The concept of covariance measures how two random variables vary together, and it is the statistical glue that connects relatives, traits, and environments in classical theory. Within a single population, the covariance between relatives arises because they share alleles that are identical by descent. But when an individual comes from one biparental cross, say an indica rice cross A/B, and another from a completely different cross, such as a japonica cross C/D, there is no established framework for asking how much their genotypic values co-vary. Rebollo and Bernardo propose an elegant answer: shared marker alleles between individuals belonging to two different populations can serve as the basis for defining and estimating a between-population covariance. In other words, even when two populations share no recent ancestry, the molecular markers they have in common provide a measurable bridge for quantifying joint genetic variation.

The analytical core of the method extends earlier formulas developed by Osthushenrich and colleagues in 2017, which estimated the genetic variance within a recombinant inbred line population from marker effects. The authors define the genotypic value of an individual as a random variable whose variance can be decomposed across chromosomes and loci. Under a purely additive genetic model, the heterozygote effect is zero, and the probability that a locus carries the alternative or reference homozygote depends on allele frequencies and the inbreeding coefficient F, which ranges from zero in an F2 population to one in fully inbred doubled haploids. Crucially, the covariance between loci depends on the linkage disequilibrium coefficient D, which in turn is a function of the effective recombination rate between markers. Working through the algebra with allele frequencies of one half, the authors arrive at a strikingly compact result: the genotypic variance equals one plus F multiplied by a sum over chromosome pairs of twice the linkage disequilibrium coefficient times the product of the two marker effects.

This compact formula carries a subtle but important feature: linkage phase is captured by the signs of the marker effects themselves, with same signs indicating coupling and opposite signs indicating repulsion, rather than being embedded in the disequilibrium coefficient as in the classical formulation of Lynch and Walsh. The authors verified that their expression reduces to the classical result when F equals zero, and they checked the algebra with small-scale numerical examples. From the variance formula, the covariance between two populations follows from a simple identity: the variance of the sum of two populations’ genotypic values equals the sum of their individual variances plus twice their covariance. By computing the variance of the combined population using summed marker effects, and subtracting the two individual variances, the between-population covariance drops out analytically.

One practical obstacle is that the method nominally requires a linkage map to compute recombination rates between markers. Maps can be inaccurate, based on the wrong population, inflated by generations of random mating, or simply unavailable when only physical positions are known. The authors provide a workaround: for fully inbred populations and for F2 populations, the disequilibrium coefficient can be estimated directly from observed marker and haplotype frequencies as the absolute difference between the observed frequency of double-minor-allele homozygous haplotypes and the product of the individual allele frequencies. When phasing information is unavailable and the population is not fully inbred, published approximations from Ragsdale and Gravel can supply an unbiased estimate. This flexibility means the framework can be applied even when map resources are poor, a meaningful advantage for orphan crops and under-resourced breeding programs.

To validate the approach, the authors simulated four nested populations of one thousand individuals each, built so that each successive population segregated for only half as many quantitative trait loci as the one before. Populations were generated at three inbreeding levels, F2, F3, and doubled haploid, using the AlphaSimR simulation package with maize-like genome parameters: ten chromosomes of two Morgans each, carrying twenty quantitative trait loci apiece. Across one thousand independent simulation repeats, the analytical estimates tracked the true variances and covariances with remarkable fidelity. The correlation between observed and estimated variance exceeded 0.8 in every repeat, with a mean of 0.99, whether the known linkage map or the estimated disequilibrium coefficients were used. As theory predicted, variances and covariances in F2 populations were half those of doubled haploids, and F3 populations fell at seventy-five percent, confirming that the framework behaves correctly across the full range of inbreeding.

The real-world test came from a commercial maize breeding program at AgReliant Genetics. Four biparental populations of doubled haploid lines, each crossed to two of four tester inbreds, were evaluated in multi-environment field trials across four to ten locations in the U.S. Corn Belt over three years. Five traits were recorded: grain yield, grain moisture at harvest, plant height, ear height, and test weight. Marker effects for each population were estimated separately using ridge regression best linear unbiased prediction, and the authors discovered that the choice of shrinkage factor mattered enormously. When shrinkage was based on the raw number of markers, analytical variance estimates were consistently biased downward; when based on the number of fifty-centimorgan chromosome segments, they were biased upward. Using the effective number of independent tests, calculated from the eigenvalues of the marker correlation matrix following Li and Ji, yielded near-unbiased estimates that correlated strongly, above 0.97, with classical restricted maximum likelihood estimates.

The estimated covariances between maize populations revealed trait-specific and population-specific structure. For grain yield, covariances between population pairs ranged from negative values up to 9.52 square tonnes per hectare, with a mean of 1.67. The correlations between marker effects across populations were generally low to moderate, averaging 0.10 for yield, and the highest covariances appeared, unsurprisingly, between the same population tested with different testers. The lowest covariances occurred between populations sharing no parents. Notably, the correlation between the estimated genetic covariance and the marker-effect correlation was significant for most traits, reaching 0.77 for test weight, underscoring that between-population covariance reflects both relatedness and the genetic architecture of each trait. The authors also showed that their approach scales gracefully: where a joint multitrait analysis of fifty populations with five thousand markers would require estimating 250,000 marker effects and 1,275 covariance parameters simultaneously, their method decomposes the problem into fifty independent analyses.

The implications stretch well beyond the maize field. In breeding, heterogeneous genetic covariances could be folded into prediction models to improve accuracy when training and target populations differ, a chronic challenge in genomic selection. In quantitative trait locus mapping, the framework may allow genetically related populations to be pooled into larger discovery panels, boosting the power and resolution of gene detection. Beyond agriculture, the authors point to ecology and evolutionary genetics, where quantifying genetic covariance between subpopulations could illuminate hybrid zones, inform conservation of genetic diversity, and sharpen studies of population structure and adaptation. The method does carry limitations: it assumes a uniform inbreeding level within each population, and it requires marker effects, which are population-specific and depend on phenotypic data that may not yet exist for untested or uncreated populations.

What makes the study conceptually significant is that it fills a gap that has persisted since the founding of quantitative genetics. Covariance within populations has been understood since Fisher; covariance between traits and environments since the 1930s and 1950s. But the covariance between populations, the quantity that governs how much genetic information flows across breeding program boundaries, subspecies divides, and heterotic group walls, had no formal definition or estimator. By grounding between-population covariance in shared marker alleles and validating it against both simulated truth and commercial field data, Rebollo and Bernardo have given geneticists a new measurement tool that is analytically tractable, computationally scalable, and applicable at any level of inbreeding. As genomic datasets continue to accumulate across crops, livestock, and wild species, a framework for measuring how populations co-vary genetically may prove as foundational as the one Fisher built for measuring how relatives do.

Subject of Research: Estimation of genetic covariance between plant biparental populations using molecular marker effects

Article Title: Genetic covariance between plant biparental populations: concept and estimation

Article References: Rebollo, I., & Bernardo, R. (2026). Genetic covariance between plant biparental populations: concept and estimation. Heredity. https://doi.org/10.1038/s41437-026-00887-w

Image Credits: AI Generated

DOI: 10.1038/s41437-026-00887-w

Keywords: genetic covariance, quantitative genetics, plant breeding, biparental populations, marker effects, linkage disequilibrium, maize, genomic prediction, inbreeding coefficient, doubled haploids, RR-BLUP, Heredity

The research team, led by Roberto Fritsche-Neto of North Carolina State University together with colleagues in Brazil, used stochastic computer simulations to model a rice breeding program over 30 years. Rather than relying on historical field data, which reflects past conditions and cannot easily be re-run under alternative strategies, the researchers built a virtual breeding pipeline in which they could systematically vary the selection method while holding everything else constant. The simulation was anchored in the real-world parameters of the Louisiana State University AgCenter rice breeding program and conducted with the AlphaSimR package, giving the results a practical grounding that purely theoretical exercises often lack.

The simulated program targeted three traits simultaneously: increasing grain yield, reducing the chalkiness rate that degrades rice grain quality, and keeping plant height stable to avoid lodging. Grain yield received the heaviest weight, at 70 percent of the selection objective, while chalkiness and plant height each received 15 percent. The team compared seven selection pipelines built around three classical indices: an empirical index with fixed weights, the Smith–Hazel index that uses genetic and phenotypic covariance matrices, and the Pesek–Baker index, which is constructed from desired gains rather than economic weights. Each index was tested under traditional phenotypic selection and under genomic selection, with the genomic component applied either before index construction or after the index had already been calculated from phenotypes.

The technical distinction between those two architectures turned out to be central. In the first approach, univariate genomic prediction models were fitted separately for each trait using ridge regression best linear unbiased prediction, generating genomic estimated breeding values for yield, chalkiness, and height individually before combining them into an index. In the second, the multi-trait index was first computed from phenotypic data and then treated as a single synthetic trait for genomic prediction. The researchers also refreshed the training population each cycle using a grandparents-parents-offspring strategy, one of the most effective approaches for maintaining prediction accuracy across recurrent cycles, and ran 100 independent replicates of every scenario to capture the stochastic uncertainty inherent in real breeding.

When the 30-year trajectories were compared, no single strategy dominated every trait, but a clear pattern emerged. Applying genomic selection directly to a phenotypically constructed Pesek–Baker index produced the highest gains for grain yield, reaching values close to 10.3 on the standardized scale. When genomic prediction was performed trait by trait before index construction, the Smith–Hazel index delivered the best yield response, near 10.0. For chalkiness, however, the trait-specific prediction approach paired with the Pesek–Baker index was the clear winner, driving the population mean down to roughly negative 4.0, the most favorable reduction observed, compared with only about negative 1.6 under the traditional empirical index.

Plant height told a subtler story. Because the breeding objective was stability rather than directional change, strategies that pushed height upward were actually performing poorly against the stated goal. The Smith–Hazel index combined with genomic selection produced the largest increases in plant height, which ran counter to the program’s intent. The Pesek–Baker index, whether used traditionally or with trait-specific genomic prediction, kept height changes moderate, plateauing near 3.5, while the empirical index was the most conservative of all. Taken together, the results underscore a principle that breeders know intuitively but rarely test so rigorously: the best pipeline is the one that matches the full set of objectives, not the one that maximizes a single headline trait.

The dynamics of selection accuracy and additive genetic variance added a deeper layer to the findings. As expected in recurrent selection, accuracy generally declined over cycles as favorable alleles fixed and the population became more uniform. But the rate and pattern of that decline depended sharply on how genomic information was deployed. When genomic prediction was applied directly to an already aggregated index, accuracy for chalkiness and plant height collapsed dramatically, approaching zero within the first breeding cycle before a partial recovery. The authors attribute this to information compression: folding three traits with different heritabilities, correlations, and selection directions into one value before prediction strips away the trait-specific genetic signals the model needs to learn.

In contrast, estimating genomic values for each trait separately before building the index preserved that information and produced responses more consistent with the multi-trait objectives, with gentler accuracy losses and greater stability after the initial cycles. Additive genetic variance told a parallel story, with genomic scenarios generally depleting variability faster than traditional ones, and the greatest losses appearing in the Smith–Hazel and index-first genomic pipelines. The researchers caution that residual variance alone is not proof of future gain, but the joint pattern of eroding accuracy and shrinking variance in the index-first scenarios points to a genuine methodological limitation rather than simple exhaustion of genetic variation.

The practical implications reach well beyond rice. The study’s authors argue that genomic selection should not be viewed as a wholesale replacement for phenotypic selection but as a tool whose value depends on where it is inserted in the pipeline. For programs juggling traits that must move in opposite directions, such as raising yield while lowering chalkiness, the trait-specific prediction route followed by index construction appears the more reliable architecture. The Pesek–Baker index, built explicitly around desired gains, proved the most faithful vehicle for translating a breeder’s ideotype into a selection criterion, delivering strong yield gains, the deepest chalkiness reduction, and minimal unwanted shifts in plant height. Similar logic applies in crops like malting barley, where quality traits must be held within narrow thresholds rather than pushed to extremes.

The authors acknowledge limitations that shape how far these conclusions can be generalized. The simulations used a fixed marker density of 540 SNPs, and prediction accuracy depends on linkage disequilibrium between markers and causal loci, so different genotyping platforms could shift the magnitude of the observed responses. The genetic architecture was purely additive, with no dominance or epistasis, and only three traits were modeled. Future work, the team notes, should expand the trait set and validate the findings in real breeding populations. Still, for a field racing to feed a growing population under a changing climate, the message is concrete and actionable: in multi-trait breeding, the sequence in which genomic prediction and selection indices are combined is not a technical footnote but a decision that can determine both the magnitude and the direction of genetic progress.

Subject of Research: Integration of genomic prediction with multi-trait selection indices in simulated rice breeding programs

Article Title: Combining genomic prediction and multi-trait indices through stochastic simulations: do index type and deployment order affect genetic gain?

Article References: Fritsche-Neto, R., Queiroz, L. G. C., Viana, J., Gupta, K., Grover, K., & DoVale, J. C. (2026). Combining genomic prediction and multi-trait indices through stochastic simulations: do index type and deployment order affect genetic gain?. Theoretical and Applied Genetics, 139(10), Article 284. https://doi.org/10.1007/s00122-026-05398-0

Image Credits: AI Generated

DOI: 10.1007/s00122-026-05398-0

Keywords: genomic selection, selection index, plant breeding, rice, stochastic simulation, Pesek–Baker index, Smith–Hazel index, genetic gain, grain yield, chalkiness, additive genetic variance, RR-BLUP