Tropical Maize Haploid Inducers Prove Season-Dependent in Multi-Year Field Trial

Doubled haploid technology is one of the most powerful shortcuts in modern maize breeding, allowing breeders to fix elite genetics in fully homozygous lines within a fraction of the generations required by conventional inbreeding. The engine of this technology is the haploid inducer, a specialized line that, when crossed as a pollinator onto a maternal plant, produces seeds in which the maternal genome has been lost. Yet most of the inducers that power breeding pipelines worldwide, such as the well-known temperate line BHI306, were developed under temperate conditions and may not perform reliably under the heat, humidity, and pest pressure of tropical savanna environments. A new open-access study published in BMC Plant Biology by Arnat Thawarorit of Khon Kaen University and colleagues, including Thomas Lübberstedt of Iowa State University, sets out to answer a deceptively simple question: can tropical-adapted inducer lines deliver stable, season-proof haploid induction performance?

The research team evaluated fourteen advanced F₅ tropical-derived haploid inducer lines alongside the temperate standard BHI306 across three distinct growing seasons at Khon Kaen University in Thailand: the 2023/24 dry season, the 2024/25 dry season, and the 2025 rainy season. Pacific789 served as the maternal tester for all crosses. The traits tracked included the putative haploid induction rate, or HIR, marker misclassification rates with and without the flat-embryo criterion, phenotypically confirmed haploid recovery, red root uniformity, pollen production capacity, and a suite of flowering and plant architecture traits. This design allowed the researchers to separate genuine genetic differences among inducer lines from the noise imposed by shifting seasonal conditions, a distinction that is critical for any breeding program hoping to deploy these lines commercially.

The statistical machinery behind the study was deliberately rigorous. The team applied fixed- and random-genotype REML models, the former for entry comparisons and the latter for estimating entry-mean broad-sense heritability, a measure of how much of the observed variation in a trait is genetically repeatable across environments. The results were striking in their unevenness. Entry-mean broad-sense heritability ranged from zero to 0.74 depending on the trait. Marker misclassification with the flat-embryo criterion, denoted MCR_F, posted the highest estimate at 0.74, with a 95 percent confidence interval spanning 0.30 to 0.89. The haploid induction rate and red root uniformity showed moderate point estimates of 0.59 and 0.60 respectively, but both carried wide confidence intervals that signal considerable uncertainty. Marker misclassification without the flat-embryo criterion, MCR_NF, hit a boundary estimate of zero, meaning essentially none of its variation could be attributed to stable genetic differences among lines.

That heritability pattern carries an important practical message for breeders. When a trait such as MCR_F shows high heritability, selecting inducer lines based on a single season of data can be reasonably reliable, because the ranking of lines is likely to hold up across future environments. Conversely, when a trait like MCR_NF shows a heritability estimate of zero, a line that appears to perform well in one season may be mediocre the next, and the apparent superiority is little more than environmental noise. The authors report that significant overall genotype effects were detected for HIR, MCR_F, red root uniformity, and tassel length, whereas MCR_NF, phenotypically confirmed haploid induction rates, and most agronomic traits showed no significant overall genotypic differentiation. In other words, some components of induction performance are genuinely heritable while others are not, and knowing which is which changes how selection should be conducted.

Perhaps the most consequential finding concerned genotype-by-season interaction, the statistical term for situations in which different genetic lines win and lose under different environmental conditions. Significant genotype-by-season variation occurred for most induction-related and agronomic traits measured in the study. This was not a marginal effect. The relative rankings of the inducer lines shifted from one season to the next, meaning that a line crowned the best performer in a dry-season trial could fall from favor in a rainy-season planting. For tropical breeding programs that may run crossing nurseries in multiple seasons per year, this instability translates directly into operational risk: a single-season evaluation could systematically favor lines that happen to suit one set of conditions rather than lines that will serve the breeding pipeline year-round.

To capture this instability quantitatively, the researchers examined genotype-level correlations among HIR, marker misclassification, phenotypically confirmed haploid recovery, and plant architecture traits. They found that the relationships among these traits varied substantially in both magnitude and direction from season to season, with several associations that appeared in one season simply failing to reproduce in another. This lack of consistency matters because breeders often rely on indirect selection, using one easily measured trait as a proxy for another that is expensive or slow to assess. If the correlation between, say, induction rate and confirmed haploid recovery flips between dry and rainy seasons, then a proxy strategy validated in one season may actively mislead decision-making in another. The study’s evidence suggests that no single trait measured once can substitute for a proper multi-season assessment.

The study also probed the reliability of the diagnostic markers used to identify haploid seeds. Two complementary systems are commonly employed in maize doubled haploid work: the R1-nj anthocyanin marker, which colors the embryo and endosperm of non-haploid kernels so that putative haploids can be culled visually, and the Pl-1 gene, which confers purple pigmentation in the roots of diploid seedlings, allowing haploids to be flagged after germination. The researchers assessed red root uniformity as a measure of how consistently the Pl-1 system expresses, and analyzed misclassification rates for the kernel-based marker both with and without requiring the embryo shape criterion. Notably, the team found that Pl-1 root expression showed weak and inconsistent relationships with R1-nj kernel misclassification, supporting their complementary use in haploid screening rather than reliance on either marker system alone. For programs running large-scale haploid identification, this argues for keeping both diagnostic layers in place.

Reproductive performance provided one area where the tropical-derived material clearly outshone its temperate benchmark. The tropical-derived lines showed greater pollen-shedding capacity than BHI306, a trait with direct operational value, since inducer lines with robust pollen production can pollinate more maternal rows and reduce the land, labor, and seed costs per haploid produced. Meanwhile, most plant architectural traits showed no consistent overall genotypic differentiation, meaning that on traits such as plant height and general stature, the inducer lines were largely interchangeable. This combination of strong pollen performance with architecture that is not a differentiator simplifies some aspects of selection but also underscores that the traits that actually matter for induction quality are the ones where the genotype-by-season noise was loudest.

To synthesize these threads into an actionable ranking, the authors constructed a predefined weighted selection index combining HIR, MCR_NF, and pollen production capacity, and combined it with hierarchical clustering to group lines with similar overall profiles. The outcome was telling: different lines led in different seasons. K8/BHI306-68 topped the index in the 2023/24 dry season, K8/BHI306-53 led in the 2024/25 dry season, and K8/BHI306-61 took the top spot in the 2025 rainy season. Yet one entry stood apart from this churning leaderboard. K8/BHI306-61 was the only line ranked among the top three in all three seasons, and it was also assigned to Cluster B in the hierarchical clustering analysis. That combination of consistent operational ranking and a coherent cluster profile marks it as the most promising candidate to emerge from the trial.

The study’s conclusions are measured rather than triumphant. The tropical-derived inducer lines retained substantial phenotypic induction capacity under tropical savanna conditions, which is a genuine achievement given the temperate origins of the founding inducer germplasm. But their relative performance was strongly trait- and season-dependent, and the authors conclude that reliable selection requires integrating induction capacity, diagnostic reliability, reproductive performance, and cross-season consistency rather than leaning on a single trait or analytical approach. K8/BHI306-61, with its three-season consistency, now warrants further validation across multiple locations, multiple years, and different maternal genetic backgrounds before it can be endorsed for routine deployment. For a field in which every season of testing delays a breeding cycle, the study delivers an uncomfortable but valuable truth: in tropical maize doubled haploid production, the season itself is a first-class variable, and any pipeline that ignores it does so at its own cost.

Subject of Research: Seasonal performance and stability of tropical-adapted maize haploid inducer lines for doubled haploid breeding

Article Title: Seasonal performance and stability of tropical-adapted maize haploid inducers

Article References: Thawarorit, A., Lübberstedt, T., Dermail, A., Duangpapeng, P., Ruanjaichon, V., Chankaew, S., & Suriharn, K. (2026). Seasonal performance and stability of tropical-adapted maize haploid inducers. BMC Plant Biology. https://doi.org/10.1186/s12870-026-10019-5

Image Credits: AI Generated

DOI: 10.1186/s12870-026-10019-5

Keywords: maize, doubled haploids, haploid inducers, tropical maize, genotype-by-environment, heritability, R1-nj marker, Pl-1 root marker, pollen production, plant breeding, seasonal adaptation, BMC Plant Biology