- Best Linear Unbiased Prediction (BLUP)-based evaluation is one of the most important statistical methods used in modern animal breeding and quantitative genetics. BLUP, a method developed to estimate the genetic merit of animals using performance records, pedigree relationships, and appropriate statistical models. It helps breeders identify animals with superior inherited potential for economically important traits such as milk production, growth rate, fertility, disease resistance, feed efficiency, carcass quality, and longevity. By separating genetic influences from systematic environmental effects, BLUP-based evaluation supports more accurate comparisons among animals and contributes to sustainable genetic improvement in livestock populations.
- The primary objective of BLUP-based evaluation is to predict the breeding values of animals as accurately as possible using the information available. A breeding value represents the additive genetic contribution an animal is expected to pass to its offspring for a particular trait. Because an animal’s true breeding value cannot normally be observed directly, it must be estimated from performance records and information about genetically related animals. The resulting estimated breeding value, or EBV, provides a basis for ranking animals and making selection decisions. Animals with favorable EBVs are more likely, on average, to transmit desirable genetic characteristics to their offspring, although the performance of individual offspring will also depend on the other parent and environmental conditions.
- BLUP has become widely used because it provides a consistent framework for evaluating animals with different amounts of information. Some animals have numerous performance records and many recorded relatives, while others may have only a single observation or no direct performance records. BLUP can incorporate these different sources of information into a common statistical analysis. For example, a young bull may be evaluated using its pedigree, genomic information where available, and records from related animals before it has produced offspring. An older bull may have additional information from the performance of its daughters. BLUP combines relevant evidence according to the statistical model rather than relying only on an animal’s own observed performance.
- One of the major challenges in animal breeding is that observed performance reflects both genetic and environmental influences. A dairy cow’s milk yield, for instance, depends not only on inherited genetic potential but also on nutrition, age, health, lactation stage, housing, and management. Similarly, the growth rate of beef cattle can be affected by feeding systems, climate, disease exposure, and herd management. If these environmental influences are ignored, animals raised under favorable conditions may appear genetically superior to animals raised under less favorable conditions. BLUP-based evaluation addresses this problem by including appropriate environmental factors in the statistical model, helping breeders make fairer comparisons among animals.
- The statistical foundation of BLUP is the linear mixed model. A general animal breeding model can be expressed as:
- y = Xb + Za + e
- In this equation, y represents the vector of observed performance records, b represents the vector of fixed effects, a represents the vector of random additive genetic effects or breeding values, and e represents residual effects. The matrices X and Z connect the observations to the corresponding fixed and genetic effects. Fixed effects may include herd, year, season, age group, sex, or other systematic factors, depending on the trait and evaluation system. The additive genetic effects represent the inherited differences among animals, while residual effects account for variation not explained by the model.
- The term “best” in BLUP refers to obtaining predictions with minimum prediction error variance among linear unbiased predictors under the specified model and assumptions. “Linear” means that the predictions are linear functions of the observed data. “Unbiased” means that the expected predictions are correct across repeated samples under the model assumptions, rather than systematically overestimating or underestimating the target effects. “Prediction” refers to estimating random effects, such as breeding values, rather than estimating only fixed parameters. The method uses information about genetic relationships and the variance components of the trait to determine how much weight should be given to different sources of evidence.
- A central component of BLUP-based genetic evaluation is the relationship matrix. In pedigree-based animal models, the numerator relationship matrix, usually denoted by A, describes the expected additive genetic relationships among animals based on their ancestry. Parent-offspring pairs have an expected relationship of 0.5 in a standard non-inbred population with unrelated founders, and full siblings also have an expected relationship of 0.5. Half-siblings have an expected relationship of 0.25. These values describe expected relationships, not the exact proportion of DNA shared by every individual pair. The relationship matrix allows records from parents, siblings, progeny, and other relatives to contribute to the evaluation of an animal.
- The animal model is one of the most widely used forms of BLUP in animal breeding. It treats the additive genetic effect of each animal as a random effect and uses pedigree relationships to estimate breeding values across generations. A record from one animal can provide information about its relatives, while records from several relatives can contribute to the evaluation of an animal without direct performance data. Animal models are particularly useful in large breeding populations because they allow genetic evaluations to incorporate records from multiple generations and account for relationships throughout the population. Depending on the trait, models may also include maternal genetic effects, permanent environmental effects, common litter effects, or other relevant sources of variation.
- BLUP-based evaluation can be applied to a wide range of livestock species and traits. In dairy cattle, it is commonly used to evaluate milk yield, fat and protein production, fertility, udder health, and longevity. In beef cattle, it may be used for growth rate, birth weight, calving ease, carcass traits, and feed efficiency. In sheep and goats, BLUP can support the evaluation of wool quality, milk production, litter size, growth, and parasite resistance. Pig breeding programs use BLUP for traits such as daily weight gain, feed conversion, litter performance, and meat quality. Poultry breeding programs may apply related statistical approaches to egg production, growth, feed efficiency, and survival. The specific model must be appropriate for the species, trait, data structure, and breeding objective.
- The quality of BLUP evaluations depends heavily on the quantity and quality of available data. Accurate animal identification, complete pedigree records, consistent trait definitions, and reliable performance measurements are essential. Missing records, incorrect parentage, measurement errors, and unrecorded management differences can reduce the usefulness of the evaluation. Data should be collected under standardized procedures, and important environmental effects should be recorded so that they can be incorporated into the model. Large datasets are not automatically better if they contain substantial errors or inconsistent measurements. Reliable records and appropriate statistical assumptions are crucial for obtaining meaningful breeding value estimates.
- Heritability and genetic variance also influence BLUP-based evaluation. Heritability describes the proportion of phenotypic variance attributable to additive genetic variance in a particular population and environment. When a trait has relatively high heritability, an animal’s own performance may provide substantial information about its breeding value. When heritability is low, environmental influences account for a larger share of the observed variation, and records from relatives or progeny may be particularly important. The accuracy of an evaluation also depends on the number of informative records, the relationships among animals, the structure of the population, and the suitability of the statistical model. Heritability alone does not determine how reliable an individual breeding value will be.
- Accuracy and reliability are important measures for interpreting BLUP results. Accuracy describes the correlation between an estimated breeding value and the true breeding value, while reliability is commonly expressed as the square of accuracy. For example, an accuracy of 0.80 corresponds to a reliability of 0.64, or 64%. An animal with a high-reliability EBV generally has a more stable genetic estimate than an animal whose EBV is based on limited information. However, reliability is not the same as the probability that an animal will produce a particular result. Even highly reliable breeding values describe expected genetic differences, and offspring performance remains subject to inheritance and environmental variation.
- BLUP-based evaluation is particularly useful for comparing animals across different herds and management conditions. Mixed models can account for systematic differences among contemporary groups, such as animals raised in the same herd during the same year and season. This helps prevent environmental advantages from being incorrectly attributed to genetics. However, fair comparisons require appropriate connections among herds and groups through shared sires, relatives, or other genetic links. If groups are completely disconnected, or if important environmental effects are not recorded, the ability to compare animals across those groups may be limited. The quality of the evaluation therefore depends not only on the statistical method but also on the design and connectedness of the recording system.
- Genomic information has expanded the application of BLUP in modern breeding programs. Traditional pedigree-based BLUP uses expected relationships derived from ancestry, whereas genomic approaches can use DNA marker data to estimate relationships and predict genetic merit. Genomic BLUP methods use genomic relationship information in the evaluation process, while single-step genomic BLUP combines pedigree, genomic, and phenotypic information in a unified framework. These methods can improve the evaluation of young animals and allow DNA information to contribute alongside performance records and pedigree relationships. Their success depends on accurate genotyping, suitable reference populations, appropriate statistical models, and consistent performance recording.
- Progeny information remains valuable in BLUP-based evaluation, particularly for traits that cannot be measured directly in breeding animals or that require observations from offspring. For example, a dairy bull’s breeding value for milk production can be estimated using the milk records of his daughters. The number of daughters, the reliability of their records, and the environments in which they produce milk influence the information available. BLUP combines these records with pedigree relationships and other relevant information to estimate the bull’s genetic merit. Although genomic selection has made earlier evaluation possible for many traits, progeny data continue to be useful for validating predictions and improving genetic evaluations across generations.
- BLUP can also be adapted to traits with different biological and statistical characteristics. Standard linear mixed models are suitable for many continuously measured traits, including body weight and milk production. Threshold models may be used for categorical traits such as calving difficulty or certain disease outcomes. Random regression models are useful for repeated measurements over time, such as milk production throughout lactation or growth during development. Survival models may be used for longevity or time-to-event traits. These approaches extend the principles of mixed-model genetic evaluation to different data structures. Selecting an appropriate model helps ensure that predictions reflect the biology of the trait and the way observations are collected.
- One important limitation of BLUP is that its performance depends on the assumptions and structure of the statistical model. If relevant environmental effects are omitted, genetic parameters are poorly estimated, or data are systematically biased, breeding value predictions may be less accurate. BLUP does not automatically correct errors in animal identification, unreliable measurements, or incomplete recording. It also cannot guarantee that genetic rankings will remain identical across different environments when genotype-by-environment interactions are important. Breeding organizations should therefore validate evaluation models, monitor genetic trends, review data quality, and update evaluations as new information becomes available.
- The practical use of BLUP-based evaluation begins with defining the breeding objective and identifying the traits that matter for production, animal health, welfare, and economic performance. Relevant performance and pedigree records are then collected and checked for quality. Statistical models are fitted to estimate breeding values, and animals are ranked according to their genetic merit for the traits of interest. Breeders can use these estimates to choose parents for the next generation, design mating plans, and monitor progress over time. Selection indexes can combine multiple EBVs into an overall measure of merit, helping breeders balance production, fertility, disease resistance, longevity, and other objectives rather than selecting solely for one trait.
- BLUP also contributes to the management of genetic diversity. Selecting only a small number of top-ranked animals may accelerate improvement for specific traits in the short term, but it can increase inbreeding and reduce genetic variation. Pedigree and genomic relationship information can be used to identify related animals and design matings that balance expected genetic gain with the control of inbreeding. Breeding programs should consider both the genetic merit of individual animals and their contribution to the diversity of the population. This is particularly important for small breeds and populations in which the loss of genetic variation may limit future improvement.
- In conclusion, BLUP-based evaluation is a cornerstone of modern animal breeding because it provides a systematic method for estimating breeding values from performance records, pedigree relationships, and, in advanced systems, genomic data. By accounting for relevant environmental effects and genetic relationships, BLUP improves the fairness and consistency of animal comparisons. Its effectiveness depends on reliable data, suitable statistical models, appropriate genetic parameters, and regular validation. When integrated with balanced breeding objectives, genomic technologies, and responsible inbreeding management, BLUP-based evaluation helps breeders achieve sustainable genetic progress and improve the productivity, health, and adaptability of livestock populations.
