Development of a new modular software tool for agroecological food web model development: Tiko'n

Agroecological food web modelling, which models trophic relationships between the species present in agricultural fields, has significant potential uses in the fields of theoretical and practical agroecology, specifically regarding the planning and evaluation of integrated pest management and biocontrol strategies. However, the practical use of such models with field data has been hindered by their complexity and difficulty in calibration, as well as the lack of software tools to streamline and standardise their development and evaluation. In this research we present a new modular software tool, Tiko'n, which allows users to rapidly (within a few lines of Python code) develop, calibrate and validate agroecological food web population dynamics models based on observed field population data. We then demonstrate its use in developing a food web model of a coconut pest in Sri Lanka and apply the model to assessing the potential impact of alternative biocontrol strategies. The software allowed for very rapid model development and adjustment and showed that smaller parasitoid releases may be more efficient than a single, larger application. It is the authors' hope that this tool will contribute to a more widespread use of agroecological food web modelling in research and agronomic practice.