Ruminal microbiome changes across lactation in primiparous Holstein cows with varying methane intensity: Heritability assessment.

Methane is a potent greenhouse gas produced during the ruminal fermentation and is associated with a loss of feed energy. Therefore, efforts to reduce methane emissions have been ongoing in the last decades. Methane production is highly influenced by factors such as the ruminal microbiome and host genetics. Previous studies have proposed to use the ruminal microbiome to reduce long-term methane emissions, as ruminal microbiome composition is a moderately heritable trait and genetic improvement accumulates over time. Lactation stage is another important factor that might influence methane production but potential associations with the ruminal microbiome have not been evaluated previously. This study sought to examine the changes in ruminal microbiome over the lactation period of primiparous Holstein cows differing in methane intensity and estimate the heritability of the abundance of relevant microorganisms. Ruminal content samples from 349 primiparous Holstein cows with 14 - 378 d in milk were collected from May 2018 to June 2019. Methane intensity (MI) of each cow was calculated as methane concentration/milk yield. Up to 64 taxonomic features (TF) from 20 phyla had a significant differential abundance between cows with low and high MI early in lactation, 16 TF during mid lactation, and none late in lactation. Taxonomical features within the Firmicutes, Proteobacteria, Melainabacteria, Cyanobacteria, Bacteroidetes and Actinobacteria phyla were associated to low MI, whereas eukaryotic TF and those within the Euryarchaeota, Verrucomicrobia, Kiritimatiellaeota, Lentisphaerae phyla were associated to high MI. Out of the 60 TF that were found to be differentially abundant between early and late lactation in cows with low MI, 56 TF were also significant when cows with low and high MI were compared in the first third of the lactation. In general, microbes associated with low MI were more abundant early in lactation (e.g., Acidaminococcus, Aeromonas and Weimeria genera) and showed low to moderate heritabilities (0.03 to 0.33). These results suggest some potential to modulate the rumen microbiome composition through selective breeding for lower MI. Differences in the ruminal microbiome of cows with extreme MI levels likely result from variations in the ruminal physiology of these cows and were more noticeable early in lactation probably due to important interactions between the host phenotype and environmental factors associated to that period. Our results suggest that the ruminal microbiome evaluated early in lactation may be more precise for MI difference, and hence, this should be considered to optimize sampling periods to establish a reference population in genomic selection scenarios.

The oral microbiome as a proxy for feed intake in dairy cattle.

Genetic material from rumen microorganisms can be found within the oral cavity, and hence there is potential in using the oral microbiome as a proxy of the ruminal microbiome. Feed intake (FI) influences the composition of rumen microbiota and might directly influence the oral microbiome in dairy cattle. Ruminal content samples (RS) from 29 cows were collected at the beginning of the study and also 42 d later (RS0 and RS42, respectively). Additionally, 18 oral samples were collected through buccal swabbing at d 42 (OS42) from randomly selected cows. Samples were used to characterize and compare the taxonomy and functionality of the oral microbiome using Nanopore sequencing and to evaluate the feasibility of using the oral microbiome to estimate FI. Up to 186 taxonomical features were found differentially abundant (DA) between RS and OS42. Similar results were observed when comparing OS42 to RS collected at different days. Microorganisms associated with the liquid fraction of the rumen were less abundant in OS42 as these were probably swallowed after regurgitation. Up to 1,102 KEGGs were found to be DA between RS and OS42 and these results differed when comparing time of collection, but differentially abundant KEGGs were mainly associated to metabolism in both situations. Models based on the oral microbiome and rumen microbiome differed in their selection of microbial groups and biological pathways to predict FI. In the rumen, fiber-associated microorganisms are considered suitable indicators of feed intake. On the other hand, biofilm formers like Gammaproteobacteria or Bacteroidia classes are deemed appropriate proxies for predicting feed intake from oral samples. Models from RS exhibited some predictive ability to estimate FI, but OS significantly outperformed them. The best lineal model to estimate FI was obtained with the relative abundance of taxonomical feature at genera level, achieving an average R2 equal to 0.88 within the training data, and a root mean square error equal to 3.46 ± 0.83 (standard deviation; SD) kg of DM/ as well as a Pearson correlation coefficient between observed and estimated FI of 0.48 ± 0.30 in the test data. The results from this study suggest that oral microbiome has potential to predict FI in dairy cattle, and it encourages validating this potential in other populations.

Estimation of genetic parameters for feed efficiency traits using random regression models in dairy cattle.

Feed efficiency has become an increasingly important research topic in recent years. As feed costs rise and the environmental impacts of agriculture become more apparent, improving the efficiency with which dairy cows convert feed to milk is increasingly important. However, feed intake is expensive to measure accurately on large populations, making the inclusion of this trait in breeding programs difficult. Understanding how the genetic parameters of feed efficiency and traits related to feed efficiency vary throughout the lactation period is valuable to gain understanding into the genetic nature of feed efficiency. This study used 121,226 dry matter intake (DMI) records, 120,500 energy-corrected milk (ECM) records, and 98,975 metabolic body weight (MBW) records, collected on 7,440 first-lactation Holstein cows from 6 countries (Canada, Denmark, Germany, Spain, Switzerland, and the United States), from January 2003 to February 2022. Genetic parameters were estimated using a multiple-trait random regression model with a fourth-order Legendre polynomial for all traits. Weekly phenotypes for DMI were re-parameterized using linear regressions of DMI on ECM and MBW, creating a measure of feed efficiency that was genetically corrected for ECM and MBW, referred to as genomic residual feed intake (gRFI). Heritability (SE) estimates varied from 0.15 (0.03) to 0.29 (0.02) for DMI, 0.24 (0.01) to 0.29 (0.03) for ECM, 0.55 (0.03) to 0.83 (0.05) for MBW, and 0.12 (0.03) to 0.22 (0.06) for gRFI. In general, heritability estimates were lower in the first stage of lactation compared with the later stages of lactation. Additive genetic correlations between weeks of lactation varied, with stronger correlations between weeks of lactation that were close together. The results of this study contribute to a better understanding of the change in genetic parameters across the first lactation, providing insight into potential selection strategies to include feed efficiency in breeding programs.

Invited review: Recursive models in animal breeding: Interpretation, limitations, and extensions.

Structural equation models allow causal effects between 2 or more variables to be considered and can postulate unidirectional (recursive models; RM) or bidirectional (simultaneous models) causality between variables. This review evaluated the properties of RM in animal breeding and how to interpret the genetic parameters and the corresponding estimated breeding values. In many cases, RM and mixed multitrait models (MTM) are statistically equivalent, although subject to the assumption of variance-covariance matrices and restrictions imposed for achieving model identification. Inference under RM requires imposing some restrictions on the (co)variance matrix or on the location parameters. The estimates of the variance components and the breeding values can be transformed from RM to MTM, although the biological interpretation differs. In the MTM, the breeding values predict the full influence of the additive genetic effects on the traits and should be used for breeding purposes. In contrast, the RM breeding values express the additive genetic effect while holding the causal traits constant. The differences between the additive genetic effect in RM and MTM can be used to identify the genomic regions that affect the additive genetic variation of traits directly or causally mediated for another trait or traits. Furthermore, we presented some extensions of the RM that are useful for modeling quantitative traits with alternative assumptions. The equivalence of RM and MTM can be used to infer causal effects on sequentially expressed traits by manipulating the residual (co)variance matrix under the MTM. Further, RM can be implemented to analyze causality between traits that might differ among subgroups or within the parametric space of the independent traits. In addition, RM can be expanded to create models that introduce some degree of regularization in the recursive structure that aims to estimate a large number of recursive parameters. Finally, RM can be used in some cases for operational reasons, although there is no causality between traits.

Random regression models to estimate genetic parameters for milk yield, fat, and protein contents in Tunisian Holsteins.

This study aimed to find the parsimonious random regression model (RRM) to evaluate the genetic potential for milk yield (MY), fat content (FC), and protein content (PC) in Tunisian Holstein cows. For this purpose, 551,139; 331,654; and 302,396 test day records for MY, FC, and PC were analysed using various RRMs with different Legendre polynomials (LP) orders on additive genetic (AG) and permanent environmental (PE) effects, and different types of residual variances (RV). The statistical analysis was performed in a Bayesian framework with Gibbs sampling, and the model performances were assessed, mainly, on the predictive ability criteria. The study found that the optimal model for evaluating these traits was an RRM with a third LP order and nine classes of heterogeneous RV. In addition, the study found that heritability estimates for MY, FC, and PC ranged from 0.11 to 0.22, 0.11 to 0.17, and 0.12 to 0.18, respectively, indicating that genetic improvement should be accompanied by improvements in the production environment. The study also suggested that new selection rules could be used to modify lactation curves by exploiting the canonical transformation of the random coefficient covariance (RC) matrix or by using the combination of slopes of individual lactation curves and expected daily breeding values.

Mitigation of greenhouse gases in dairy cattle via genetic selection: 2. Incorporating methane emissions into the breeding goal.

The objective of this study was to analyze the impact of incorporating enteric methane into the breeding objective of dairy cattle in Spain, and to evaluate both genetic and economic response of traits in the selection index under 4 scenarios: (1) the current ICO (Spanish total merit index), used as benchmark; (2) a hypothetical penalization of methane emissions through a carbon tax; (3) considering methane as a net energy loss for the animal; and (4) desired genetic response to reduce methane production by 20% in 10 yr. A bio-economic model was developed to derive the economic values for production and methane traits in each scenario. The estimated economic values for methane were estimated at -€1.21/kg and -€0.32/kg for scenarios 2 and 3, respectively. When merged with other traits in the selection index, methane had less economic importance (1-5%) than milk protein yield (39-42%) or milk fat yield (27-28%). Under these scenarios, selection resulted in an unfavorable response in methane emissions when it was included with an economic weight, with an increase in methane estimated from 0.52 to 0.60 kg/cow per year. Small differences in total profit per cow per year were observed between indices. The incorporation of methane production into the breeding objective had a negligible effect on production, with minor reductions in the expected genetic gain for fat and protein yields and in total economic benefits. However, total methane emissions in the dairy industry in Spain were estimated to decrease between 2 and 5% in the next 10 yr due to positive genetic trends for milk yield and an expected decrease in the total number of dairy cows. Additionally, methane intensity per 1 billion liters of milk would decrease in all scenarios. The uncertainty in the genetic parameters of methane and in carbon prices were tested in a sensitivity analysis, resulting in small deviations from the benchmark scenario. A major effect was observed only under the desired genetic response scenario. In this case, it was possible to achieve a 20% reduction of methane production in 10 yr via selective breeding but at the expense of a larger ad hoc weight (33%) of methane in the selection index and decelerating the genetic gain for production traits from 6 to 18%. This study shows the potential of including environmental traits in the selection indices while retaining populations profitable for producers.

Mitigation of greenhouse gases in dairy cattle via genetic selection: 1. Genetic parameters of direct methane using noninvasive methods and proxies of methane.

Records of methane emissions from 1,501 cows on 14 commercial farms in 4 regions of Spain were collected from May 2018 to June 2019. Methane concentrations (MeC) were measured using a nondispersive infrared methane detector installed within the feed bin of the automatic milking system during 14- to 21-d periods. Rumination time (RT; min/d) was collected using collars with a tag that registered time (minutes) spent eating and ruminating. The means of MeC and methane production (MeP) were 1,254.28 ppm and 182.49 g/d, respectively; mean RT was 473.38 min/d. Variance components for MeC, MeP, and RT were estimated with REML using pedigree and genomic information in a single-step model. Heritabilities for MeC and MeP were 0.11 and 0.12, respectively. Rumination time showed a slightly larger heritability estimate (0.17). The genetic correlation between MeP and MeC was high (>0.95), suggesting that selection on either trait would lead to a positive correlated response on the other. Negative correlations were estimated between RT and MeC (-0.24 ± 0.38) and MeP (-0.43 ± 0.35). Methane concentration and MeP had slightly positive correlations with milk yield (0.17 ± 0.39 and 0.21 ± 0.36), protein percentage (0.08 ± 0.32 and 0.30 ± 0.45), protein yield (0.22 ± 0.41 and 0.31 ± 0.35), fat percentage (0.02 ± 0.40 and 0.27 ± 0.36), and fat yield (0.27 ± 0.28 and 0.29 ± 0.28) from bivariate analyses. Rumination time had positive correlations with milk yield (0.41 ± 0.75) and protein yield (0.26 ± 0.57) and negative correlations with fat yield (-0.45 ± 0.32), protein percentage (-0.15 ± 0.38), and fat percentage (-0.40 ± 0.47). A positive approximated genetic correlation was estimated between fertility and MeC (0.10 ± 0.05) and MeP (0.18 ± 0.05), resulting in slightly higher CH4 production when selecting for better fertility [days open estimated breeding values (EBV) are expressed with mean 100 and SD 10, inversely related to days from calving to conception; that is, greater days open EBV implies better fertility]. Positive correlations were also estimated for stature with MeC and MeP (0.30 ± 0.04 and 0.43 ± 0.04, respectively). Other type traits (chest width, udder depth, angularity, and capacity) were positively correlated with methane traits, possibly because of higher milk yield and higher feed intake from these animals. Rumination time showed positive EBV correlations with production traits and type traits, and negative correlations with somatic cell count and body condition score. Based on the genetic correlations and heritabilities estimated in this study, methane is measurable and heritable, and estimates of genetic correlations suggest no strong opposition to current breeding objectives in Spanish Holsteins.

Changes in the rumen and colon microbiota and effects of live yeast dietary supplementation during the transition from the dry period to lactation of dairy cows.

The first objective of this study was to evaluate the dynamics and their potential association with animal performance of the microbiota in both the rumen and colon of dairy cows as they move from a nonlactation to a lactation ration. The second objective was to assess the potential effects on the microbiota of live yeast supplementation. Twenty-one Holstein cows were split in 2 treatments consisting of 1 × 1010 cfu/d of live yeast (LY; n = 10) or no supplementation (control; n = 11) starting 21 d before until 21 d after calving. At 14 d before and 7 and 21 d after calving, samples of rumen and colon digesta were obtained from each cow using an endoscope. Total DNA was extracted and submitted to high-throughput sequencing. Shannon diversity index, in both the rumen and colon, was unaffected by LY; however, in the rumen it was lowest 7 d after calving and returned to precalving values at 21 d in milk, whereas in the colon it was greatest 14 d before calving but decreased after calving. In the rumen, LY supplementation increased the relative abundance (RA) of Bacteroidales (group UCG-001), Lachnospiracea (groups UCG-002 and UCG-006), and Flexilinea 14 d before calving, and increased RA of Streptococcus 21 d after calving compared with control cows. However, changes in the ruminal microbiota were more drastic across days relative to calving than as influenced by the dietary treatment, and the effect of LY in the colon was milder than in the rumen. The ruminal RA of several genera was associated with postcalving DMI, and that of Gastranaerophilales was the only order positively associated with milk yield. Several genera were positively correlated with feed efficiency, with Clostridiales (unclassified) being the only genus negatively associated with feed efficiency. In the colon, Prevotellaceae (group Ga6A1) was the only genus positively associated with feed efficiency. The ruminal RA of Prevotella 7 and Ruminobacter 14 d precalving was negatively correlated with dry matter intake and milk yield postcalving. The RA of Parabacteroides in the colon 14 d before calving was negatively correlated with milk yield, whereas the RA of Eggerthellaceae (unclassified) and Erysipelotrichaceae (groups c and unclassified) were positively correlated with feed efficiency. Interestingly, LY supplementation doubled the RA of Eggerthellaceae (unclassified) in the colon. It is concluded that microbial diversity in the rumen experiences a transient reduction after calving, whereas in the colon, the reduction is maintained at least until 21 d in milk. Most of the effects of LY on rumen microbiota were observed before calving, whereas in the colon, LY effects were more moderate but consistent and independent of the stage of production. The microbial community of the rumen after calving is more associated with feed intake, milk yield, and feed efficiency than that of the colon. However, the colon microbiota before calving is more associated with feed efficiency after calving than that of the rumen.

Short communication: Signs of host genetic regulation in the microbiome composition in 2 dairy breeds: Holstein and Brown Swiss.

This study aimed to evaluate whether the host genotype exerts any genetic control on the microbiome composition of the rumen in cattle. Microbial DNA was extracted from 18 samples of ruminal content from 2 breeds (Holstein and Brown Swiss). Reads were processed using mothur (https://www.mothur.org/) in 16S and 18S rRNA gene-based analyses. Then, reads were classified at the genus clade, resulting in 3,579 operational taxonomic units (OTU) aligned against the 16S database and 184 OTU aligned against the 18S database. After filtering on relative abundance (>0.1%) and penetrance (95%), 25 OTU were selected for the analyses (17 bacteria, 1 archaea, and 7 ciliates). Association with the genetic background of the host animal based on the principal components of a genomic relationship matrix based on single nucleotide polymorphism markers was analyzed using Bayesian methods. Fifty percent of the bacteria and archaea genera were associated with the host genetic background, including Butyrivibrio, Prevotella, Paraprevotella, and Methanobrevibacter as main genera. Forty-three percent of the ciliates analyzed were also associated with the genetic background of the host. In total, 48% of microbes were associated with the host genetic background. The results in this study support the hypothesis and provide some evidence that there exists a host genetic component in cattle that can partially regulate the composition of the microbiome.

Including nonadditive genetic effects in mating programs to maximize dairy farm profitability.

We compared the outcome of mating programs based on different evaluation models that included nonadditive genetic effects (dominance and heterozygosity) in addition to additive effects. The additive and dominance marker effects and the values of regression on average heterozygosity were estimated using 632,003 single nucleotide polymorphisms from 7,902 and 7,510 Holstein cows with calving interval and production (milk, fat, and protein yields) records, respectively. Expected progeny values were computed based on the estimated genetic effects and genotype probabilities of hypothetical progeny from matings between the available genotyped cows and the top 50 young genomic bulls. An index combining the traits based on their economic values was developed and used to evaluate the performance of different mating scenarios in terms of dollar profit. We observed that mating programs with nonadditive genetic effects performed better than a model with only additive effects. Mating programs with dominance and heterozygosity effects increased milk, fat, and protein yields by up to 38, 1.57, and 1.21 kg, respectively. The inclusion of dominance and heterozygosity effects decreased calving interval by up to 0.70 d compared with random mating. The average reduction in progeny inbreeding by the inclusion of nonadditive genetic effects in matings compared with random mating was between 0.25 to 1.57 and 0.64 to 1.57 percentage points for calving interval and production traits, respectively. The reduction in inbreeding was accompanied by an average of A$8.42 (Australian dollars) more profit per mating for a model with additive, dominance, and heterozygosity effects compared with random mating. Mate allocations that benefit from nonadditive genetic effects can improve progeny performance only in the generation where it is being implemented, and the gain from specific combining abilities cannot be accumulated over generations. Continuous updating of genomic predictions and mate allocation programs are required to benefit from nonadditive genetic effects in the long term.

Improving the reliability of female fertility breeding values using type and milk yield traits that predict energy status in Australian Holstein cattle.

The objectives of this study were (1) to propose changing the selection criteria trait for evaluating fertility in Australia from calving interval to conception rate at d 42 after the beginning of the mating season and (2) to use type traits as early fertility predictors, to increase the reliability of estimated breeding values for fertility. The breeding goal in Australia is conception within 6 wk of the start of the mating season. Currently, the Australian model to predict fertility breeding values (expressed as a linear transformation of calving interval) is a multitrait model that includes calving interval (CVI), lactation length (LL), calving to first service (CFS), first nonreturn rate (FNRR), and conception rate. However, CVI has a lower genetic correlation with the breeding goal (conception within 6 wk of the start of the mating season) than conception rate. Milk yield, type, and fertility data from 164,318 cow sired by 4,766 bulls were used. Principal component analysis and genetic correlation estimates between type and fertility traits were used to select type traits that could subsequently be used in a multitrait analysis. Angularity, foot angle, and pin set were chosen as type traits to include in an index with the traits that are included in the multitrait fertility model: CVI, LL, CFS, FNRR, and conception rate at d 42 (CR42). An index with these 8 traits is expected to achieve an average bull first proof reliability of 0.60 on the breeding objective (conception within 6 wk of the start of the mating season) compared with reliabilities of 0.39 and 0.45 for CR42 only or the current 5-trait Australian model. Subsequently, we used the first eigenvector of a principal component analysis with udder texture, bone quality, angularity, and body condition score to calculate an energy status indicator trait. The inclusion of the energy status indicator trait composite in a multitrait index with CVI, LL, CFS, FNRR, and CR42 achieved a 12-point increase in fertility breeding value reliability (i.e., increased by 30%; up to 0.72 points of reliability), whereas a lower increase in reliability (4 points, i.e., increased by 10%) was obtained by including angularity, foot angle, and pin set in the index. In situations when a limited number of daughters have been phenotyped for CR42, including type data for sires increased reliabilities compared with when type data were omitted. However, sires with more than 80 daughters with CR42 records achieved reliability estimates close to 80% on average, and there did not appear to be a benefit from having daughters with type records. The cost of phenotyping to obtain such reliabilities (assuming a cost of AU$14 per cow with type data and AU$5 per cow with pregnancy diagnosed) is lower if more pregnancy data are collected in preference to type data. That is, efforts to increase the reliability of fertility EBV are most cost effective when directed at obtaining a larger number of pregnancy tests.

Hot topic: Definition and implementation of a breeding value for feed efficiency in dairy cows.

A new breeding value that combines the amount of feed saved through improved metabolic efficiency with predicted maintenance requirements is described. The breeding value includes a genomic component for residual feed intake (RFI) combined with maintenance requirements calculated from either a genomic or pedigree estimated breeding value (EBV) for body weight (BW) predicted using conformation traits. Residual feed intake is only available for genotyped Holsteins; however, BW is available for all breeds. The RFI component of the "feed saved" EBV has 2 parts: Australian calf RFI and Australian lactating cow RFI. Genomic breeding values for RFI were estimated from a reference population of 2,036 individuals in a multi-trait analysis including Australian calf RFI (n=843), Australian lactating cow RFI (n=234), and UK and Dutch lactating cow RFI (n=958). In all cases, the RFI phenotypes were deviations from a mean of 0, calculated by correcting dry matter intake for BW, growth, and milk yield (in the case of lactating cows). Single nucleotide polymorphism effects were calculated from the output of genomic BLUP and used to predict breeding values of 4,106 Holstein sires that were genotyped but did not have RFI phenotypes themselves. These bulls already had BW breeding values calculated from type traits, from which maintenance requirements in kilograms of feed per year were inferred. Finally, RFI and the feed required for maintenance (through BW) were used to calculate a feed saved breeding value and expressed as the predicted amount of feed saved per year. Animals that were 1 standard deviation above the mean were predicted to eat 66 kg dry matter less per year at the same level of milk production. In a data set of genotyped Holstein sires, the mean reliability of the feed saved breeding value was 0.37. For Holsteins that are not genotyped and for breeds other than Holsteins, feed saved is calculated using BW only. From April 2015, feed saved has been included as part of the Australian national selection index, the Balanced Performance Index (BPI). Selection on the BPI is expected to lead to modest gains in feed efficiency.

On the value of the phenotypes in the genomic era.

Genetic improvement programs around the world rely on the collection of accurate phenotypic data. These phenotypes have an inherent value that can be estimated as the contribution of an additional record to genetic gain. Here, the contribution of phenotypes to genetic gain was calculated using traditional progeny testing (PT) and 2 genomic selection (GS) strategies that, for simplicity, included either males or females in the reference population. A procedure to estimate the theoretical economic contribution of a phenotype to a breeding program is described for both GS and PT breeding programs through the increment in genetic gain per unit of increase in estimated breeding value reliability obtained when an additional phenotypic record is added. The main factors affecting the value of a phenotype were the economic value of the trait, the number of phenotypic records already available for the trait, and its heritability. Furthermore, the value of a phenotype was affected by several other factors, including the cost of establishing the breeding program and the cost of phenotyping and genotyping. The cost of achieving a reliability of 0.60 was assessed for different reference populations for GS. Genomic reference populations of more sires with small progeny group sizes (e.g., 20 equivalent daughters) had a lower cost than those reference populations with either large progeny group sizes for fewer genotyped sires, or female reference populations, unless the heritability was large and the cost of phenotyping exceeded a few hundred dollars; then, female reference populations were preferable from an economic perspective.

Incorporating heifer feed efficiency in the Australian selection index using genomic selection.

The economic benefit of expanding the Australian Profit Ranking (APR) index to include residual feed intake (RFI) was evaluated using a multitrait selection index. This required the estimation of genetic parameters for RFI and genetic correlations using single nucleotide polymorphism data (genomic) correlations with other traits. Heritabilities of RFI, dry matter intake (DMI), and all the traits in the APR (milk, fat, and protein yields; somatic cell count; fertility; survival; milking speed; and temperament), and genomic correlations between these traits were estimated using a Bayesian framework, using data from 843 growing Holstein heifers with phenotypes for DMI and RFI, and bulls with records for the other traits. Heritability estimates of DMI and RFI were 0.44 and 0.33, respectively, and the genomic correlation between them was 0.03 and nonsignificant. The genomic correlations between the feed-efficiency traits and milk yield traits were also close to zero, ranging between -0.11 and 0.10. Positive genomic correlations were found for DMI with stature (0.16) and with overall type (0.14), suggesting that taller cows eat more as heifers. One issue was that the genomic correlation estimates for RFI with calving interval (ClvI) and with body condition score were both unfavorable (-0.13 and 0.71 respectively), suggesting an antagonism between feed efficiency and fertility. However, because of the relatively small numbers of animals in this study, a large 95% probability interval existed for the genomic correlation between RFI and ClvI (-0.66, 0.36). Given these parameters, and a genetic correlation between heifer and lactating cow RFI of 0.67, inclusion of RFI in the APR index would reduce RFI by 1.76 kg/cow per year. Including RFI in the APR would result in the national Australian Holstein herd consuming 1.73 × 10(6) kg less feed, which is worth 0.55 million Australian dollars (A$) per year and is 3% greater than is currently possible to achieve. Other traits contributing to profitability, such as milk production and fertility, will also improve through selection on this index; for example, ClvI would be reduced by 0.53 d/cow per year, which is 96% of the gain for this trait that is achieved without RFI in the APR.

Prediction of liveweight of cows from type traits and its relationship with production and fitness traits.

Liveweight (LWT) data for Australian Holstein cows was predicted from different type traits based on actual LWT and type data of 932 cows collected from 20 different herds over a 3-yr period. In addition to LWT measured using scales, visual estimates of LWT were also available on 90% of the cows with LWT data. The future predictive ability of different models was assessed using 10-fold cross-validation. The relationships between LWT and selected type traits, including body condition score (BCS), were also calculated to assess the usefulness of various traits to predict LWT genetically. The relationships of predicted LWT measures with production and fitness traits were also estimated in an attempt to assess the consequence of continuous selection on an economic index that includes predicted LWT with a negative economic value. The heritability of LWT was 0.4. Among type traits, stature, chest width, bone quality, BCS, udder depth, central ligament, and muzzle width were correlated with LWT both genetically and phenotypically and were used to predict LWT of cows. Predicted LWT measures, using several sets of traits and visually estimated LWT were genetically highly correlated with each other (>0.73). Phenotypically, visually estimated LWT of cows was slightly more correlated with actual LWT than that predicted from type traits, but genetically both approaches gave the same accuracy. The predicted estimates of LWT were also positively genetically correlated with energy-corrected milk yield and had near zero correlation with survival. The correlations of different measures of LWT with fertility traits were unfavorable or near zero, suggesting that selection for reduced LWT may not cause deterioration in fertility traits. However, it may be useful to consider broadening the breeding objective to include traits that are associated with energy balance, particularly if traits such as BCS and bone quality are included in the set of traits used to predict LWT. Based on the results from this study, the inclusion of predicted LWT with negative economic values into the breeding objective would have no negative effect on fitness traits.

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.

Assets of imputation to ultra-high density for productive and functional traits.

The aim of this study was to evaluate different-density genotyping panels for genotype imputation and genomic prediction. Genotypes from customized Golden Gate Bovine3K BeadChip [LD3K; low-density (LD) 3,000-marker (3K); Illumina Inc., San Diego, CA] and BovineLD BeadChip [LD6K; 6,000-marker (6K); Illumina Inc.] panels were imputed to the BovineSNP50v2 BeadChip [50K; 50,000-marker; Illumina Inc.]. In addition, LD3K, LD6K, and 50K genotypes were imputed to a BovineHD BeadChip [HD; high-density 800,000-marker (800K) panel], and with predictive ability evaluated and compared subsequently. Comparisons of prediction accuracy were carried out using Random boosting and genomic BLUP. Four traits under selection in the Spanish Holstein population were used: milk yield, fat percentage (FP), somatic cell count, and days open (DO). Training sets at 50K density for imputation and prediction included 1,632 genotypes. Testing sets for imputation from LD to 50K contained 834 genotypes and testing sets for genomic evaluation included 383 bulls. The reference population genotyped at HD included 192 bulls. Imputation using BEAGLE software (http://faculty.washington.edu/browning/beagle/beagle.html) was effective for reconstruction of dense 50K and HD genotypes, even when a small reference population was used, with 98.3% of SNP correctly imputed. Random boosting outperformed genomic BLUP in terms of prediction reliability, mean squared error, and selection effectiveness of top animals in the case of FP. For other traits, however, no clear differences existed between methods. No differences were found between imputed LD and 50K genotypes, whereas evaluation of genotypes imputed to HD was on average across data set, method, and trait, 4% more accurate than 50K prediction, and showed smaller (2%) mean squared error of predictions. Similar bias in regression coefficients was found across data sets but regressions were 0.32 units closer to unity for DO when genotypes were imputed to HD density. Imputation to HD genotypes might produce higher stability in the genomic proofs of young candidates. Regarding selection effectiveness of top animals, more (2%) top bulls were classified correctly with imputed LD6K genotypes than with LD3K. When the original 50K genotypes were used, correct classification of top bulls increased by 1%, and when those genotypes were imputed to HD, 3% more top bulls were detected. Selection effectiveness could be slightly enhanced for certain traits such as FP, somatic cell count, or DO when genotypes are imputed to HD. Genetic evaluation units may consider a trait-dependent strategy in terms of method and genotype density for use in the genome-enhanced evaluations.

Reaction norm of fertility traits adjusted for protein and fat production level across lactations in Holstein cattle.

A total of 304,001 artificial insemination outcomes in up to 7 lactations from 142,389 Holstein cows, daughters of 5,349 sires and 101,433 dams, calving between January 1995 and December 2007 in 1,347 herds were studied by a reaction norm model. The (co)variance components for days to first service (DFS), days open, nonreturn rate in the first service (NRFS), and number of services per conception were estimated by 6 models: 3 Legendre polynomial degrees for the genetic effects and adjustment or not for the level of fat plus protein (FP) production recorded at day closest to DFS. For all traits and type of FP adjustment, a second degree polynomial showed the best fit. The use of the adjusted FP model did not increase the level of genetic (co)variance components except for DFS. The heritability for each of the traits was low in general (0.03-0.10) and increased from the first to fourth calving; nevertheless, very important variability was found for the estimated breeding value (EBV) of the sires. The genetic correlations (rg) were close to unity between adjacent calvings, but decreased for most distant parities, ranging from rg=0.36 (for DFS) to rg=0.63 (for NRFS), confirming the existence of heterogeneous genetic (co)variance components and EBV across lactations. The results of the eigen decomposition of rg shows that the first eigenvalue explained between 82 to 92% and the second between 8 to 14% of the genetic variance for all traits; therefore, a deformation of the overall mean trajectory for reproductive performance across the trajectory of the different calving could be expected if selection favored these eigenfunctions. The results of EBV for the 50 best sires showed a substantial reranking and variation in the shape of response across lactations. The more important aspect to highlight, however, is the difference between the EBV of the same sires in different calvings, a characteristic known as plasticity, which is particularly important for DFS and NRFS. This component of fertility adds another dimension to selection for fertility that can be used to change the negative genetic progress of reproductive performance presented in this population of Holstein cows. The use of a reaction norm model should allow producers to obtain more robust cows for maintenance of fertility levels along the whole productive life of the cows.

The gradient boosting algorithm and random boosting for genome-assisted evaluation in large data sets.

In the next few years, with the advent of high-density single nucleotide polymorphism (SNP) arrays and genome sequencing, genomic evaluation methods will need to deal with a large number of genetic variants and an increasing sample size. The boosting algorithm is a machine-learning technique that may alleviate the drawbacks of dealing with such large data sets. This algorithm combines different predictors in a sequential manner with some shrinkage on them; each predictor is applied consecutively to the residuals from the committee formed by the previous ones to form a final prediction based on a subset of covariates. Here, a detailed description is provided and examples using a toy data set are included. A modification of the algorithm called "random boosting" was proposed to increase predictive ability and decrease computation time of genome-assisted evaluation in large data sets. Random boosting uses a random selection of markers to add a subsequent weak learner to the predictive model. These modifications were applied to a real data set composed of 1,797 bulls genotyped for 39,714 SNP. Deregressed proofs of 4 yield traits and 1 type trait from January 2009 routine evaluations were used as dependent variables. A 2-fold cross-validation scenario was implemented. Sires born before 2005 were used as a training sample (1,576 and 1,562 for production and type traits, respectively), whereas younger sires were used as a testing sample to evaluate predictive ability of the algorithm on yet-to-be-observed phenotypes. Comparison with the original algorithm was provided. The predictive ability of the algorithm was measured as Pearson correlations between observed and predicted responses. Further, estimated bias was computed as the average difference between observed and predicted phenotypes. The results showed that the modification of the original boosting algorithm could be run in 1% of the time used with the original algorithm and with negligible differences in accuracy and bias. This modification may be used to speed the calculus of genome-assisted evaluation in large data sets such us those obtained from consortiums.