening cattle when analysed during early fattening stages and to develop diet-specific prediction equations for residual feed intake (RFI). From a total of 364 Charolais bulls across seven cohorts, we selected 64 animals with extreme RFI values. The animals were fed either a corn‑ or grass-silage diets. These animals were chosen from four out of the available seven cohorts. Animals from three cohorts (24 high-RFI and 24 low-RFI, having a mean RFI difference of 1.48 kg/d) were used for biomarker confirmation and prediction model training. Animals from a fourth cohort (8 high-RFI and 8 low-RFI, having a mean RFI difference of 0.98 kg/d) were used for model external validation. Blood samples were collected at the beginning of the feed efficiency test (333 ± 20 days), and plasma underwent targeted metabolomic for 630 metabolites, natural abundance of 15N (δ15N), insulin, and IGF-1 analysis. Seven previously identified plasma biomarkers for RFI in late-fattening beef cattle still kept their capability for discriminating low and high RFI animals when analysed during early fattening stages (P < 0.05). Among these confirmed biomarkers, five were common for both grass- and corn-fed animals (creatinine, β-alanine, triglyceride TG18:0_34:2, symmetric dimethyl-arginine and phosphatidylcholine PC aa C30:2) while two were diet-specific (IGF-1 for grass silage-based diet, and isoleucine for corn silage-based diet. No new plasma biomarkers of RFI were identified at early-fattening stages (false discovery rate  > 0.05). Prediction models were developed based on seven confirmed RFI biomarkers analysed during early-fattening. Two logistic regression models incorporating creatinine and either IGF-1 (for grass silage-based diet) or PC aa C30:2 (for corn silage-based diet) effectively distinguished between high- and low-RFI animals with high sensitivity and specificity (area under the curve > 0.80). The biomarkers used in the models showed moderate to high repeatability between early and late fattening stages (0.45 < r < 0.65). The models were successfully externally validated, with more than 85% of animals from the fourth cohort correctly classified. Once validated in larger cohorts and utilising cost-effective and rapid analytical methods, these models could support precision feeding and breeding programmes, aiming to reduce the cost of raising beef cattle.

Some plasma biomarkers of residual feed intake in beef cattle remain consistent regardless of intake level

This study investigated whether plasma biomarkers of residual feed intake (RFI), identified under ad libitum feeding conditions in beef cattle, remained consistent during feed restriction. Sixty Charolais crossbred young bulls were divided into two groups for a crossover study. Group A was initially fed ad libitum (first test) and then restricted (second test) on the same diet, while Group B experienced the opposite sequence. Blood samples were collected from the 12 most divergent RFI animals in each group at the end of the first test and again after the second test. 12 plasma variables consistently increased, while three consistently decreased during feed restriction (FDR < 0.05). Only two metabolites, α-aminoadipic acid for Group A and 5-aminovaleric acid for Group B, were associated with RFI independent of feed intake level (FDR < 0.05), demonstrating moderate-to-high repeatability across feeding levels (intraclass correlation coefficient ≥ 0.59). Notably, both metabolites belong to the same metabolic pathway: lysine degradation. These metabolites consistently correlated with RFI, irrespective of fluctuations in feed intake, indicating a connection to individual metabolic processes influencing feed efficiency. These findings suggest that a portion of RFI phenotypic variance is inherent to an individual's metabolic efficiency beyond variations in feed intake.

Relationship between residual feed intake and digestive traits of fattening bulls fed grass silage- or maize silage-based diets

Several studies tried to identify digestive determinants of individual variation in feed efficiency between fattening bulls, because of their importance for breeding and management strategies. Most studies focused on single traits or single diet. Little is known about diet-dependent differences in digestive determinants and on their relative importance in distinguishing divergent residual feed intake (RFI) bulls. This research aimed (i) to identify digestive traits that differed between bulls diverging in RFI and fed a maize silage- or a grass silage-based diets; (ii) to highlight the relationships between RFI and digestive traits, and (iii) to explore the hierarchy among digestive traits in discriminating RFI divergent bulls. After an initial RFI test of 84 days on 100 Charolais growing bulls fed two different diets based on grass silage (GS), or maize silage (MS), the 32 most RFI divergent bulls were selected (eight efficient RFI- and eight inefficient RFI+ bulls per diet) and measured thereafter for total tract apparent digestibility and transit rate, enteric gas emissions (CH4 and H2), rumen pH, and feeding behaviour. Rumen particle size and visceral organ and reticulo-omasal orifice (ROO) sizes and rumen and ileum histology were measured at slaughter on the 32 selected extreme RFI bulls. Irrespective of the diet, efficient bulls (RFI-) had lower rumen size, CH4 yield (g/kg DM intake; tendency), lower number of cells in the ileal crypts, tended to have longer time of rumen pH below 5.8 and lower proportion of small size particles in rumen content than non-efficient bulls (RFI+). A long-term test for feed efficiency (197 d on average) was performed on the whole experimental period until slaughter for the 100 animals. The long-term RFI value was negatively related to time spent in activity other than ingestion, rumination, and resting, and positively related (tendency) to the duration of ingestion events, to rumen and abomasum size, irrespective of the diet. Diet-dependent effects were noted: with GS, efficient (RFI-) bulls showed a slower transit rate, whereas with MS, efficient (RFI-) bulls tended to have shorter resting events and a smaller ROO than inefficient bulls (RFI+). The transit rate and the ROO size tended to be positively related, while total tract apparent digestibility of nitrogen was negatively related to long-term RFI value, but only in GS. Rumen size appeared as the most discriminating digestive variable between RFI divergent bulls, but this result should be validated on a larger number of animals and diets.

Comparative analysis of signalling pathways in tissue protein metabolism in efficient and non-efficient beef cattle: acute response to an identical single meal size

Protein turnover has been associated to residual feed intake (RFI) in beef cattle. However, this relationship may be confounded by feeding level and affected by the composition of the diet being fed. Our aim was to assess postmortem the protein metabolism signalling pathways in skeletal muscle and liver of 32 Charolais young bulls with extreme RFI phenotypes. Bulls were fed two contrasting diets during the whole fattening period but were subjected to a similar and single nutritional stimulus, induced by their respective concentrate, just prior to slaughter. The key targets were protein degradation (autophagy and ubiquitin) and synthesis signalling pathways through western-blot analysis, as well as hepatic transaminase activity. To ensure a precise assessment of all animals at the same postprandial time, they were provided with a test meal (2.5 kg of either a high-starch and high-protein concentrate or high-fibre and low-protein concentrate) 3 hours prior to slaughter, irrespective of their RFI grouping. Blood and tissues were sampled at the slaughterhouse (3 h and 3 h30 postprandially, respectively). In response to an identical single meal size, efficient RFI animals showed higher (P < 0.05) postprandial plasma β-hydroxybutyrate concentrations and insulinemia (only with the high-starch concentrate) than non-efficient animals. Moreover, efficient RFI bulls had lower muscle (P = 0.04) and liver (P = 0.08) ubiquitin protein abundance (degradation pathway) and tended to have lower alanine transaminase activity in the liver (P = 0.06) compared to non-efficient bulls, regardless of diet. A positive correlation between protein degradation potential and amino acid catabolism was identified in this study (r = 0.52, P = 0.004), which was interpreted as being biologically linked to the RFI phenotype. Efficient RFI bulls also had a faster potential for protein synthesis in the muscle, as indicated by their greater ratio of phosphorylated to total form of ribosomal protein S6 kinase (P = 0.05), regardless of diet. Results on protein synthesis pathway in muscle and plasma metabolite concentrations suggested that efficient RFI cattle may have a faster nutrient absorption and insulin responsiveness after feeding than inefficient cattle. We did not find significant differences in hepatic protein synthesis pathways between the two RFI groups (P > 0.05). Our findings suggest that, in response to an identical single meal size, efficient RFI animals exhibited lower activation of tissue protein degradation pathways and faster muscle protein synthesis activation compared to their inefficient counterparts. This pattern was observed regardless of the composition of the tested meals.

The yield of adipose tissue and by-products in the course of the slaughter of inbred and outbred bulls of the Ukrainian beef breed

The research focuses on analysing and generalising the distribution of internal adipose tissue and organs that are not part of the carcasses of inbred and outbred bulls of the Ukrainian beef breed. Animal stock inbreeding was determined based on five breeding records according to Wright’s method modified by Kyslovskyi. Two experimental groups of 5 bulls were formed. The average inbreeding coefficient for inbred bulls was 3.43%. Animals were bred up to 18 months of age. Following slaughter, the mass and the yield of the head, liver, lungs, heart, kidneys, and brain were determined, and 4 types of fat were separated and weighed: perirenal, from the stomach, intestines, and pericardial. Inbred animals are more prone to the accretion of internal adipose tissue. Inbred bulls have 1.8 points more of it. Fat is more intensely accumulated around inbred bulls' multichambered stomachs and kidneys. Intensive fat accumulation was observed around the hearts and intestines of outbred bulls. Adipose tissue around the heart and intestines is more variable in inbred and outbred animals – from the forestomach and kidneys. The weight of inbred bulls’ liver is less by 22.4%, kidneys – by 62.5%, heart – by 11.1%, and head – by 23.8% compared to outbred ones. The weight of their lungs is more by 10.5%. At the same time, inbred bulls tend to have brain weight gain of 12.5% and testicles – by 8.3%. Thus, inbreeding application in Ukrainian beef breeds with a small population size affects the growth of internal organs and the intensity of accumulation and distribution of interior fat. Due to more intensive accumulation of internal adipose tissue, inbred bulls have increased expenditure of forage energy for its formation. They are characterized by an increased yield of low-value raw fat, making them less efficient than outbred bulls for beef production.     

Protein metabolism, body composition and oxygen consumption in young bulls divergent in residual feed intake offered two contrasting forage-based diets

Protein metabolism and body composition have been identified as major determinants of residual feed intake (RFI) in beef cattle fed high-starch fattening diets. This study aimed to evaluate if these two identified RFI determinants in beef cattle are the same across two contrasting silage-based diets. During two consecutive years, an 84-day feed efficiency test (Test A) immediately followed by a second 112-day feed efficiency test (Test B) was carried out using a total of 100 animals offered either one of two diets (either corn silage- or grass silage-based) over 196 days. At the end of Test A, the 32 animals most divergent for RFI (16 extreme RFI animals per diet, eight low RFI and eight high RFI) were identified and evaluated during Test B for their i) N use efficiency (NUE; N retention/N intake) calculated either from a 10-d nitrogen balance trial or from estimations based on body composition changes occurring during the whole experiment (Test A and Test B; 196 days), ii) carcass and whole-body protein turnover rates analysed through the 3-methyl-histidine urinary excretion and the N isotopic turnover rates of urine, respectively, and iii) body composition measured at the slaughterhouse at the end of Test B. Oxygen consumption was measured during Test B for the 100 animals by two GreenFeed systems. Irrespective of the diet, efficient RFI animals tended (P = 0.08) to improve their NUE when N retention was estimated for 196 days or when considering their lower urinary urea-N to total N ratio (P = 0.03). In contrast, NUE calculated during the 10-d N balance showed no differences (P = 0.65) across RFI groups suggesting that this method may not be suitable to capture small NUE differences. Efficient RFI individuals presented higher dressing percentage and muscle deposition in the carcass (P = 0.003) but lighter rumen (P = 0.001), and a trend for lower oxygen consumption (P = 0.08) than inefficient RFI animals irrespective of the diet. Lower protein degradation rates of skeletal muscle and lower protein synthesis rates of plasma proteins were found in efficient RFI cattle but only with the corn silage-based diet (RFI × Diet; P = 0.02). The higher insulinaemia associated with the corn silage-based diet (P = 0.001) seemed to be a key metabolic feature explaining the positive association between protein turnover and RFI only in this diet. Feed N was more efficiently used for growth by efficient RFI animals regardless of the diet but lower protein turnover rates in efficient RFI animals were only observed with corn silage-based diets.

Non-invasive metabolomics biomarkers of production efficiency and beef carcass quality traits

The inter-cattle growth variations stem from the interaction of many metabolic processes making animal selection difficult. We hypothesized that growth could be predicted using metabolomics. Urinary biomarkers of cattle feed efficiency were explored using mass spectrometry-based untargeted and targeted metabolomics. Feed intake and weight-gain was measured in steers (n = 75) on forage-based growing rations (stage-1, 84 days) followed by high-concentrate finishing rations (stage-2, 84 days). Urine from days 0, 21, 42, 63, and 83 in each stage were analyzed from steers with the greater (n = 14) and least (n = 14) average-daily-gain (ADG) and comparable dry-matter-intake (DMI; within 0.32 SD of the mean). Steers were slaughtered after stage-2. Adjusted fat-thickness and carcass-yield-grade increased in greater-ADG-cattle selected in stage-1, but carcass traits did not differ between ADG-selected in stage-2. Overall 85 untargeted metabolites segregated greater- and least-ADG animals, with overlap across diets (both stages) and breed type, despite sampling time effects. Total 18-bile acids (BAs) and 5-steroids were quantified and associated with performance and carcass quality across ADG-classification depending on the stage. Stepwise logistic regression of urinary BA and steroids had > 90% accuracy identifying efficient-ADG-steers. Urine metabolomics provides new insight into the physiological mechanisms and potential biomarkers for feed efficiency.

Common and diet-specific metabolic pathways underlying residual feed intake in fattening Charolais yearling bulls

Residual feed intake (RFI) is one of the preferred traits for feed efficiency animal breeding. However, RFI measurement is expensive and time-consuming and animal ranking may depend on the nature of the diets. We aimed to explore RFI plasma biomarkers and to unravel the underlying metabolic pathways in yearling bulls fed either a corn-silage diet rich in starch (corn diet) or a grass-silage diet rich in fiber (grass diet). Forty-eight extreme RFI animals (Low-RFI, n = 24, versus High-RFI, n = 24, balanced per diet) were selected from a population of 364 Charolais bulls and their plasma was subjected to a targeted LC-MS metabolomic approach together with classical metabolite and hormonal plasma analyses. Greater lean body mass and nitrogen use efficiency, and lower protein turnover were identified as common mechanisms underlying RFI irrespective of the diet. On the other hand, greater adiposity and plasma concentrations of branched-chain amino acids (BCAA) together with lower insulin sensitivity in High-RFI animals were only observed with corn diet. Conversely, greater plasma concentrations of BCAA and total triglycerides, but similar insulin concentrations were noted in efficient RFI cattle with grass diet. Our data suggest that there are diet-specific mechanisms explaining RFI differences in fattening Charolais yearling bulls.

Plasma proteins δ15N vs plasma urea as candidate biomarkers of between-animal variations of feed efficiency in beef cattle: Phenotypic and genetic evaluation

Identifying animals that are superior in terms of feed efficiency may improve the profitability and sustainability of the beef cattle sector. However, measuring feed efficiency is costly and time-consuming. Biomarkers should thus be explored and validated to predict between-animal variation of feed efficiency for both genetic selection and precision feeding. In this work, we aimed to assess and validate two previously identified biomarkers of nitrogen (N) use efficiency in ruminants, plasma urea concentrations and the ¹⁵N natural abundance in plasma proteins (plasma δ¹⁵N), to predict the between-animal variation in feed efficiency when animals were fed two contrasted diets (high-starch vs high-fibre diets). We used an experimental network design with a total of 588 young bulls tested for feed efficiency through two different traits (feed conversion efficiency [FCE] and residual feed intake [RFI]) during at least 6 months in 12 cohorts (farm × period combination). Animals reared in the same cohort, receiving the same diet and housed in the same pen, were considered as a contemporary group (CG). To analyse between-animal variations and explore relationships between biomarkers and feed efficiency, two statistical approaches, based either on mixed-effect models or regressions from residuals, were conducted to remove the between-CG variability. Between-animal variation of plasma δ¹⁵N was significantly correlated with feed efficiency measured through the two criteria traits and regardless of the statistical approach. Conversely, plasma urea was not correlated to FCE and showed only a weak, although significant, correlation with RFI. The response of plasma δ¹⁵N to FCE variations was higher when animals were fed a high-starch compared to a high-fibre diet. In addition, we identified two dietary factors, the metabolisable protein to net energy ratio and the rumen protein balance that influenced the relation between plasma δ¹⁵N and FCE variations. Concerning the genetic evaluation, and despite the moderate heritability of the two biomarkers (0.28), the size of our experimental setup was insufficient to detect significant genetic correlations between feed efficiency and the biomarkers. However, we validated the potential of plasma δ¹⁵N to phenotypically discriminate two animals reared in identical conditions in terms of feed efficiency as long as they differ by at least 0.049 g/g for FCE and 1.67 kg/d for RFI. Altogether, the study showed phenotypic, but non-genetic, relationships between plasma proteins δ¹⁵N and feed efficiency that varied according to the efficiency index and the diet utilised.

Performance of Angus weaner heifers varying in residual feed intake-feedlot estimated breeding values grazing severely drought-affected pasture

Context Beef industry productivity and profitability would be enhanced by improved efficiency at pasture. Our research is evaluating performance at pasture of Angus heifers divergent in estimated breeding values for residual feed intake determined from feedlot data (RFI-f-EBV) under a range of grazing conditions. Aims To determine whether Low- and High-RFI-f-EBV cattle differ in their growth response when pasture quality and availability become limiting to performance. Methods Eight-month-old heifers (n = 40) weaned at 6 months of age grazed within two replicates of 20, each with 10 low feedlot-efficiency (High-RFI-f-EBV) and 10 high feedlot-efficiency (Low-RFI-f-EBV) heifers. Each replicate grazed each of eight 1.25-ha paddocks comprising severely drought-affected, low-quality (mean dry-matter (DM) digestibility 44.1%, crude protein 7.3% DM, and 6.1 MJ metabolisable energy/kg DM) mixed perennial and annual native temperate grasses at 7-day intervals during repeated 28-day cycles, with Phase 1 with 2834 kg DM/ha and Phase 2 with 1890 kg DM/ha mean starting biomass. Heifers were yard-weighed weekly on nine occasions during the 8-week study. Key results During Phase 1 of grazing, the heifers gained 6.2 kg liveweight (LW) and during Phase 2 of grazing they lost 10 kg LW on average. Differences in LW between the RFI-f-EBV groups were not evident at the start or end of the study. However, over the 56 days of study, average daily change in LW calculated from the difference between starting and final LW was higher for Low-RFI-f-EBV heifers than for High-RFI-f-EBV heifers (–33 vs –127 g/day, s.e.m. = 41 g/day, P = 0.026). A similar result was evident when average daily LW change was determined from regression of LW on the day of study (–6 vs –96 g/day, s.e.m. = 41 g/day, P = 0.033). No significant interactions between grazing Phase and RFI-EBV group were evident for the growth responses. Conclusions Higher feedlot-efficiency (Low-RFI-f-EBV) weaner heifers maintained LW somewhat better than lower feedlot-efficiency (High-RFI-f-EBV) heifers, as the nutritional availability at pasture became more limiting. Implications Low-RFI-f-EBV weaner heifers may be more nutritionally resilient than are High-RFI-f-EBV heifers under drought conditions and, hence, may require less supplementary feed to maintain growth performance.

Gene networks for three feed efficiency criteria reveal shared and specific biological processes

Background

French beef producers suffer from the decrease in profitability of their farms mainly because of the continuous increase in feed costs. Selection for feed efficiency in beef cattle represents a relevant solution to face this problem. However, feed efficiency is a complex trait that can be assessed by three major criteria: residual feed intake (RFI), residual gain (RG) and feed efficiency ratio (FE), which involve different genetic determinisms. An analysis that combines phenotype and whole-genome sequence data provides a unique framework for genomic studies. The aim of our study was to identify the gene networks and the biological processes that are responsible for the genetic determinism that is shared between these three feed efficiency criteria.

Results

A population of 1477 French Charolais young bulls was phenotyped for feed intake (FI), average daily gain (ADG) and final weight (FW) to estimate RFI, RG and FE. A subset of 789 young bulls was genotyped on the BovineSNP50 single nucleotide polymorphism (SNP) array and imputed at the sequence level using RUN6 of the 1000 Bull Genomes Project. We conducted a genome-wide association study (GWAS) to estimate the individual effect of 8.5 million SNPs and applied an association weight matrix (AWM) approach to analyse the results, one for each feed efficiency criterion. The results highlighted co-association networks including 626 genes for RFI, 426 for RG and 564 for FE. Enrichment assessment revealed the biological processes that show the strongest association with RFI, RG and FE, i.e. digestive tract (salivary, gastric and mucin secretion) and metabolic processes (cellular and cardiovascular). Energetic functions were more associated with RFI and FE and cardio-vascular and cellular processes with RG. Several hormones such as apelin, glucagon, insulin, aldosterone, the gonadotrophin releasing hormone and the thyroid hormone were also identified, and these should be tested in future studies as candidate biomarkers for feed efficiency.

Conclusions

The combination of network and pathway analyses at the sequence level led to the identification of both common and specific mechanisms that are involved in RFI, RG and FE, and to a better understanding of the genetic determinism underlying these three criteria. The effects of the genes involved in each of the identified processes need to be tested in genomic evaluations to confirm the potential gain in reliability of using functional variants to select animals for feed efficiency.

Feed efficiency and carcass metrics in growing cattle1

Some definitions of feed efficiency such as residual energy intake (REI) and residual gain (RG) may not truly reflect production efficiency. The energy sinks used in the derivation of the traits include metabolic live-weight; producers finishing cattle for slaughter are, however, paid on the basis of carcass weight, as opposed to live-weight. The objective of the present study was to explore alternative definitions of REI and RG which are more reflective of production efficiency, and quantify their relationship with performance, ultrasound, and carcass traits across multiple breeds and sexes of cattle. Feed intake and live-weight records were available on 5,172 growing animals, 2,187 of which also had information relating to carcass traits; all animals were fed a concentrate-based diet representative of a feedlot diet. Animal linear mixed models were used to estimate (co)variance components. Heritability estimates for all derived REI traits varied from 0.36 (REICWF; REI using carcass weight and carcass fat as energy sinks) to 0.50 (traditional REI derived with the energy sinks of both live-weight and ADG). The heritability for the RG traits varied from 0.24 to 0.34. Phenotypic correlations among all definitions of the REI traits ranged from 0.90 (REI with REICWF) to 0.99 (traditional REI with REI using metabolic preslaughter live-weight and ADG). All were different (P < 0.001) from one suggesting reranking of animals when using different definitions of REI to identify efficient cattle. The derived RG traits were either weakly or not correlated (P > 0.05) with the ultrasound and carcass traits. Genetic correlations between the REI traits with carcass weight, dressing difference (i.e., live-weight immediately preslaughter minus carcass weight) and dressing percentage (i.e., carcass weight divided by live-weight immediately preslaughter) implies that selection on any of the REI traits will increase carcass weight, lower the dressing difference and increase dressing percentage. Selection on REICW (REI using carcass weight as an energy sink), as opposed to traditional REI, should increase the carcass weight 2.2 times slower but reduce the dressing difference 4.3 times faster. While traditionally defined REI is informative from a research perspective, the ability to convert energy into live-weight gain does not necessarily equate to carcass gain, and as such, traits such as REICW and REICWF provide a better description of production efficiency for feedlot cattle.

Methane and Carbon Dioxide Emission of Beef Heifers in Relation with Growth and Feed Efficiency

Simple Summary

For sustainable meat production, beef farmers must make the best use of grass and roughage while limiting the carbon footprint of their herds. The genetic improvement in feed efficiency and enteric methane production of replacement heifers is possible if the recorded phenotypes are available. Intuitively, the relationship between the two traits should be negative, i.e., favorable, since the energy lost with the methane is not available for heifer metabolism. The measurement of feed efficiency requires several weeks of feed intake recording. The enteric methane emission rate can also be recorded over several weeks. The two traits of 326 beef heifers from two experimental farms were measured simultaneously for 8 to 12 weeks. The correlations between roughage intake, daily gain, and methane were all positive. The enteric methane emission rate was positively related to body weight, daily gain, and dry matter intake. The relationship with feed efficiency was slightly positive, i.e., unfavorable. Therefore, the two traits should be recorded simultaneously to evidence low-emitting and efficient heifers. This study also showed that replacing the feed intake recording with the carbon dioxide emission rate appeared potentially beneficial for selecting these low-emitting and efficient heifers.

Abstract

Reducing enteric methane production and improving the feed efficiency of heifers on roughage diets are important selection objectives for sustainable beef production. The objective of the current study was to assess the relationship between different methane production and feed efficiency criteria of beef heifers fed ad libitum roughage diets. A total of 326 Charolais heifers aged 22 months were controlled in two farms and fed either a grass silage (n = 252) or a natural meadow hay (n = 74) diet. Methane (CH₄) and carbon dioxide (CO₂) emission rates (g/day) were measured with GreenFeed systems. The dry matter intake (DMI), average daily gain (ADG), CH₄ and CO₂ were measured over 8 to 12 weeks. Positive correlations were observed among body weight, DMI, ADG, CH₄ and CO₂. The residual feed intake (r_wgDMI) was not related to CH₄ or residual methane (r_wiCH₄). It was negatively correlated with methane yield (CH₄/DMI): R_p = −0.87 and −0.83. Residual gain (r_wiADG) and ADG/DMI were weakly and positively related to residual methane (r_wiCH₄): R_p = 0.21 on average. The ratio ADG/CO₂ appeared to be a useful proxy of ADG/DMI (R_p = 0.64 and 0.97) and CH₄/CO₂ a proxy of methane yield (R_p = 0.24 and 0.33) for selecting low-emitting and efficient heifers.