Predicting life-history traits in a stored bean petst beetle Callosobruchus chinensis (Coleoptera: Chrysomelidae: Bruchinae) using machine learning
Life-history traits play an important role in insect population dynamics and ecological processes. The azuki bean beetle Callosobruchus chinensis is a common pest of stored legumes and is also widely used as a model species in ecological and evolutionary research. In this study, we tested whether machine learning models could be used to estimate several traits of C. chinensis, including elytral length, development time and adult lifespan. Experimental data were obtained from laboratory populations. The dataset included biological and environmental variables such as strain, treatment condition, developmental day, sex, temperature, and CO2. Six different machine learning models were tested, including linear regression, random forest, support vector machine (SVM), neural network, gradient boosting and AdaBoost. Model performance was evaluated using cross-validation. The coefficient of determination (R2) and root mean square error (RMSE) were used to measure prediction accuracy. Prediction accuracy differed among traits. Elytral length showed relatively higher predictability than the other traits, while development time was difficult to estimate in most models. Lifespan was easier to predict than the other traits, and the neural network produced one of the highest prediction accuracies among the tested models. Feature importance analysis also showed that factors such as sex and treatment condition contributed to variation in several traits. Machine learning models therefore helped reveal relationships among biological variables and life-history traits in C. chinensis. Combining ecological experiments with machine learning analysis may help improve our understanding of insect traits and may support future studies in insect ecology and pest management.