نوع مقاله : مقاله پژوهشی
نویسندگان
1 دانشیار، بخش تحقیقات اصلاح ذرت و گیاهان علوفهای، موسسه تحقیقات اصلاح و تهیه نهال و بذر، سازمان تحقیقات، آموزش و ترویج کشاورزی، کرج، ایران.
2 استادیار، بخش تحقیقات علوم زراعی و باغی، مرکز تحقیقات و آموزش کشاورزی و منابع طبیعی فارس، سازمان تحقیقات، آموزش و ترویج کشاورزی، شیراز، ایران.
3 استادیار، بخش تحقیقات علوم زراعی و باغی، مرکز تحقیقات و آموزش کشاورزی و منابع طبیعی جنوب کرمان، سازمان تحقیقات، آموزش و ترویج کشاورزی، جیرفت، ایران.
چکیده
کلیدواژهها
عنوان مقاله [English]
نویسندگان [English]
Objective
This study was conducted to evaluate the efficiency of different spatial correction methods and their impact on multi-environment analyses in maize breeding programs. Maize, as one of the most strategic crops, plays a crucial role in food security and livestock feed supply. Therefore, the selection of high-yielding and stable hybrids is a primary objective in breeding programs. However, spatial heterogeneity in field experiments can increase experimental error and reduce the accuracy of genotype effect estimation, an issue that becomes even more critical in multi-environment analyses.
Materials and Methods
In this study, 102 new maize hybrids along with three check cultivars were evaluated across three environments (Karaj, Jiroft, and Shiraz) using an alpha-lattice design with two replications. Data analysis was performed using a two-stage approach, in which weights derived from the inverse of the prediction error variance of the spatial models were incorporated into the multi-environment analysis to account for heterogeneity in observation precision. In the first stage, to address spatial heterogeneity, data from each environment were analyzed separately using three models: SpATS (Spatial Analysis of Trials with Splines), the first-order separable autoregressive model (AR1×AR1), and their combination (SpATS + AR1×AR1). In the second stage, multi-environment analysis was conducted using the Factor Analytic model to investigate the structure of genotype × environment interaction (GEI). Additionally, WAASB and WASSBY indices were calculated to assess genotype stability and performance.
Results
The results showed that the SpATS model, due to its high ability to capture large-scale spatial variation, significantly reduced prediction error compared to raw data and other models. In contrast, the AR1×AR1 model provided only limited improvement, and its combination with SpATS did not show a significant advantage over using SpATS alone. Although the alpha-lattice design partially controlled spatial variation, the application of spatial models, particularly SpATS, resulted in greater improvement in estimation accuracy and genotype ranking.
Multi-environment analysis revealed that the GEI exhibited a structured, interpretable pattern, and that the environments differed in their ability to discriminate among genotypes. Based on the WASSBY index, which integrates yield and stability, hybrids H31, H23, H22, H27, and H53 were identified as superior genotypes. Despite the superiority of these five hybrids, to maintain genetic diversity and provide opportunities for future breeding, approximately 20% of the top-performing genotypes based on yield and stability indices were advanced to the next stage of the breeding program.
Conclusion
Overall, the findings of this study indicate that the use of spatial models, particularly SpATS, combined with multi-environment analysis based on the factor analytic model provides a precise and reliable framework for improving genotype selection. This approach can significantly enhance decision-making efficiency in maize breeding programs.
کلیدواژهها [English]
)