Project
LWV23117 INnovations for Smart Plant Insect Resistance Evaluation and testing (INSPIRE) - BO-67-001-032
The usage of pest quantification spans not only for pest monitoring, and crop protection, but also for plant breeding and variety testing. Previous research has demonstrated the potential of quantitative plant-insect resistance data to accelerate selection for resistant plants in plant breeding and testing biological control agents' effectivity; however, this requires conducting insect assays for evaluation and selection. On the other hand, current variety testing protocols for insect resistance are mostly done manually and data is analysed categorically, which leads to subjectiveness and errors.
Presently, phenotyping for insect resistance predominantly requires manual processes, and demand substantial time for data acquisition, incurring substantial costs. While automated phenotyping methods have improved the evaluation of plant physiological and morphological traits, the quantification of plant-insect resistance traits using fast and easy to use tools still has often been overlooked due to its challenges. This encompasses not only plant morphological traits like damage response (e.g., lesions and spectral alterations), oviposition, survival rates, and developmental stages, but also the appearance of plant’s physical defensive structures (e.g., trichomes). Additionally, the evaluation of plant-insect resistance is often done by scoring through categorical data, which restricts the depth of analysis as it may not capture small variations between categories. Moreover, ensuring data quality consistency often requires trained personnel.
Incorporating artificial intelligence (AI) offers the potential for rapid and objective data generation, enhancing both plant selection and resistance analysis processes. Through AI, it becomes possible to overcome obstacles in quantifying complex plant-insect interactions on images, thereby potentially opening new possibilities for enhanced pest management strategies that are both more efficient and accurate.
This project focuses on developing phenotyping technologies using mobile phone imaging for enhancing insect resistance testing, thus contributing to crop monitoring, crop protection, and plant breeding programs. Software developed based on AI-based algorithms will be coupled to different image acquisition hardware setups. The different tools will accelerate quantification of insects and determination of plant-insect resistance. More importantly, data generated from this project will be used as essential tools for exploration of new insect resistance traits for biological control agent efficacy testing, plant selection, and diagnostics. In conclusion, having rapid phenotyping tools can largely benefit by reducing labour costs and accelerating data acquisition and analysis, thereby meeting the insect-related challenges dealt by climate change.