Genetic control of residual variance offers opportunities to increase uniformity and resilience of livestock and aquaculture species. Improving uniformity and resilience of animals will improve health and welfare of animals and lead to more homogenous products. Our aims in this review were to summarize the current models and methods to study genetic control of residual variance, genetic parameters and genomic results for residual variance and discuss future research directions. Typically, the genetic coefficient of variation is high (median = 0.27; range 0-0.86) and the heritability of residual variance is low (median = 0.01; range 0-0.10). Higher heritabilities can be achieved when increasing the number of records per animal. Divergent selection experiments have supported the feasibility of selecting for high or low residual variance. Genomic studies have revealed associations in regions related to stress, including those from the heat shock protein family. Although the number of studies is growing, genetic control of residual variance is still poorly understood, but big data and genomics offer great opportunities. K E Y W O R D Sgenetic control of residual variance, genetic heterogeneity of residual variance, micro-environmental sensitivity, resilience, uniformity 264 | IUNG et al.

Measurements of milk yield (MY), somatic cell score (SCS), percentage of fat (FP), protein (PP), lactose (LP), casein (CP) and percentage of palmitic (C16:0), stearic (C18:0), oleic (C18:1), total saturated (SFA), unsaturated (UFA), monounsaturated (MUFA) and polyunsaturated (PUFA) fatty acids in milk from 5,224 Holstein cows were evaluated as a function of a temperature and humidity index (THI). Legendre orthogonal polynomials from second to seventh order were tested. The best fit order for MY, PP and C18:0 was the third, whereas the second for all other traits. The heritability estimates decreased for MY (0.31 to 0.14), FP (0.28 to 0.16), LP (0.43 to 0.30), SCS (0.14 to 0.09), SFA (0.33 to 0.22) and C16:0 (0.31 to 0.26), whereas increased for CP (0.32 to 0.42), MUFA (0.08 to 0.13), UFA (0.07 to 0.11) and C18:1 (0.07 to 0.11) as the THI level increased. For PP, heritabilities (0.26 to 0.39) presented larger values in intermediate THI. For PUFA and C18:0, heritabilities were approximately constant (0.13 to 0.14 and 0.15, respectively). However, the greatest variations may have been the result of the limitations of Legendre polynomials at the extreme points of the curve, and the pattern of heritabilities curves was approximately constant for the evaluated traits. Spearman's rank correlations between breeding values in extreme THI levels were greater than 0.80 for all traits considering all animals, only cows and only bulls. When considering the top 1% and the top 50% animals (only cows, only bulls and all), Spearman correlations smaller than 0.70 were found, suggesting reranking of the animals. Although there was little variation in the variance components over THI, it is possible that there is no heat stress in the animals studied, because, on average, there was no great impact of the thermal load on the traits. One possible explanation is the use of herds with little climatic difference among herds, as well as the use of fans and sprinklers into the barns. However, the THI levels may be important factors in the selection process, as reranking of animals was verified.

BackgroundIn livestock, residual variance has been studied because of the interest to improve uniformity of production. Several studies have provided evidence that residual variance is partially under genetic control; however, few investigations have elucidated genes that control it. The aim of this study was to identify genomic regions associated with within-family residual variance of yearling weight (YW; N = 423) in Nellore bulls with high density SNP data, using different response variables. For this, solutions from double hierarchical generalized linear models (DHGLM) were used to provide the response variables, as follows: a DGHLM assuming non-null genetic correlation between mean and residual variance (rmv ≠ 0) to obtain deregressed EBV for mean (dEBVm) and residual variance (dEBVv); and a DHGLM assuming rmv = 0 to obtain two alternative response variables for residual variance, dEBVv_r0 and log-transformed variance of estimated residuals (ln_).ResultsThe dEBVm and dEBVv were highly correlated, resulting in common regions associated with mean and residual variance of YW. However, higher effects on variance than the mean showed that these regions had effects on the variance beyond scale effects. More independent association results between mean and residual variance were obtained when null rmv was assumed. While 13 and 4 single nucleotide polymorphisms (SNPs) showed a strong association (Bayes Factor > 20) with dEBVv and ln_, respectively, only suggestive signals were found for dEBVv_r0. All overlapping 1-Mb windows among top 20 between dEBVm and dEBVv were previously associated with growth traits. The potential candidate genes for uniformity are involved in metabolism, stress, inflammatory and immune responses, mineralization, neuronal activity and bone formation.ConclusionsIt is necessary to use a strategy like assuming null rmv to obtain genomic regions associated with uniformity that are not associated with the mean. Genes involved not only in metabolism, but also stress, inflammatory and immune responses, mineralization, neuronal activity and bone formation were the most promising biological candidates for uniformity of YW. Although no clear evidence of using a specific response variable was found, we recommend consider different response variables to study uniformity to increase evidence on candidate regions and biological mechanisms behind it.Electronic supplementary materialThe online version of this article (10.1186/s12864-018-5003-4) contains supplementary material, which is available to authorized users.

Fatty acids (FA) have been related to effects on human health, sensory quality and shelf life of dairy products, cow's health and methane emission. However, despite their importance, they are not regularly measured in all dairy herds yet, which can affect the accuracy of estimated breeding values (EBV) for these traits. In this case, an alternative is to use genomic selection. Thus, the aim was to assess the use of genomic information in the genetic evaluation for milk traits in a tropical Holstein population.Monthly records (n = 36,457) of milk FA percentage, daily milk yield and quality traits from 4,203 cows as well as the genotypes of 755 of these cows for 57,368 single nucleotide polymorphisms (SNP) were used. Polygenic and genomic-polygenic models were applied for EBV prediction, and both models were compared through the EBV accuracy calculated from the prediction error and Spearman's correlation among EBV rankings. Prediction accuracy was assessed by using cross-validation.In this case, the accuracy was the correlation between the genomic breeding values (GEBV) obtained as the sum of SNP effects and the EBV obtained in the polygenic model in each validation group. For all traits, the use of the genomic-polygenic model did not alter the animals' ranking, with correlations higher than 0.87. Nevertheless, through this model, the accuracy increased from 1.5% to 6.8% compared to the polygenic model. The correlations between GEBV and EBV varied from 0.52 to 0.68. Therefore, the use of a small group of genotyped cows in the genetic evaluation can increase the accuracy of EBV for milk FA and other traditional milk traits.