Search

International audience; Involuntary temporary feed restriction on commercial cattle is likely to become more frequent with forage shortages in the context of climate change. If general consequences of feed restriction have been the subject of an abundant scientific literature, focus on the inter-individual variability of response is scarce. Here, we explore the response profile in terms of BW, body condition score, milk production, calf weight and cyclicity resumption of 293 lactations from 169 Charolais cows during a winter feed restriction in early lactation and its subsequent recovery at grazing using a principal component analysis followed by a hierarchical clustering on principal component. Results show a very continuous range of response profiles that was divided into three clusters: one with light animals having an intermediate response in terms of milk production and body maintenance, one with animals prioritising body maintenance and cyclicity resumption over milk production and calf weight, and the last one with animals prioritising milk production and calf weight over the rest. Among the animals performing more than one lactation, 57% remain in the same cluster on two successive lactations. This work highlights that an average group response to feed restriction may hide various resilience individual profiles. Further studies are required to determine the existence of a genetic component as well as the consequences of not taking this phenomenon into consideration with the regular use of feed restriction in commercial farms.

Improving feed efficiency is an important issue for beef producers. To study its feasibility, 588 Charolaisyoung bulls were fattened in four experimental stations, from 2015 to 2019, with either corn silage or grasssilage. Feed conversion efficiency (FCE) was higher for animals fattened with corn silage (0,164 kg/kgMS vs 0,146 kg/kg MS). Efficient animals better used diet nitrogen and had a lower natural abundance of15N in plasma. Genetic study showed that feed efficiency is heritable, but some correlations wereunfavorable with performance traits. A genetic selection of beef cattle is possible with a selection indexthat combined both feed efficiency and production.; Sélectionner sur l’efficience alimentaire est un enjeu important pour les éleveurs bovins allaitants. Pourétudier sa faisabilité, 588 jeunes bovins ont été engraissés dans quatre stations expérimentales, de 2015à 2019, avec de l’ensilage de maïs ou d’herbe ad libitum. L’efficience de conversion des aliments (EA)était supérieure chez les animaux engraissés avec l’ensilage de maïs (0,164 kg/kg MS contre0,146 kg/kg MS) lié à des croissances élevées. Les animaux efficients valorisaient mieux l’azotealimentaire et avaient une abondance naturelle en 15N dans le plasma moindre. Les analyses génétiquesont montré que l’efficience alimentaire est héritable, mais certaines corrélations sont non favorables avecles performances. Une amélioration génétique du troupeau allaitant est possible via un index de sélectioncombinant efficience alimentaire et production.

The GreenFeed (GF) system spot samples the breath of individual animals for short periods (3–7 min), at various times of the day, over several days, weeks, or months to measure methane production. To obtain a diurnal CH4 emission pattern from the GF system, the animal visits need to be appropriately distributed over a 24 h feeding cycle. This can be difficult to achieve as the number of visits is greater during certain hours of the day. However, visits can be manipulated to some extent by the type of ration provided. Arthur performed a field study using beef cattle, and concluded that 30 observations per animal (15 d, cattle fed an ad libitum roughage diet) were sufficient to produce a reliable CH4 emission outcome from the GF system, if the durations of each visit are sufficient (minimum of 3 min). Others have stated that 50 spot measurements per animal were needed over a 17 d period to obtain a high repeatability (R=0.7) for dairy cows. Manafiazar concluded that a minimum of 20 spot samples over a 7- to 14-d period would be highly repeatable as well as highly correlated with feed intake. Ultimately, accurate outcomes from the GF system can be obtained when the operator strictly controls the number and timing of animal visits. This can be easily managed in tie-stalls but is more difficult in grazing systems, Chapter of the Living Handbook Methods in cattle physiology and behaviour – Recommendations from the SmartCow consortium

International audience; Genetic selection on feed efficiency could increase the profitability of French beef farms. The BEEFALIM 2020project aims to study the genetic determinism of this trait measured from 2016 to 2019 on 588 young Charolais bullsin four experimental farms. After weaning in commercial farms, the young bulls entered the fattening barns at 303(±26) days of age and were fed ad libitum either with grass or corn silage. During the 29 (±3) weeks of the test period,dry matter intake (DMI) was recorded daily and animals were weighted every 14 days. Average daily gain (ADG)and final body weight (FW) were calculated using regression of body weight on time. Three feed efficiency criteriawere used: residual feed intake (RFI), residual gain (RG) and feed conversion efficiency (FCE). Genetic parameterswere estimated using WOMBAT software with an animal model containing the contemporary group as fixed effect,age at the start of the test as covariate and farm origins × year, genetic and residual as random effects. During thetest period, DMI averaged 9.65 (±1.13) kg/day, ADG 1.51 (±0.29) kg/day and FW 698 (±74) kg. The three traitswere closely correlated phenotypically: 0.67 between FI and ADG, 0.77 between FW and ADG and 0.82 betweenFI and FW. Feed intake and ADG were more heritable (0.34) than FW (0.20). The three feed efficiency traits weremoderately heritable (0.22 for RFI and RG; 0.18 for FCE). No genetic correlation was found between RFI and RGor FCE while a strong genetic correlation of 0.99 was estimated between RG and FCE. Residual feed intake wasgenetically correlated with DMI (0.82), ADG (0.52) and FW (0.49). Both RG and FCE were closely correlated withADG (0.79 and 0.75, respectively) and moderately with DMI (0.31 and 0.23, respectively) or FW (0.33 and 0.25,respectively). Even if RFI and RG were phenotypically independent with traits used in both regressions, geneticrelationships were found between them. Genetic improvement is possible through a selection index combining bothfeed efficiency and production traits.

International audience; Beef cattle industry could benefit from the identification of individuals with superior feed efficiency. Thus, biomarkersshould be developed and validated to contribute to this objective. This study aimed to explore plasma biomarkersof residual feed intake (RFI) in fattening beef cattle through a targeted metabolomics approach. For this, 167 youngCharolais bulls were tested for RFI in 3 independent cohorts during at least 6 months. Animals within each cohortwere evenly assigned to either a high-starch corn silage diet or a high-fibre grass silage diet. Blood was sampled fromeach animal one month before the end of the test. Plasma from the 48 most extreme animals in terms of RFI (24 lowvs 24 high RFI; balanced within cohorts and diets) were analysed for 630 metabolites by FIA- or LC- tandem massspectrometry. Metabolites were analysed by ANOVA separately for each diet including the fixed effect of cohort, RFIgroup and their interaction. In the corn silage diet, plasma concentration of the 3 branched-chain amino acids (BCAA;Leu, Ile and Val) was lower (-15%; P≤0.01) in low-compared to high-RFI, likely reflecting their higher metabolicuse for protein synthesis and muscle accretion. In contrast, creatinine and hydroxyl-proline, markers of muscle andcollagen body mass respectively, showed higher plasma concentration (+18-20%; P≤0.01) in low- compared tohigh-RFI animals. In the grass silage diet, most of the RFI discriminant metabolites belonged to different classes oftriglycerides and their plasma concentrations were always higher (+50% on average; P=0.003) in low- comparedto high RFI animals, indicating either their lower incorporation into fat cells or their lower use as energy substrateby the muscle. Only creatinine was a common RFI discriminant metabolite in the two diets and Val, though alsodiscriminant, showed opposite trends. Our results suggest that discriminant metabolites of RFI are not the same acrosstwo contrasting diets. However, differences in muscle body mass revealed by plasma creatinine concentration seemsto be a determinant of RFI regardless the diet.

Multiblock PLS-DA on fecal and plasma visible-near-infrared spectra for discriminating young bulls according to their efficiency. Preliminary results. Conférence Chimiométrie 2020

Improving the ability of animals to convert feed resources into food for humans is needed for more sustainable livestock systems. Genetic selection for animals eating less while maintaining their performance (i.e. low residual feed intake) appears a smart strategy but its effectiveness relies on high-throughput animal phenotyping. Here, we explored plasma nitrogen (N) isotope ratios in an attempt to identify easily superior young bulls in terms of residual feed intake (RFI). For this, 48 Charolais young bulls fed two contrasting diets (corn vs grass silage diets) were selected from a larger population as extreme RFI animals (24 low-RFI vs 24 high-RFI) and their plasma analyzed for natural 15N abundance (δ15N) in the whole protein (bulk protein) and in the individual protein-bound amino acids (PbAA). For the first time, we showed that the δ15N in plasma bulk protein differed (P = 0.007) between efficient (Low-RFI) and inefficient (High-RFI) cattle regardless of diet. Furthermore, most analyzed PbAA followed the same trend as the bulk protein, with lower (P < 0.05) δ15N values in more efficient (Low-RFI) compared with less efficient (High-RFI) cattle, again regardless of diet. The only three exceptions were Phe, Met and Lys (P > 0.05) for which the first metabolic reaction before being catabolized does not involve transamination, a pathway known naturally to enrich AAs in 15N. The contrasted isotopic signatures across RFI groups only in those PbAA undergoing transamination are interpreted as differences in transamination rates and N-use efficiency between Low- and High- RFI phenotypes. Natural isotopic N signatures in bulk proteins and specific PbAA can be proposed as biomarkers of RFI in growing beef cattle fed different diets. However, the current study cannot delineate whether this effect only occurs post-absorption or to some extent also in the rumen. Our data support the conclusion that most efficient cattle in terms of RFI upregulate N conservation mechanisms compared with less efficient cattle, and justify future research on this topic.

Additional file 5: Fig. S2. Network interaction between GO terms, KEGG pathways and genes from the AWM analysis with RG as key phenotype. Fig. S3. Network interaction between GO terms, KEGG pathways and genes from the AWM analysis with FE as key phenotype.

Additional file 1. List of breeds included in the reference population panel for whole-genome imputation.

Additional file 2: Fig. S1. Manhattan plots of residual feed intake (a), residual gain (b) and feed efficiency ratio (c).

Cheap and rapid methods to predict between-animal variability of feed efficiency (FE) are needed. We evaluated, by using a large dataset on beef cattle (n = 355) fed contrasting diets (grass vs corn based diets), the potential of plasma natural 15N abundance (δ15N) as a biomarker of between-animal variation in FE, measured either as feed conversion efficiency (FCE) or residual feed intake (RFI). Data were analyzed by mixed models, with diet and δ15N as a fixed effect and experimental farm, batch and pen as random effects. A negative and significant correlation was found between FCE and δ15N within each contemporary group, with a lower response (slope) however in grass vs corn diets (δ15N × diet interaction; P = 0.002). Likewise, we found a significant but positive relationship between δ15N and RFI, but without the influence of diet or any tested random effect. Our results indicate that δ15N is a promising biomarker of between-animal variation of feed efficiency (FCE and RFI) in growing beef cattle.

International audience; Slaughter traits have a main economic value in beef cattle. They are also an important information source on the physiological variability among animals. To investigate the genomic control of some of these traits, a sequencebased GWAS was performed on 578 adult Charolais cows. Animals were slaughtered at seven years of age, after a fattening period. They were phenotyped for 18 traits: empty body weight (EBW), carcass yield, percentage of muscle, percentage of fat, 5th quarter, 5th quarter’s fat, leather, rumen, omasum, abomasum, intestines, total digestive tract, total reproductive tract, liver, lungs, heart, kidneys and spleen. The carcass yield and all 5th quarter and organ traits were expressed relative to the EBW. Cows were genotyped with the BovineSNP50 SNPchip (Illumina Inc., San Diego), and their genotypes were imputed up to the sequence level based on the run6 of the 1000 bull genomes project (2,333 animals including 128 Charolais bulls), using successively the FImpute and Minimac software. GWAS analyses were performed using the GCTA software. All traits together, we identified 19 significant QTL at the stringent sequence based Bonferroni threshold (-logP>8.19). The carcass yield, the percentages of muscle and fat, the 5th quarter and the 5th quarter’s fat were all found highly associated with the same region of the BTA2. These QTL reflect the well-known effect of the myostatin major gene on the carcass composition. Even though the relative weight of the 5th quarter shows a strong association with this gene, no QTL was observed at this position for the organs taken individually. Other QTL were detected in association with the kidneys (several QTL on BTA 18 and 24), with the lungs (two QTL on BTA 19), with the rumen (BTA 2 and 28), with the heart (BTA5) and with the abomasum (BTA 28). Further investigation will be made in parallel with other phenotypes such as feed efficiency.

International audience

National audience

International audience; The microbiota in the rumen fluid of 65 Holstein cows was characterized through the 16S rRNA (bacteria) targeted re-sequencing. Methane yields (CH4y = CH4/DMI) and dry matter intake (DMI) were individually measured during 3 weeks with the GreenFeed and Calan gate systems, respectively. Twenty of these cows were measured (CH4y) and rumen contents sampled twice in two successive months. A combination of multivariate, clustering and microbial co-abundance network analysis was implemented to identify a set of OTUs jointly associated with CH4y and the discrimination of samples according to the structure of ruminal microbial ecosystem. In addition, repeated measures were used to provide an estimation of the 16S ruminal stability within a month, which was R=0.64. Three ruminotype-like clusters (R1, R2 and R3) were identified; and R2 was associated with higher CH4y emission. Sparse partial least squares discriminant analysis (sPLS-DA) was done based on samples classification following two criteria: ruminotype-like cluster assignation and CH4y phenotype classification (low, high and intermediate). A sPLS regression model using the phenotype variation of CH4y emission was also implemented. Overall, 28 OTUs, which explain the 52% of the phenotype variation of CH4y, were identified by combining the approaches. The taxonomic classification of these OTUs included families linked to CH4 emission such as Christensenellaceae, Lachnospiraceae, and Ruminococcaceae. According to the interaction patterns, 83% (25/28) of these OTUs belong to the same sub-network module. In summary, our results suggest a relative stability of the rumen microbiota as well as a common set of interacting OTUs simultaneously linked to CH4y emission and microbial ruminal ecosystem structure.

International audience