Simultaneous estimation of genotype by environment interaction accounting for discrete and continuous environmental descriptors in Irish dairy cattle

Genotype by environment interaction due to heat stress in Brown Swiss cattle

Due to its geographical position and a highly variable orography, Italy is characterized by several climatic areas and thus, by many different dairy cow farming systems. Brown Swiss cattle, in this context, are a very appreciated genetic resource for their adaptability and low metabolic requirement. The significant heterogeneity in farming systems may consist of genotype by environment (G × E) interactions with neglected changes in animals' rank position. The objective of this study was to investigate G × E for heat tolerance in Brown Swiss cattle for several production traits (milk, fat, and protein yield in kilograms; fat, protein, and cheese yield in percentage) and 2 derivate traits (fat-corrected milk and energy-corrected milk). We used the daily maximum temperature-humidity index (THI) range, calculated according to weather stations' data from 2008 to 2018 in Italy, and 202,776 test-day records from 23,396 Brown Swiss cows from 639 herds. Two different methodologies were applied to estimate the effect of the environmental variable (THI) on genetic parameters: (1) the reaction norm model, which uses a continuous random covariate to estimate the animal additive effect, and (2) the multitrait model, which splits each production pattern as a distinct and correlated trait according to the first (a thermal comfort condition), third (a moderate heat stress condition), and fifth (a severe heat stress condition) mean THI value quintile. The results from the reaction norm model showed a descending trend of the additive genetic effect until THI reached the value of 80. Then we recorded an increase with high extreme THI values (THI 90). Permanent environmental variance at increasing THI values revealed an opposite trend: The plot of heritability and the ratio of animal permanent environmental variance to phenotypic variance showed that when the environmental condition worsens, the additive genetic and permanent environmental component for production traits play a growing role. The negative additive genetic correlation between slope and linear random coefficient indicates no linear relationship between the production traits or under heat stress conditions, except for milk yield and protein yield. In tridimensional wireframe plots, the extreme margin decreases until a minimum of ∼0.90 of genetic correlation in the ECM trait, showing that the magnitude of G × E interaction is greater than the other traits. Genetic correlation values in Brown Swiss suggest the possibility of moderate changes in animals' estimated breeding value in heat stress conditions. Results indicated a moderate G × E interaction but significant variability in sire response related to their production level.

Genotype by Environment Interaction and Selection Response for Milk Yield Traits and Conformation in a Local Cattle Breed Using a Reaction Norm Approach

Local breeds are often reared in various environmental conditions (EC), suggesting that genotype by environment interaction (GxE) could influence genetic progress. This study aimed at investigating GxE and response to selection (R) in Rendena cattle under diverse EC. Traits included milk, fat, and protein yields, fat and protein percentage, and somatic cell score, three-factor scores and 24 linear type traits. The traits belonged to 11,085 cows (615 sires). Variance components were estimated in a two-step reaction norm model (RNM). A single trait animal model was run to obtain the solutions of herd-EC effect, then included in a random regression sire model. A multivariate response to selection (R) in different EC was computed for traits under selection including beef traits from a performance test. GxE accounted on average for 10% of phenotypic variance, and an average rank correlation of over 0.97 was found between bull estimated breeding values (EBVs) by either including or not including GxE, with changing top ranks. For various traits, significantly greater genetic components and R were observed in plain farms, loose housing rearing system, feeding total mixed ration, and without summer pasture. Conversely, for beef traits, a greater R was found for mountain farms, loose housing, hay-based feeding and summer pasture.

Genotype by environment interaction for female fertility traits under conventional and organic production systems in Danish Holsteins

Conventional and organic production systems mainly differ in feeding strategies, outdoor and pasture access, and the use of antibiotic treatments. These environmental differences could lead to a genotype by environment interaction (G × E) and a requirement for including G × E in breeding decisions. The objectives of this study were to estimate variance components and heritabilities for conventional and organic production systems and investigate G × E under these 2 production systems for female fertility traits in Danish Holsteins. The analyzed traits included the interval from calving to first insemination (ICF), the interval from first to last insemination, number of inseminations per conception (NINS), and non-return rate within 56 d after the first insemination. Records of female fertility in heifers and the first 3 lactations in cows as well as grass ratio of feed at herd level were collected during the period from 2011 to 2016. The performances of a trait in heifers and cows (lactation 1 to 3) were considered as different traits. The (co)variance components and the resulting heritabilities and genetic correlations were estimated using 2 models. One was a bivariate model treating performances of a trait under organic and conventional production systems as 2 different traits using a reduced data set, and the other was a reaction norm model with random regression on the production system and the grass ratio of feed using a full data set. The full data set comprised records of 37,836 females from 112 organic herds and 513,599 females from 1,224 conventional herds, whereas the reduced data set comprised records from all these 112 organic herds and 92,696 females from 185 convention herds extracted from the full data set with grass ratio of feed lower than 0.20. All female fertility performances of the organic production system were superior to those of the conventional production system. Besides, heterogeneities in additive genetic variances and heritabilities were observed between conventional and organic production systems for all traits. Furthermore, genetic correlations between these 2 production systems ranged from 0.607 to 1.000 estimated from bivariate models and from 0.848 to 0.999 estimated from reaction norm models. Statistically significant G × E were observed for NINS in heifers, non-return rate within 56 d after the first insemination in heifers, and ICF from the bivariate model, and for ICF and NINS in cows from the reaction norm model.

Estimation of genetic variation for macro- and micro-environmental sensitivities of milk yield and composition in Holstein cows using double hierarchical generalized linear models

The aim of this study was to estimate genetic parameters for environmental sensitivities in milk yield and composition of Iranian Holstein cows using the double hierarchical generalized linear model (DHGLM) method. Data set included test-day productive records of cows which were provided by the Animal Breeding Center and Promotion of Animal Products of Iran during 1983 to 2014. In the DHGLM method, a random regression model was fitted which included two parts of mean and residual variance. A random regression model (mean model) and a residual variance model were used to study the genetic variation of micro-environmental sensitivities. In order to consider macro-environmental sensitivities, DHGLM was extended using a reaction norm model, and a sire model was applied. Based on the mean model, additive genetic variances for the mean were 38.25 for milk yield, 0.23 for fat yield and 0.03 for protein yield in the first lactation, respectively. Based on the residual variance model, additive genetic variances for residual variance were 0.039 for milk yield, 0.030 for fat yield and 0.020 for protein yield in the first lactation, respectively. Estimates of genetic correlation between milk yield and macro- and micro-environmental sensitivities were 0.660 and 0.597 in the first lactation, respectively. The results of this study indicated that macro- and micro-environmental sensitivities were present for milk production traits of Iranian Holsteins. High genetic coefficient of variation for micro-environmental sensitivities indicated the possibility of reducing environmental variation and increase in uniformity via selection.

Genotype by environment (climate) interaction improves genomic prediction for production traits in US Holstein cattle

Genotype by environment interaction (G × E) in dairy cattle productive traits has been shown to exist, but current genetic evaluation methods do not take this component into account. As several environmental descriptors (e.g., climate, farming system) are known to vary within the United States, not accounting for the G × E could lead to reranking of bulls and loss in genetic gain. Using test-day records on milk yield, somatic cell score, fat, and protein percentage from all over the United States, we computed within herd-year-season daughter yield deviations for 1,087 Holstein bulls and regressed them on genetic and environmental information to estimate variance components and to assess prediction accuracy. Genomic information was obtained from a 50k SNP marker panel. Environmental effect inputs included herd (160 levels), geographical region (7 levels), geographical location (2 variables), climate information (7 variables), and management conditions of the herds (16 total variables divided in 4 subgroups). For each set of environmental descriptors, environmental, genomic, and G × E components were sequentially fitted. Variance components estimates confirmed the presence of G × E on milk yield, with its effect being larger than main genetic effect and the environmental effect for some models. Conversely, G × E was moderate for somatic cell score and small for milk composition. Genotype by environment interaction, when included, partially eroded the genomic effect (as compared with the models where G × E was not included), suggesting that the genomic variance could at least in part be attributed to G × E not appropriately accounted for. Model predictive ability was assessed using 3 cross-validation schemes (new bulls, incomplete progeny test, and new environmental conditions), and performance was compared with a reference model including only the main genomic effect. In each scenario, at least 1 of the models including G × E was able to perform better than the reference model, although it was not possible to find the overall best-performing model that included the same set of environmental descriptors. In general, the methodology used is promising in accounting for G × E in genomic predictions, but challenges exist in identifying a unique set of covariates capable of describing the entire variety of environments.

Genotype by environment interactions on culling rates and 305-day milk yield of Holstein cows in 3 US regions

The objective of this study was to investigate genotype by environment interactions for culling rates and milk production in large and small dairy herds in 3 US regions, using genotypes, pedigree, and phenotypes. Single nucleotide polymorphism (SNP) marker variances were also estimated in these different environments. Culling rates including cow mortality were based on 6 Dairy Herd Improvement termination codes reported by dairy producers. Separate data sets for culling rates and 305-d milk yield were created for large and small dairy herds in the US regions of the Southeast (SE), Southwest (SW), and Northeast (NE) for the first 3 lactation cows that calved between 1999 and 2008. Genomic information from 42,503 SNP markers on 34,506 bulls was included in the analysis to predict genomic estimated breeding value (GEBV) of culling rates and 305-d milk yield with a single-step genomic BLUP using a bivariate threshold-linear model. Cow replacement rates in large SE and NE herds were higher. Heritability estimates of culling rates ranged from 0.03 to 0.11, but the differences were small between large and small herds and among the 3 US regions. Genetic correlations between culling rates and 305-d milk yield were medium to high for cows sold for poor production and reproduction problems. Correlations of GEBV for culling rates among the 3 US regions ranged from 0.34 to 0.92 and were lower between the SW and the other regions, especially in small herds. Correlations of GEBV between large and small herds ranged from 0.44 to 0.90 and were lower in the SW. These results indicate genotype by environment interactions of cow culling rate between the US regions and between large and small herds. Correlations of top 30 SNP marker effects for culling rates between 2 US regions ranged from 0.64 to 0.98 and were higher than those of more SNP marker effects except for a culling reason "sold for dairy purpose." Those correlations between large and small herds ranged from 0.67 to 0.98. High correlations of top SNP marker effects on culling reasons between the US regions and between large and small herds suggest that major markers can be useful for selection in different environments. The SNP variance shown in a marker gene segment on chromosome 14 was strongly associated with milk production in large and small herds in the NE but not in the SE and SW. Marker genes on chromosome 14 also showed a strong association with cow culling rates due to poor production and mortality in large herds in the NE.

Multi-trait linear reaction norm model to describe the pattern of phenotypic expression of some economic traits in beef cattle across a range of environments

The multi-trait reaction norm (MTRN) model was extended to beef cattle reared under tropical conditions with the following objectives: to compare multi-trait (MT) and MTRN models regarding the genetic parameters obtained; and to characterize G × E, the pattern of phenotypic expression, and the environmental sensitivity of animals for postweaning weight gain (PWG), scrotal circumference (SC), and annual average productivity of the cow (PRODAM). There was divergence in the estimates between the MT and MTRN models when the posterior probability intervals of additive genetic variances and heritability coefficients of PWG and PRODAM were analyzed. The MTRN model indicated an increase in heritability for PWG and PRODAM with improvement of the environmental conditions. For SC, heritability was practically the same, irrespective of the environmental conditions. The genetic correlations between the traits studied were low but varied over environments by the MTRN model. Considering genetic correlations obtained by the MTRN model for the same trait, lower estimates were obtained between extreme favorable and unfavorable environments. This finding suggest re-ranking of breeding values in different environments mainly for PWG and PRODAM. Thus, G × E is more important for PWG and PRODAM than for SC and should be included in the genetic evaluation of these traits. The traits PWG and PRODAM can be considered plastic traits, whereas SC is poorly plastic. The genetic trends in individual animal slopes indicate that the population is moving towards greater plasticity. This could be a matter of concern for breeders since greater plasticity seems to limit heritability and, consequently, the responses to selection.

Estimation of genetic variance for macro- and micro-environmental sensitivity using double hierarchical generalized linear models

BACKGROUND

Genetic variation for environmental sensitivity indicates that animals are genetically different in their response to environmental factors. Environmental factors are either identifiable (e.g. temperature) and called macro-environmental or unknown and called micro-environmental. The objectives of this study were to develop a statistical method to estimate genetic parameters for macro- and micro-environmental sensitivities simultaneously, to investigate bias and precision of resulting estimates of genetic parameters and to develop and evaluate use of Akaike's information criterion using h-likelihood to select the best fitting model.

METHODS

We assumed that genetic variation in macro- and micro-environmental sensitivities is expressed as genetic variance in the slope of a linear reaction norm and environmental variance, respectively. A reaction norm model to estimate genetic variance for macro-environmental sensitivity was combined with a structural model for residual variance to estimate genetic variance for micro-environmental sensitivity using a double hierarchical generalized linear model in ASReml. Akaike's information criterion was constructed as model selection criterion using approximated h-likelihood. Populations of sires with large half-sib offspring groups were simulated to investigate bias and precision of estimated genetic parameters.

RESULTS

Designs with 100 sires, each with at least 100 offspring, are required to have standard deviations of estimated variances lower than 50% of the true value. When the number of offspring increased, standard deviations of estimates across replicates decreased substantially, especially for genetic variances of macro- and micro-environmental sensitivities. Standard deviations of estimated genetic correlations across replicates were quite large (between 0.1 and 0.4), especially when sires had few offspring. Practically, no bias was observed for estimates of any of the parameters. Using Akaike's information criterion the true genetic model was selected as the best statistical model in at least 90% of 100 replicates when the number of offspring per sire was 100. Application of the model to lactation milk yield in dairy cattle showed that genetic variance for micro- and macro-environmental sensitivities existed.

CONCLUSION

The algorithm and model selection criterion presented here can contribute to better understand genetic control of macro- and micro-environmental sensitivities. Designs or datasets should have at least 100 sires each with 100 offspring.

Integration of epidemiology into the genetic analysis of mastitis in Swedish Holstein

Heritability of mastitis (and diseases in general) tends to be low. One possible cause is that no clear distinction can be made between resistant and nonresistant animals, because healthy animals include animals that have not been exposed to pathogens and resistant animals. To account for this, we quantified the prevalence of clinical mastitis (CM) and subclinical mastitis (SCM) in 2,069 Swedish Holstein herds as a measure of exposure. Herd prevalence averaged 26.5% for SCM and 6.4% for CM; 61% of the first lactations of 177,309 cows were classified as having at least one case of SCM and 10% as having CM. In a reaction norm approach, heritability of (S)CM was quantified as a function of herd prevalence of (S)CM. The best-fitting model was a second-order polynomial of first-lactation cow SCM as a function of herd prevalence SCM, and a first-order (linear) polynomial of first-lactation cow CM as a function of CM herd prevalence. Heritability for SCM ranged from 0.069 to 0.105 and for CM from 0.016 to 0.032. For both, we found no clear effect of herd prevalence on their heritability. Genetic correlations within traits across herd prevalences were all greater than 0.92. Whether relationships among prevalence, exposure, disease, and genetics were as expected is a matter of discussion, but reaction norm analyses may be a valuable tool for epidemiological genetics.

The genetic analysis of tolerance to infections: a review

Tolerance to infections is defined as the ability of a host to limit the impact of a given pathogen burden on host performance. Uncoupling resistance and tolerance is a challenge, and there is a need to be able to separate them using specific trait recording or statistical methods. We present three statistical methods that can be used to investigate genetics of tolerance-related traits. Firstly, using random regressions, tolerance can be analyzed as a reaction norm slope in which host performance (y-axis) is regressed against an increasing pathogen burden (x-axis). Genetic variance in tolerance slopes is the genetic variance for tolerance. Variation in tolerance can induce genotype re-ranking and changes in genetic and phenotypic variation in host performance along the pathogen burden trajectory, contributing to environment-dependent genetic responses to selection. Such genotype-by-environment interactions can be quantified by combining random regressions and covariance functions. To apply random regressions, pathogen burden of individuals needs to be recorded. Secondly, when pathogen burden is not recorded, the cure model for time-until-death data allows separating two traits, susceptibility and endurance. Susceptibility is whether or not an individual was susceptible to an infection, whereas endurance denotes how long time it took until the infection killed a susceptible animal (influenced by tolerance). Thirdly, the normal mixture model can be used to classify continuously distributed host performance, such as growth rate, into different sub-classes (e.g., non-infected and infected), which allows estimation of host performance reduction specific to infected individuals. Moreover, genetics of host performance can be analyzed separately in healthy and affected animals, even in the absence of pathogen burden and survival data. These methods provide novel tools to increase our understanding on the impact of parasites, pathogens, and production diseases on host traits.

Inbreeding-stress interactions: evolutionary and conservation consequences

The effect of environmental stress on the magnitude of inbreeding depression has a long history of intensive study. Inbreeding-stress interactions are of great importance to the viability of populations of conservation concern and have numerous evolutionary ramifications. However, such interactions are controversial. Several meta-analyses over the last decade, combined with omic studies, have provided considerable insight into the generality of inbreeding-stress interactions, its physiological basis, and have provided the foundation for future studies. In this review, we examine the genetic and physiological mechanisms proposed to explain why inbreeding-stress interactions occur. We specifically examine whether the increase in inbreeding depression with increasing stress could be due to a concomitant increase in phenotypic variation, using a larger data set than any previous study. Phenotypic variation does usually increase with stress, and this increase can explain some of the inbreeding-stress interaction, but it cannot explain all of it. Overall, research suggests that inbreeding-stress interactions can occur via multiple independent channels, though the relative contribution of each of the mechanisms is unknown. To better understand the causes and consequences of inbreeding-stress interactions in natural populations, future research should focus on elucidating the genetic architecture of such interactions and quantifying naturally occurring levels of stress in the wild.