Reliability of breeding values for feed intake and feed efficiency traits in dairy cattle: When dry matter intake recordings are sparse under different scenarios

Currently, routine recordings of dry matter intake (DMI) in commercial herds are practically nonexistent. Recording DMI from commercial herds is a prerequisite for the inclusion of feed efficiency (FE) traits in dairy cattle breeding goals. To develop future on-farm phenotyping strategies, recording strategies that are low cost and less demanding logistically and that give relatively accurate estimates of the animal's genetic merit are therefore needed. The objectives of this study were (1) to estimate genetic parameters for daily DMI and FE traits and use the estimated parameters to simulate daily DMI phenotypes under different DMI recording scenarios (SCN) and (2) to use the simulated data to estimate for different scenarios the associated reliability of estimated breeding value and accuracies of genomic prediction for varying sizes of reference populations. Five on-farm daily DMI recording scenarios were simulated: once weekly (SCN1), once monthly (SCN2), every 2 mo (SCN3), every 3 mo (SCN4), and every 4 mo (SCN5). To estimate reliability of estimated breeding values, DMI and FE observations and true breeding values were simulated based on variance components estimated for daily observations of Nordic Red cows. To emulate realistic on-farm recording, 5 data set replicates, each with 36,037 DMI and FE records, were simulated for real pedigree and data structure of 789 Holstein cows. Observations for the 5 DMI recording scenarios were generated by discarding data in a step-wise manner from the full simulated data per the scenario's definitions. For each of these scenarios, reliabilities were calculated as correlation between the true and estimated breeding values. Variance components and genetic parameters were estimated for daily DMI, residual feed intake (RFI), and energy conversion efficiency (ECE) fitting the random regression model. Data for variance components were from 227 primiparous Nordic Red dairy cows covering 8 to 280 d in milk. Lactation-wise heritability for DMI, RFI, and ECE was 0.33, 0.12, and 0.32, respectively, and daily heritability estimates during lactation ranged from 0.18 to 0.45, 0.08 to 0.32, and 0.08 to 0.45 for DMI, RFI, and ECE, respectively. Genetic correlations for DMI between different stages of lactation ranged from ?0.50 to 0.99. The comparison of different on-farm DMI recording scenarios indicated that adopting a less-frequent recording scenario (SCN3) gave a similar level of accuracy as SCN1 when 17 more daughters are recorded per sire over the 46 needed for SCN1. Such a strategy is less demanding logistically and is low cost because fewer observations need to be collected per animal. The accuracy of genomic predictions associated with the 5 recording scenarios indicated that setting up a relatively larger reference population and adopting a less-frequent DMI sampling scenario (e.g., SCN3) is promising. When the same reference population size was considered, the genomic prediction accuracy of SCN3 was only 5.0 to 7.0 percentage points lower than that for the most expensive DMI recording strategy (SCN1). We concluded that DMI recording strategies that are sparse in terms of records per cow but with slightly more cows recorded per sire are advantageous both in genomic selection and in traditional progeny testing schemes when accuracy, logistics, and cost implications are considered.     

Short communication: Validation of genomic breeding value predictions for feed intake and feed efficiency traits

Validating genomic prediction equations in independent populations is an important part of evaluating genomic selection. Published genomic predictions from 2 studies on (1) residual feed intake and (2) dry matter intake (DMI) were validated in a cohort of 78 multiparous Holsteins from Australia. The mean realized accuracy of genomic prediction for residual feed intake was 0.27 when the reference population included phenotypes from 939 New Zealand and 843 Australian growing heifers (aged 5–8 mo) genotyped on high density (770k) single nucleotide polymorphism chips. The 90% bootstrapped confidence interval of this estimate was between 0.16 and 0.36. The mean realized accuracy was slightly lower (0.25) when the reference population comprised only Australian growing heifers. Higher realized accuracies were achieved for DMI in the same validation population and using a multicountry model that included 958 lactating cows from the Netherlands and United Kingdom in addition to 843 growing heifers from Australia. The multicountry analysis for DMI generated 3 sets of genomic predictions for validation animals, one on each country scale. The highest mean accuracy (0.72) was obtained when the genomic breeding values were expressed on the Dutch scale. Although the validation population used in this study was small (n = 78), the results illustrate that genomic selection for DMI and residual feed intake is feasible. Multicountry collaboration in the area of dairy cow feed efficiency is the evident pathway to achieving reasonable genomic prediction accuracies for these valuable traits.     

Direct and indirect predictions of enteric methane daily production,yield, and intensity per unit of milk and cheese,from fatty acids and milk Fourier-transform infrared spectra

Mitigating the dairy chain's contribution to climate change requires cheap, rapid methods of predicting enteric CH₄ emissions (EME) of dairy cows in the field. Such methods may also be useful for genetically improving cows to reduce EME. Our objective was to evaluate different procedures for predicting EME traits from infrared spectra of milk samples taken at routine milk recording of cows. As a reference method, we used EME traits estimated from published equations developed from a meta-analysis of data from respiration chambers through analysis of various fatty acids in milk fat by gas chromatography (FA_GC). We analyzed individual milk samples of 1,150 Brown Swiss cows from 85 farms operating different dairy systems (from very traditional to modern), and obtained the cheese yields of individual model cheeses from these samples. We also obtained Fourier-transform infrared absorbance spectra on 1,060 wavelengths (5,000 to 930 waves/cm) from the same samples. Five reference enteric CH₄ traits were calculated: CH₄ yield (CH₄/DMI, g/kg) per unit of dry matter intake (DMI), and CH₄ intensity (CH₄/CM, g/kg) per unit of corrected milk (CM) from the FA_GC profiles; CH₄ intensity per unit of fresh cheese (CH₄/CY_CURD, g/kg) and cheese solids (CH₄/CY_SOLIDS, g/kg) from individual cheese yields (CY); and daily CH₄ production (dCH₄, g/d). Direct infrared (IR) calibrations were obtained by BayesB modeling; the determination coefficients of cross-validation varied from 0.36 for dCH₄ to 0.57 for CH₄/CM, and were similar to the coefficient of determination values of the equations based on FA_GC used as the reference method (0.47 for CH₄/DMI and 0.54 for CH₄/CM). The models allowed us to select the most informative wavelengths for each EME trait and to infer the milk chemical features underlying the predictions. Aside from the 5 direct infrared prediction calibrations, we tested another 8 indirect prediction models. Using IR-predicted informative fatty acids (FA_IR) instead of FA_GC, we were able to obtain indirect predictions with about the same precision (correlation with reference values) as direct IR predictions of CH₄/DMI (0.78 vs. 0.76, respectively) and CH₄/CM (0.82 vs. 0.83). The indirect EME predictions based on IR-predicted CY were less precise than the direct IR predictions of both CH₄/CY_CURD (0.67 vs. 0.81) and CH₄/CY_SOLIDS (0.62 vs. 0.78). Four indirect dCH₄ predictions were obtained by multiplying the measured or IR-predicted daily CM production by the direct or indirect CH₄/CM. Combining recorded daily CM and predicted CH₄/CM greatly increased precision over direct dCH₄ predictions (0.96–0.96 vs. 0.68). The estimates obtained from the majority of direct and indirect IR-based prediction models exhibited herd and individual cow variability and effects of the main sources of variation (dairy system, parity, days in milk) similar to the reference data. Some rapid, cheap, direct and indirect IR prediction models appear to be useful for monitoring EME in the field and possibly for genetic/genomic selection, but future studies directly measuring CH₄ with different breeds and dairy systems are needed to validate our findings.     

Genetic parameters across lactation for feed intake,fat- and protein-corrected milk,and liveweight in first-parity Holstein cattle

Breeding values for dry matter intake (DMI) are important to optimize dairy cattle breeding goals for feed efficiency. However, generally, only small data sets are available for feed intake, due to the cost and difficulty of measuring DMI, which makes understanding the genetic associations between traits across lactation difficult, let alone the possibility for selection of breeding animals. However, estimating national breeding values through cheaper and more easily measured correlated traits, such as milk yield and liveweight (LW), could be a first step to predict DMI. Combining DMI data across historical nutritional experiments might help to expand the data sets. Therefore, the objective was to estimate genetic parameters for DMI, fat- and protein-corrected milk (FPCM) yield, and LW across the entire first lactation using a relatively large data set combining experimental data across the Netherlands. A total of 30,483 weekly records for DMI, 49,977 for FPCM yield, and 31,956 for LW were available from 2,283 Dutch Holstein-Friesian first-parity cows between 1990 and 2011. Heritabilities, covariance components, and genetic correlations were estimated using a multivariate random regression model. The model included an effect for year-season of calving, and polynomials for age of cow at calving and days in milk (DIM). The random effects were experimental treatment, year-month of measurement, and the additive genetic, permanent environmental, and residual term. Additive genetic and permanent environmental effects were modeled using a third-order orthogonal polynomial. Estimated heritabilities ranged from 0.21 to 0.40 for DMI, from 0.20 to 0.43 for FPCM yield, and from 0.25 to 0.48 for LW across DIM. Genetic correlations between DMI at different DIM were relatively low during early and late lactation, compared with mid lactation. The genetic correlations between DMI and FPCM yield varied across DIM. This correlation was negative (up to −0.5) between FPCM yield in early lactation and DMI across the entire lactation, but highly positive (above 0.8) when both traits were in mid lactation. The correlation between DMI and LW was 0.6 during early lactation, but decreased to 0.4 during mid lactation. The highest correlations between FPCM yield and LW (0.3–0.5) were estimated during mid lactation. However, the genetic correlations between DMI and either FPCM yield or LW were not symmetric across DIM, and differed depending on which trait was measured first. The results of our study are useful to understand the genetic relationship of DMI, FPCM yield, and LW on specific days across lactation.     

Blood plasma traits associated with genetic merit for feed utilization in Holstein cows

The objective of this study was to evaluate the potential of selection for feed utilization on associated blood plasma metabolite and hormone traits. Dry matter intake (DMI) was recorded in 970 Holsteins from 11 commercial farms in Pennsylvania and used to derive dry matter efficiency (DME; fat-corrected milk yield/DMI), crude protein efficiency (CPE; protein yield/crude protein intake), and residual feed intake (RFI, defined as actual feed intake minus expected feed intake for maintenance and milk production, based on calculation of DMI adjusted for yield, body weight, and body condition score). Estimated breeding values for the 4 feed utilization traits (DMI, DME, CPE, and RFI), yield traits, body traits, and days open were standardized according to their respective genetic standard deviations. Up to 631 blood samples from 393 cows from 0 to 60 d in milk (DIM) were evaluated for blood plasma concentrations of glucose, nonesterified fatty acids (NEFA), β-hydroxybutyrate (BHB), creatinine, urea, growth hormone (GH), 3,5,3′-triiodothyronine (T3), and other parameters. Blood plasma traits were regressed on DIM, lactation number, herd, and standardized genetic merit. Cows with higher genetic merit for yield had significantly higher concentrations of GH, NEFA (milk and protein yield), and BHB (fat yield) from 31 to 60 DIM, but lower concentrations of glucose from 0 to 30 DIM, and T3 (milk yield, 0–60 DIM). The high GH–low glucose–low T3 concentration pattern was further accentuated for cows with genetic merit for enhanced feed efficiency (higher DME and lower RFI). Cows with a genetic tendency to be thin (low body condition score) also had elevated GH concentrations, but lower blood glucose, creatinine, and T3 concentrations. Those characteristics associated with enhanced feed efficiency (higher GH and lower glucose and T3 concentrations) were unfavorably associated with fertility, as indicated by elevated days open. Elevated NEFA and BHB concentrations were also associated with extended days open. Consideration of metabolic profiles when evaluating feed efficiency might be a method of maintaining high levels of health and reproductive fitness when selecting for feed efficiency.     

Genetic analysis of efficiency traits in Austrian dairy cattle and their relationships with body condition score and lameness

This study is part of a larger project whose overall objective was to evaluate the possibilities for genetic improvement of efficiency in Austrian dairy cattle. In 2014, a 1-yr data collection was carried out. Data from 6,519 cows kept on 161 farms were recorded. In addition to routinely recorded data (e.g., milk yield, fertility, disease data), data of novel traits [e.g., body weight (BW), body condition score (BCS), lameness score, body measurements] and individual feeding information and feed quality were recorded on each test-day. The specific objective of this study was to estimate genetic parameters for efficiency (related) traits and to investigate their relationships with BCS and lameness in Austrian Fleckvieh, Brown Swiss, and Holstein cows. The following efficiency (related) traits were considered: energy-corrected milk (ECM), BW, dry matter intake (DMI), energy intake (INEL), ratio of milk output to metabolic BW (ECM/BW^0.75), ratio of milk output to DMI (ECM/DMI), and ratio of milk energy output to total energy intake (LE/INEL, LE = energy in milk). For Fleckvieh, the heritability estimates of the efficiency (related) traits ranged from 0.11 for LE/INEL to 0.44 for BW. Heritabilities for BCS and lameness were 0.19 and 0.07, respectively. Repeatabilities were high and ranged from 0.30 for LE/INEL to 0.83 for BW. Heritability estimates were generally lower for Brown Swiss and Holstein, but repeatabilities were in the same range as for Fleckvieh. In all 3 breeds, more-efficient cows were found to have a higher milk yield, lower BW, slightly higher DMI, and lower BCS. Higher efficiency was associated with slightly fewer lameness problems, most likely due to the lower BW (especially in Fleckvieh) and higher DMI of the more-efficient cows. Body weight and BCS were positively correlated. Therefore, when selecting for a lower BW, BCS is required as additional information because, otherwise, no distinction between large animals with low BCS and smaller animals with normal BCS would be possible.     

Identification of biological traits associated with differences in residual energy intake among lactating Holstein cows

Residual feed intake, which is usually used to estimate individual variation of feed efficiency, requires frequent and accurate measurements of individual feed intake to be carried out. Developing a breeding scheme based on residual feed intake in dairy cows is therefore complicated, especially because feed intake is not measurable for a large population. Another solution could be to focus on biological determinants of feed efficiency, which could potentially be directly and broadband measurable on farm. Several phenotypes have been identified in literature as being associated with differences in feed efficiency. The present study therefore aims to identify which biological mechanisms are associated with residual energy intake (REI) differences among dairy cows. Several candidate phenotypes were recorded frequently and simultaneously throughout the first 238 d in milk for 60 Holstein cows fed on a constant diet based on maize silage. A multiple linear regression of the 238 d in milk average of net energy intake was fitted on the 238 d in milk averages for milk energy output, metabolic body weight, the sum over the 238 d in milk of both, body condition score loss and gain, and the residuals were defined as REI. A partial least square regression was fitted over all biological traits to explain REI variability. Linear multiple regression explained 93.6% of net energy intake phenotypic variation, with 65.5% associated with lactation requirement, 23.2% with maintenance, and 4.9% with body reserves change; the 6.4% residuals represented REI. Overall, measured biological traits contributed to 58.9% of REI phenotypic variability, which were mainly explained by activity (26.5%) and feeding behavior (21.3%). However, apparent confounding was observed between behavior, activity, digestibility, and rumen-temperature variables. Drawing a conclusion on biological traits that explain feed efficiency differences among dairy cows was not possible due to this apparent confounding between the measured variables. Further investigation is needed to validate these results and to characterize the causal relationship of feed efficiency with feeding behavior, digestibility, body reserves change, activity, and rumen temperature.     

Genetic parameters of rumination time and feed efficiency traits in primiparous Holstein cows under research and commercial conditions

Feed efficiency has the potential to be improved both through feeding, management, and breeding. Including feed efficiency in a selection index is limited by the fact that dry matter intake (DMI) recording is only feasible under research facilities, resulting in small data sets and, consequently, uncertain genetic parameter estimates. As a result, the need to record DMI indicator traits on a larger scale exists. Rumination time (RT), which is already recorded in commercial dairy herds by a sensor-based system, has been suggested as a potential DMI indicator. However, RT can only be a DMI indicator if it is heritable, correlates with DMI, and if the genetic parameters of RT in commercial herd settings are similar to those in research facilities. Therefore, the objective of our study was to estimate genetic parameters for RT and the related traits of DMI in primiparous Holstein cows, and to compare genetic parameters of rumination data between a research herd and 72 commercial herds. The estimated heritability values were all moderate for DMI (0.32–0.49), residual feed intake (0.23–0.36), energy-corrected milk (ECM) yield (0.49–0.70), and RT (0.14–0.44) found in the research herd. The estimated heritability values for ECM were lower for the commercial herds (0.08–0.35) than that for the research herd. The estimated heritability values for RT were similar for the 2 herd types (0.28–0.32). For the research herd, we found negative individual level correlations between RT and DMI (?0.24 to ?0.09) and between RT and RFI (?0.34 to ?0.03), and we found both positive and negative correlations between RT and ECM (?0.08 to 0.09). For the commercial herds, genetic correlations between RT and ECM were both positive and negative (?0.27 to 0.10). In conclusion, RT was not found to be a suitable indicator trait for feed intake and only a weak indicator of feed efficiency.     

Evaluation of different feed intake models for dairy cows

The objective of the current study was to evaluate feed intake prediction models of varying complexity using individual observations of lactating cows subjected to experimental dietary treatments in periodic sequences (i.e., change-over trials). Observed or previous period animal data were combined with the current period feed data in the evaluations of the different feed intake prediction models. This would illustrate the situation and amount of available data when formulating rations for dairy cows in practice and test the robustness of the models when milk yield is used in feed intake predictions. The models to be evaluated in the current study were chosen based on the input data required in the models and the applicability to Nordic conditions. A data set comprising 2,161 total individual observations was constructed from 24 trials conducted at research barns in Denmark, Finland, Norway, and Sweden. Prediction models were evaluated by residual analysis using mixed and simple model regression. Great variation in animal and feed factors was observed in the data set, with ranges in total dry matter intake (DMI) from 10.4 to 30.8 kg/d, forage DMI from 4.1 to 23.0 kg/d, and milk yield from 8.4 to 51.1 kg/d. The mean biases of DMI predictions for the National Research Council, the Cornell Net Carbohydrate and Protein System, the British, Finnish, and Scandinavian models were −1.71, 0.67, 2.80, 0.83, −0.60 kg/d with prediction errors of 2.33, 1.71, 3.19, 1.62, and 2.03 kg/d, respectively, when observed milk yield was used in the predictions. The performance of the models were ranked the same, using either mixed or simple model regression analysis, but generally the random contribution to the prediction error increased with simple rather than mixed model regression analysis. The prediction error of all models was generally greater when using previous period data compared with the observed milk yield. When the average milk yield over all periods was used in the predictions of feed intake, the increase in prediction error of all models was generally less than when compared with previous period animal data combined with current feed data. Milk yield as a model input in intake predictions can be substantially affected by current dietary factors. Milk yield can be used as model input when formulating rations aiming to sustain a given milk yield, but can generate large errors in estimates of future feed intake and milk production if the economically optimal diet deviates from the current diet.     

Breeding goal traits accounting for feed intake capacity and roughage or concentrate intake separately

Current breeding tools aiming to improve feed efficiency use definitions based on total dry matter intake (DMI); for example, residual feed intake or feed saved. This research aimed to define alternative traits using existing data that differentiate between feed intake capacity and roughage or concentrate intake, and to investigate the phenotypic and genetic relationships among these traits. The data set contained 39,017 weekly milk yield, live weight, and DMI records of 3,164 cows. The 4 defined traits were as follows: (1) Feed intake capacity (FIC), defined as the difference between how much a cow ate and how much she was expected to eat based on diet satiety value and status of the cow (parity and lactation stage); (2) feed saved (FS), defined as the difference between the measured and the predicted DMI, based on the regression of DMI on milk components within experiment; (3) residual roughage intake (RRI), defined as the difference between the measured and the predicted roughage intake, based on the regression of roughage intake on milk components and concentrate intake within experiment; and (4) residual concentrate intake (RCI), defined as the difference between the measured and the predicted concentrate intake, based on the regression of concentrate intake on milk components and roughage intake within experiment. The phenotypic correlations were ?0.72 between FIC and FS, ?0.84 between FS and RRI, and ?0.53 between FS and RCI. Heritability of FIC, FS, RRI, and RCI were estimated to be 0.21, 0.12, 0.15, and 0.03, respectively. The genetic correlations were ?0.81 between FS and FIC, ?0.96 between FS and RRI, and ?0.25 between FS and RCI. Concentrate intake and RCI had low heritability. Genetic correlation between DMI and FIC was 0.98. Although the defined traits had moderate phenotypic correlations, the genetic correlations between DMI, FS, FIC, and RRI were above 0.79 (in absolute terms), suggesting that these traits are genetically similar. Therefore, selecting for FIC is expected to simply increase DMI and RRI, and there seems to be little advantage in separating concentrate and roughage intake in the genetic evaluation, because measured concentrate intake was determined by the feeding system in our data and not by the genetics of the cow.     

Weather influences feed intake and feed efficiency in a temperate climate

A key goal for livestock science is to ensure that food production meets the needs of an increasing global population. Climate change may heighten this challenge through increases in mean temperatures and in the intensity, duration, and spatial distribution of extreme weather events, such as heat waves. Under high ambient temperatures, livestock are expected to decrease dry matter intake (DMI) to reduce their metabolic heat production. High yielding dairy cows require high DMI to support their levels of milk production, but this may increase susceptibility to heat stress. Here, we tested how feed intake and the rate of converting dry matter to milk (feed efficiency, FE) vary in response to natural fluctuations in weather conditions in a housed experimental herd of lactating Holstein Friesians in the United Kingdom. Cows belonged to 2 lines: those selected for high genetic merit for milk traits (select) and those at the UK average (control). We predicted that (1) feed intake and FE would vary with an index of temperature and humidity (THI), wind speed, and the number of hours of sunshine, and that (2) the effects of (1) would depend on the cows' genetic merit. Animals received a mixed ration, available ad libitum, from automatic feed measurement gates. Using >73,000 daily feed intake and FE records from 328 cows over 8 yr, we found that select cows produced more fat- and protein-corrected milk, and had higher DMI and FE than controls. Cows of both lines decreased DMI and fat- and protein-corrected milk but, importantly, increased FE as THI increased. This suggests that improvements in the efficiency of converting feed to milk may partially offset the costs of reduced milk yield owing to a warmer climate, at least under conditions of mild heat stress. The rate of increase in FE with THI was steeper in select cows than in controls, which raises the possibility that select cows use more effective coping tactics. This is, to our knowledge, the first longitudinal study on the effects of weather on FE. Understanding how weather influences feed intake and efficiency can help us to develop management and selection practices that optimize productivity under unfavorable weather conditions. This will be an important aspect of climate resilience in future.     

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

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.     

Predicting dry matter intake in Canadian Holstein dairy cattle using milk mid-infrared reflectance spectroscopy and other commonly available predictors via artificial neural networks

Dry matter intake (DMI) is a fundamental component of the animal's feed efficiency, but measuring DMI of individual cows is expensive. Mid-infrared reflectance spectroscopy (MIRS) on milk samples could be an inexpensive alternative to predict DMI. The objectives of this study were (1) to assess if milk MIRS data could improve DMI predictions of Canadian Holstein cows using artificial neural networks (ANN); (2) to investigate the ability of different ANN architectures to predict unobserved DMI; and (3) to validate the robustness of developed prediction models. A total of 7,398 milk samples from 509 dairy cows distributed over Canada, Denmark, and the United States were analyzed. Data from Denmark and the United States were used to increase the training data size and variability to improve the generalization of the prediction models over the lactation. For each milk spectra record, the corresponding weekly average DMI (kg/d), test-day milk yield (MY, kg/d), fat yield (FY, g/d), and protein yield (PY, g/d), metabolic body weight (MBW), age at calving, year of calving, season of calving, days in milk, lactation number, country, and herd were available. The weekly average DMI was predicted with various ANN architectures using 7 predictor sets, which were created by different combinations MY, FY, PY, MBW, and MIRS data. All predictor sets also included age of calving and days in milk. In addition, the classification effects of season of calving, country, and lactation number were included in all models. The explored ANN architectures consisted of 3 training algorithms (Bayesian regularization, Levenberg-Marquardt, and scaled conjugate gradient), 2 types of activation functions (hyperbolic tangent and linear), and from 1 to 10 neurons in hidden layers). In addition, partial least squares regression was also applied to predict the DMI. Models were compared using cross-validation based on leaving out 10% of records (validation A) and leaving out 10% of cows (validation B). Superior fitting statistics of models comprising MIRS information compared with the models fitting milk, fat and protein yields suggest that other unknown milk components may help explain variation in weekly average DMI. For instance, using MY, FY, PY, and MBW as predictor variables produced a predictive accuracy (r) ranging from 0.510 to 0.652 across ANN models and validation sets. Using MIRS together with MY, FY, PY, and MBW as predictors resulted in improved fitting (r = 0.679–0.777). Including MIRS data improved the weekly average DMI prediction of Canadian Holstein cows, but it seems that MIRS predicts DMI mostly through its association with milk production traits and its utility to estimate a measure of feed efficiency that accounts for the level of production, such as residual feed intake, might be limited and needs further investigation. The better predictive ability of nonlinear ANN compared with linear ANN and partial least squares regression indicated possible nonlinear relationships between weekly average DMI and the predictor variables. In general, ANN using Bayesian regularization and scaled conjugate gradient training algorithms yielded slightly better weekly average DMI predictions compared with ANN using the Levenberg-Marquardt training algorithm.     

Short communication: Interactions of milk, fat, and protein yield genotypes with herd feeding characteristics

Data from 879 Holstein cows from 11 tie-stall herds in Pennsylvania were analyzed to determine the effects of nutritional management practices on the level of genetic expression for milk, fat, and protein yields. Environments were defined according to the amount of dry matter refusals at the end of 24h for the average cow (DMR), diet crude protein percentage (CP), and diet NE(L) concentration. Sire predicted transmitting ability (PTA) was available for all cows, whereas 775 cows were genotyped and received a molecular breeding value (MBV) for milk, fat, and protein yields. Milk, fat, and protein yields were regressed on sire PTA and cow MBV independently in addition to combined breeding values (CBV) of sire PTA and cow MBV. Four-trait animal models with fat-corrected milk yield in high and low environments plus either body weight or body condition score in high and low environments treated as separate traits were also evaluated. Regressions on sire PTA (0.31 for fat yield to 0.54 for milk yield) were significantly lower in the 5 herds that had the lowest average DMR than in the 6 herds with highest average DMR (0.82 for fat yield to 1.11 for protein yield). The regressions of milk and protein yield on CBV were also significantly lower in the 5 herds with low NE(L) concentration in the ration than in herds that had high NE(L) concentration. Genetic correlations from animal models showed that large cows were more affected by low DMR, CP, and NE(L) concentration than smaller cows. Efforts to minimize feed wastage must ensure that cows receive adequate nutrient intake to avoid suppression of genetic potential for yield, particularly for larger cows.     

Strategies for implementing genomic selection for feed efficiency in dairy cattle breeding schemes

Alternative genomic selection and traditional BLUP breeding schemes were compared for the genetic improvement of feed efficiency in simulated Norwegian Red dairy cattle populations. The change in genetic gain over time and achievable selection accuracy were studied for milk yield and residual feed intake, as a measure of feed efficiency. When including feed efficiency in genomic BLUP schemes, it was possible to achieve high selection accuracies for genomic selection, and all genomic BLUP schemes gave better genetic gain for feed efficiency than BLUP using a pedigree relationship matrix. However, introducing a second trait in the breeding goal caused a reduction in the genetic gain for milk yield. When using contracted test herds with genotyped and feed efficiency recorded cows as a reference population, adding an additional 4,000 new heifers per year to the reference population gave accuracies that were comparable to a male reference population that used progeny testing with 250 daughters per sire. When the test herd consisted of 500 or 1,000 cows, lower genetic gain was found than using progeny test records to update the reference population. It was concluded that to improve difficult to record traits, the use of contracted test herds that had additional recording (e.g., measurements required to calculate feed efficiency) is a viable option, possibly through international collaborations.     

Dietary restriction improved feed efficiency of inefficient lactating cows

The aim of this study was to reduce voluntary dry matter intake (DMI) to increase feeding efficiency of preclassified inefficient (INE) dairy cows through restricted feeding. We studied the effects of dietary restriction on eating behavior, milk and energy-corrected milk (ECM) production, in vivo digestibility, energy balance, and measures of feed efficiency [residual feed intake (RFI) and ECM/DMI]. Before the experiment, 12 pairs of cows were classified as INE. The 2 dietary treatments consisted of ad libitum feeding versus restricted feeding of the same total mixed ration containing 36.5% roughage. Inefficient cows fed the restricted total mixed ration had a shorter eating time and lower meal and visit frequency, but a similar rate of eating, meal size, and meal duration compared with INE cows fed ad libitum. Compared with the INE cows fed ad libitum, restricted INE cows had 12.8% lower intake, their dry matter and neutral detergent fiber digestibility remained similar, and their ECM yield was 5.3% lower. Feed efficiency, measured as RFI, ECM/DMI, and net energy retained divided by digestible energy intake, was improved in the restricted INE cows as compared with the ad libitum cows. Our results show that moderate DMI restriction has the potential to improve feed efficiency of preclassified INE cows.     

Liability to diseases and their relation to dry matter intake and energy balance in German Holstein and Fleckvieh dairy cows

Dairy cow efficiency is increasingly important for future breeding decisions. The efficiency is determined mostly by dry matter intake (DMI). Reducing DMI seems to increase efficiency if milk yield remains the same, but resulting negative energy balance (EB) may cause health problems, especially in early lactation. Objectives of this study were to examine relationships between DMI and liability to diseases. Therefore, cow effects for DMI and EB were correlated with cow effects for 4 disease categories throughout lactation. Disease categories were mastitis, claw and leg diseases, metabolic diseases, and all diseases. In addition, this study presents relative percentages of diseased cows per days in milk (DIM), repeatability, and cow effect correlations for disease categories across DIM. A total of 1,370 German Holstein (GH) and 287 Fleckvieh (FV) primiparous and multiparous dairy cows from 12 dairy research farms in Germany were observed over a period of 2 yr. Farm staff and veterinarians recorded health data. We modeled health and production data with threshold random regression models and linear random regression models. From DIM 2 to 305 average daily DMI was 22.1 kg/d in GH and 20.2 kg/d in FV. Average weekly EB was 2.8 MJ of NE_L/d in GH and 0.6 MJ of NE_L/d in FV. Most diseases occurred in the first 20 DIM. Multiparous cows were more susceptible to diseases than primiparous cows. Relative percentages of diseased cows were highest for claw and leg diseases, followed by metabolic diseases and mastitis. Repeatability of disease categories and production traits was moderate to high. Cow effect correlations for disease categories were higher for adjacent lactation stages than for more distant lactation stages. Pearson correlation coefficients between cow effects for DMI, as well as EB, and disease categories were estimated from DIM 2 to 305. Almost all correlations were negative in GH, especially in early lactation. In FV, the course of correlations was similar to GH, but correlations were mostly more negative in early lactation. For the first 20 DIM, correlations ranged from ?0.31 to 0.00 in GH and from ?0.42 to ?0.01 in FV. The results illustrate that future breeding for dairy cow efficiency should focus on DMI and EB in early lactation to avoid health problems.     

Genome-wide association study for milk infrared wavenumbers

Individual wavenumbers of the infrared (IR) spectra of bovine milk have been shown to be moderately to highly heritable. The objective of this study was to identify genomic regions associated with individual milk IR wavenumbers. This is expected to provide information about the genetic background of milk composition and give insight in the relation between IR wavenumbers and milk components. For this purpose, a genome-wide association study was performed for a selected set of 50 individual IR wavenumbers measured on 1,748 Dutch Holstein cows. Significant associations were detected for 28 of the 50 wavenumbers. In total, 24 genomic regions distributed over 16 bovine chromosomes were identified. Major genomic regions associated with milk IR wavenumbers were identified on chromosomes 1, 5, 6, 14, 19, and 20. Most of these regions also showed significant associations with fat, protein, or lactose percentage. However, we also identified some new regions that were not associated with any one of these routinely collected milk composition traits. On chromosome 1, we identified 2 new genomic regions and hypothesized that they are related to variation in milk phosphorus content and orotic acid, respectively. On chromosome 20, we identified a new genomic region that seems to be related to citric acid. Identification of genomic regions associated with milk phosphorus content, orotic acid, and citric acid suggest that the milk IR spectra contain direct information on these milk components. Consequently milk IR analyses probably can be used to predict these milk components, which have low concentrations in milk; this can lead to novel applications of milk IR spectroscopy for dairy cattle breeding and herd management.     

The effect of steam-flaked or dry ground corn and supplemental phytic acid on phosphorus partitioning and ruminal phytase activity in lactating cows

The effect of starch source and phytic acid (PA) supplementation on phosphorus (P) partitioning and ruminal phytase activity was evaluated in eight midlactation cows (four ruminally cannulated). Cows were randomly assigned to treatments in replicated 4 x 4 Latin squares with four 18-d periods. Diets included dry ground corn (DG) or steam-flaked corn (SF), with no supplemental P (L; 0.33% P) or supplemental purified PA (0.44% P) to provide additional P from a nonmineral source. Total collection of milk, urine, and feces was conducted on d 16 to 18 of each period. Ruminal fluid was sampled and ruminal pH measured every 8 h on d 17 and 18. Milk yield was unaffected by starch source, despite lower DMI by cows fed SF. Cows fed SF had increased DM digestibility compared with those fed DG, and tended to have higher efficiency of milk yield (1.40 vs. 1.35 kg of milk/kg of DMI). Intake and fecal excretion of P was lower in cows fed SF than in cows fed DG. In cows fed SF, milk P as a percentage of P intake increased compared with cows fed DG. Ruminal pH was unaffected by diet, but milk fat content was lower for cows fed SF. Milk yield, DMI, and feed to milk ratio were not affected by supplementation with PA. Although cows fed PA had increased P intake compared with cows fed low P diet, increased P excretion resulted in no differences in apparent P digestibility. Phosphorus balance tended to be higher in cows fed PA, but milk P as a percentage of intake was reduced. The interaction of starch source and PA affected ruminal phytase activity. Altering starch source to improve efficiency of milk yield in lactating dairy cows may help reduce P losses from dairy farms.