Predicting Insect Pest Phenology: calibrating laboratory-based model parameter estimates from field-based monitoring of agroecosystems – PI2P

Worldwide, insect pests are responsible for potential crop losses up to 80%, so that crop protection plays a central role in safeguarding crop productivity. Recent studies estimate that invasive insect species cost a minimum of US$70.0 billion per year globally, while substantial savings could be achieved. It is thus critical to develop models that accurately predict insect pest phenology to implement control practices and interventions at the right time. Despite the well-known sensitivity of insects to temperature, their potential response to ongoing climate change together with the one from their natural enemies, remains insufficiently documented.
Building models that accurately predict insect phenology and that are transferable to field conditions is thus a critical issue. The project’s objectives are to develop phenological models based on the available literature, to improve their performance by providing parameter estimates calibrated from field insect monitoring and taking into account agroecosystems parameters, and to use these models to study insect responses to climate change.
Phenological models can be built from development rate models that describe the time needed for a species in a life stage to reach the next one (e.g., egg to larva), as a function of temperature. The phenological model results from the combination of development rate models for each life stage. Temperature-dependent development rate is typically characterized under laboratory conditions in rearing units. Generally, the development rate is null under a critical minimum temperature, then increases slowly up to an optimal temperature for development, and then decreases rapidly to a critical maximum temperature. There are almost 40 development rate mathematical models in the literature and at least 3000 articles on the relationship between temperature and development rate. Development rate mathematical model choice is however still an issue with profound consequences as different models leads to different results. Also, whether a model is more appropriate than another or if different models should be used for different species is still a matter of debate.
PI2P will gather into a database published results from available articles in the literature and will provide guidance for development rate model choice, and new ways of characterizing development rate with applications in the face of ongoing climate change. PI2P will use the database to evaluate insect responses to climate change by using adequate mathematical models and new performance metrics (as opposed to traditional metrics such as thermal safety margin). PI2P will build phenological model taking into account agroecosystems parameters using an individual-based approach to take into account intraspecific variability, and use Internet of things sensors to get environmental variables close to those experienced by the insects in different ecosystems worldwide (Eastern Africa, the Andes, France).
The expected outputs are a better understanding of the relationship between temperature and development rate, a better understanding of phenology and its drivers, and a set of methods and tools from raw dataset to development rate models, new performance metrics, and phenology models with applications for agriculture and crop protection.