KASP-Based Genotyping Reveals Super-Early Maturity Allele Diversity in High-Latitude Soybean Germplasm from Mohe, Northeast China (>53° N)

Abstract

Soybean (Glycine max) is a critically important crop for oil, protein, feed, and food security in China. Expanding soybean cultivation into high-latitude regions represents one of the most direct and effective strategies to increase total production. In the present study, we employed KASP (Kompetitive Allele-Specific PCR) marker technology to systematically analyze 18 variant loci across 14 flowering-time genes in 443 soybean germplasm accessions adapted to high-latitude conditions in Arctic Village (Beiji Cun), Mohe City (>53° N), northeastern China. Our results revealed clear functional-tier-dependent selection gradients: key mutation sites (frequency > 96%) in upstream photoreceptors and core circadian clock genes, such as E2 and GmPRR3a, were nearly fixed in the population, whereas downstream flowering genes such as GmFT5b and GmFT2b remained under dynamic selection. Combinatorial analysis of early-maturity allelic variants identified 178 distinct genotype combinations, including six dominant types (n ≥ 10). Field phenotypic analysis demonstrated that the cumulative number of early-maturity alleles was significantly negatively correlated with flowering time, with specific allele combinations such as FT5a^A + FKF1b-hap3^T exhibiting particularly strong flower-promoting effects. A set of 80 highly enriched super-early-maturity accessions, including extreme materials such as MHL22002, were identified, providing valuable genetic resources and a theoretical framework for elucidating the flowering regulatory mechanisms of high-latitude soybean and for breeding super-early-maturing varieties.

1. Introduction

Soybean (Glycine max) is a globally important crop that serves as a primary source of plant protein, edible oil, and a key component of high-quality protein in animal feed. It is also rich in vitamins and minerals, conferring high nutritional and economic value [1]. Although China is the center of soybean origin and domestication, it currently accounts for only approximately 5% of global production. Given continuously rising domestic demand, increasing soybean yield has become a core objective of Chinese agriculture. Expanding cultivation northward into high-latitude regions is among the most direct and effective strategies to boost total soybean production.

Northeast China represents the core region for domestic soybean production. Heilongjiang Province consistently contributes over 40% of national output and leads the country in both planting area and total yield. According to data from the National Bureau of Statistics [2] and the Heilongjiang Bureau of Statistics [3], the soybean planting area in the Heihe region (47°42′–51°03′ N, 124°45′–129°18′ E) reached 1359.1 thousand hectares in 2025, representing 12.98% of the national total (10,474 thousand hectares). Nevertheless, the soybean planting area in territories north of Heihe (high-latitude regions, predominantly >50° N) was only 165.01 thousand hectares in 2023 merely 12.14% of the Heihe area. By contrast, FAO data [4] indicate that the 2023 soybean harvested area in the Russian Federation was 3500 thousand hectares, suggesting considerable untapped potential in China’s high-latitude regions. The ecological environment of these regions characterized by low temperatures and extreme long-day photoperiods poses significant challenges to soybean, which is an obligate short-day plant. The markedly extended summer day length and lower temperatures inhibit flowering and pod setting, thereby restricting yield formation [5]. Consequently, the development of soybean varieties with super-early maturity adapted to the ecological constraints of high-latitude regions has become a central breeding objective. Notably, Russia, as a major global producer, also faces similar bottlenecks in its vast high-latitude territories, where extreme long days and low temperatures severely limit yield potential. Therefore, dissecting the adaptive mechanisms of Chinese ultra-early germplasm can provide elite genetic resources and practical breeding references not only for China’s frontier regions but also for Russia and other high-latitude countries worldwide.

Flowering time is a critical determinant of the soybean growth cycle and directly influences final yield. To avoid yield losses due to developmental mismatches with the short frost-free season, plants must achieve timely flowering within the limited growing window. Substantial progress has been made in elucidating the genetic regulation of flowering in soybean. The E1–E4 genes are key regulators of photoperiodic flowering and maturity. The E1 gene (Glyma.06g207800) suppresses flowering under long-day conditions, and its recessive alleles (e.g., e1-as, e1-fs, and e1-nl) markedly promote flowering [6]. The GmGIa (E2) gene (Glyma.10g221500) encodes a GIGANTEA homolog that regulates FT gene expression and functions as a flowering suppressor to delay floral transition in soybean [7]. E3 (Glyma.19G224200) and E4 (Glyma.20G090000) encode phytochrome A homologs, GmphyA3 and GmphyA2, respectively, which are involved in light perception and signal transduction [8,9]. Different allele combinations at these loci exert synergistic or additive effects, providing the genetic basis for adaptation to diverse environments [10]. The GmELF3/E6/J gene (Glyma.04G050200), also known as the “long juvenility” gene, harbors a functional deletion allele that causes a frameshift mutation via single-base deletion, resulting in protein truncation and delayed flowering an adaptation enabling the plant to thrive in tropical environments [11].

The FLOWERING LOCUS T (FT) gene family also plays a pivotal role in regulating flowering in soybean. This superfamily comprises ten members forming five pairs of homologs: GmFT1a/b, GmFT2a/b, GmFT3a/b, GmFT5a/b, and GmFT4/6. These genes are functionally classified as either flowering promoters (GmFT2a/b, GmFT3a/b, and GmFT5a/b) or inhibitors (GmFT1a/b and GmFT4/6) [12]. GmFT2a (E9, Glyma.16G150700) and GmFT5a (Glyma.16G044100) are major promoters highly expressed under short-day conditions to induce flowering [13]. In natural populations, the e9 allele containing a SORE-1 retrotransposon reduces GmFT2a expression, thereby delaying flowering [14]. Conversely, GmFT1a (Glyma.18G298900), GmFT1b (Glyma.18G299000), and GmFT4 (Glyma.08G363100) are highly expressed under long-day conditions but barely detectable under short-day conditions; they are regulated by E1 and suppress flowering. GmFT6 (Glyma.08G363200) is likewise considered a repressor, as its overexpression in Arabidopsis significantly delays flowering [15,16,17,18].

Several additional genes play important roles in high-latitude adaptation. GmFUL2a (Glyma.05G018800, Tof5) promotes flowering under long-day conditions and improves adaptation to high latitudes [19]. Within the photoperiod regulatory network, GmPRR3a (Glyma.U034500, Tof11) and GmPRR3b (Glyma.12G073900, Tof12) positively regulate LHY homologs to facilitate E1 expression, ultimately inhibiting flowing in long-day environments. The tof12-1 allele was domesticated to shorten the growth period, while the tof11-1 allele further enhances early maturity and adaptation to high latitudes [20,21]. Tof7 (Glyma.07G048500, E11) is a single locus regulating early maturity, with allelic differences causing 12–15 days variations in maturity [22]. GmEID1 modulates photoperiod sensitivity and yield by forming a complex with E3/E4 and the evening complex [23]. Collectively, the soybean flowering network centered on E genes, FT genes, and clock genes coordinately regulates flowering time through downstream meristem identity genes such as AP1 and LFY [24].

Kompetitive Allele-Specific PCR (KASP) is a high-throughput technology for genotyping single-nucleotide polymorphisms (SNPs) and insertions/deletions (InDels) based on allele-specific primers and fluorescent detection [25]. KASP utilizes primers with SNP-specific bases at the 3′ end and universal tags at the 5′ end, coupled with fluorescent dyes (e.g., FAM, HEX), to accurately distinguish alleles. Each KASP assay employs two allele-specific primers and one common reverse primer; only perfectly matched primers amplify efficiently, generating allele-specific fluorescence. KASP technology has been widely applied in molecular breeding programs for crops including wheat [26], maize [27], rice [28], tomato [29], rapeseed [30], and potato [31], demonstrating broad utility for marker-assisted selection of traits related to disease resistance, quality, and environmental adaptation. In soybean, KASP markers have been developed for early maturity-related genes such as E1, E2 (GmGIa), E3 (GmphyA3), E4 (GmphyA2), J (GmELF3), GmPRR3a, GmPRR3b, GmFUL2a, GmLHY1a/1b, and GmSOC1a [10,32]. Compared with other molecular marker technologies such as SNP arrays and TaqMan probes, KASP exhibits distinct advantages, including low development cost, simple experimental procedures, and a short detection turnaround cycle. However, a systematic analysis of key maturity genotypes in high-latitude soybean germplasm combined with a comprehensive molecular marker system for all relevant loci has yet to be established, representing a critical bottleneck limiting breeding progress.

The present study aims to systematically characterize the genetic variation in flowering-related genes in super-early maturing soybean germplasm from high latitudes. By integrating existing KASP markers with newly developed markers for key functional loci, we optimized and established a molecular marker system suitable for detecting early-flowering variations in soybean germplasm. This system provides reliable technical support and theoretical reference for marker-assisted breeding of super-early maturing soybean adapted to high-latitude conditions.

2. Materials and Methods

2.1. Plant Materials

A total of 443 soybean germplasm accessions adapted to the ecological conditions of Arctic Village (Beiji Cun) (122°21′–122°21′ E, 53°27′–53°33′ N), North Pole Town, Mohe City, Heilongjiang Province, China, and capable of normal maturation under local conditions were selected for this study. All accessions were initially grown under standardized conditions in a climate chamber with a 16 h light/8 h dark photoperiod at a constant temperature of 26 °C. Three biological replicates were used per accession. Young trifoliolate leaves were sampled seven days after emergence for genomic DNA extraction. Field trials were also conducted at the Shunyi Experimental Station (116°34′12″ E, 40°14′24″ N) in Beijing to systematically record the flowering time (R1 growth stage) of each accession.

2.2. Methods

2.2.1. DNA Extraction

Genomic DNA was extracted from approximately 100 mg of fresh leaf tissue harvested from seedlings at the seven-day stage using the cetyltrimethylammonium bromide (CTAB) method following standard protocols. DNA quality and concentration were assessed by agarose gel electrophoresis and spectrophotometry prior to downstream applications.

2.2.2. KASP Primer DesignPreviously Developed KASP Markers for Early Maturity-Related Genes

KASP primer sequences for the flowering genes E1, E2 (GmGIa), E3 (GmphyA3), E4 (GmphyA2), and GmPRR3a were obtained from previously published studies [10,32] (Table S1).

The PCR reaction program performed on the Quantstudio 7 Flex was as follows (

Table 1. KASP Reaction System.

Step	Temperature	Time	Cycles
1	95 °C	10 min	1 cycle
2	95 °C	20 s	10 cycles
	61–55 °C (drop 0.6 per cycle)	40 s
3	95 °C	20 s	30–45 cycles
	55 °C	40 s
4	25 °C	forever	1 cycle

):

Design of Novel KASP Markers for Additional Early Maturity-Related Genes

A clear stepwise strategy was adopted for flowering genes without established KASP markers. First, for genes with functionally validated polymorphic sites, allele-specific primers were designed directly based on the published variant positions. Second, for genes without characterized mutations, the HapSnap tool in the SoyOmics database (https://ngdc.cncb.ac.cn/soyomics/index, accessed on 20 September 2025) was used to identify loci differentiating high-latitude accessions from mid- and low-latitude ones as the KASP genotyping loci for these genes, according to the Chinese geographic grouping results in the database [33,34]. Then, targeted KASP primers were designed for these selected informative loci (Table S1). All newly synthesized primer sets were individually validated for amplification specificity prior to large-scale application.

2.2.3. Genotyping and Data Analysis

KASP reactions were performed according to the manufacturer’s standard protocol for each marker. Genotyping data and allele information were compiled and organized in Microsoft Excel. OriginPro 2026 10.3 software (OriginLab, Northampton, MA, USA) was used to process the genotyping results and generate figures illustrating the distribution of different KASP marker genotypes across the accession panel. Statistical analyses, including analysis of variance (ANOVA) and least significant difference (LSD) tests (p < 0.05), were performed to assess differences in flowering time among haplotype groups.

3. Results

3.1. Upstream Photoreceptor Genes Display Divergent Selection Patterns

In this study, we systematically genotyped early-maturity allelic variants of the upstream photoreceptor genes GmphyA3 (E3), GmphyA2 (E4), and GmFKF1b in 443 soybean germplasm accessions using KASP marker technology (Figure 1 and Figure S1). For the E3 gene, the frequencies of the early-maturity alleles e3-fs and e3-ns were 40.18% (n = 178) and 0.68% (n = 3), respectively. Notably, a high proportion of heterozygotes (CT genotype) was observed at the e3-ns locus (14.67%, n = 65), suggesting ongoing segregation at this site. For the E4 gene, the e4-kes mutation frequency was 64.56% (n = 286), while the frequency of the large-fragment insertion allele e4-sore was considerably lower (6.32%). Within the blue-light receptor gene GmFKF1b (Tof8), two variant types were detected: the early-maturity haplotype FKF1b-hap3^T was found at a high frequency of 72.00%, substantially exceeding the frequency of FKF1b-H3^T (1.58%).

3.2. Core Circadian Clock Genes Exhibit Multi-Level Allele Frequency Distributions

Using KASP molecular markers, we systematically characterized 443 high-latitude soybean accessions at five circadian clock-related loci (Figure 2 and Figures S2–S4). The early-maturity mutation LHY1a^T of GmLHY1a (Tof16) was detected at a relatively low frequency (14.45%), indicating limited selection of this allele within the high-latitude population. By contrast, the early-maturity mutations PRR3a^A and PRR5c^A of GmPRR3a and GmPRR5c were present at frequencies of 96.84% and 17.16%, respectively. The PRR3a^A allele displayed near-complete fixation and strong positive selection in the population. The recessive early-maturity allele e2-ns of the E2 gene reached a frequency of 98.19%, representing a near-fixed state and serving as the overwhelmingly dominant allele at this locus. The early-maturity mutation ELF3^C of GmELF3 (J) was present at an intermediate frequency of 38.83%. These differential allele frequency distributions among circadian clock genes collectively define the genetic architecture underlying early-maturity adaptation in high-latitude soybean.

3.3. Downstream Signaling Genes Show Dynamic Selection Patterns in High-Latitude Soybean Populations

KASP analysis of downstream flowering pathway genes revealed two recessive, early-maturing mutation types e1-as and e1-fs within the core floral repressor E1, which functions as the terminal node of the photoperiodic signaling cascade. The selection frequency of the e1-as allele (60.27%) was substantially higher than that of e1-fs (3.84%), indicating preferential selection of the former allele at high latitudes. The GG genotype of the floral integrator GmSOC1a (Tof18) was nearly fixed in the Arctic Village population (99.32%), consistent with its status as an essential component of the early-maturity genetic background. Among the florigen genes, contrasting selection patterns were observed: GmFT5b (FT5b^G, 76.52%) and GmFT2b (FT2b^T, 56.66%) were present at markedly higher frequencies than GmFT5a (FT5a^T, 24.15%) and GmFT6 (FT6^C, 23.93%). These results indicate that genes at distinct functional tiers within the soybean flowering pathway may co-evolve to jointly regulate high-latitude adaptation (Figure 3 and Figures S5–S8).

3.4. Allelic Variation Combinations Drive High-Latitude Super-Early Maturity in Soybean

Soybean materials capable of normal flowering and maturation at latitudes ≥ 53° N belong to maturity group MG0000, which represents the earliest-maturing cohort among currently available Chinese soybean germplasm. Based on KASP genotyping of 18 early-maturity loci across 14 early-maturity-related genes in the 443 Arctic Village accessions, we integrated these multi-locus genotype data and analyzed their associations with field-measured phenotypes. Although the accessions as a whole exhibited early-maturity characteristics, some non-early-maturity alleles remained segregating within the population (Table S2).

Analysis of 18 mutation loci across 14 early maturity-related genes including E1, E2 (GmGIa), E3 (GmphyA3), E4 (GmphyA2), GmELF3 (J), GmPRR3a, GmPRR5c, GmLHY1a (Tof16), GmFKF1b (Tof8), GmSOC1a (Tof18), GmFT2b, GmFT5a, GmFT5b, and GmFT6 revealed that the number of early-maturity alleles carried per accession ranged from 4 to 13. The modal class was 9 early-maturity alleles (n = 99), and 80 accessions (18.06% of the panel) carried 10 or more early-maturity allelic variants. The highest allele count was observed in MHL22129 (n = 13), while MHL22082, Fuyuan42, and Keyan104 each carried only 4 early-maturity alleles (Figure 4A; Table S2).

Combinatorial analysis of allele variation patterns identified a total of 178 distinct genotype combinations, of which six dominated the population (Figure 4B, Table S3). The most prevalent combination, Hap_008 (n = 39), exhibited a mean flowering time of 17 days, 1.6 days earlier than the population mean of 18.1 days (Figure 4B). Among the six major haplotype groups, mean flowering time gradually advanced as the number of early-maturity mutations increased (Figure 4C, Table S3). For example, haplotype Hap_009 (five mutations) showed a mean flowering time of 20.5 days, while haplotypes Hap_006 and Hap_041 (each with ten mutations) had earlier average flowering times of 17.0 days.

Distinct differences in flowering time were also observed among haplotype groups carrying equal numbers of early-maturity mutations, likely reflecting differential effects of individual alleles. Among haplotypes carrying ten mutations, Hap_006 flowered 1.8 days earlier than Hap_041. This difference was attributable to the specific allele combination FT5a^A + FKF1b-hap3^T in Hap_006, which conferred a stronger flower-promoting effect than the PRR5c^A + LHY1a^T combination characteristic of Hap_041 (Figure 4C, Tables S3 and S4).

Among the 172 rare haplotype combinations (n < 10), more clear phenotypic differences were observed. Accessions MHL22002 (Hap_062) and Shengdou062 (Hap_176), which both carried nine early-maturity alleles and shared eight core alleles, nonetheless differed in flowering time by 10.4 days (Tables S3 and S4). This difference was attributed to the stronger flower-promoting effect of the e4-sore allele in MHL22002 relative to the FT6^C allele in Shengdou062.

Accession MHL22129 (Hap_101) displayed a notable genotype–phenotype discrepancy of considerable research interest: despite harboring the greatest number of early-maturity alleles (n = 13), its measured flowering time was 21.3 days 3.2 days later than the population mean (18.1 days) and 10.0 days later than the earliest-flowering accession MHL22002 (11.3 days). Similarly, MHL22133 (Hap_103), which carried 10 early-maturity alleles, had a flowering time of 26 days 13.9 days later than the population mean. We hypothesize that these accessions harbor undetected strong late-flowering alleles that counteract the cumulative effects of the early-maturity mutations (Table S4).

4. Discussion

The limited expansion of soybean cultivation into high-latitude regions is primarily constrained by the crop’s sensitivity to photoperiod and temperature. In these environments, early flowering and early maturation significantly shorten the growth period, enabling plants to escape early autumn frost and complete reproductive development within the abbreviated frost-free season. The early-maturity phenotype in soybean is a quantitative trait governed by the combined regulatory effects of multiple early-maturity alleles [35]. In this study, we performed KASP genotyping at 17 mutation loci of 14 early maturity-regulated genes in 443 soybean accessions from Arctic Village and systematically resolved their selection patterns under extreme high-latitude conditions.

4.1. Differential Selection Frequency of Early Maturity-Related Genes at Arctic Village

Plants regulate precise flowering timing by perceiving environmental light signals. Photoreceptor genes, including the phytochrome genes GmphyA3 (E3) and GmphyA2 (E4) [36,37] and the blue-light receptor gene GmFKF1b [38,39], detect light quantity, quality, and duration to establish the primary photoperiodic signaling foundation. Meanwhile, the circadian clock genes GmLHY [40], GmPRR3a, GmPRR5c, GmGIa (E2), and GmELF3 (J) form a transcriptional–translational feedback loop that generates a near-24 h rhythm [41], enabling accurate day-length measurement and developmental synchronization. Following environmental and clock inputs, signals converge on downstream genes, including E1, GmSOC1a, and members of the FT family, which constitute a hierarchical regulatory framework for flowering induction.

Numerous studies have investigated the molecular mechanisms underlying high-latitude adaptation in soybean. The most well-characterized mechanism involves the enrichment of recessive early-maturity alleles at the E1, E2 (GmGIa), E3 (GmphyA3), and E4 (GmphyA2) loci in high-latitude soybean materials [10]. More recently, circadian clock genes such as GmPRR3a (Tof11) and GmPRR3b (Tof12) have also been implicated in high-latitude adaptation [20,21].

In this study, chi-square test and Hardy–Weinberg equilibrium (HWE) analysis were conducted on all 18 variant loci, and the results showed that the p values of all loci were extremely small. This indicated that these loci significantly deviated from the Hardy–Weinberg equilibrium in the tested high-latitude soybean germplasm population, suggesting a strong directional selection for these early-maturity-related variant loci in the population. The distribution of their gene frequencies was not formed randomly but driven by selection pressure, which confirmed that these loci are key functional sites regulating the high-latitude adaptability of soybean.

We expanded on these findings by systematically genotyping E genes, circadian clock genes, and FT family members in 443 high-latitude accessions, incorporating both published KASP markers and newly developed markers for previously uncharacterized early-maturity loci. The results revealed a clear functional-tier-dependent gradient in allele selection, reflecting differences in functional importance and evolutionary constraints on high-latitude adaptation.

Genes involved in photoperiod perception, signal integration, and core circadian clock function exhibited the strongest selective signals. The near-complete fixation of e2-ns at E2, PRR3a^A at GmPRR3a, and the GG genotype at GmSOC1a indicates that these early-maturity alleles constitute the core genetic foundation for soybean adaptation to high-latitude growing conditions. As key nodes in light signal transduction, the fixation of these variants ensures rapid initiation of the flowering program under the extreme long-day photoperiods of Arctic Village, thereby mitigating the risk of early frost damage.

In contrast, different upstream regulators exhibited distinct selection patterns. The early-maturity allele of GmLHY1a (LHY1a^T) was present at a low frequency (14.45%), whereas the FKF1b-hap3^T variant of GmFKF1b was highly prevalent (72.00%); the alternative precocious allele FKF1b-H3^T was extremely rare (1.58%), consistent with the findings of Li et al. [39]. This discrepancy may reflect the non-redundant functions of circadian clock components: LHY1a acts as a core oscillator, and mutations at this locus may be subject to stronger functional constraints to maintain rhythmic stability. The differential selection of FKF1b alleles may reflect distinct selective pressures on photoperiodic regulation of flowering time under high-latitude conditions.

Within the downstream florigen gene family, clear functional divergence was evident. The high-frequency selection of GmFT5b (FT5b^G, 76.52%) contrasted markedly with the low frequency of GmFT5a (FT5a^T, 24.15%), a pattern opposite to that observed in mid-latitude populations. This suggests that GmFT5b may be more functionally adaptive under extreme long-day photoperiods a potential consequence of gene duplication followed by subfunctionalization.

It is noteworthy that certain early-maturity alleles, such as e1-fs (3.84%), e3-ns (0.68%), and e4-sore (6.32%), were present at very low frequencies. The high heterozygosity (14.67%) observed at the e3-ns locus suggests that this site may be subject to ongoing dynamic selection, while the large-fragment insertion mode of e4-sore may directly limit its transmission efficiency. The persistence of these low-frequency alleles reveals the complex mechanisms maintaining genetic polymorphism under strong directional selection.

In this study, multiple early-maturity-related alleles showed a low-frequency distribution in soybean germplasm at 53° N high latitudes. For example, the e4-sore allele with a large fragment insertion in the E4 gene had a frequency of only 6.32% in the tested population, the e3-ns allele in the E3 gene was as low as 0.68%, and the e1-fs allele in the E1 gene was merely 3.84%. The existence of these low-frequency alleles provides unexploited genetic resources for early-maturity soybean breeding. The early-maturity allele LHY1a^T of the circadian clock gene GmLHY1a had a selection frequency of 14.45%. As a variant of the core node gene in the photoperiod signaling pathway, it has not been targeted for utilization in high-latitude soybean breeding. These currently underutilized alleles are important components of the genetic regulatory system for soybean early maturity and may serve as vital potential resources for future super-early-maturity soybean breeding at ultra-high latitudes. Marker-assisted breeding can be applied in future research to provide a new theoretical basis for expanding the soybean planting range.

Collectively, the capacity of soybean to adapt to Arctic Village reflects the synergistic action of genes spanning multiple tiers of the flowering pathway. The stable fixation of upstream core nodes ensures timely flowering initiation, while the differential selection of mid- and downstream genes enables fine-tuned responses to specific environmental stresses. This multilayered, graded selection pattern constitutes the genetic basis for high-latitude adaptation in soybean.

4.2. Super-Early Maturity Results from the Accumulation of Multiple Early Maturity Alleles

Analysis of flowering-related genetic mutations across 443 super-early-maturing soybean accessions from Arctic Village confirmed that the aggregation of multiple early-maturity alleles is a critical genetic basis for the extreme early-maturity phenotypes observed in these germplasm. A general trend showed that higher numbers of early-maturity alleles were typically associated with shorter flowering times.

For example, haplotype Hap_008 (eight early-maturity variants) flowered 1.1 days earlier than the population mean, while Hap_006 (ten variants) flowered a further 0.9 days earlier than Hap_008. Although the effects of different mutation combinations varied, the general trend was clear: a greater number of early-maturity alleles was associated with earlier flowering. This finding is consistent with observations on additive effects of flowering-time genes in Arabidopsis [42] and supports an additive model of allele-effect accumulation in soybean adaptation.

Beyond allele number, specific allele combinations also play important roles in determining flowering time. Among the six major haplotype combinations, Hap_008 (eight mutations) showed mean flowering 1.2 days earlier than Hap_047 (seven mutations). Comparative analysis revealed that the sole additional allele in Hap_008 was FT2b^T, suggesting that this variant makes a disproportionate contribution to early flowering. Specific allele interactions were also evident: Hap_006 (ten mutations) flowered 0.9 days earlier than Hap_008 (eight mutations), primarily attributed to the e3-fs + FT5a^A combination. These findings highlight the importance of both allele number and specific allele identity in determining the flowering time of soybean germplasm at high latitudes.

4.3. KASP Marker Development and Superior Early Maturity Germplasm Are Core Supports for High-Latitude Breeding

Several accessions in our study exhibited apparent genotype–phenotype discordance. The haplotype Hap_002 (nine mutations) had a mean flowering time 0.4 days later than Hap_008 (eight mutations), despite carrying all eight mutations of Hap_008 plus the potent early-maturity allele e3-fs. Examination of individual accessions revealed that Hap_002 included materials such as MHL22038 and MHL22039, which flowered markedly later than the population average despite harboring multiple early-maturity alleles. We hypothesize that these accessions carry undetected strong late-flowering loci that counteract the effects of the characterized early-maturity alleles. A similar explanation likely applies to MHL22129 (Hap_101), which possessed the largest number of early-maturity mutations (n = 13) yet exhibited a relatively late-flowering phenotype (21.3 days), and to MHL22133 (Hap_103), which carried ten early-maturity alleles but flowered at 26 days. These phenotypic anomalies suggest that additional, uncharacterized regulatory factors influence flowering time in certain accessions and merit further investigation.

From a practical breeding perspective, the KASP marker system newly developed for ten early-maturity-related genes in this study, together with the 80 high-allele-accumulation accessions and extreme materials such as MHL22002 and MHL22129, provide key genetic resources and technical tools for soybean improvement in high-latitude environments. Through marker-assisted selection, favorable allele combinations can be efficiently pyramided into elite genetic backgrounds to breed new soybean varieties adapted to extreme high-latitude conditions. This approach has the potential to broaden the latitudinal range of soybean cultivation and increase total soybean production in China, with important practical implications for food security.

China is the origin of soybeans and the center of their diversity. The Arctic Village super-early-maturing soybean germplasm was formed through long-term artificial selection. It has adapted to the extreme environment at about 53° N latitude, which features high latitude and strong photoperiod. These materials serve as a unique gene pool. They are expected to overcome the limitations of expanding soybean northwards. Their value to global breeding programs is invaluable.

This study focuses on maturity genes such as E3, E4 and FKF1b. These are internationally recognized core loci for photoperiod regulation. We identified dominant allele combinations in super-early-maturing soybeans. We also analyzed their frequency distribution characteristics. This provides new insights into differences in selection pressure along the latitude gradient. Using extreme environments as natural experimental systems is a valuable approach. This strategy can be adopted by other high-latitude regions. Examples include Canada, northern Europe and the Russian Far East.

Currently, climate change and food security are key drivers. Soybean cultivation continues to expand towards higher latitudes. Countries such as Canada, Scandinavia and Siberia are introducing and testing new varieties. The high-latitude adaptive allele combinations identified in this study can be used immediately. These can be directly applied in the molecular design of breeding programs in these emerging production areas.

5. Conclusions

In this study, 443 MG0000 super-early-maturity soybean accessions from Arctic Village the northernmost region of China were used to systematically analyze the genetic basis of high-latitude adaptation. Key early-maturity alleles of certain upstream photoreceptor and core circadian clock genes (e.g., E2 and GmPRR3a) were found to be nearly fixed in the population, while early-maturity variants of other genes were maintained at lower frequencies reflecting differences in functional importance and selection pressure across the flowering regulatory network.

These low-frequency early-maturity alleles, such as e1-fs, e3-ns, e4-sore, FKF1b-H3T, LHY1aT, and FT5aT, are currently underutilized in high-latitude soybean germplasm, this indicates that they still hold great breeding potential, these alleles may further shorten flowering time of super-early cultivars, and provide useful clues for expanding soybean cultivation in high-latitude regions.

The cumulative number of early-maturity alleles tended to decrease as flowering time became shorter, while specific allele combinations (e.g., FT5a^A + FKF1b-hap3^T; FT2b^T + ELF3^C) conferred stronger flower-promoting effects, demonstrating that the super-early-maturity phenotype is jointly determined by the breadth of allele polymorphism and specific dominant allele combinations.

The established KASP marker system can be applied for early-stage screening of breeding materials in the field, effectively shortening the breeding cycle. The identified superior allele combinations provide key molecular targets for marker-assisted design breeding, facilitating the development of new varieties adapted to extreme environments. The genetic characterization of MG0000 germplasm reported here provides important germplasm resources and technical support for promoting the northward expansion of the soybean cultivation boundary and safeguarding soybean production in high-latitude regions of China.