Imputation of missing milk Fourier transform mid-infrared spectra using existing milk spectral databases: A strategy to improve the reliability of breeding values and predictive models

The use of milk Fourier transform mid-infrared (FT-MIR) spectrometry to develop management and breeding tools for dairy farmers and industry is growing and supported by the availability of numerous new predicted phenotypes to assess the nutritional quality of milk and its technological properties, but also the animal health and welfare status and its environmental fingerprint. For genetic evaluations, having a long-term and representative spectral dairy herd improvement (DHI) database improves the reliabilities of estimated breeding values (EBV) from these phenotypes. Unfortunately, most of the time, the raw spectral data used to generate these estimations are not stored. Moreover, many reference measurements of those phenotypes, needed during the FT-MIR calibration step, are available from past research activities but lack spectra records. So, it is impossible to use them to improve the FT-MIR models. Consequently, there is a strong interest in imputing those missing spectra. The innovative objective of this study was to use the existing large spectral DHI database to estimate missing spectra by selecting probable spectra using, as the match criteria, common dairy traits recorded for a long time by DHI organizations. We tested 4 match criteria combinations. Combination 1 required to have equal fat and protein contents between the sample for which a spectrum was to be estimated and the reference samples in the DHI database. Combination 2 also required an equal urea content. Combination 3 requested equal fat, protein, and lactose contents. Finally, combination 4 included all criteria. When more than one spectrum was found during the search, their average was the estimated spectrum for the query sample. Concretely, this study estimated missing spectra for 1,700 samples using 2,000,000 spectral DHI records. For assessing the effect of this spectral estimation on the prediction quality, FT-MIR equations were used to predict 11 phenotypes, selected as their quantification used different FT-MIR regions. They were related to the milk fat and mineral composition, lactoferrin content, quantity of eructed methane, body weight (BW), and dry matter intake. The accuracy between predictions obtained from actual and estimated spectra was evaluated by calculating the mean absolute error (MAE). The criteria in the fourth and second combinations were too strict to estimate a spectrum for most samples. Indeed, for many samples, no spectra with the same values for those matching criteria was found. The third match criteria combination had a poorer prediction performance for all studied traits and spectral absorptions than the first combination due to fewer matched samples available to compute the missing spectrum. By allowing a range for matching lactose content (±0.1 g/dL milk), we showed that this new combination increased the number of selected samples to compute missing spectra and predict better the infrared absorption at different wavenumbers, especially those related to the lactose quantification. The prediction performance was further improved by performing queries on the entire Walloon DHI spectral database (6,625,570 spectra), and it varied among the studied phenotypes. Without considering the traits used for the matching, the best predictions were obtained for the content of saturated fatty acids (MAE = 0.15 g/dL milk) and BW (MAE = 12.80 kg). Yet, the predictions for the unsaturated fatty acids were less accurate (MAE = 0.13 and 0.018 g/dL milk for monounsaturated and polyunsaturated fatty acids), likely because of the poorer predictions of spectral regions related to long-chain fatty acids. Similarly, poorer predictions were observed for the amount of methane eructed by dairy cows (MAE = 47.02 g/d), likely because it is not directly related to fat content or composition. Prediction accuracies for the remaining traits were also low. In conclusion, we observed that increasing the number of relevant matching criteria helps improve the quality of FT-MIR predicted phenotypes and the number of spectra used during the search. So, it would be of great interest to test in the future the suitability of the developed methodology with large-scale international spectral databases to improve the reliability of EBV from these FT-MIR-based phenotypes and the robustness of FT-MIR predictive models.

Researching 100 t cows: An innovative approach to identify intrinsic cows factors associated with a high lifetime milk production

Longevity is an important trait both from an economic and social perspective. Modern dairy cows are criticized for their short productive lifespan: only a minority of animals survives to a fourth lactation, implying that most cows are culled before reaching their maximal potential. In contrast, the population of 100 t cows (HT), reaching the threshold of 100,000 kg lifetime milk yield, is growing rapidly. As these cows combine a long lifespan with high functionality, a better understanding of their intrinsic characteristics might help us to improve the overall lifespan and lifetime production in dairy cows. The aim of the present research was to compare HT with their less-producing herd mates in order to identify intrinsic cow factors associated with longevity and high lifetime production. Therefore, we matched 26,248 HT with 691,597 herd mates, born in the same year in the same herd. Data were provided by Coöperatie rundveeverbetering (CRV) and contained birth dates, calving dates, milk yield and dam information. In addition, scores for conformation traits based on classifications in the first lactation and breeding values (for milk yield, fertility, udder health and claw health) were provided. Multivariable conditional logistic regression models were built to identify factors associated with reaching a lifetime milk yield of 100,000 kg. Results revealed cows born in September and born out of heifers to have the highest odds to become a HT. When cows received a score ≥ 83 (population average 80) for udder and feet & legs conformation, they had higher odds of reaching the 100,000 kg threshold. While a greater body condition and larger rump angle increased the odds of becoming a HT, this was decreased in cows with a large body depth. Finally, breeding values for milk yield, fertility, udder health and claw health were positively associated with the likelihood of reaching a lifetime milk yield of 100,000 kg. In conclusion, to increase lifetime milk yield in dairy herds, farmers should select heifers with high scores for conformation traits like udder and feet & legs and high breeding values for milk yield, fertility and udder health. Furthermore, our data suggest that being born in September out of a heifer potentially contributes to reaching a high lifetime milk yield.

Genetic analysis of resilience indicators based on milk yield records in different lactations and at different lactation stages

Resilience is the ability of cows to cope with disturbances, such as pathogens or heat waves. To breed for improved resilience, it is important to know whether resilience genetically changes throughout life. Therefore, the aim was to perform a genetic analysis on 2 resilience indicators based on data from 3 periods of the first lactation (d 11-110, 111-210, and 211-340) and the first 3 full lactations, and to estimate genetic correlations with health traits. The resilience indicators were the natural log-transformed variance (LnVar) and lag-1 autocorrelation (r_auto) of daily deviations in milk yield from an expected lactation curve. Low LnVar and r_auto indicate low variability in daily milk yield and quick recovery, and were expected to indicate good resilience. Data of 200,084 first, 155,784 second, and 89,990 third lactations were used. Heritabilities were similar based on different lactation periods (0.12-0.15 for LnVar, 0.05-0.06 for r_auto). However, the heritabilities of the resilience indicators based on full first lactation were higher than those based on lactation periods (0.20 for LnVar, 0.08 for r_auto), due to lower residual variances. Heritabilities decreased from 0.20 in full lactation 1 to 0.19 in full lactation 3 for LnVar and from 0.08 to 0.06 for r_auto. For LnVar, as well as for r_auto, the strongest genetic correlation between lactation periods was between period 2 and 3 (0.97 for LnVar, 0.96 for r_auto) and the weakest between period 1 and 3 (0.81 for LnVar, 0.65 for r_auto). Similarly, for both traits the genetic correlation between full lactations was strongest between lactations 2 and 3 (0.99 for LnVar, 0.95 for r_auto) and weakest between lactations 1 and 3 (0.91 for LnVar, 0.71 for r_auto). For LnVar, genetic correlations with resilience-related traits, such as udder health, ketosis, and longevity, adjusted for correlations with milk yield, were almost always favorable (-0.59 to 0.02). In most cases these genetic correlations were stronger based on full lactations than on lactation periods. Genetic correlations were similar across full lactations, but the correlation with udder health increased substantially from -0.31 in lactation 1 to -0.51 in lactation 3. For r_auto, genetic correlations with resilience-related traits were always favorable in lactation period 1 and in most full lactations, but not in the other lactation periods. However, correlations were weak (-0.27 to 0.15). Therefore, as a resilience indicator for breeding, LnVar is preferred over r_auto. A multitrait index based on estimated breeding values for LnVar in lactations 1, 2, and 3 is recommended to improve resilience throughout the lifetime of a cow.

A large-scale study on the effect of age at first calving, dam parity, and birth and calving month on first-lactation milk yield in Holstein Friesian dairy cattle

Milk yield during first lactation is an important economical trait. Age at first calving (AFC) is considered an important predictor of subsequent milk yield. In addition, both season of birth, as well as season of calving, have been shown to influence milk production, with conflicting results. Finally, higher parity of the dam has been associated with a lower performance of the offspring. The aim of the present study was to assess the effect of the above-mentioned factors based on a large-scale study and to rank the most important determinants for first-lactation milk yield. Data on 3,810,678 Holstein Friesian heifers, born in Belgium and the Netherlands between 2000 and 2015, were provided by Cooperative CRV and CRV BV (Arnhem, the Netherlands) and consisted of birth dates, calving dates, and first-lactation productions. In addition, herd, sire, and dam information was provided. Linear regression models were built with herd-calving year and sire as random effects and 305-d energy-corrected milk (ECM) yield during first lactation as outcome variable. Birth month, calving month, parity of the dam, and AFC were included as fixed effects in the model and a dominance analysis was performed to rank the associated factors according to importance. Results revealed AFC to be the most important factor (R² = 0.047), with an increase in ECM up to an age of 33 mo. Calving month was a more important predictor than birth month (R² = 0.010 vs. R² = 0.002, respectively), with the highest first-lactation production in heifers calving in October to December, and the lowest in heifers calving in June and July. Birth month had a limited effect on first-lactation milk yield (R² = 0.002), potentially masked by rearing strategies during early life. Finally, parity of the dam ≥3 was associated with a reduced ECM of the offspring (R² = 0.002). In conclusion, our results show AFC to be an important determinant of milk yield during first lactation. In addition, seasonal patterns in milk production are seen, which should be further explored to identify the underlying mechanism.

Predicting survival in dairy cattle by combining genomic breeding values and phenotypic information

Advances in technology and improved data collection have increased the availability of genomic estimated breeding values (gEBV) and phenotypic information on dairy farms. This information could be used for the prediction of complex traits such as survival, which can in turn be used in replacement heifer management. In this study, we investigated which gEBV and phenotypic variables are of use in the prediction of survival. Survival was defined as survival to second lactation, plus 2 wk, a binary trait. A data set was obtained of 6,847 heifers that were all genotyped at birth. Each heifer had 50 gEBV and up to 62 phenotypic variables that became gradually available over time. Stepwise variable selection on 70% of the data was used to create multiple regression models to predict survival with data available at 5 decision moments: distinct points in the life of a heifer at which new phenotypic information becomes available. The remaining 30% of the data were kept apart to investigate predictive performance of the models on independent data. A combination of gEBV and phenotypic variables always resulted in the model with the highest Akaike information criterion value. The gEBV selected were longevity, feet and leg score, exterior score, udder score, and udder health score. Phenotypic variables on fertility, age at first calving, and milk quantity were important once available. It was impossible to predict individual survival accurately, but the mean predicted probability of survival of the surviving heifers was always higher than the mean predicted probability of the nonsurviving group (difference ranged from 0.014 to 0.028). The model obtained 2.0 to 3.0% more surviving heifers when the highest scoring 50% of heifers were selected compared with randomly selected heifers. Combining phenotypic information and gEBV always resulted in the highest scoring models for the prediction of survival, and especially improved early predictive performance. By selecting the heifers with the highest predicted probability of survival, increased survival could be realized at the population level in practice.

Comparing regression, naive Bayes, and random forest methods in the prediction of individual survival to second lactation in Holstein cattle

In this study, we compared multiple logistic regression, a linear method, to naive Bayes and random forest, 2 nonlinear machine-learning methods. We used all 3 methods to predict individual survival to second lactation in dairy heifers. The data set used for prediction contained 6,847 heifers born between January 2012 and June 2013, and had known survival outcomes. Each animal had 50 genomic estimated breeding values available at birth and up to 65 phenotypic variables that accumulated over time. Survival was predicted at 5 moments in life: at birth, at 18 mo, at first calving, at 6 wk after first calving, and at 200 d after first calving. The data sets were randomly split into 70% training and 30% testing sets to evaluate model performance for 20-fold validation. The methods were compared for accuracy, sensitivity, specificity, area under the curve (AUC) value, contrasts between groups for the prediction outcomes, and increase in surviving animals in a practical scenario. At birth and 18 mo, all methods had overlapping performance; no method significantly outperformed the other. At first calving, 6 wk after first calving, and 200 d after first calving, random forest and naive Bayes had overlapping performance, and both machine-learning methods outperformed multiple logistic regression. Overall, naive Bayes has the highest average AUC at all decision points up to 200 d after first calving. Random forest had the highest AUC at 200 d after first calving. All methods obtained similar increases in survival in the practical scenario. Despite this, the methods appeared to predict the survival of individual heifers differently. All methods improved over time, but the changes in mean model outcomes for surviving and non-surviving animals differed by method. Furthermore, the correlations of individual predictions between methods ranged from r = 0.417 to r = 0.700; the lowest correlations were at first calving for all methods. In short, all 3 methods were able to predict survival at a population level, because all methods improved survival in a practical scenario. However, depending on the method used, predictions for individual animals were quite different between methods.