| |
|
September-October 2025
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Enviromic prediction of sugarcane yield using nonlinear Gaussian kernels and climatic data
Tays Silva Batista , Wagner Barbosa, Germano Costa-Neto, Felipe Lopes da Silva, Bruno Portela Brasileiro and Luiz Alexandre Peternelli
ABSTRACT: The integration of environmental covariates into prediction models has demonstrated strong potential to increase the accuracy of genotype selection in plant breeding. When the number of phenotypic observations is limited, data simulation can be an effective strategy to expand training datasets while preserving model robustness. In this study, we tested the hypotheses that (i) incorporating environmental covariates improves the predictive performance of sugarcane yield models and (ii) increasing the number of simulated genotypes reduces prediction errors. Sugarcane yield, measured in tons of stalks per hectare (TSH), was predicted using Bayesian mixed models that incorporated Gaussian environmental kernels derived from NASA Power remote sensing data. Due to the limited availability of empirical data (11 genotypes evaluated in 8 environments), synthetic genotypes were generated using a segmented regression approach based on genotype-specific fitness and stability parameters. These genotypes were generated based on real data and validated against a subset of observed genotypes previously excluded from model training. Three validation scenarios (3G, 5G, and 7G) were implemented, with an increasing number of genotypes used in the training phase (183, 185, and 187, respectively). Three model structures were evaluated, differing in the inclusion of genotypic, environmental, and GxE interactions. The results confirmed that both environmental covariates and genotype simulation contributed significantly to the reduction of root mean square error (RMSE), with reductions of up to 82.5%. These findings highlight the relevance of enviromics and data augmentation strategies in early-stage plant breeding trials
[ FULL TEXT PDF 1-14 ] DOI: 10.22587/ajbas.2025.19.5.1
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
The Effect of Fiber-Adhesive Composite on Bond Strength of Plastic Plates- Part IV: The Effect of Temperature on the Mechanical Behavior and Mechanical Properties
Abdelaziz A. Noaman
ABSTRACT: The aim of this study is to use reinforced plastic composite for natural gas transportation at high temperature and pressure. The reinforced plastic composite is extensively used in many industries.In this work, the effect of temperature on PVC/PVC sandwich plates containing 0.95 Wt.% glass fiber and blended with different adhesives (epoxy, polyester, and polyvinyl chloride) is studied. Different temperatures (35 oC, 45 oC, 55 oC, 65 oC, and 75 oC) and three exposure times (2, 4, and 6 Hrs) are used. The mechanical properties, such as tensile strength, ultimate strength, modulus of elasticity, fracture energy, impact energy, and adhesion shear at different temperatures, are studied. The activation energy of PVC/PVC sandwich plates reinforced with glass fiber and blended with different adhesives was studied. It has been found that the mechanical properties of PVC/PVC sandwich plates containing 0.95 Wt.% glass fiber and blended with different types of adhesives are affected by temperature. In addition, it has been found that the activation energy of PVC/PVC sandwich plates reinforced with glass fiber and blended with polyvinyl acetate is the highest among epoxy and polyester. Based on the results obtained from the experiments, it has been found that the mechanical properties and high-temperature resistance of the PVC were significantly improved when adhesives and reinforcing materials were blended with the PVC.
[ FULL TEXT PDF 15-32 ] DOI: 10.22587/ajbas.2025.19.5.2
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
|
|
|