#DigitAg labelled post-doctorate
Integrative ecological modelling of Senegalese locust management
- Starting date: 1st February 2021
- Co-Supervisors: Cyril Piou, CBGP, Cirad
- Funding: US-AIDS
- #DigitAg: Axis 6: Multiscale modelling and simulation, Challenge 3: ICT and crop protection, Challenge 6: ICT and agricultural territory management, Challenge 8: ICT and agricultural development in Southern countries (Africa)
Keywords: Agent based modelling, pest management, knowledge integration, decision support system
Abstract: Locusts are among the most destructive of agricultural resources in the world. For some species, called locusts, phase polyphenism is a phenotypic plasticity that allows them to modify their behaviour and population dynamics depending on density.
Oedaleus senegalensis, the Senegalese locust, is a species of locust that can cause problems throughout the Sahel region. It presents a slight phase polyphenism, but above all a particular nutritional ecology that modifies its population dynamics depending on the content of sugar and protein of the plants on which it feed. As part of a project led by Arizona State University (Global Locust Initiative, PI Arianne Cease), experiments at the scale of agricultural plots are underway to verify how soil fertilization modify local dynamics of populations of O. senegalensis. However, since this species has significant migratory capacities, the results at the scale of a plot are not necessarily extrapolated linearly to the scale of a landscape or a region.
This locust species has already been the subject of the development of population dynamics models. However, these are not (or little) spatialized and do not take into account the local specificities of resources available for the development of populations. The objective of the project is to develop an integrative model representing the population dynamics of Oedaleus senegalensis in Senegal. This model will have to integrate the techniques of management of this species and in particular the technical itineraries of the farmers to modify the levels of proteins and sugars of the plants. Ultimately, the model will allow: 1) to extrapolate the results obtained at the plot scale to the landscape scale; and 2) to assess the necessary proportions of local fertilization to allow management of this pest avoiding significant agricultural losses.
An agent-based modelling approach is proposed. This approach should allow representing the landscapes by spatial entities in which the locust populations develop. Migrations and various traits of locust life histories will be taken into account using the literature on this species. Nevertheless, one of the difficulties of the model to be developed will be its ability to integrate data of different kinds: biological and ecological knowledge on the species, expert knowledge of farmers, models of plant growth, environmental information from remote sensing and possibly fine meteorological data. The model should also be a platform for discussion with stakeholders, so should have a user-friendly interface so that they can interact with the model and transfer the results of the simulations to real situations.
Contact: lucile.marescot [AT] cirad.fr – 07.58.74.74.99
Communications / Papers: