Improving feed efficiency of dairy cows through breeding is expected to reduce enteric methane production per unit of milk produced. This study examined the effect of 2 forage-to-concentrate ratios on methane production, rumen fermentation, and nutrient digestibility in Holstein and Jersey dairy cows divergent in residual feed intake (RFI). Before experimental onset, RFI was estimated using a random regression model on phenotypic herd data. Ten lactating Holstein and 10 lactating Jersey cows were extracted from the herd and allocated to a high or low pre-experimental RFI group of 5 animals each within breed. Cows were fed ad libitum with total mixed rations either low (LC) or high (HC) in concentrates during 3 periods in a crossover design with a back-cross and staggered approach. Forage-to-concentrate ratio was 68:32 for LC and 39:61 for HC. Cows adapted to the diets in 12 to 24 d and feces were subsequently collected on 2 d. Afterward, gas exchange was measured in respiration chambers and rumen liquid was collected once after cows exited the chambers. Pre-experimental RFI was included in the statistical analysis as a class (low and high RFI) or continuous variable. Methane per kilogram of dry matter intake (DMI) was lower for Holsteins than Jerseys and the response to increased concentrate level was more pronounced for Holsteins than Jerseys (27.2 vs.13.8%); a similar pattern was found for the acetate:propionate ratio. However, methane production per kilogram of energy-corrected milk (ECM) was unaffected by breed. Further, total-tract digestibility of neutral detergent fiber was higher for Jerseys than Holsteins. For RFI as a class variable, DMI, methane production regardless of the expression, and digestibility were unaffected by RFI. For RFI as a continuous variable, DMI was lower and methane per kilogram of DMI was higher for cows with negative (efficient) than positive (inefficient) RFI values, and neutral detergent fiber digestibility was higher for Holsteins with negative than positive RFI values, but not for Jerseys. Daily methane production and methane per kilogram of ECM were unaffected by RFI. In conclusion, methane per kilogram of DMI of Jerseys was lowered to a smaller extent in response to the HC diet than of Holsteins. When pre-experimental RFI was used as a continuous variable, higher methane per kilogram of DMI was found for cows with negative RFI than positive RFI values, but not for methane per kilogram of ECM. These findings call for validation in larger studies.

Identifying factors that influence the composition of the microbial population in the digestive system of dairy cattle will be key in regulating these populations to reduce greenhouse gas emissions. In this study, we analyzed rumen and fecal samples from five high residual feed intake (RFI) Holstein cows, five low RFI Holstein cows, five high RFI Jersey cows and five low RFI Jersey cows, fed either a high-concentrate diet (expected to reduce methane emission) or a high-forage diet. Bacterial communities from both the rumen and feces were profiled using Illumina sequencing on the 16S rRNA gene. Rumen archaeal communities were profiled using Terminal-Restriction Fragment Length Polymorphism (T-RFLP) targeting the mcrA gene. The rumen methanogen community was influenced by breed but not by diet or RFI. The rumen bacterial community was influenced by breed and diet but not by RFI. The fecal bacterial community was influenced by individual animal variation and, to a lesser extent, by breed and diet but not by RFI. Only the bacterial community correlated with methane production. Community differences seen in the rumen were reduced or absent in feces, except in the case of animal-to-animal variation, where differences were more pronounced. The two cattle breeds had different levels of response to the dietary intervention; therefore, it may be appropriate to individually tailor methane reduction strategies to each cattle breed.