Arable: Modelling the impact of cabbage stem flea beetle on oilseed rape crops

Oilseed rape, grown for the production of edible vegetable oil, biodiesel and animal feed, was until recently the UK's third most valuable crop and its principal oil crop. However, pressure from the crop pest, cabbage flea stem beetle (CSFB), has resulted in large losses in yield and profits. Furthermore, as a consequence of the war in the Ukraine, imports of edible oils from the Ukraine and Russia (the world's two biggest exporters) are at an all-time low (https://www.economist.com/business/2022/05/07/the-war-in-ukraine-is-rocking-the-market-for-edible-oils). Thus, it has never been more timely to begin measures to restore yields of oilseed rape crops in the UK.

This proposal focuses on mitigating the damaging effects of the CSFB by developing a model to predict the occurrence and abundance of the pest so that prevention measures can be taken to protect the crop when it is most vulnerable. The CSFB has several life stages, e.g. egg, larvae, pupae, adult, which affect the crop in different ways. For example, the adults eat the newly planted crops in Autumn, whereas larvae burrow into the leaf stems causing damage to plants over winter and into the following spring. Furthermore, the timing of the life stages depends on local weather conditions. Air temperature is known to affect the developmental rates - e.g. the time taken for the eggs to hatch, the larvae to mature to adulthood etc, and wind speed and temperature affect the ability of the adults to migrate to new crops.
Our model is process-based, meaning that when we run the model it steps forward through time and tells us the current life stage and abundance of the pest at every moment in time. Currently we only have a simple prototype model of the CSFB life cycle which does not contain information on how local weather affects the pest phenology. One aspect of this proposal is to code weather dependency into the model so that we can predict year-to-year variability and the effects of climate change on pest damage to oilseed rape. Another aspect of the proposal is collating all data currently available for CSFB and using this for building or validating our new model.

The long-term aim is that our model will be developed into a decision support system (DSS) that can be used by farmers and agronomists for crop management. To this end, knowledge transfer is a key component of our project. A stakeholder group will be formed and will meet at least twice during the lifetime of the project; once at the start of the project and once in the second half of the project. The stakeholder group will be made up of representatives from the industry partners, BASF and Frontier and other interested parties. These stakeholders represent important industry actors that are influential in the development and use of DSS for pest management. The initial workshop will enable stakeholders to influence the design of the model from an early stage; the second workshop will allow stakeholders to trial the model in its development and hence inform on how the model can be further developed be applicable in real situations. Via ADAS, the model will be socialised with growers and advisors attending ADAS Farming Association conferences. There are several ADAS Farming Associations across England, and their membership consists of local farmers and agronomists. As such, these events represent an ideal opportunity to receive feedback from those tackling the issue of CSFB first-hand.

Thus, this proposal combines industry partners (BASF and Frontier), biologists with knowledge of CSFB and its chemical and non-chemical control (HAU and ADAS) along with mathematical modellers and statisticians (BioSS) to provide a fully inter-disciplinary approach to restoring yields of oilseed rape through mitigation of the CSFB.

We propose the development of a process-based, mathematical model utilising stage-structured modelling approaches such as those pioneered by Nisbet and Gurney (1983; Theoretical Population Biology, 23 (1), 114-135) to enable forecasting of the population dynamics and phenology of the distinct life stages of the cabbage stem flea beetle (CSFB), e.g., egg, larva, pupa and adult. This is a continuous time model based on the numerical solution of delay differential equations. Our prototype model has been built using software Kettle previously developed for generic stage-structured organisms (R package, stagePop (Kettle & Nutter 2015; Meth. in Ecol. & Evol. 6:1484-1490). This project involves incorporating weather dependency into the model which will result in a system of integro-differential equations but these can also be solved using the existing software, once the relationships have been added into the code. Many of these relationships are already published but where they need to be derived from data we will use the latest statistical methods. Thus the final model will be a combination of process-based modelling and data-based modelling to enable prediction of year-to-year variability and the effects of climate change in CSFB phenology. The model will be validated against abundance data (larval and adults) as well as unit tested and documented and made publicly available following BioSS's commitment to Open Science.