Introduction

Sugar beet, a biennial herbaceous plant primarily cultivated in arid and semi-arid regions 1. Despite a cultivation history spanning roughly 200 years2, this “sugar-producing treasure” still harbors many enigmatic aspects. Predominantly cultivated in North and Northeast China, the latter region notably dominated its cultivation3,4. Sugar beet exhibits high sensitivity to soil conditions, making it an ideal candidate for soil nutrient testing. Moreover, it possesses the remarkable ability to draw nitrogen from deep soil layers to the surface, thereby enhancing nitrogen utilization efficiency, and serving as an excellent rotation crop.

The sugar beet yield is influenced by a combination of factors, including genotype (G), environment (E), and genotype by environment interaction (GEI). A higher GEI effect suggests greater sensitivity to environmental conditions and less stability for the variety5,6. Continuous enhancement of varieties is crucial for sugar beet production and development. Regional trials effectively evaluate yield, stability, and adaptation of candidate varieties, playing a vital role in selection. This experiment aimed to identify sugar beet varieties suitable for North and Northeast China cultivation by analyzing genotype-environment interactions. Statistical methods like the Additive Main Effects and Multiplicative Interaction (AMMI) model integrate Analysis of Variance (ANOVA) and Principal Component Analysis (PCA), improving prediction accuracy by decomposing interaction principal components7. Gao Hua et al. employed the AMMI model to assess the average yields of new national sugar beet varieties across different trial locations, indicating that integrating variance analysis and PCA within the AMMI model enhances the visualization of variety stability and adaptability across trial sites8. Similarly, domestic scholars like Bai Zhangcai et al. evaluated the stability of national sugar beet varieties during the 2015 regional trials using the crop variety regional trial variance analysis method introduced by Mr. Mo Huidong9,10. Yan et al. utilized Singular Value Decomposition (SVD) of the multi-environment trial (MET) data focused on the environment to generate the first two symmetric scaling principal components (PC1 and PC2), resulting in a GGE biplot, which effectively addresses vital issues concerning genotype and test environment evaluation11. Furthermore, Zivko et al. applied the GGE biplot method to elucidate the influence of genotype, environment, and GEI on root yield and sugar content, while assessing the stability of seven sugar beet hybrid varieties in the Serbian region, thereby affirming the practicality of the GGE biplot method in regional trials of sugar beet varieties12. Hassani et al. have shown in their research findings utilizing the AMMI and GGE biplot methods that they can aid breeders in saving time and reducing costs, assist in the selection of suitable and discriminating environments for sugar beet variety trials, and facilitate genotype selection for trait enhancement and the development of new variety13. Furthermore, the evaluation of target traits typically revolves around yield potential. To facilitate comparisons among different varieties and multiple traits, the genotype by trait (GT) biplot method has been widely employed for classifying and visually illustrating correlations among various agronomic traits. However, there are limitations in comprehensively evaluating multiple traits of varieties using this method. Furthermore, the assessment of target traits is primarily grounded on yield potential. To address the shortcomings and limitations of the GT biplot method in conducting comprehensive evaluations of crop multi-trait assessments, Yan Weikai introduced the genotype by yield × trait biplot (GYT). GYT biplot utilizes the combination of yield and target traits for variety classification and comprehensive assessment of various measured agronomic traits, boasting stronger relevance and broader applicability14;15,16 Yue H W utilized regional experimental data of the Huang-Huai-hai summer maize set spanning 2016 to 2017. They employed the GYT biplot technique to evaluate GYT interactions in different eco-environments. Various traits including grain yield, growth period, plant height, ear height, lodging rate, empty stalk rate, ear length, bald tip, the row number of ears, grain weight per ear, 100-grain weight per ear, stem rot, and smut were analyzed to establish a theoretical foundation for the comprehensive evaluation of variety traits in the Huanghuai sea summer maize region17. Ma Z X conducted a thorough evaluation and classification of multiple traits of broomcorn millet using GYT biplot analysis, concluding that the results obtained through GYT biplot analysis were more reliable than those from GGE and GT biplot. They identified the japonica rice variety Yi 11–02-92–4 and the glutinous variety 0515–2-2 as having the best comprehensive performance, suitable for cultivation in spring glutinous millet area in Northern China and the spring and summer glutinous millet area in the Loess Plateau18. Additionally, Yang et al. utilized the AMMI model, GGE biplot, and GYT biplot analysis to identify the best pea genotypes in diverse environmental experiments. Their findings revealed significant differences among genotypes, environments, and interactions, and Yunwan52 emerged as the most stable and optimal genotype in the tested environments. Its exceptional yield performance and wide adaptability to various environments contribute significantly to the development of new nutritious pea varieties and the enhancement of pea yields19.

This study examined the agronomic traits, yield, and related yield traits of 16 introduced sugar beet varieties across seven different ecological pilot areas through regional experiments. Based upon single-trait GGE biplot analysis, a comprehensive evaluation of the abundance, stability, and adaptability of these varieties was conducted using the GT and GYT biplot methods for multi-trait analysis. While the GT biplot effectively demonstrated the strengths of varieties in target traits and their correlations, it falls short in providing a comprehensive evaluation based on multiple traits. The GYT biplot method addresses this limitation20. Moreover, compared to conventional seed selection methods, GYT biplot analysis offers several advantages: 1) it is more professional, being grounded in scientific experiments and practical applications of specialized technology, allowing for more refined evaluation and screening of varieties; 2) it is more efficient, enabling simultaneous comprehensive analysis of multiple varieties and rapid evaluation of their adaptability and advantages, thus enhancing seed selection efficiency; and 3) it is more rational, as it considers the impact of environmental factors on variety performance and conducts evaluations based on scientific criteria, ensuring a more reasonable evaluation process. GYT biplot analysis was utilized to comprehensively evaluate and classify multiple traits of the tested varieties. Through PCA, joint analysis of variance, and correlation analysis, a comprehensive assessment was conducted on yield, yield stability, adaptability, and performance variations under specific environmental conditions. Sugar beet varieties with high and stable yields were identified through field trials, with the potential for targeted improvements in sugar beet production in specific areas. In future production practice, if the focus is on achieving consistent performance across a wide range of environmental conditions, then GGE biplot analysis is more relevant. However, if the emphasis is on performance differences under specific environmental conditions and the selection and improvement of varieties based on specific traits, then GYT biplot analysis is more meaningful. It is important to note that the final decision should consider other factors such as the economic benefits of the variety and risk management.

Materials and methods

Varieties and test sites

Tables 1 present the primary characteristics of the species and locations tested in this study. The study encompassed 16 newly introduced varieties and was carried out across seven pilot areas in North and Northeast China. The geographical distribution of test sites is illustrated in Fig. 1. Notably, two pilot projects were conducted in the Hulan District of Harbin City, Heilongjiang Province, comprising one “disease” area and one "non-disease" area. In the “disease” area, the test site has undergone rotational cultivation for numerous years, resulting in a uniform distribution of soil pathogens, which constitutes the principal cause of root rot and brown spot disease in beet. Other locations included Fanjiatun Town in Gongzhuling City, Jilin Province; Shuangyashan Hongxinglong Farm; Zhangbei County in Zhangjiakou City, Hebei Province; Nenjiang City Jiusan Agricultural Science Institute; and Hohhot City.

Table 1 Test varieties, pilot names, and corresponding codes.
Full size table
Fig. 1
figure 1

Location distribution of the pilots.(Where the graph is the author created, using Adobe Photoshop software to complete 2023, URL: https://ps.hanqiwk.cn/hKqjF?bd_vid=11825174183880544993).

Full size image

Experimental design

Each experiment utilized a completely randomized block design. KWS series sugar beet varieties imported from Germany were used as experimental materials, the beet species were planted in four identical sets of two rows, each spanning ten meters, with a ridge spacing of 0.6 m. Each plot encompassed a total area of 633.6 m2 and maintained a planting density of approximately 82,500 plants/ha. All other cultivation management procedures were carried out by experimental requirements. The agronomic traits assessed in 2023 included three quality traits (root yield, sugar content, and sugar yield) along with five growth traits including plant height, leaf length, leaf width, leaf stalk thickness, and SPAD value, as well as two disease traits (root rot and brown spot). Leaf length, leaf width, and plant height were measured using a ruler, while leaf stalk width was measured with a vernier caliper. SPAD value was determined using a specialized SPAD meter, and root yield was quantified at harvest time. Sugar content was assessed using Sugarbeet Quality Analysis Apparatus, and the incidence of root rot was calculated by dividing the number of diseased plants by the total number of plants at harvest time. Additionally, The incidence grade of brown spot was observed during the onset period.

Root yield: root rot was identified and counted in the test plot, and root yield was measured by weighing method.

Sugar Production: Sugar Production is the product of root yield and sugar content.

Sugar content: After harvesting the root of the beet, it was sent to the beet quality supervision, inspection and testing center of the Ministry of Agriculture for the determination of sugar content.

Root rot: When beet is harvested, the number of root rot plants and harvested plants in each plot is counted, and the incidence rate of root rot (%) = the number of root rot plants/the total number of plants *100. The average of three repetitions of each treatment is taken as the final result.

Brown spot: Brown spot disease of sugar beet was investigated during the root tuber growth period (August 1, 2022) and sugar accumulation period (August 15, 2022). Brown spot disease was divided into different grades according to the degree of disease. Brown spot disease was rated on a community basis according to the evaluation criteria of grade 0.5 to 4.5, and grade 0.5 was the lightest. 4.5 disease is the most serious, according to the disease, growth, overall score.

Plant height: Select three beets that can represent the average plant height every repeat, set the leaves upright, measure from the ground to the highest leaf with a ruler, and take the average value three times.

Leaf length: Three beet plants representing average plant height were selected for each repeat, and the leaves of medium size were taken and measured with a ruler.

Leaf width: Three beet plants representing average plant height were selected for each repeat, and the leaves of medium size were taken and measured with a ruler.

Leaf stalk thickness: Three beet plants representing average plant height were selected per repeat, and leaf stalks of medium thickness were measured near the ground part.

SPAD value: Ten healthy beet leaves were selected and measured by SPAD instrument.

AMMI model analysis

The AMMI model integrates analysis of variance and PCA to visualize the stability and suitability of different varieties across individual trial sites. By segregating model error and interference from the residual of the additive model, this model enhances estimation accuracy. Within the AMMI model framework, total variation is partitioned into genotypic main effects, environmental main effects, and GEI, all contributing to variation. The interaction effects between genotype and environment are further reduced into several principal components, commonly analyzed using PC1 and PC2. PC1 represents the response ratio of the variety to the environment in the genotype-environment interaction, illustrating the environmental influence in the interaction. PC2 captures the portion of the varietal-environment interaction not attributable to environmental impact. The AMMI model enables the assessment of characteristics and adaptability of tested varieties, thereby providing a theoretical foundation for the promotion of sugar beet cultivation in Northeast and North China.

GGE biplot

GGE biplots are frequently employed for analyzing data derived from multi-environment trials (MET) in crop research (Hasani et al. 2021). When examining genotype-environment interactions, GGE biplots offer a visual representation of the G main effects and GEI within the dataset. In this study, the mean matrix of the GGE biplot was constructed based on the environmental factors, with SVD being the primary component. Subsequently, a graphical representation was generated using the first two principal component scores (PC1 and PC2). This approach allows for the analysis of yield data from tested varieties regarding both yield and stability, aiding in the identification of optimal planting areas for each variety.

GT biplot

The GT biplot serves as a tool for analyzing the quality of target traits and their correlations. In this study, the GT analysis encompassed 16 sugar beet varieties, evaluating ten agronomic traits across seven pilot projects, including plant height, leaf length, leaf width, petiole diameter, SPAD value, brown spot disease, root rot disease, root yield, sugar content, and sugar yield. Construction of the GT biplot commenced with singular value decomposition of the standardized variety × trait data table (Scaling = 1 and Centering = 2)14. The first two principal components (PC1 and PC2) of varieties and traits were determined with singular values prioritizing traits (SVP = 2). These components were then scaled to create two-dimensional scatter plots with horizontal and vertical coordinates, forming the GT biplot. The vector linking the origin point to the trait icon illustrated the relationship between traits in the biplot, with the angle size between trait vectors representing the strength of their correlation. A smaller angle demonstrated a stronger correlation. Additionally, the projected distance from a cultivar to a trait vector indicated the cultivar’s relative value on that trait. Since traits had different units, they were standardized as follows:

$$Z=\frac{X-\mu }{\sigma }$$

where Z is the standardized value, X is the initial data of the trait, μ is the mean value of the trait, and σ is the standard deviation of the trait21.

GYT biplot

The GYT biplot is a graphical representation similar to the GT biplot but based on standardized data for variety, yield, and trait combinations. The yield-trait combination data in the GYT table were listed, and the necessary transformation of the trait data was performed. For attributes such as plant height (PH), leaf length (LL), leaf width (LW), thick petiole (TP), sugar content (SC), and sugar production (SP), the value of the yield-trait combination was obtained by multiplying the yield value (Y) by the trait value for each genotype (e.g., YxPH). A larger yield-trait combination value was considered better for this class. Conversely, for agronomic traits such as brown spot (BS) and root rot (RR), a smaller yield-trait combination value was preferred. Therefore, the yield-trait combination value was obtained by dividing the yield value by the trait value of each genotype. Subsequently, the yield and trait combination data in the GYT table were standardized using the "Z-standardization" method. Finally, the GYT biplot is constructed according to the standardized GYT data table. The principle remained the same as that of the GT biplot, with the traits in the GT biplot being replaced by yield-trait combination. The average of the standardized yield-trait combination values for each variety was defined as the superiority index (Sl) of the variety. The ideal index was used not only to rank genotypes but also to understand the strengths and weaknesses of each genotype through the GYT table. The ideal index is calculated as follows:

$$SI=\frac{\varepsilon Pij}{N}$$

where SI is the ideal index, Pij is the standardized value of genotype i for the yield-trait combination j in standardization, and n is the number of the yields-trait combination.

Statistical analysis

Microsoft Excel 2020 was used for data entry and arrangement, DPS V19.05 was employed for AMMI model analysis, and the GGEBiplot-GUI software package22 in R 4.2.2 was utilized for the GGE biplot. GT and GYT biplot analyses were conducted using the GGEbiplot package.

Results

AMMI model analysis

AMMI analysis was conducted on ten agronomic traits assessed in 2023 (Table 2). The results indicated that the E impacts on all the tested agronomic traits, excluding plant height, leaf length, and SPAD value, reached a significant level (p < 0.05). Notably, the environmental effects on traits like sugar content, root rot, leaf width, and petiole thickness reached a very significant level (p < 0.01). These environmental effects, encompassing traits such as sugar content, root rot, plant height, leaf length, leaf width, and petiole diameter, accounted for the highest proportion of variation, at 63.28, 69.60, 55.81, 59.85, 35.46, and 49.77. Furthermore, the interaction effects of yield and sugar content accounted for 53.93 and 56.64%, respectively. The genotypes of brown spot and SPAD exhibited the highest proportions at 65.05 and 48.48%. These interaction effects were reduced to PC1 and PC2. Significance testing revealed that, apart from leaf width and SPAD value, PC1 and PC2 of other agronomic traits reached highly significant levels. Root rot exhibited the largest percentage of the first two mutual principal components in the square sum of interaction effects (94.94%), while leaf width showed the smallest (58.61%).

Table 2 Analysis of variance and AMMI for related traits of tested varieties.
Full size table

GGE biplot analysis

The GGE-biplot model was employed to conduct PCA for G and GEI variation. PC1 and PC2 accounted for 71.31 and 18.96% of the total variation, respectively, collectively explaining 90.27% of the influence of genotype and GEI. To evaluate the regional adaptability of the varieties, a polygonal view of the GGE biplot was generated (Fig. 2), and the top-performing genotypes were identified in each environment. Notably, the same variety exhibited diverse performance characteristics across different regions, highlighting its adaptability to specific environments. In a biplot, polygons were formed by connecting markers representing varieties furthest from the origin, encompassing all other test varieties within. Additionally, the map featured vertical lines dividing the bi-standard map into sectors, each representing distinct regions. The variety situated at the apex of each polygon was deemed the best variety for that particular area, signifying its highest yield potential. In 2023, the seven trial pilots were categorized into three ecological regions, with E6, E1, E3, E4, and E2 consisting of one region. In this region, KWS4606 (G13) and KWS7748 (G1) exhibited the best performance. Meanwhile, E7 and E5 formed separate regions, with KWS8036 (G6) and KWS1130 (G3) emerging as the top-performing varieties in each respective area.

Fig. 2
figure 2

Adaptability of tested varieties. The horizontal and vertical axes represent the first principal component (PC1) and the second principal component (PC2), respectively, respectively. G1 to G16 represent species and E1 to E7 represent environment. The same below.

Full size image

GYT biplot analysis

In 2023, the initial two principal components of the GT biplot for 16 test varieties in North and Northeast China captured 53.04% of the variation sources of the interaction effect (PC1 = 29.78%, PC2 = 23.26%). Diverse vectors emerged by connecting each trait from the origin, and the correlation between each trait was assessed based on the angle between the trait vectors (Fig. 3). Root rot and brown spot exhibited negative correlations with yield, while leaf width demonstrated a negative correlation with yield. SPAD value was positively correlated with sugar content and leaf width, Leaf length and plant height were positively correlated with root rot and brown spot.

Fig. 3
figure 3

GT biplot of the tested varieties.( GY represent root yield.SC represent sugar content. SP represent sugar production.BS represent brown spot. RR represent root rot.PH represent plant height. LL represent leaf length. LW represent leaf width.TP represent thick petiole).

Full size image

The first two principal components of the GYT biplot of 16 tested varieties in North China and Northeast China in 2023 encompassed 93.04% of the variation sources of the interaction effect (PC1 = 80.19%, PC2 = 12.31%) (Fig. 4). This approximation could delineate the correlation between tested traits and the interaction between genotype and form of sugar beet varieties in North China and Northeast China. Each trait was connected from the origin to form a distinct vector, and the correlation between each trait was evaluated based on the angle between the trait vectors. As each vector-trait combination encompassed yield components, yield exhibited a positive correlation with each trait combination, aligning with the findings of Pearson correlation analysis (Fig. 5). Among them, the Angle between GY*SPAD and GY*PH and GY*LL is the smallest, and the significance is most correlated. The GYT biplot facilitated the display of yield-trait interaction among varieties. Notably, KWS4606 (G13) demonstrated superior performance in the GY-RR (GY*RR (-1)) combination, while KWS7748 (G1) excelled in the GY-BS, GY-SC, GY-SP, and GY-SPAD combinations. Additionally, KWS4310 (G11) exhibited commendable performance in the GY-LW and GY-TP combinations. The data indicated that compared to the GT biplot, the initial two principal components of the GYT biplot offered a more comprehensive explanation with better fitting, rendering more convincing results.

Fig. 4
figure 4

GYT biplot of tested varieties.(GY represent root yield.SC represent sugar content. SP represent sugar production.BS represent brown spot. RR represent root rot.PH represent plant height. LL represent leaf length. LW represent leaf width.TP represent thick petiole.)

Full size image
Fig. 5
figure 5

Pearson correlation coefficient analysis of yield and trait combinations. Red represents a positive correlation, blue represents a negative correlation, and the darker the color, the higher the correlation. GY represent root yield.SC represent sugar content. SP represent sugar production.BS represent brown spot. RR represent root rot.PH represent plant height. LL represent leaf length. LW represent leaf width.TP represent thick petiole.

Full size image

Comprehensive evaluation of test varieties

The outermost varieties were linked to form an irregular polygon, delineated by the vertical lines extending from the origin. Varieties positioned at the apex of the polygon represented those with superior performance within each region, aligning well with the traits prevalent in that specific region. G1, G13, G10, G4, G9, and G11 emerged as the top-performing varieties, with KWS7748 (G1) exhibiting a strong positive correlation with leaf width, yield, SPAD value, and other traits, making it the most suitable variety (Fig. 6). Among these, KWS4009 (G13) displayed the best compatibility with root rot and brown spot, establishing itself as the most resistant variety.

Fig. 6
figure 6

Function diagram of "Which-wins-where" of GYT biplot of test varieties.( GY represent root yield.SC represent sugar content. SP represent sugar production.BS represent brown spot. RR represent root rot.PH represent plant height. LL represent leaf length. LW represent leaf width.TP represent thick petiole.)

Full size image

A vertical line extending from the tested varieties and the average environmental axis (AEA) was drawn, with the length of the line indicating the stability of each variety. Shorter lines denoted greater stability, while varieties closer to the positive direction of the AEA axis indicated higher yields. KWS4606 (G13) and KWS7748 (G1), positioned closest to the leftmost end of the AEA axis, emerged as the top-yielding varieties (Fig. 7). However, while G13 lagged behind G1 in terms of stability, G1 emerged as the superior variety due to its optimal combination of yield and stability. Conversely, KWS8038 (G9) and KWS4315 (G7), positioned farthest to the right of the AEA axis, exhibited long vertical lines, signifying poor yields and stability among the tested varieties. The ideal index of the tested varieties, representing the average value of their standardized yield and trait combinations, corresponded to their coordinates on the AEA axis, with the magnitude of the ideal index reflecting the comprehensive evaluation of each variety. KWS4606 (G13) and KWS7748 (G1) boasted the highest ideal index values, standing at 1.48 and 1.31, respectively (Table 3).

Fig. 7
figure 7

Function diagram of "high yield and stability" of the GYT biplot.( GY represent root yield.SC represent sugar content. SP represent sugar production.BS represent brown spot. RR represent root rot.PH represent plant height. LL represent leaf length. LW represent leaf width.TP represent thick petiole.)

Full size image
Table 3 GYT data and ideal index of test varieties following standardization.
Full size table

Discussion

Given the diverse ecological conditions across regions, it is crucial to comprehensively assess the yield, stability, and applicability of new sugar beet varieties.The GYT biplot emerges as a novel statistical approach for comprehensive trait evaluation. This method enables the synchronous selection and comprehensive evaluation of yield and key traits through the yield × trait combination15,23,24, diverging from traditional single-trait assessments. Additionally, the GYT biplot underscores the primacy of yield in variety selection amidst multi-trait evaluations, with other traits only considered relevant under high-yield conditions. For example, while a sugar beet variety may exhibit robust disease resistance, its low yield renders it suitable only as a breeding parent rather than for direct production application. Thus, in sugar beet variety selection and assessment, prioritizing yield × trait combinations holds more significant value than single-trait selection. Consequently, the GYT biplot method proves more apt for variety classification and comprehensive evaluation, boasting enhanced practicality and scientific rigor.

Based on the analysis conducted in 2023 using the AMMI model, it was found that environmental factors predominantly influenced traits such as sugar content, root rot, plant height, leaf length, leaf width, and petiole thickness, serving as the primary sources of variation. Research on beet diseases has primarily has focused on root rot, brown spot, and root cluster diseases25, which collectively lead to an annual loss of approximately 16% of beet output26. Beetroot rot is intricately linked to local climate, soil conditions, and soil bacteria, primarily invading through beetroot tuber wounds. In addressing root rot prevention, conventional planting methods have proven more effective than relying solely on improved varieties, offering a direct and immediate solution. In regions with high soil pathogen prevalence, cultivating disease-resistant varieties while concurrently enhancing soil quality can significantly reduce root rot incidence and enhance yield. Furthermore, the interaction effect between yield and sugar production emerged as the predominant source of variation, impacting the breeding process by weakening the phenotype-genotype association and reducing heritability. Hence, when devising strategies to boost sugar beet yield, it becomes imperative to account for genotype-environment interaction27. Tailoring variety selection to specific environmental conditions can enhance the sugar content of sugar beet, aligning with the results of previous studies by Khodadad et al.28.Brown spot and SPAD genotypes represented the predominant sources of variation, emphasizing their significance in trait variation.

In this study, the AMMI model analysis revealed that, except for plant height, leaf length, and SPAD value, the environmental effect of the remaining traits reached a significant level (p < 0.05), affirming the substantial impact of the environment on trait expression. This result aligned with previous studies29,30,31,32. Given the unpredictable climatic conditions in North and Northeast China, expecting selected sugar beet varieties to adapt to every region and year is impractical. However, by delineating ecological zones, the unique adaptability of varieties to specific ecological conditions within certain regions can be elucidated. The GT biplot, a variant of the GGE biplot used for analyzing diverse agronomic traits, reveals correlations between different traits and characterizes a variety of traits. Nevertheless, its limitation lies in the challenge of conducting a comprehensive analysis of varieties based on multiple traits. At the same time, emphasizing high yield becomes imperative when evaluating the target traits, as assessing stability in isolation from crop yield holds little significance. To address the limitations of the GT biplot, this study introduced a novel approach called the variety-yield × trait or GYT biplot, aiming for a comprehensive assessment of tested varieties. GYT biplot integrated variety, yield, and target traits graphically, allowing for the simultaneous depiction of variety strengths and weaknesses. This methodology has gained popularity among researchers worldwide33,34. By leveraging GGE biplot-based ecological region divisions alongside GYT biplot technology to gauge the yield, stability, and trait requirements of tested varieties, precise positioning of promoted varieties can be achieved. In this study, although KWS8036 (G6) and KWS1130 (G3) did not exhibit particularly prominent ideal index values or positions on the AEA axis, their performance in the Hohhot pilot (E5) surpassed that of other varieties, demonstrating superior yield and stability.

Conclusions

In this study, the AMMI model, GGE biplot, and GYT biplot were employed to enhance comprehension of the interplay between environment and genotype. The primary objective of this study was to attain satisfactory selection advantages for the agronomic traits under scrutiny and to discern the performance disparities among various varieties across specific environmental conditions, ultimately identifying high-yielding and stable sugar beet varieties well-suited for cultivation in North and Northeast China. The AMMI model revealed that, apart from plant height, leaf length, and SPAD value, the environmental effects of the remaining traits reached were significant (p < 0.05), with sugar content, root rot, leaf width, and petiole thickness exhibiting exceptionally significant environmental effects (p < 0.01). The GYT biplot technique was employed to analyze the pairwise correlation between yield and trait combination of the tested varieties. The principal components PC1 and PC2 explained 80.19 and 12.31% of the genotype-trait interaction effect, respectively, demonstrating a predominantly positive correlation among all combined traits. Through the GYT biplot technique, sugar beet varieties with favorable yield, stability, and adaptability were identified. These findings align with practical production scenarios, affirming the suitability of this method for the regional evaluation of sugar beet varieties. Notably, KWS7748 (G1) exhibited high yield and stability, achieving the highest ideal index, offering valuable insights for the comprehensive evaluation and categorization of sugar beet varieties in North and Northeast China.