- Combined selection is a method of artificial selection in which information from multiple sources is used together to evaluate breeding animals and make selection decisions. Instead of relying only on an animal’s own phenotype, combined selection can integrate individual performance, family information, pedigree records, progeny performance, and, in modern breeding programs, genomic information. The central principle is that different sources of information provide complementary evidence about an animal’s underlying genetic merit, and combining them can increase selection accuracy and improve the efficiency of genetic improvement.
- The need for combined selection arises because no single source of information provides a complete measure of an animal’s breeding value. An individual’s phenotype is influenced by both genetic and environmental effects, while family performance provides information from genetically related animals. Progeny records provide evidence about genes transmitted by a parent, and genomic information provides direct information about patterns of DNA variation. By combining these sources appropriately, breeders can obtain a more reliable estimate of breeding value than would usually be possible from any one source alone.
- A simple representation of phenotype is P = G + E, where P is the observed phenotype, G represents genetic effects, and E represents environmental effects. The objective of selection is generally not to select animals with the highest observed phenotype alone but to identify animals with high genetic merit that can transmit favorable alleles to future generations. Combined selection provides a framework for using several forms of evidence to estimate that genetic merit.
- The genetic component of a quantitative trait includes additive genetic effects, dominance effects, and epistatic effects. Additive genetic effects are particularly important in selection because they contribute to the animal’s breeding value and can be transmitted predictably to offspring. Combined selection is therefore designed primarily to improve the accuracy with which additive genetic merit is estimated.
- One of the simplest forms of combined selection combines an animal’s own performance with the performance of its relatives. For example, an animal may have an individual record for growth while its parents, full-sibs, half-sibs, or other relatives also have growth records. Instead of choosing the animal solely on its own phenotype, the breeder considers the complete body of family information.
- This approach is useful because an individual record may be strongly influenced by environmental conditions. If an animal performs exceptionally well because it received better nutrition or experienced fewer health problems, its phenotype may overestimate its genetic merit. If its relatives also perform well under comparable conditions, confidence in its genetic superiority increases. Conversely, poor individual performance may be interpreted differently if family members consistently show high genetic performance.
- Heritability influences the value of individual performance. Narrow-sense heritability can be expressed as:
- h² = σ²_A / σ²_P
- where σ²_A is additive genetic variance and σ²_P is phenotypic variance. For traits with high heritability, individual performance can provide substantial information about breeding value. For traits with low heritability, environmental effects may obscure genetic differences, making family, progeny, or genomic information more valuable.
- Combined selection therefore allows different information sources to compensate for one another. If individual performance is weakly informative because of low heritability, family and progeny records may increase accuracy. If family information is limited, individual and genomic records may provide additional evidence. The exact weighting of each information source depends on the genetic and statistical properties of the trait.
- Selection accuracy is a central concept in combined selection. It represents the correlation between an animal’s estimated genetic merit and its true breeding value. Increasing selection accuracy generally increases the expected response to selection because breeders are more likely to choose animals that genuinely possess superior genetic merit.
- The simple response to selection can be expressed as:
- R = h² × S
- where R is selection response and S is the selection differential. This equation is useful for understanding the basic relationship between heritability and response, although modern combined selection involves multiple sources of information and is usually evaluated using more comprehensive statistical models.
- The expected rate of genetic improvement can be expressed approximately as:
- ΔG/year = i × r × σ_A / L
- where i is selection intensity, r is selection accuracy, σ_A is additive genetic standard deviation, and L is generation interval. Combined selection can increase genetic progress primarily by improving selection accuracy. It may also allow earlier selection and thereby contribute to a shorter generation interval.
- One important form of combined selection is the integration of individual and family information. Individual records provide direct evidence about the candidate’s phenotype, while family records provide information about related genetic backgrounds. The combination can be more informative than either source alone, provided that the records are appropriately weighted.
- For example, consider a group of full-sib animals with different growth rates. One animal has an excellent individual growth record, while its siblings have average performance. Another animal has a moderate individual record but comes from a family with consistently high performance. Combined selection can consider both pieces of information rather than making the decision based on only one measurement.
- Within-family selection can also be incorporated. Individuals can be evaluated relative to their family members while their overall family performance is considered simultaneously. This allows breeders to capture both between-family and within-family genetic differences.
- This is important because the best individual within a family is not necessarily the best individual in the entire population. Similarly, the best family may contain individuals that are genetically inferior to some individuals from other families. Combined selection provides a framework for integrating these levels of information.
- Pedigree information is another important component. Pedigree records establish relationships among animals and allow information from relatives to contribute to genetic evaluation. Parents, grandparents, full-sibs, half-sibs, offspring, and more distant relatives can all provide information about the expected genetic merit of an animal.
- Traditional pedigree-based genetic evaluation commonly uses an additive relationship matrix, known as the A matrix. This matrix describes expected additive genetic relationships based on pedigree. Combined selection can use these relationships to determine how much information from each relative should contribute to the evaluation of a candidate.
- Pedigree information becomes particularly useful when animals have limited individual records. Young animals, for example, may have little or no information about traits that are expressed later in life. Their parents and relatives may already have extensive records, allowing the young animal to receive an estimated breeding value before its own performance is available.
- Progeny information provides another major source of evidence. An animal’s offspring receive genes from that animal, so the performance of its progeny can provide information about the genetic effects transmitted by the parent. This is the basis of progeny testing.
- Combined selection can therefore integrate individual performance, family records, and progeny records. This is especially valuable for traits that are sex-limited, late-expressed, difficult to measure, or strongly influenced by environmental conditions.
- For example, a young dairy bull cannot directly express milk production. His breeding value for milk traits can be estimated from his pedigree, genomic information, relatives, and eventually the performance of his daughters. A combined evaluation can therefore begin before progeny records exist and become more informative as additional data accumulate.
- Genomic information has transformed combined selection. Genomic selection uses large numbers of DNA markers distributed throughout the genome to predict genetic merit. Genomic information can be combined with individual phenotypes, pedigree records, family performance, and progeny information to produce Genomic Estimated Breeding Values (GEBVs).
- The genomic relationship matrix, commonly called the G matrix, describes realized genetic similarity among animals based on genomic markers. This can complement the pedigree-based A matrix. Two animals with the same expected pedigree relationship may have different realized genomic relationships because they inherit different chromosome segments.
- This is especially important for within-family selection. Full-sibs have the same parents but inherit different combinations of parental DNA. Genomic information can therefore help distinguish genetic differences among siblings and improve the accuracy of selection within families.
- Combined selection is also closely related to BLUP, or Best Linear Unbiased Prediction. BLUP animal models can combine records from the individual, relatives, and progeny while accounting for systematic environmental effects. This makes BLUP one of the principal statistical tools for implementing combined selection in modern animal breeding.
- Rather than calculating simple averages, an animal model estimates breeding values by considering the complete relationship structure of the population. The model can account for herd, flock, farm, year, season, sex, age, management group, parity, and other systematic environmental effects. This helps separate genetic differences from environmental variation.
- Modern evaluations can also include maternal effects, permanent environmental effects, and common environmental effects. These are particularly important for traits such as birth weight, early growth, survival, and weaning weight. Without accounting for these effects, family or individual performance can be incorrectly interpreted as evidence of direct genetic merit.
- Maternal effects occur when the dam influences offspring performance through both genetic and environmental pathways. For example, milk production, maternal behavior, uterine environment, and prenatal nutrition can influence offspring growth. Combined selection models can distinguish the direct genetic effect of the offspring from maternal genetic and environmental effects when sufficient data are available.
- Common environmental effects can also influence groups of relatives. Animals raised in the same pen, litter, hatch, herd, or management group may share environmental conditions. Combined selection can account for these effects statistically rather than assuming that all similarities among relatives are genetic.
- The method is especially valuable for low-heritability traits. When individual phenotypes contain substantial environmental noise, information from many relatives can improve breeding value prediction. Health, fertility, survival, disease resistance, longevity, and behavioural traits may benefit from this approach.
- For disease resistance, combined selection can use disease records from the individual and relatives, pedigree information, genomic markers, and other health indicators. This can improve selection accuracy without requiring every candidate to undergo direct disease challenge testing.
- Disease traits may be recorded as binary outcomes, such as affected versus unaffected. These traits can often be analyzed using threshold models, which assume an underlying continuous liability even though the observed phenotype is categorical. Combined family and genomic information can contribute to estimating this underlying genetic liability.
- Selection for disease resistance should also be distinguished from selection for disease tolerance. Resistance relates to limiting infection or pathogen burden, whereas tolerance concerns maintaining health or performance despite infection. These traits may have different genetic architectures and genetic correlations, so combined selection should be based on clearly defined breeding objectives.
- Combined selection is equally relevant to reproductive traits. Fertility, conception rate, pregnancy success, litter size, embryo survival, age at sexual maturity, reproductive longevity, and maternal ability can all be difficult to evaluate early in life. Information from relatives, progeny, and genomic data can allow selection decisions to be made before the candidate has complete reproductive records.
- For sex-limited traits, combined selection is particularly powerful. Male animals can be evaluated using records from female relatives and offspring, while females can benefit from information from male relatives for traits expressed primarily in males. This allows genetic selection for traits that cannot be directly measured in every selection candidate.
- Combined selection is also useful for production traits such as milk production, meat production, egg production, wool production, fiber quality, growth, feed efficiency, and carcass characteristics. Individual records can be combined with family and progeny information, allowing breeding values to be predicted even when candidates have incomplete records.
- For feed efficiency, for example, individual measurement may be expensive because animals must undergo controlled feed intake testing. Family information can provide additional evidence about genetic merit, reducing the need to directly measure every candidate. Genomic information can further increase the accuracy of early selection.
- Combined selection can also support improvement of animal welfare and behavioural traits. Temperament, stress response, handling behavior, aggression, resilience, injury risk, and other welfare-related traits can be influenced by genetics and environment. Combining individual observations with family and genomic information can improve genetic evaluation while reducing reliance on a single subjective measurement.
- The method also has applications in adaptation and climate resilience. Traits such as heat tolerance, disease resilience, survival under challenging conditions, and performance in low-input environments can be evaluated using records collected from relatives and across environments. Genotype–environment interaction (G×E) should be considered because an animal’s genetic performance may change across different production systems or climatic conditions.
- Combined selection becomes particularly important in multi-trait breeding programs. Selecting for a single trait can cause undesirable correlated responses because traits may be genetically correlated. For example, selection for rapid growth may affect mature size, feed intake, fertility, longevity, or metabolic health.
- Genetic correlation describes the extent to which genetic effects influencing two traits are associated. A breeding program therefore needs to consider several traits simultaneously when defining selection objectives.
- A selection index can combine information from multiple traits and assign appropriate economic or biological weights to each. For example, a breeding objective may include production, fertility, disease resistance, survival, welfare, feed efficiency, and adaptation. Combined selection can provide information for each trait and integrate it into an overall selection criterion.
- The distinction between a selection criterion and a breeding objective is important. The breeding objective defines the traits that the breeding program ultimately wants to improve, while the selection criterion consists of the information used to predict genetic merit for those traits. Combined selection improves the selection criterion by integrating multiple information sources.
- Combined selection can also improve the management of genetic diversity. When breeding decisions are based only on the highest individual phenotype or EBV, a small number of animals may become disproportionately represented. This can increase genetic concentration and reduce effective population size.
- A simplified relationship between the rate of increase in inbreeding and effective population size is:
- ΔF ≈ 1 / (2Ne)
- where ΔF is the expected increase in inbreeding per generation and Ne is effective population size. Although simplified, this relationship illustrates why genetic contribution must be considered alongside genetic merit.
- The popular sire effect is one example of how high genetic merit can become concentrated. If a small number of elite males are used extensively through artificial insemination or other reproductive technologies, their genes can rapidly dominate the population. Combined selection should therefore be integrated with population-management strategies.
- Mate allocation can help balance these objectives. Once animals with high estimated breeding values have been identified, mating plans can be designed to minimize excessive relatedness while retaining desirable genetic combinations.
- Expected offspring inbreeding can be estimated from parental relatedness. If φ(sire, dam) represents kinship or coancestry:
- E(F_offspring) = φ(sire, dam)
- If r(sire, dam) represents the additive genetic relationship:
- E(F_offspring) = r(sire, dam) / 2
- These relationships allow breeders to incorporate expected inbreeding into mating decisions.
- Optimal Contribution Selection (OCS) extends this principle by optimizing the genetic contribution of selected animals to the next generation. OCS can balance expected genetic gain against the increase in relatedness and inbreeding. This is especially important in small, closed, or highly selected populations.
- Combined selection should also consider genetic load and potentially harmful genetic variants. Selection for desirable traits can unintentionally increase the frequency of harmful alleles if those variants are linked to favorable traits or remain undetected. Genetic testing and genomic information can help identify known deleterious variants and support more responsible breeding decisions.
- Heterozygosity and genetic diversity are important components of long-term breeding sustainability. A breeding program should not simply maximize the genetic merit of the current generation. It should preserve sufficient genetic variation to allow future improvement and adaptation.
- Combined selection is therefore compatible with conservation breeding and genetic-resource management. In small populations, breeders can use pedigree and genomic information to identify genetically valuable animals while controlling relatedness and maintaining representation of diverse lineages.
- The method can also help manage genetic concentration. If several sources of information are considered, breeders can identify animals with high genetic merit that are not already heavily represented in the population. This provides opportunities to increase genetic gain without relying exclusively on a small number of popular families or sires.
- Combined selection can also be applied across generations. As additional records accumulate, the estimated breeding value of an animal may change. A young animal may initially receive an evaluation based largely on pedigree and genomic information. As its own records become available, individual performance contributes additional evidence. Later, the performance of its offspring may further update the estimate.
- This dynamic updating is one of the strengths of modern genetic evaluation. Breeding values are predictions, not permanent labels. They should be updated as new information becomes available and as the reference population and breeding objectives change.
- Combined selection also depends heavily on data quality. Incorrect pedigree information, inconsistent phenotyping, missing records, inaccurate measurements, and systematic differences between management groups can reduce the reliability of genetic evaluations. Accurate identification and standardized recording are therefore essential.
- Genomic parentage verification can help correct pedigree errors and improve the accuracy of relationship information. This is particularly valuable when large breeding populations contain complex family structures or when natural mating makes parentage uncertain.
- Phenotypic recording remains essential even in genomic breeding programs. Genomic prediction depends on reference populations containing animals with both genotype and high-quality phenotype records. Combined selection therefore does not make traditional phenotyping obsolete. Instead, it increases the value of well-designed phenotypic recording.
- The relative importance of each information source depends on the trait and breeding population. For a highly heritable trait with accurate individual measurement, the individual’s own performance may receive substantial weight. For a low-heritability trait, family and progeny records may become more important. For a young animal without a phenotype, genomic and pedigree information may dominate the evaluation.
- The amount of information available also affects selection decisions. An animal with many accurately recorded relatives and progeny may have a more reliable breeding value than an animal with only a single individual record. Statistical models account for this difference in information quantity and quality.
- Combined selection is particularly useful when generation interval is important. Genomic and family information can allow young animals to be evaluated before they have complete lifetime records. Earlier selection can reduce generation interval and potentially increase the annual rate of genetic gain.
- However, faster genetic gain is not always desirable if it is achieved through excessive inbreeding or loss of diversity. The breeding objective should therefore include long-term sustainability. Genetic gain, fertility, health, survival, welfare, adaptation, and diversity should be considered together.
- The main advantage of combined selection is therefore not simply that it uses more data. Its greater value comes from combining complementary sources of information in a statistically appropriate way. Individual records reveal how an animal performs; family records provide information from relatives; progeny records reveal transmitted genetic effects; pedigree records describe expected relationships; and genomic information reveals realized genetic relationships.
- The main limitation is that combining information does not automatically improve accuracy. Poor-quality or strongly biased data can reduce the reliability of genetic evaluation. Environmental confounding, incorrect pedigrees, small reference populations, measurement errors, and inappropriate statistical assumptions can all reduce the benefits of combined selection.
- Another limitation is that highly correlated information sources do not provide completely independent evidence. Ten records from closely related animals do not necessarily contain ten times as much information as one record. Modern quantitative genetic models account for relationships and correlations when weighting records.
- Combined selection also requires appropriate computational and statistical infrastructure. Large-scale animal breeding programs may need databases, pedigree systems, genomic laboratories, genetic evaluation software, and trained quantitative geneticists. However, many of these technologies are increasingly accessible to commercial breeding organizations and research programs.
- The method can be implemented at different levels of complexity. A small breeding program may combine individual and family records using relatively simple methods. Larger programs can use BLUP animal models, genomic prediction, selection indexes, and optimization algorithms to integrate thousands or millions of records.
- Combined selection is therefore not a single fixed statistical formula. It is a general breeding strategy in which multiple sources of information are used together to improve genetic decisions. The exact methodology depends on the species, trait, population size, recording system, reproductive system, and breeding objective.
- The approach is particularly valuable when the breeding program seeks simultaneous improvement in multiple economically and biologically important traits. A modern selection program may combine production records, growth measurements, fertility records, health information, behavioural observations, pedigree, genomic data, and progeny performance within one genetic evaluation system.
- The ultimate objective is to predict which animals are most valuable as parents of the next generation. The best breeding animal is not necessarily the animal with the highest phenotype for one trait. It is the animal whose overall genetic merit is favorable for the breeding objective and whose use is compatible with genetic diversity, health, fertility, welfare, and long-term population sustainability.
- Combined selection therefore represents an important transition from simple selection methods toward modern data-integrated animal breeding. It builds upon individual selection, family selection, within-family selection, and progeny testing, while incorporating pedigree and genomic information to produce increasingly accurate genetic evaluations.
- When combined with BLUP, genomic selection, selection indexes, genetic testing, mate allocation, and optimal contribution selection, it can increase the accuracy and efficiency of genetic improvement while reducing some of the risks associated with excessive selection intensity and genetic concentration.
- Ultimately, successful combined selection requires a balance between information, genetic gain, and population management. Breeders should use the best available evidence to identify animals with high breeding value, but they should also maintain genetic diversity, manage inbreeding, protect fertility and health, and consider future environmental challenges. In this integrated form, combined selection becomes a powerful foundation for efficient, accurate, and sustainable animal breeding.
