Project

Optimizing biological control through individual based modelling and genetic algorithms

At the moment, pesticides are the standard protocol for controlling insect pests in many parts of the world. The undesirable side effects of pesticides have caused many producers to increasingly adopt practices that suppress the pest populations by natural pest control mechanisms (e.g. biological control). Biological control makes use of natural enemies to suppress insect pest levels in the field or greenhouse. The impact of a release of natural enemies is not the simple accumulation of small-scale behavior of each natural enemy, but rather the linked behavior and interactions of all individuals in the agricultural system. Analyzing and exploring the effects of biological control practices is therefore very hard to accomplish in the field, providing a niche for research by ecological models. Ecological modeling can fulfill a predictive role by analyzing the impacts of natural enemies releases on both target and non-target populations.

Aim

The focus of this project is on the tri-tropic interactions between natural enemies, insect pests and plants on the population and landscape level. A better understanding of the factors that affect these interactions is essential in improving the current biological control practices in both natural and laboratory settings. Our aim is to provide better guidelines for companies, farmers and researchers.

Approach

This is achieved by the construction of individual based models that simulate the behavior of the organisms in a artificial landscape. The population dynamics in individual based models are the results of the interactions of organisms on the individual scale, which allows unlimited customization of the behavior of every individual in the model. This scaling up from individual behavior to population dynamics is key to our research question. The other important player in our approach is the genetic algorithm, a form of evolutionary computing that finds the solutions of optimization problems by making use of bio-inspired operators. It can find the target points for further improvement of biological control practices, by running multiple simulations of the individual based model.

Student Opportunities

We are open to applications for thesis projects! We have different thesis topics available.

  • Thesis Projects
  • If you are interested, feel free to send me an email at Bart.Pannebakker@wur.nl.