If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The objective of the present study was to investigate factors related to variation in feed efficiency (FE) among cows. Data included 841 cow/period observations from 31 energy metabolism studies assembled across 3 research stations. The cows were categorized into low-, medium-, and high-FE groups according to residual feed intake (RFI), residual energy-corrected milk (RECM), and feed conversion efficiency (FCE). Mixed model regression was conducted to identify differences among the efficiency groups in animal and energy metabolism traits. Partial regression coefficients of both RFI and RECM agreed with published energy requirements more closely than cofficients derived from production experiments. Within RFI groups, efficient (Low-RFI) cows ate less, had a higher digestibility, produced less methane (CH4) and heat, and had a higher efficiency of metabolizable energy (ME) utilization for milk production. High-RECM (most efficient) cows produced 6.0 kg/d more of energy-corrected milk (ECM) than their Low-RECM (least efficient) contemporaries at the same feed intake. They had a higher digestibility, produced less CH4 and heat, and had a higher efficiency of ME utilization for milk production. The contributions of improved digestibility, reduced CH4, and reduced urinary energy losses to increased ME intake at the same feed intake were 84, 12, and 4%, respectively. For both RFI and RECM analysis, increased metabolizability contributed to approximately 35% improved FE, with the remaining 65% attributed to the greater efficiency of utilization of ME. The analysis within RECM groups suggested that the difference in ME utilization was mainly due to the higher maintenance requirement of Low-RECM cows compared with Medium- and High-RECM cows, whereas the difference between Medium- and High-RECM cows resulted mainly from the higher efficiency of ME utilization for milk production in High-RECM cows. The main difference within FCE (ECM/DMI) categories was a greater (8.2 kg/d) ECM yield at the expense of mobilization in High-FCE cows compared with Low-FCE cows. Methane intensity (CH4/ECM) was lower for efficient cows than for inefficient cows. The results indicated that RFI and RECM are different traits. We concluded that there is considerable variation in FE among cows that is not related to dilution of maintenance requirement or nutrient partitioning. Improving FE is a sustainable approach to reduce CH4 production per unit of product, and at the same time improve the economics of milk production.
The productivity of dairy cattle has risen considerably due to advances in nutrition, genetics, and management. Feed efficiency (FE) is an important trait under practical conditions, as it has a major influence on profitability and environmental efficiency in the dairy industry.
estimated that FE of dairy cows in North America has doubled in the past 50 yr, largely as a consequence of selecting and managing cows for increased productivity. With increased milk production, a greater proportion of feed energy is used for milk instead of cow maintenance, resulting in the dilution of maintenance. However, in intensive production systems, further improvements in FE from increased productivity are marginal. Breeding for increased production has been associated with larger body size, leading to an increased ME requirement per unit of metabolic BW (
), alleviating the intake effects on dietary ME concentration.
An increase in productivity has limited potential for further improvements for FE, and thus alternative approaches for increasing FE are needed. Feed conversion efficiency (FCE), expressed as ECM/DMI, is a traditional measure of FE in growing and lactating animals. The use of FCE as a selection criterion, however, has many limitations. For instance, selection for greater milk output increases the cow's energy requirement more than intake potential, resulting in mobilization of body tissues to support the increased energy demand of lactation (
). Residual feed intake (RFI) is a common measure of FE in dairy cattle in recent literature. It is calculated as a difference between DMI or energy intake and predicted intake estimated by regression models from energy sinks (production, maintenance, and body weight change). A negative RFI value means feed intake is less than expected and indicates an efficient animal, whereas a positive value indicates an inefficient animal.
proposed residual solids production as an alternative approach to estimating FE in lactating cows. This method estimates the difference between actual and predicted production for a given DMI, body size, and BCS. A positive value indicates greater efficiency and therefore, is more easily understood by producers than negative RFI values that indicate high efficiency.
cautioned that in some cases RFI may not represent true FE, which partly can be due to biases in estimating partial regression coefficients in the model. Furthermore, relationships between DMI and energy sink variables (milk energy; metabolic BW, MBW; and BW change, ΔBW) can be highly heterogeneous across different trials (
). Often, the partial regression coefficients of RFI models are biologically meaningless and different from expected values. This can be attributed to short measurement periods and inaccuracies in estimating energy balance from ΔBW. The mechanisms associated with variation in FE among cows are poorly understood. The variation in true FE (not related to dilution of maintenance requirement or partitioning nutrients between milk production and body tissues) can be related to converting dietary GE to ME, and subsequently converting ME to net energy (NE). The former includes variation in fecal, CH4, and urinary energy (UE) losses, and the latter includes heat production (HP) associated with body maintenance and HP from milk synthesis. Therefore, analysis of data from respiration calorimetry studies could be useful in elucidating the mechanisms behind variation in FE among dairy cows. The objective of the present study was to quantify the effects of different energy losses on different FE traits. The cows were categorized into low-, medium-, and high-FE groups on the basis of 3 different efficiency traits. The results of factors influencing energy losses, variance components, and correlations are presented in our companion paper (
A database containing energy balance data for 841 dairy cow observations was established from calorimetry studies conducted at Agri-Food and Biosciences Institute–AFBI (Hillsborough, UK), The Danish Cattle Research Centre (AU Foulum, Denmark), and Natural Resources Institute Finland–LUKE (Jokioinen, Finland). The list of studies used in the current study is given (Supplemental Data File S1; https://doi.org/10.3168/jds.2020-18259). The range of calorimetric data included in the database is summarized in Table 1. Further details of conducted experiments, calculations, outlier detection, diet composition, and energy metabolism traits were reported in our companion paper (
Feed efficiency in this study was evaluated as RFI, residual ECM production (RECM), and FCE. The MIXED procedure (SAS Institute Inc., Cary, NC) was used to predict DMI for each cow by fitting the following regression model:
where b0 is the intercept, b1 is the partial regression coefficient of ECM yield (kg/d), b2 is the partial regression coefficient of MBW (kg), b3 is the partial regression coefficient of positive energy balance (EBp, MJ/d), b4 is the partial regression coefficient of negative energy balance (EBn, MJ/d), and Exp, Diet(Exp), and Period(Exp) are random effects of experiment, diet within experiment, and period within experiment, respectively. Residual feed intake was calculated as the difference between actual DMI and predicted DMI. Because RFI represents a difference between actual and predicted DMI, a low or negative RFI indicates high efficiency.
Instead of milk solids production (fat + protein) used by
A Nordic proposal for an energy corrected milk (ECM) formula.
in: Gaillon P. Chabert Y. European Association for Animal Production Publication, Performance Recording of Animals: State of the Art, 1990; 27th Biennial Session of the International Committee for Animal Recording. Centre for Agricultural Publishing and Documentation,
Paris, France1990: 156-192
) as a measure of milk production. We used GE instead of DMI in predicting ECM yield to account for the possible effects of fat supplementation on energy metabolism. Using GE instead of determined ME intake in predicting ECM yield accounts for energy losses in the conversion of GE to ME (fecal, CH4, and UE losses). The following model was used to predict ECM yield:
where b0 is the intercept, b1 is the partial regression coefficient of GE intake (MJ/d), b2 is the partial regression coefficient of MBW (kg), b3 is the partial regression coefficient of EBp (MJ/d), b4 is the partial regression coefficient of EBn (MJ/d), and Exp, Diet(Exp), and Period(Exp) are random effects of experiment, diet within experiment, and period within experiment, respectively. Residual ECM yield was calculated as the difference between actual and predicted ECM yield. Because RECM represents a difference between actual and predicted ECM yield, a high or positive RECM indicates high efficiency.
Feed conversion efficiency was calculated as FCE = ECM (kg)/DMI (kg) without taking into account the effects of BW change that is not usually recorded in short-term respiration chamber studies.
The efficiency of ME utilization for milk production (kl) was calculated as
kl = El(0)/(ME intake – ME requirement for maintenance),
all expressed as MJ/d, where El(0) is milk energy corrected for zero energy balance calculated as milk energy + (1/0.95) × EBp or milk energy + 0.84 × EBn (
).Cows were categorized into 3 groups of approximately equal sizes (n = 278–284) by RFI value and classified as high RFI (High-RFI; RFI >0.72), medium RFI (Medium-RFI; RFI −0.39 to 0.72) or low RFI (Low-RFI; RFI < −0.39). Similarly, they were grouped by RECM value and classified as high RECM (High-RECM; RECM >1.2), medium RECM (Medium-RECM; RECM −1.32 to 1.2) or low RECM (Low-RECM; RECM < −1.32). Cows with FCE below 1.28 were categorized to group low FCE (Low-FCE), cows with FCE 1.28 to 1.51 were categorized to group medium FCE (Medium-FCE) and cows with FCE above 1.51 were categorized to high-FCE (High-FCE). The effects of RFI, RECM, and FCE groups on intake, production, and energy metabolism variables were determined using the MIXED Procedure of SAS according to a model that included the fixed effect of RFI, RECM, or FCE group and random effects of Exp, Diet(Exp), and Period(Exp). Further pairwise comparisons of least squares means among the efficiency groups were performed using the PDIFF option in the LSMEANS statement. Mixed model regression analysis with random effects of Exp, Diet(Exp), and Period(Exp) was used to evaluate quantitative relationships between variables.
RESULTS
The following equations were derived from the data for RFI and RECM:
where RFI, DMI, and ECM are expressed in kg/d; MBW as kg; and EBp and EBn as MJ/d. Excluding the intercept, the contributions of ECM, MBW, EBp and EBn on predicted intake were 53, 44, 5, and −2%, respectively;
where RECM and ECM are as kg/d, MBW as kg, and GE intake, EBp, and EBn as MJ/d.
Means and standard error of the mean of different animal and energy metabolism variables for RFI groups are presented in Table 2. The difference in RFI between Low- and High-RFI cows was 2.6 kg/d (P < 0.001), but ECM yield was similar among RFI groups. However, RECM was 4.6 kg/d (P < 0.001) greater for Low- than for High-RFI cows and Medium-RFI cows were intermediate. Feed conversion efficiency in terms of ECM/DMI followed the same pattern. Medium-RFI cows were 11 kg heavier than Low- and High-RFI cows. Energy losses in feces, CH4, and urine increased (P < 0.001) with increasing RFI. As a result of reduced energy losses (21 MJ/d) with higher efficieccy (Low-RFI), the difference in ME intake (25 MJ/d) between Low- and High-RFI groups was about 45% of the difference in GE intake (46 MJ/d). Greater (25 MJ/d; P < 0.001) ME intake of High-RFI cows was counterbalanced by a greater HP (21 MJ/d; P < 0.001), resulting in no differences in energy balance. Lower (P < 0.001) HP in Low- compared with High-RFI cows was associated with improved (P < 0.001) efficiency of ME utilization for milk production.
Table 2Production and energy metabolism characteristics of low (n = 279), medium (n = 282), and high (n = 280) residual feed intake (RFI) in respiration chamber studies (total n = 841)
Gross energy digestibility (digestible energy/GE) was 24 g/kg higher (P < 0.001) for Low- than for High-RFI cows. Methane and UE, as proportions of GE intake, were not different between the efficiency groups, but their combined effect resulted in a 6 kJ/MJ higher (P = 0.002) ME/digestible energy ratio for Low- than High-RFI cows. Metabolizability of GE (ME/GE) was 4.0% (P < 0.001) greater for Low- vs. High-RFI group. Methane intensity (g of CH4/kg of ECM) was 11% lower (P < 0.001) for Low-RFI cows compared with High-RFI cows.
High-RECM cows produced 6.0 kg of ECM /d more (P < 0.001) than Low-RECM cows at the same DMI (Table 3). Feed conversion efficiency (ECM/DMI) was 26% higher (P < 0.001) for High- compared with Low-RECM cows. Gross energy losses as feces (P < 0.001) and as CH4 emission were lower (P < 0.001) for High-RECM (efficient cows) when compared with their Low-RECM counterparts. The High-RECM cows had a higher (P < 0.02) ME intake and produced less (P < 0.001) heat than Low-RECM cows. The contributions of improved digestibility, and reduced CH4 and urinary losses to greater ME intake of High- compared with Low-RECM group were 83, 12, and 4%, respectively. Methane yield was higher in Low-RECM cows compared with other groups. Diet digestibility and metabolizability were 23 and 28 kJ/MJ higher (P < 0.001) in High- compared with Low-RECM cows. The cows in the High-RECM group had a higher (P < 0.001) efficiency of ME utilization (kl) than cows in Medium- and Low-RECM groups. The contribution of increased ME intake to RECM, calculated using the difference in ME intake and observed kl-value, of the Medium-RECM group was 2.2 kg/d; that is, about 37% of the observed difference between High- and Low-RECM groups. Consequently, 63% of the difference in RECM was attributed to the higher kl of High-RECM cows. When RECM was estimated using observed ME intake, the difference in RECM between High-RECM and Low-RECM groups was 4.1 kg/d. Methane production per kg of ECM was reduced (P < 0.001) with improved efficiency, at 24% lower in High- compared with Low-RECM cows.
Table 3Production and energy metabolism characteristics of low (n = 279), medium (n = 278), and high (n = 284) residual ECM production (RECM) in respiration chamber studies (total n = 841)
The following linear regressions of milk energy-corrected for zero energy balance (MJ/d) on ME intake (MJ/d), both scaled to MBW, were estimated for different RECM groups:
Low-RECM: milk energy = −0.510 ± 0.0279 + 0.658 ± 0.0152 × ME intake,
[5]
Medium-RECM: milk energy = −0.432 ± 0.0297 + 0.645 ± 0.0157 × ME intake,
[6]
High-RECM: milk energy = −0.434 ± 0.0279 + 0.674 ± 0.0142 × ME intake.
[7]
The intercept (maintenance energy requirement) tended to be greater (P = 0.06) for Low- than for Medium- and High-RECM cows. Numerically, kl (slope) was greater for High- compared with Medium-RECM cows.
Table 4 summarizes the production and energy metabolism characteristics of the Low-, Medium-, and High-FCE groups. High-FCE cows produced 44% more (P < 0.001) ECM per kg of DMI than Low-FCE cows. The difference was mainly due to differences in partitioning nutrients between milk production and body energy retention, and partly due to lower BW of High- compared with Low-FCE cows. No differences between the groups were observed in converting GE to ME or the utilization of ME. Methane intensity was 28.6% (5.6 g/kg of ECM; P < 0.001) greater for low- than for high-FE cows with the difference entirely attributed to greater ECM yield of the latter group.
Table 4Production and energy metabolism characteristics of low (n = 279), medium (n = 282), and high (n = 280) feed conversion efficiency (ECM/DMI) cows in respiration chamber studies (total n = 841)
The residuals of regressions of RECM on RFI were positively related to DMI and ECM yield (Figure 1). Methane yield was negatively related to the residuals (Figure 2). The effects of other variables (BW, digestibility, energy balance, and kl) on the residual were small.
Figure 1The effects of DMI and ECM yield on the residuals of regressions of residual ECM on residual feed intake (n = 841).
Figure 2The effects of methane yield on the residuals of regressions of residual ECM on residual feed intake (n = 841). CH4E = methane energy; GE = gross energy.
). Traditionally, efficiency is expressed as a ratio between product and feed intake in the form of mass or energy value of milk per unit (mass or energy) of intake. A major shortcoming of this definition is that it does not fully account for body tissue mobilization patterns, especially during early lactation (
). Residual ECM production is an alternative approach to assessing FE in dairy cows. Residual feed intake correlates positively with feed intake, but not with milk yield, MBW, or ΔBW, suggesting that efficient (i.e., low RFI) cows eat less. In contrast, efficient cows with high RECM have higher milk yields at similar feed intake and MBW than low RECM cows. While the RFI approach focuses on production cost, dividing it over a range of energy sinks, the RECM approach changes focus from cost to income (
). Residual feed intake or RECM on their own, however, do not measure production efficiency. This is because large animals that eat more than small animals, but have the same ECM yield, may have the same RFI or RECM, but clearly different production efficiency (ECM/DMI) or milk income over feed cost. Economically, RECM as a measure of efficiency is more favorable than RFI; assuming that relative price of ECM is 2-fold compared with DMI, the difference in milk income over feed cost between the most and the least efficient groups would be about 4 times greater for RECM than for RFI. According to our knowledge, this was the first time that RFI or RECM were estimated from respiration chamber data. Partial regression coefficients of DMI or ECM on energy sinks agreed reasonably well with energy requirements based on respiration chamber studies (e.g.,
Residual analysis of the model predicting RECM from RFI (Figure 1) indicated that these 2 variables are not the same traits; at a given RFI, the cows with higher DMI or ECM yield had a higher RECM. In other studies (
reported for full lactation RFI in Holstein cows (SD = 16.4 MJ of ME). For DMI prediction, the partial regression coefficient for ECM (0.347; Eq. 3) was slightly lower than that reported by
. Using the mean ME concentration of 11.8 MJ/kg of DM and a value of 0.64 for kl would give a value of 0.417 kg DM/kg of ECM. The average coefficient of DMI on ECM for cows in 50 to 150 DIM was 0.37 (range of 0.29–0.47) from 12 research stations in different countries (Templeman et al., 2015). In the studies of
, the coefficients of DMI on ECM were markedly lower, ranging from below 0.05 to 0.25 at different stages of lactation. The minimum coefficient of ECM would be 0.17 assuming no losses of dietary GE in digestion and metabolism. In the study of
, the overall partial regression coefficient of ME intake on ECM (2.67 MJ/kg of ECM) was lower than the energy content of ECM (3.14 MJ/kg). They also reported almost 3-fold differences in their coefficient of ECM at different stages of lactation. In the present data, DIM was negatively associated with kl, but quantitatively the effect was small (0.017 units, a 2.7% change, per 100 d). The variation in kl during lactation contradicts with the constant kl estimated using the
In the present study, the regression coefficient of DMI on MBW (0.0645; Eq. 3) is markedly greater than the corresponding coefficients (0.044 and 0.055) calculated using maintenance requirements of
, the partial regression coefficient of DMI on MBW averaged 0.092 (range from 0.06–0.16), which, for a 600 kg cow, corresponds to 11 kg of DMI to meet the maintenance requirements.
reported a value of 0.81 MJ of ME/kg of MBW (0.0686 kg of DM when ME = 11.8 MJ/kg of DM) ranging from 0.48 to 1.09 at different stages of lactation.
Contrary to other studies investigating RFI, we used determined EB as an energy sink instead of ΔBW. The coefficient of DMI on positive energy balance (0.0704 kg/MJ; Eq. 3) corresponds to 1.36 and 1.57 kg of DM per kg of ΔBW, calculated using the energy values of ΔBW in the FiM (
systems (19.3 and 22.3 MJ/kg respectively; BCS = 3.0). Using these ΔBW energy values for the efficiency of ME utilization for body tissue gain (0.65 and 0.75 in FiM and NRC, respectively), the expected coefficient of DMI on ΔBW were 2.50 kg/kg for both systems. Random errors in estimating EB could, at least partly, explain the discrepancy between observed and expected values. In respiration chamber studies, EB is calculated as a difference between GE intake and all energy sinks, and therefore all measurement errors are accumulated in EB. The coefficient was numerically greater for positive than negative energy balance, which is consistent with the greater ME requirement for body gain than the energy supply from mobilization. The ratio of these coefficients was consistent with the
). This can reflect that ΔBW is a poor proxy of EB. Random variation in ΔBW measurements is large, especially when experimental periods are short. Even if ΔBW was measured accurately, it may not reflect changes in EB correctly. The energy content of ΔBW could be highly variable. In early lactation, the cows can be in negative energy balance at zero ΔBW, when high energy fat tissues are mobilized and low energy visceral tissues are growing (
). It could improve accuracy to include BCS in RFI and RECM models because BCS affects both maintenance requirement per kg of MBW and the energy value of ΔBW. Even when based on whole lactation data, the partial regression coefficients obtained in models for predicting ME intake were not biologically meaningful based on
). Their coefficients for MBW were too high and too low for ECM, meaning that the model overestimated the variability in energy intake assigned to MBW and underestimated variability assigned to ECM.
Overall, we can conclude that our partial regression coefficients of DMI on energy sinks were more consistent with energy requirements in different systems (e.g.,
). Inconsistent coefficients, especially in early lactation, can be related to the poor relationship between DMI and ECM yield due to the mobilization of body tissues, and to poor predictions of true EB from ΔBW.
Residual ECM Model
Partial regression coefficients of ECM yield on GE intake, MBW, EBp, and EBn were consistent with energy requirements in the FiM (
systems. Using the average GE concentration (18.4 MJ/kg of DM), the calculated increase in ECM yield was 2.0 kg/kg of DMI, approximately 85% of expected ECM per DMI according to the energy systems stated above. The effect of MBW as an energy sink corresponded to 10 kg of ECM for a 600 kg cow. According to the
system, the maintenance requirement of a 600 kg cow is equivalent to requirements of 12 to 13 kg of ECM. The partial regression coefficients of ECM on EBp and on EBn were about 55% of the coefficients presented by
. When determined ME intake was used in the RECM model (results not shown), the partial regression coefficients were closer to those based on energy requirement: ME intake = 90%, maintenance requirement = 12.5 kg of ECM, and EBp and EBn approximately 60% of requirements (
). Closer agreement with the ME model is likely because it takes into account the metabolizability of the diet (ME/GE), but the model based on GE intake analysis allowed us to evaluate the effects of fecal, CH4 and UE losses on the efficiency variables. Partial regression coefficients of ECM on intake and other energy sinks are seldom reported.
reported values ranging from 0.4 to 1.0 kg of DM per 1 kg of ECM at different stages of lactation; these values were markedly lower than expected according to energy systems.
Factors Influencing Residual Feed Intake and Residual ECM Yield
The major components affecting FE were as follows: (1) factors that alter the dilution of maintenance, i.e., the proportion of NE that is captured in milk instead of used for maintenance and (2) factors that alter the conversion of GE to NE; that is, energy losses in digestion and metabolism of nutrients. As a measure of FE, RFI is independent of production level, BW, and ΔBW (or energy balance in the present study), whereas RECM is independent of intake, BW, and ΔBW. However, selection for RFI and RECM may be difficult because they require accurate measures of individual feed intake, which is seldom recorded on commercial farms. As such, indirect selection using related component traits may be helpful in understanding the expected effect of genetic selection for FE on these traits.
Digestibility
In the present study, diet digestibility was positively related to improved FE, expressed as either RFI or RECM (Table 2, Table 3). Increased fecal energy losses (20.4 MJ/d) accounted for 44% of the greater (46 MJ/d) GE intake of High-RFI cows compared with Low-RFI cows. Reduced DMI and improved digestibility contributed to 42 and 58% of the lower fecal energy losses, respectively, in Low- compared with High-RFI cows. We calculated that the difference in digestibility between Low- and High-RECM cows accounted for 30% (1.8 kg of ECM) of the difference in RECM. Our findings are consistent with other studies that demonstrated negative relationships between diet digestibility and RFI, although these were not always significant. Based on limited data for beef cattle, the contribution of the variation in digestive efficiency to the differences in RFI among cows was 10 to 20% (
found that the differences in DM digestibility accounted for 14% of the difference in RFI between the 2 groups of cattle. In lactating Holstein cows (n = 109) fed 2 diets,
Associations between residual feed intake and apparent nutrient digestibility, in vitro methane-producing activity, and volatile fatty acid concentrations in growing beef cattle.
Associations between residual feed intake and apparent nutrient digestibility, in vitro methane-producing activity, and volatile fatty acid concentrations in growing beef cattle.
In the present study, digestibility estimated by linear regression among RFI groups was reduced by 8.8 g/kg of DMI (results not shown), agreeing with the value (7.2) reported by
suggested that that overall higher digestibility in low RFI cattle might be the consequence of lower DMI. However, in the present study digestibility was not significantly influenced by DMI (
from a meta-analysis examining the effects of intake and diet composition on digestibility. Overall, the differences in digestibility between RFI groups found in the study by
were greater than could be predicted from DMI effects, resulting from increased digesta passage rate for high-RFI group. This suggests that RFI groups are more divergent in digestibility than could be expected from differences in digesta passage rate. The differences in digestibility between efficiency groups might also be associated with the longer rumination time per kilogram of DM in high-efficient cows (
), but this may not necessarily cause digestibility differences.
Methane
Considering the concerns that enteric CH4 production from ruminants is contributing to climate change, improving FE could help mitigate CH4 production, while sustaining current levels of milk production (
Relationships of feedlot feed efficiency, performance, and feeding behavior with metabolic rate, methane production, and energy partitioning in beef cattle.
). When the cows were grouped according to RFI, CH4 production was 3.4 MJ/d lower in Low- than in High-RFI cows, which equals 7% of the difference in GE intake between the groups. The lack of differences in CH4 yield (kJ of CH4E/MJ of GE) between RFI groups was unexpected because CH4 yield is normally negatively associated with intake (
) predicted 5 kJ/MJ higher CH4 yield for low-RFI cows than for high-RFI cows, but there were no differences in CH4 yield between RFI groups in our study.
When the cows were grouped according to RECM, the lower (1.3 MJ/d) CH4 production contributed to 12% of the greater ME intake of High- compared with Low-RECM cows at the same DMI. Methane yield was the highest in the Low-RECM cows, despite having the lowest digestibility. However, within each RECM group, there was a positive (P < 0.001) association between digestibility and CH4 yield ranging from 0.06 to 0.09 kJ/MJ of GE per kJ/MJ difference in GE digestibility. Predicted (
) CH4 yield was 1.5 kJ/MJ greater for High- than for Low-RECM cows but in contrast, the observed difference was −3.8 kJ/MJ in the current study. Our findings do not fully agree with
, who suggested that if the efficiency improves as a result of increased metabolic efficiency, the CH4 yield would not decrease. However, if the improved efficiency is due to an increase in digestibility, CH4 production may increase. In an earlier study (
), efficiency variables were negatively related to CH4 yield when the statistical model included study effect, but not diet nor period effects.
Different relationships between digestibility and CH4 yield on individual cow and efficiency group basis is difficult to explain. One reason could be a rumen fermentation pattern that favors metabolic efficiency and low CH4 production. However,
Methane production, rumen fermentation, and diet digestibility of Holstein and Jersey dairy cows being divergent in residual feed intake and fed at 2 forage-to-concentrate ratios.
did not observe differences in rumen fermentation patterns between low- and high-RFI cows. Similarly, no differences in rumen fluid VFA were observed between RFI groups in feedlot cattle (
, the differences between RFI groups in rumen VFA patterns were inconsistent between low- and high-RFI beef cattle. In addition, variation in rumen fermentation pattern between cows fed the same diet was small (
) suspected that Methanobrevibacter spp. might use acetate as a substrate for CH4 production, possibly leading to greater energy loss. However, it could be questioned if a difference in the methanogenic population was a causative factor for differences in FE. Even if the CH4 energy loss between Low- and Medium-RECM (0.9 MJ/d) was due to greater utilization of acetate for methanogenesis in Low-RECM cows, the quantitative effect on ME supply is insufficient to account for any major part of the differences in RECM between the efficiency groups. Overall, the contributions of reduced CH4 production to improved efficiency was only 7% (RFI) or 5% (RECM), indicating that if differences in the methanogenic population were the causative factor, the mechanism was not energy-sparing from methanogenesis. Because of the relatively small contribution to ME supply, the possible effects of methanogenic populations on the efficiency should, therefore, be mediated via body metabolism.
Methane intensity (g of CH4/kg of ECM) clearly improved with increased efficiency, as the differences between low and high-efficiency groups were greater for RECM than for RFI (4.8 vs. 2.0 g/kg). Based on Akaike's information criteria and residual variance, RECM and ECM were better predictors of CH4 intensity than total CH4 production or CH4 yield. Changes in CH4 intensity were mainly due to reduced DMI (RFI model) or increased production (RECM model), with only minor effects assigned to changes in CH4 yield. Our results suggest that selecting efficient animals is the most sustainable and efficient way to reduce CH4 production per unit of product, and does not require any measurements of CH4 production, which is challenging under commercial conditions.
The Efficiency of ME Utilization
The efficiency of ME utilization had a greater influence than the metabolizability of the diet (ME/GE) on both FE variables (i.e., RFI and RECM). The greater ME intake (20.5 MJ/d) of High- compared with Low-RFI cows was counterbalanced by an equivalent loss as heat. As heat is produced only from ME and body energy mobilization, it can be calculated that the difference in HP accounts for 1.7 kg of DM using the average dietary ME concentration (20.5 MJ of ME/11.8 MJ of ME/kg of DM = 1.7 kg of DM). Therefore, 65% of the difference in RFI between low- and high-efficiency cows could be attributed to a more efficient metabolic processes, i.e., converting ME to milk energy. Similarly, when using RECM as an efficiency trait improved metabolic efficiency accounted for 64% of the higher efficiency of High- compared with Low-RECM cows.
estimated that the contribution to RFI of various biological processes in cattle was 37% tissue metabolism, 9% heat increment of feeding, 10% activity, and 5% body composition. In total, the processes that are related to differences in the metabolism of absorbed nutrients contributed 61% to the variation in RFI (
We calculated kl using the classical approach by regressing milk energy-corrected for zero energy balance against ME intake per MBW. With this approach, almost all variation in the efficiency is attributed to metabolic efficiency of converting ME above maintenance to milk, whereas maintenance requirement is only influenced by ME/GE (q-value) that varies marginally between cows fed the same diet. However, when milk energy at zero energy balance was regressed against ME intake separately for each RECM group, the difference between Low- vs. Medium- and High-RECM cows was in the intercept (i.e., maintenance requirement), whereas the difference between Medium- and High-RECM cows was mainly due to the higher slope (i.e., kl) of High-RECM cows. According to
, the efficiency of energy utilization in the mammary gland is rather constant and the variation in maintenance requirement is the main cause of differences in the efficiency.
stated that resting energy expenditures could vary over a 2- to 3-fold range in animals of equal weight. The coefficient of variation of fasting HP of 14 dairy cows was 10.4% (mean = 0.42 MJ/kg of MBW) when measured 31 d after lactation ceased (
The metabolisable energy requirement for maintenance and the efficiency of utilisation of metabolisable energy for lactation by dairy cows offered grass silage-based diets.
found that the ME requirement for maintenance varied by 8 to 10% between cows of similar size. The results from fasting HP studies suggest that the variation in the maintenance requirement can have a large effect on between-cow variation in FE, attributable to the tissue metabolism component.
Differences in activity and feeding behavior could contribute to variation in the efficiency of dairy cows. However, in the studies of
, RFI was not related to the activity estimated by a pedometer. This is not surprising considering a low (140 steps/h; SD ∼30) overall activity rate of dairy cows in barn conditions (
, ME requirement for 1 km of walking for a 600-kg cow is about 1.6 MJ of ME, equal to the requirement of 0.3 kg of ECM (5% of the difference between Low- and High-RECM). Differences between high- and low-efficient cows in feeding behavior have been inconsistent. Eating time was positively correlated with RFI in the study of
. In contrast, studies of Conner et al. (2013) and Ben Geir et al. (2018) showed eating time was similar for low- and high-RFI cows. Eating rate (Conner et al., 2013; Ben Geir et al., 2018;
). Considering the differences in HP between low and high-efficiency cows (21 MJ/d for High-RFI vs. Low-RFI; 9 MJ/d for High-RECM vs. Low-RECM), it is unlikely that differences in feeding behavior have any major contribution to the differences in FE.
found a positive phenotypic correlation between SCC and RFI, suggesting that increased SCC might partly explain variation in the efficiency of feed conversion among cows. In agreement, increased SCC was associated with decreases in ECM yield and DMI, but as the relative decrease was greater for ECM yield, FE expressed as ECM/DMI decreased (
found that more efficient cows exhibited differences in genes associated with immunity and the inflammatory response, which could affect their ability to elicit a response to an immune challenge. The effect of SCC on FE is likely related to increased energy expenditure associated with inflammation.
Energy Balance.
Increasing negative energy balance is a major concern when using FE expressed as an input/output ratio. In the current analysis, differences in partitioning nutrients between milk production and body energy resources explained most of the differences in ECM/DMI between FE groups without any differences in metabolizability of GE or in kl. However, both RFI and RECM indicated differences in the efficiency between FE groups. Overestimation of kl for cows with positive EB could be one reason for this discrepancy. We used the
). Using the efficiency values for energy retention and mobilization estimated from the current data, the kl-values for Low-, Medium- and High-FCE groups were 0.624, 0.640, and 0.664 (P < 0.001), respectively. This indicated that there were true differences in the efficiency of ME utilization between the FE groups. The differences in kl between RFI and RECM groups were similar whether El(0) was calculated using coefficients of
The results of the current study indicate that there are large differences in the efficiency of feed conversion among cows that are not related to dilution of maintenance requirement or repartitioning of nutrients between milk production and body tissues. It also showed that between-cow differences in converting dietary GE to ME and the efficiency of ME utilization had a strong influence on efficiency traits. The current study was based on respiration chamber data that most likely give a more accurate estimate of energy balance than estimates from ΔBW, especially if the measurement periods are short. However,
found no differences in plasma nonesterified fatty acid concentrations between low- and high-RFI cows, indicating that higher efficiency was not due to differences in mobilization. In the present study, the residuals were not related to energy balance when RECM was predicted from RFI, suggesting that energy balance influences these 2 efficiency traits in the same way.
, found no correlation between RFI and fertility, but reported a positive correlation between residual milk solids production, and fertility, suggesting that the mechanisms are different. It is possible that efficient cows have more resources available for other functions, such as reproduction.
CONCLUSIONS
Data from respiration chamber studies showed considerable variation in FE among cows when expressed as either RFI or RECM. The partial regression coefficients of energy sinks for predicting DMI or ECM were biologically meaningful. About 65% of the difference between low- and high-efficiency cows, irrespective of efficiency trait (RFI or RECM), was derived from improved utilization of ME, and 35% assigned to greater metabolizability of GE. Improved digestibility and reduced CH4 production accounted for 83 and 12% of increased ME intake, respectively. Regression analysis within each RECM group suggested that the difference between Low- vs. Medium- and High-RECM groups was mainly due to the higher maintenance requirement in Low-RECM cows, while the difference between Medium- and High-RECM groups resulted mainly from improved kl. Variation among cows in FCE was mainly due to differences in partitioning energy between milk production and body tissues when milk energy at zero energy balance was estimated using
coefficients. Methane production per kg of ECM reduced with improved efficiency, with a greater difference between RECM groups than between RFI groups. Increased ECM yield from the same DMI (RECM) or reduced DMI at the same ECM yield (RFI) contributed to most of the differences in CH4/ECM among the efficiency groups, while differences in CH4/DMI had only minor (RECM) or no (RFI) effects on CH4/ECM. Therefore, improving FE is a sustainable way to reduce CH4 production per unit of product.
ACKNOWLEDGMENTS
This study was funded by the Faculty of Animal Science and Veterinary Medicine of Swedish University of Agricultural Sciences (SLU) with co-funding from Swedish Farmers' Foundation for Agricultural Research (SLF), Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (FORMAS), and the Feed Utilization in Nordic Cattle (FUNC) project (Denmark). The authors thank their colleagues from Agri-Food and Biosciences Institute (Hillsborough, UK), Agri-Food Research in Finland (currently Natural Resources Institute Finland(LUKE), Jokioinen, Finland), and Aarhus University (Aarhus, Denmark), for the experimental data used in the present study. The authors have not stated any conflicts of interest.
Associations between residual feed intake and apparent nutrient digestibility, in vitro methane-producing activity, and volatile fatty acid concentrations in growing beef cattle.
Relationships of feedlot feed efficiency, performance, and feeding behavior with metabolic rate, methane production, and energy partitioning in beef cattle.
Methane production, rumen fermentation, and diet digestibility of Holstein and Jersey dairy cows being divergent in residual feed intake and fed at 2 forage-to-concentrate ratios.
A Nordic proposal for an energy corrected milk (ECM) formula.
in: Gaillon P. Chabert Y. European Association for Animal Production Publication, Performance Recording of Animals: State of the Art, 1990; 27th Biennial Session of the International Committee for Animal Recording. Centre for Agricultural Publishing and Documentation,
Paris, France1990: 156-192
The metabolisable energy requirement for maintenance and the efficiency of utilisation of metabolisable energy for lactation by dairy cows offered grass silage-based diets.
A meta-analysis based on an individual-cow data set was conducted to investigate between-cow variations in the components and measurements of feed efficiency (FE) and to explore the associations among these components. Data were taken from 31 chamber studies, consisting of a total of 841 cow/period observations. The experimental diets were based on grass or corn silages, fresh grass, or a mixture of fresh grass and straw, with cereal grains or by-products as energy supplements, and soybean or canola meal as protein supplements.
30452-5&id=gr1.jpg){kind=link}
30452-5&id=gr2.jpg){kind=link}