Research Projects

Abstract

This project addresses the critical challenge of managing Rhagoletis completa (walnut husk fly), a pest responsible for up to 80% yield loss of walnuts worldwide. Traditional monitoring methods involving labour-intensive yellow sticky traps are time-consuming, unsafe for operators to apply and prone to low accuracy and delays in detecting the pest, resulting in missed opportunities for pest control and therefore reduced yields. In addition, there is a lack of many insights that are crucial to our understanding of pest bionomics and their efficient control that have not yet been explored.

OBJECTIVE AND INNOVATION

The project aims to develop a self-sufficient, solar-powered automatic monitoring system equipped with high-resolution cameras, AI-based pest recognition, and a self-cleaning adhesive tape. The system will be integrated with a mobile and web app to provide real-time pest data, threshold alerts and predictive modelling. In addition, the traps will be compared to traditional manual monitoring traps to determine their efficiency.

The objectives include:

• Optimisation of automated traps for efficient pest detection, easy installation and maintenance.
• Comparison of automated systems with conventional methods across multiple orchards and trap placement heights to determine effectiveness.
• Establishing a national monitoring network to map pest distribution and dynamics under different climatic conditions.In purely scientific field, we will answer the missing questions about walnut husk fly that have not yet been researched:
• Identifying the specific factors attracting female pests to walnut fruit (e.g., volatile compounds, fruit shape, color, size, or metabolic composition).
• Determining thresholds for early and late infestations and their impact on fruit growth and development.
• Description of the effects of walnut husk fly infestation on the volatile components of the kernel and shell and the effects on the sensory and metabolic properties of the kernel.
• Investigation of the metabolic and sensory effects of walnut husk fly infestation on walnut oil from differently infested kernels (are damaged kernels are usable and how does that affects the walnut oil)

EXPECTED OUTCOMES

• A robust automated monitoring system that reduces labour by 70% and achieves 80% accuracy in pest detection.
• Accurate pest detection and pest dynamics record through the operation and establishment of a network of monitoring stations in different regions of the country
• Insight into the optimal placement of traps and pest dynamics for effective management.
• A significantly deepened understanding of pest thresholds and the impact of pests on daily fruit growth, which has improved yield assessment and fruit growing technologies, as well as the timing and applications of sprays.
• A better understanding of the metabolic changes underlying the attraction of pests to walnut fruit and the chemicals involved in this process.
• Improved understanding of walnut husk fly thresholds, the impact of pests on fruit growth and their exo-morphological characteristics
• Impact of the walnut husk fly on walnut and oil quality, focussing on volatile compounds, sensory properties and metabolites.
• Metabolites and volatiles will be identified to provide the scientific community and the public with new information on the quality of kernels and walnut oil.

With the successful implementation of this project, we hope to revolutionise pest monitoring and management in walnut orchards and ensure higher yields with lower and more targeted use of pesticides. In addition, a much deeper knowledge of the pest and its impact on yield and options for farmers with infested fruit will be gained and made available to the scientific community, farmers, legislators, students and the public, which will provide us with a better understanding of the pest and options for more efficient pest monitoring and management in the future.

Researchers

• External link to University of Ljubljana, Biotechnical Faculty Open in new window(Department of agronomy, Chair for fruit growing, viticulture and vegetable growing – Programe group Hortikultura= P4-0013) 
  (Aljaž Medič, Robert Veberič, Tomaž Pliberšek, Tilen Zamljen, Mariana Cecilia Grohar, Petra Kunc, Maja Mikulič Petkovšek, Jerneja Jakopič in Kris Pirih)
   
• External link to EFOS informacijske rešitve d.o.o Open in new window. 
  (Mateja Štefančič, Matej Štefančič in Katarina Mele)
   
• External link to Kmetijsko gozdarska zbornica Slovenije Kmetijsko gozdarski zavod Maribor Open in new window / Institute of Agriculture 
  (Rok Veber, Jože Miklavc, Boštjan Matko, Miroslav Mešl, Leonida Lešnik, Urška Škrabar in Anja Preložnik)

The phases of the project and their realization

WP1: Improvement of the automated system for monitoring the walnut husk fly, together with a mobile and web application for users

A self-sufficient automated device will be designed with high-resolution cameras, self-cleaning sticky tapes, and an artificial intelligence–based pest recognition tool. This will enable the identification of Rhagoletis completa with more than 80% accuracy. The system will be accessible through a user-friendly application that will generate alerts when pest thresholds are exceeded, provide real-time images, weather data, and predictive models, i.e. pest dynamics.

Expected outcome: This includes a fully functional automated system for monitoring the walnut husk fly. It will significantly reduce fieldwork and enable optimal use of pesticides. A comprehensive analysis will document the requirements for the automated system and highlight differences compared to classical methods. A trap design will be developed/modified for effective capture of the walnut husk fly, ensuring easy maintenance even when traps are installed high in tree canopies. An algorithm or artificial intelligence–based network (deep learning) will be developed for efficient recognition of the target insect from images generated by automated traps. In addition, modifications will be implemented in the core platform/application to fully integrate walnut husk fly monitoring. This will include changes to the web and mobile applications and adjustments of machine-learning models for predicting pest population dynamics. After two years of optimization, adaptation, and testing, the application together with the automated system will be fully functional and ready for commercial production and use. Real-time data will enable the application of proactive management strategies, while automated walnut husk fly monitoring will reduce fieldwork, improve worker safety, and increase the accuracy of records.

WP2: Comparison of the efficiency of the automated system with classical walnut husk fly monitoring

Automated traps will be compared with classical methods of walnut husk fly monitoring in several orchards, with the aim of optimizing their placement and efficiency. A national network of automated traps will be established in five Slovenian regions, allowing monitoring of walnut husk fly dynamics under different climatic conditions. The results will positively influence walnut husk fly management strategies and enable a reduction in yield losses.

Expected outcome: Successful implementation of this work package will lead to highly efficient and accurate recording of walnut husk fly catches and population dynamics across different regions of the country. This will significantly improve overall walnut husk fly management in orchards, reduce yield losses and the proportion of non-marketable fruit. Through access to real-time data and automated recognition of the walnut husk fly, conventional monitoring practices will be transformed and orchard management efficiency increased, resulting in higher yields.

WP3: Validation of automated traps compared to classical walnut husk fly monitoring

Within this work package, we will measure and compare exomorphological fruit traits, the content of volatile compounds, and other metabolites in infested and healthy walnut fruits. We will determine the damage threshold for the walnut husk fly and assess the impact of infestation at different stages of fruit development to determine how the timing of attack affects fruit development. We will evaluate walnut oil produced from kernels that are differently affected by the walnut husk fly. Specifically, we will test the sensory properties of the oil and determine volatile and metabolic compounds in the oil, thereby establishing whether affected kernels are still suitable for processing into walnut oil.

Expected outcomes: Through the implementation of this work package, we will deepen knowledge and understanding of the impact of the walnut husk fly on fruit growth. We will improve understanding of metabolic processes in walnuts and the chemical compounds involved, which are key in attracting the walnut husk fly to the fruit. In addition, we will determine the impact of damaged kernels on walnut oil quality. We will also identify different metabolites and volatile compounds both in the walnut kernel itself and in walnut oil.

With the successful implementation of this project, we aim to revolutionize the monitoring and management of the walnut husk fly in walnut orchards and ensure higher yields with reduced and more targeted pesticide use. Furthermore, we will gain deeper insight into the pest, its impact on yield, and the possibilities for processing infested fruits. All findings will be freely available under open access to the scientific community, farmers, policymakers, students, and the general public, enabling better understanding of the pest and more effective monitoring and management strategies for the walnut husk fly in the future.
 

Project partners

• Chamber of Agriculture and Forestry of Slovenia (KGZS) – Institute of Agriculture and Forestry Maribor
• EFOS d.o.o.