Towards Automatic Monitoring and Population Dynamics Modelling of Microscopic Insect Pest-predator Interactions in Horticultural Crops

Technology plays a vital role in meeting global food demand, which is forecasted to increase by 56% in 2050. Controlling insect pests is one of the major challenges that can be addressed using technology, as insect pests are one of the primary causes of yield losses. Typically, insect pests are controlled using insecticides, which have negative impacts on the ecosystem and human health. Thus, Integrated pest management (IPM) has been employed in many crop fields for effective and sustainable pest management. Some key components of IPM are (1) the use of biological control agents (i.e., natural enemies of insect pests including predatory insects and parasitoids), (2) effective pest monitoring; and (3) insect pest population modelling and forecasting. Such techniques are mostly exercised using manual techniques, which are labour-intensive and time-consuming. Computer science technology can potentially be applied to automate such current manual practices.

This research aims to improve the current practice of insect pest monitoring and population modelling for better IPM. In the first stage, we conduct a Systematic Literature Review (SLR) on existing image-based insect classification methods to identify key research gaps. To bridge these research gaps, the second stage of this research develops an automated system to distinguish morphologically-close microscopic species (i.e., species-level insect classification). In the next stage, we study and model the complex relationship between insect pests with their natural enemies, climate conditions and the growth stage of the host plant. Subsequently, the modelled relationships are applied to simulate the temporal population dynamics of insect pests to help decision support systems in crop fields. As a case study, Western Flower Thrips (WFTs), a major worldwide crop pest, are targeted as the insect pest in this research. This is due to: (1) the huge number of thrip species; (2) their small size (~1mm); (3) the similar morphological characteristics among thrip species; (4) their great resistance to most chemicals; and (5) the feasibility of accessing data related to WFTs in Austrailia strawberry farms. We believe that the outcomes of this research will help to alleviate the drawbacks in current IPM practices and boost crop yield and quality with reduced dependence on insecticides.

Publications

  • [1] Amarathunga, Don Chathurika, et al. "Methods of insect image capture and classification: A Systematic literature review." Smart Agricultural Technology 1 (2021): 100023.
  • [2] Amarathunga, Don Chathurika, et al. "Fine-grained image classification of microscopic insect pest species: Western Flower thrips and Plague thrips." Computers and Electronics in Agriculture 203 (2022): 107462.
  • [3] Ratnayake, Malika Nisal, et al. "Spatial monitoring and insect behavioural analysis using computer vision for precision pollination." International Journal of Computer Vision 131.3 (2023): 591-606.
  • [4] Amarathunga, Don Chathurika Kshanthi and Parry, Hazel and Grundy, John and Dorin, Alan, A Predator-Prey Population Dynamics Simulation for Biological Control Of Frankliniella Occidentalis (Western Flower Thrips) By Orius Laevigatus In Strawberry Plants. Available at SSRN: https://ssrn.com/abstract=4548656 or http://dx.doi.org/10.2139/ssrn.4548656

Project Lead

Don Chathurika Amarathunga (PhD Candidate)

Main Supervisor

A/Prof Alan Dorin

Co Supervisors

Prof John Grundy, Dr Hazel Parry

Thrips in a strawberry field

Figure 01: Thrips in a strawberry field

The main research areas of the project

Figure 02: The main research areas of the project