Use of genotype × environment interaction model to accommodate genetic heterogeneity for residual feed intake, dry matter intake, net energy in milk, and metabolic body weight in dairy cattle

Heterogeneity of variance and genetic parameters for milk production in cattle, using Bayesian inference

The goal of this study was to verify the effect of heterogeneity of variance (HV) on milk production in up to 305 days of lactation (L305) of daughters of Girolando, Gir and Holstein sires, as well as in the genetic evaluation of these sires and their progenies. in Brazil. The model included contemporary groups (consisting of herd, year and calving season) as a fixed effect, cow age at calving (linear and quadratic effects) and heterozygosity (linear effect) as covariates, in addition to the random effects of direct additive genetic and environmental, permanent and residual. The first analysis consisted of the single-trait animal model, with L305 records (disregarding HV). The second considered classes of standard deviations (SD): two-trait model including low and high classes (considering HV), according to the standardized means of L305 for herd-year of calving. The low SD class was composed of herds with SD equal to or less than zero and the high class with positive SD values. Estimates of (co)variance components and breeding values were obtained separately for each scenario using Bayesian inference via Gibbs sampling. Different heritability was estimated. Higher for the high DP class in the Gir (0.20) and Holstein (0.15) breeds, not occurring the same in the Girolando breed, with a lower value among the classes for the high DP (0.10). High values of genetic correlations were also found between low and high SD classes (0.88; 0.85 and 0.79) for the Girolando, Gir and Holstein breeds, respectively. Like the order correlations (Spearman) which were also high for the three breeds analyzed (equal to or above 0.92). Thus, the presence of HV had a smaller impact for L305 and did not affect the genetic evaluation of sires.

Fitting Genomic Prediction Models with Different Marker Effects among Prefectures to Carcass Traits in Japanese Black Cattle

We fitted statistical models, which assumed single-nucleotide polymorphism (SNP) marker effects differing across the fattened steers marketed into different prefectures, to the records for cold carcass weight (CW) and marbling score (MS) of 1036, 733, and 279 Japanese Black fattened steers marketed into Tottori, Hiroshima, and Hyogo prefectures in Japan, respectively. Genotype data on 33,059 SNPs was used. Five models that assume only common SNP effects to all the steers (model 1), common effects plus SNP effects differing between the steers marketed into Hyogo prefecture and others (model 2), only the SNP effects differing between Hyogo steers and others (model 3), common effects plus SNP effects specific to each prefecture (model 4), and only the effects specific to each prefecture (model 5) were exploited. For both traits, slightly lower values of residual variance than that of model 1 were estimated when fitting all other models. Estimated genetic correlation among the prefectures in models 2 and 4 ranged to 0.53 to 0.71, all <0.8. These results might support that the SNP effects differ among the prefectures to some degree, although we discussed the necessity of careful consideration to interpret the current results.

Identification of candidate genes on the basis of SNP by time-lagged heat stress interactions for milk production traits in German Holstein cattle

The aim of this study was to estimate genotype by time-lagged heat stress (HS) variance components as well as main and interaction SNP-marker effects for maternal HS during the last eight weeks of cow pregnancy, considering milk production traits recorded in the offspring generation. The HS indicator was the temperature humidity index (THI) for each week. A dummy variable with the code = 1 for the respective week for THI ≥ 60 indicated HS, otherwise, for no HS, the code = 0 was assigned. The dataset included test-day and lactation production traits from 14,188 genotyped first parity Holstein cows. After genotype quality control, 41,139 SNP markers remained for the genomic analyses. Genomic animal models without (model VC_nHS) and with in-utero HS effects (model VC_wHS) were applied to estimate variance components. Accordingly, for genome-wide associations, models GWA_nHS and GWA_wHS, respectively, were applied to estimate main and interaction SNP effects. Common genomic and residual variances for the same traits were very similar from models VC_nHS and VC_wHS. Genotype by HS interaction variances varied, depending on the week with in-utero HS. Among all traits, lactation milk yield with HS from week 5 displayed the largest proportion for interaction variances (0.07). For main effects from model GWA_wHS, 380 SNPs were suggestively associated with all production traits. For the SNP interaction effects from model GWA_wHS, we identified 31 suggestive SNPs, which were located in close distance to 62 potential candidate genes. The inferred candidate genes have various biological functions, including mechanisms of immune response, growth processes and disease resistance. Two biological processes excessively represented in the overrepresentation tests addressed lymphocyte and monocyte chemotaxis, ultimately affecting immune response. The modelling approach considering time-lagged genotype by HS interactions for production traits inferred physiological mechanisms being associated with health and immunity, enabling improvements in selection of robust animals.

Climate sensitivity of milk production traits and milk fatty acids in genotyped Holstein dairy cows

The aim of this study was the evaluation of climate sensitivity via genomic reaction norm models [i.e., to infer cow milk production and milk fatty acid (FA) responses on temperature-humidity index (THI) alterations]. Test-day milk traits were recorded between 2010 and 2016 from 5,257 first-lactation genotyped Holstein dairy cows. The cows were kept in 16 large-scale cooperator herds, being daughters of 344 genotyped sires. The longitudinal data consisted of 47,789 test-day records for the production traits milk yield (MY), fat yield (FY), and protein yield (PY), and of 20,742 test-day records for 6 FA including C16:0, C18:0, saturated fatty acids (SFA), unsaturated fatty acids (UFA), monounsaturated fatty acids (MUFA), and polyunsaturated fatty acids (PUFA). After quality control of the genotypic data, 41,057 SNP markers remained for genomic analyses. Meteorological data from the weather station in closest herd distance were used for the calculation of maximum hourly daily THI. Genomic reaction norm models were applied to estimate genetic parameters in a single-step approach for production traits and FA in dependency of THI at different lactation stages, and to evaluate the model stability. In a first evaluation strategy (New_sire), all phenotypic records from daughters of genotyped sires born after 2010 were masked, to mimic a validation population. In the second strategy (New_env), only daughter records of the new sires recorded in the most extreme THI classes were masked, aiming at predicting sire genomic estimated breeding values (GEBV) under heat stress conditions. Model stability was the correlation between GEBV of the new sires in the reduced data set with respective GEBV estimated from all phenotypic data. Among all test-day production traits, PY responded as the most sensitive to heat stress. As observed for the remaining production traits, genetic variances were quite stable across THI, but genetic correlations between PY from temperate climates with PY from extreme THI classes dropped to 0.68. Genetic variances in dependency of THI were very similar for C16:0 and SFA, indicating marginal climatic sensitivity. In the early lactation stage, genetic variances for C18:0, MUFA, PUFA, and UFA were significantly larger in the extreme THI classes compared with the estimates under thermoneutral conditions. For C18:0 and MUFA, PUFA, and UFA in the middle THI classes, genetic correlations in same traits from the early and the later lactation stages were lower than 0.50, indicating strong days in milk influence. Interestingly, within lactation stages, genetic correlations for C18:0 and UFA recorded at low and high THI were quite large, indicating similar genetic mechanisms under stress conditions. The model stability was improved when applying the New_env instead of New_sire strategy, especially for FA in the first stage of lactation. Results indicate moderately accurate genomic predictions for milk traits in extreme THI classes when considering phenotypic data from a broad range of remaining THI. Phenotypically, thermal stress conditions contributed to an increase of UFA, suggesting value as a heat stress biomarker. Furthermore, the quite large genetic variances for UFA at high THI suggest the consideration of UFA in selection strategies for improved heat stress resistance.

Spatial modelling improves genetic evaluation in smallholder breeding programs

BACKGROUND

Breeders and geneticists use statistical models to separate genetic and environmental effects on phenotype. A common way to separate these effects is to model a descriptor of an environment, a contemporary group or herd, and account for genetic relationship between animals across environments. However, separating the genetic and environmental effects in smallholder systems is challenging due to small herd sizes and weak genetic connectedness across herds. We hypothesised that accounting for spatial relationships between nearby herds can improve genetic evaluation in smallholder systems. Furthermore, geographically referenced environmental covariates are increasingly available and could model underlying sources of spatial relationships. The objective of this study was therefore, to evaluate the potential of spatial modelling to improve genetic evaluation in dairy cattle smallholder systems.

METHODS

We performed simulations and real dairy cattle data analysis to test our hypothesis. We modelled environmental variation by estimating herd and spatial effects. Herd effects were considered independent, whereas spatial effects had distance-based covariance between herds. We compared these models using pedigree or genomic data.

RESULTS

The results show that in smallholder systems (i) standard models do not separate genetic and environmental effects accurately, (ii) spatial modelling increases the accuracy of genetic evaluation for phenotyped and non-phenotyped animals, (iii) environmental covariates do not substantially improve the accuracy of genetic evaluation beyond simple distance-based relationships between herds, (iv) the benefit of spatial modelling was largest when separating the genetic and environmental effects was challenging, and (v) spatial modelling was beneficial when using either pedigree or genomic data.

CONCLUSIONS

We have demonstrated the potential of spatial modelling to improve genetic evaluation in smallholder systems. This improvement is driven by establishing environmental connectedness between herds, which enhances separation of genetic and environmental effects. We suggest routine spatial modelling in genetic evaluations, particularly for smallholder systems. Spatial modelling could also have a major impact in studies of human and wild populations.

On Hadamard and Kronecker products in covariance structures for genotype × environment interaction

When including genotype × environment interactions (G × E) in genomic prediction models, Hadamard or Kronecker products have been used to model the covariance structure of interactions. The relation between these two types of modeling has not been made clear in genomic prediction literature. Here, we demonstrate that a certain model based on a Hadamard formulation and another using the Kronecker product lead to exactly the same statistical model. Moreover, we illustrate how a multiplication of entries of covariance matrices is related to modeling locus × environmental-variable interactions explicitly. Finally, we use a wheat and a maize data set to illustrate that the environmental covariance E can be specified easily, also if no information on environmental variables - such as temperature or precipitation - is available. Given that lines have been tested in different environments, the corresponding environmental covariance can simply be estimated from the training set as phenotypic covariance between environments. To achieve a high level of increase in predictive ability, the environmental covariance has to be defined appropriately and records on the performance of the lines of the test set under different environmental conditions have to be included in the training set.

Enhancing genomic selection by fitting large-effect SNPs as fixed effects and a genotype-by-environment effect using a maize BC1F3:4 population

The popularity of genomic selection (GS) has increased owing to its prospects in commercial breeding. It is necessary to enhance GS to increase its efficiency. In this study, a maize BC₁F_3:4 population, consisting of 481 families, was evaluated for days to anthesis in four environments, and genotyped with DNA chips including 55,000 single nucleotide polymorphisms (SNPs). This population was used to investigate whether GS could be enhanced by borrowing information from the genetic basis and genotype-by-environment (G × E) interaction. The results showed that: 1) fitting the top four large-effect SNPs as fixed effects could increase prediction accuracy, including three minor-effect SNPs explaining less than 10% phenotypic variance; 2) the increase of prediction accuracy when fitting large-effect SNPs as fixed effects was related to the decrease of genetic variance; 3) generally, the GS model fitting large-effect SNPs as fixed effects and G × E component enhanced GS. Therefore, we propose fitting large-effect markers as fixed effects and G × E effect for crop breeding projects in order to obtain accurately predicted phenotypic data and conduct efficient selection of desired plants.

Inclusion of herdmate data improves genomic prediction for milk-production and feed-efficiency traits within North American dairy herds

Genomic data are widely available in the dairy industry and provide a cost-effective means of predicting genetic merit to inform selection decisions and increase genetic gains. As more dairy farms adopt genomic selection practices, dairy producers will soon have genomic data available on all of the animals within their herds. This is a very rich, but currently underused, source of information. Herdmates provide an excellent indication of how a selection candidate's genetics will perform within a given herd, noting that herdmates often include close relatives that share a similar environment. The study objective was to evaluate the utility of incorporating herdmate data into genomic predictions in a data set composed of 3,303 Holsteins from one herd in Canada and 6 herds throughout the United States. Within-herd prediction accuracy was assessed for milk-production and feed-efficiency traits determined from genomic best linear unbiased prediction under 4 different scenarios. Scenario 1 did not include herdmates in the training population. Scenarios 2 through 4 included herdmates in the training population, and scenarios 3 and 4 also included modeling of herd-specific marker effects. Leave-one-out cross validation was used to maximize the number of herdmates in the training population in scenarios 2 through 4, while maintaining constant training population size with scenario 1. Results from the present study reveal the importance of incorporating herdmate data into genomic evaluations. Inclusion of herdmates in the training population improved mean within-herd prediction accuracy for milk-production traits (± standard error) by 0.08 ± 0.03 (milk yield), 0.07 ± 0.03 (fat percentage), and 0.05 ± 0.01 (protein percentage) and feed-efficiency traits by 0.07 ± 0.02 (milk energy), 0.03 ± 0.02 (DMI), and 0.08 ± 0.01 (metabolic body weight). Modeling herd-specific marker effects further improved mean within-herd prediction accuracy for milk yield and energy by 0.03 ± 0.01 and 0.02 ± 0.01, respectively. Herds with higher within-herd heritability and low genomic correlation with the remaining herds benefitted most from the inclusion of herdmate data.

Lactation curves and model evaluation for feed intake and energy balance in dairy cows

A good health status of high-performing dairy cows is essential for successful production. Feed intake affects the metabolic stability of dairy cows and can be used as a measurement for energy balance. By implementing feed intake and energy balance into the breeding goal, these traits provide great potential for an improvement in the health of dairy cows by breeders. In this study, fixed and random regression models were tested to establish appropriate models for a further analysis of this approach. A total of 1,374 Holstein-Friesian cows and 327 Simmental cows (SI) from 12 German research farms participating in a collaboration called optiKuh were phenotyped. Feed intake data recording was standardized across farms, and energy balance was calculated using phenotypic information on milk yield, milk ingredients, live weight, gestation stage, and feed intake. The phenotypic data set consisted of a total of 40,012 Holstein-Friesian and 16,996 SI with average weekly dry matter intakes of 21.8 ± 4.3 and 20.2 ± 3.6 kg/d, respectively. Observations of days in milk 1 to 350 were used to evaluate the best-fitting models and to estimate the repeatability and correlations between cow effects at different stages for feed intake and energy balance. Four parametric functions (Ali and Schaeffer and Legendre polynomials of second, third, and fourth degree) were compared to model the lactation curves. Based on the corrected Akaike information criterion and the Bayesian information criterion, the goodness of fit was evaluated to choose the best-fitting model for the finest description of lactation curves for the traits energy balance and feed intake. Legendre polynomial fourth degree was the best-fitting model for random regression models. In contrast, Ali and Schaeffer was the best choice for fixed regression models. Feed intake and energy balance acted as expected: the feed intake increased slowly at the beginning of lactation and the negative energy balance switched to a positive range around 40 to 80 d of lactation. The repeatabilities of both traits were quite similar and the repeatabilities for SI were the highest for both traits. Additionally, correlations between cow effects were closest between early days in milk. These results emphasize the possibility that the unique optiKuh data set can be used for further genetic analyses to enable genomic selection for feed intake or energy balance.

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.