Maize (Zea mays L.) is one of the most important agricultural crops in the world; however, its production is often threatened by several harmful insects, one of which is the European corn borer, Ostrinia nubilalis (Hübner). This study aimed to examine the efficacy of several biological control methods against this pest. A randomized block design was used in the study, which included three treatments: parasitic wasps (Trichogramma brassicae), common green lacewings (Chrysoperla carnea), a combination of both agents and a control. The results showed that the treatment with T. brassicae wasps, as well as their combination with C. carnea predators, significantly improved the maize grain yield over the control treatment. The same combined treatment significantly improved the yield over treatments with only Trichogramma and only C. carnea predators, suggesting a higher efficacy of the combination of both agents in improving grain yield. Fewer tunnels and larvae in stalks, compared to other treatments, including the control, were also observed in this treatment. The reduced number of tunnels and larvae in stalks directly enabled plants to allocate more resources into grain development, which contributed to improved grain yield. In light of the growing focus on minimizing insecticide applications to mitigate environmental impacts, a combination of parasitic wasp and lacewings predator could be a suitable biological alternative to the use of chemical insecticides.

We evaluated a multi-output neural network framework for jointly analyzing maize grain yield (GY) and root lodging percentage (LP) using above-ground morphological traits measured under defined environmental conditions. To address model robustness, the multi-output neural network was compared with linear regression, elastic net, random forest, and XGBoost using repeated five-fold cross-validation, an 80/20 holdout split, and independent year-wise validation. Under repeated cross-validation, XGBoost provided the strongest average predictive performance for both traits, with R² values of 0.57 for GY and 0.67 for LP. The multi-output neural network showed moderate performance, with R² values of 0.49 for GY and 0.57 for LP. Final holdout performance for the neural network for GY and LP was R² = 0.64 and R² = 0.92, respectively. Year-wise validation showed weak temporal transferability because the two seasons differed not only in environmental conditions, but also in lodging mechanism. Repeated permutation importance identified ear width (EW), kernel row number (RNE), thousand kernel mass (KM1000), and kernel number per ear (KNE) as important predictors of GY, while LP prediction was most strongly associated with internode major diameter (IDmajor), ear length (EL), and the number of green leaves (NGL). Across both permutation importance and SHAP, only RNE and NGL were consistently shared between GY and LP. Supplementary ALE diagnostics indicated that RNE showed increasing model-estimated effects for both predicted GY and LP, whereas NGL showed a positive association with predicted GY but a decreasing or nonlinear association with predicted LP. These results show that joint modeling can support exploratory trait interpretation, but the predictive relationships remain environment-specific and should not be interpreted as causal or broadly transferable without further multi-environment validation.