Lihua Liu

Data imbalance is common in multi-environment trials (METs) and complicates genotype and environment evaluations. To address this, linear mixed models (LMMs) and factor-analytic linear mixed models (FA-LMMs) were employed to estimate best linear unbiased predictions (BLUPs) for genotype effects. The highly unbalanced wheat grain yield data from the 2021–2025 national regional trials in the Northern Winter Wheat Region (NWWR) served as our case study. These BLUP values were then integrated into the genotype plus genotype-by-environment interaction (GGE) biplot framework to construct BLUP-GGE biplots. Subsequently, these biplots were systematically compared in terms of genotype evaluation, environment evaluation, and mega-environment delineation. Based on Akaike Information Criterion (AIC) and cross-validation, FA2 was chosen as the representative FA-LMM for comparison with LMM. For genotype evaluation, the two biplots produced highly consistent rankings. Based on the FA-LMM output, the top five genotypes were BH5318, BH2605, BH7868, JD39, and ZM8729. For mega-environment delineation, both approaches consistently identified one major and one minor mega-environment, with eight and three environments, respectively. The FA-LMM-based biplot provided clearer separation between them. For environment evaluation, the two biplots showed highly significant correlations for discriminating ability, representativeness, and desirability index. Luannan ranked first, followed by Zunhua, Baodi, Gu’an, Changping, and Shunyi, Xushui and Wuqing were intermediate, while Taiyuan, Tunliu, and Qixian performed relatively poorly. In summary, both methods are effective and reliable for analyzing highly unbalanced MET data, and show high consistency in genotype and environment evaluation. However, the FA-LMM-based BLUP-GGE biplot provides superior goodness-of-fit and more precise mega-environment delineation.

The accurate delineation of mega-environments (MEs) and the rigorous evaluation of test locations are critical for optimizing regional variety trial schemes, particularly when addressing unbalanced datasets from multi-year, multi-location wheat (Triticum aestivum L.) trials. This study aimed at refining the regional wheat trial framework in the Huanghuai Winter Wheat Region (HWWR) of China using an integrated BLUP-GGE biplot approach, which combines best linear unbiased prediction (BLUP) values with genotype main effect plus genotype-by-environment interaction (GGE) biplot analysis to account for temporal variability and experimental error. We systematically evaluated the BLUP-GGE biplot approach, focusing on its goodness of fit and its ability to resolve inter-location relationships. We further assessed test location representativeness, discriminating ability, and overall desirability via the BLUP-GGE biplot, and contrasted ME delineation outcomes between the traditional “which-won-where” polygon method and the test location clustering-based approach. The BLUP-GGE biplot explained 72.9% of total phenotypic variation, with all location vectors displaying positive correlations (maximum angle = 88.8°), confirming the ecological homogeneity of the target region and yielding robust evaluation results. Based on the ideal tester view, Puyang was identified as the most desirable location, followed by Zhumadian, Shangqiu, and Huixian, while Lianyungang and Suqian exhibited relatively poor comprehensive performance. MEs delineated by the “which-won-where” method showed strong inter-ME correlations and insufficient differentiation, whereas the location clustering-based method markedly enhanced inter-ME discrimination (maximum vector angle > 60°), stably partitioning the HWWR into three distinct MEs with clear cultivar–ME interaction patterns: ME1 (Lianyungang, Suqian, Fuyang, Suzhou, Guoyang, Huixian, Huai’an, Xinmaqiao, Huayin, and Yangling), ME2 (Luoyang, Xinxiang, Zhumadian, Shangqiu, Puyang, and Luohe), and ME3 (Baoji, Xuzhou, Yuanyang, Sheyang, and Xingyang). This study confirms the superiority of the BLUP-GGE biplot for analyzing unbalanced multi-year multi-environment trial data and validates a robust clustering strategy for ME delineation. The findings provide a scientific basis for optimizing wheat regional trial systems and facilitating precise cultivar deployment in the HWWR, and offer a reference for analogous studies on other crops or ecological regions.

Accurate delineation of mega-environments (MEs) and rigorous evaluation of test lo-cations are critical for optimizing regional variety trial schemes, particularly when addressing unbalanced datasets from multi-year, multi-location wheat (Triticum aestivum L.) trials. This study aimed to refine the regional wheat trial framework in the Huanghuai winter wheat region (HWWR) of China using an integrated BLUP-GGE biplot approach, which combines best linear unbiased prediction (BLUP) values with genotype main effect plus genotype-by-environment interaction (GGE) biplot analysis to account for temporal variability and experimental error. We systematically com-pared GGE biplots constructed from raw phenotypic data and BLUP values in terms of goodness of fit and their ability to resolve inter-location relationships. We further assessed test location representativeness, discriminating ability, and overall desirability via the BLUP-GGE biplot, and contrasted ME delineation outcomes between the traditional “which-won-where” polygon method and the test location clustering-based approach. The BLUP-GGE biplot explained 81.1% of total phenotypic variation, a substantial improvement over the conventional raw-data GGE biplot (62.4%). Raw-data GGE biplots exhibited highly complex inter-location correlations, distorted by unaccounted year effects and environmental noise, which hindered reliable location evaluation and ME classification. In contrast, all location vectors in the BLUP-GGE biplot displayed positive correlations (maximum angle = 83.9°), confirming the ecological homogeneity of the target region and yielding robust evaluation results. Based on the ideal tester view, ZMD was identified as the most desirable location, followed by SQU and PY, while LYG, BJ, SQ, and XY exhibited relatively poor comprehensive performance. MEs delineated by the “which-won-where” method showed strong inter-ME correlations and insufficient differentiation, whereas the location clustering-based method markedly enhanced inter-ME discrimination (maximum vector angle ≈70°), stably partitioning the HWWR into three distinct MEs with clear cultivar–ME interaction patterns: ME1 (HX, SZ, FY, SQ, LYG), ME2 (ZMD, SQU, GY, XX, HA, LH, XMQ, LY, HY, YL, PY, XZ), and ME3 (SY, YY, XY, BJ). This study confirms the superiority of the BLUP-GGE biplot for analyzing unbalanced multi-year multi-environment trial data and validates a robust clustering strategy for ME delineation. The findings provide a scientific basis for optimizing wheat regional trial systems and facilitating precise cultivar deployment in the HWWR, and offer a reference for analogous studies on other crops or ecological regions.

Molecular marker identification is a technique used to ensure fairness when submitting innovative crop varieties. Based on data from a collaborative experiment, our study appraised the reliability of the SNP molecular marker method applied simultaneously to China’s five major crops (wheat, rice, maize, cotton, and soybean) and proposed improved methods to (1) estimate detection uncertainty statistics and (2) graphically represent the detection results. We found that the detection method was quite reliable, as the average trueness rates for wheat, rice, cotton, soybean, and maize were 99.5%, 99.2%, 98.1%, 97.2%, and 96.2%, respectively, in sequence. The laboratory effect, genotype effect, and the effect of the interaction between the laboratory and the genotype all reached highly significant levels in the detection results, but their significance differed between crops. The proposed multi-genotype method confirmed the overestimation of the uncertainty statistics by up to 13% through the single-genotype method currently recommended in ISO 5725. The proposed Lab plus Lab by Genotype interaction (LLG) biplot provides a simple, intuitive, and efficient way to present, in two distinct and complementary biplot views, the trueness–precision of the detection results and the accuracy of the laboratories involved, respectively.