Year 4 progress highlights:

This project is developing modern quantitative methods to measure model and predict biodiversity and ecosystem response to threats from pest and disease and other environmental drivers including climate change. Progress is good with a slight course adjustment for WS 2.1 following staff changes.

• Mathematical models for natural pest management.  Objectives are to extend models of cereal aphid disease transmission to investigate interactions between aphid species, aphid genetic diversity and viral diversity and the impact on disease spread. A model for two aphid species R. padi and S. avenae was calibrated and a paper describing the impacts and pinpointing critical life-history data-gaps was published The model has been updated to allow discrete time spatial movement and parallelized to improve efficiency to allow extensive scenario testing.
• Modelling tools for spatio-temporal population and community dynamics.  Objectives are to develop and apply spatio-temporal modelling and inference tools for population and community dynamics. We have developed Bayesian inference tools to model the spread of the Great Spruce bark beetle assessing risks to the West of Scotland Pest Free Area, and are exploring extension of these tools to community dynamics. Spatio-temporal quantitative genetic, population and individual based models were developed to account for local adaptation, phenotypic plasticity and environmental change to explore management strategies including the potential for local adaptation and assisted migration to increase tree population resilience to climate change.
• Process-based life cycle modelling.  Aims to enhance control of crop pests and animal health through the development of methods for stage structured modelling. Applications include potato cyst nematode (PCN) and modelling migration and life cycle of the cabbage stem flea beetle, a serious pest of oilseed rape in England, and likely to become so  in Scotland. A data-driven model using local weather conditions to predict beetle arrival after the crop is sown and a process-based model predicting how different life stage durations will change with temperature, were completed and published in peer-reviewed journals.
• Hierarchical spatio-temporal modelling of aphid dynamics.  Objectives are to develop hierarchical spatio-temporal models for aphid dynamics and to build apps for non-expert end-users to guide actions.  A statistical methodology paper is in review. A prototype app for in-season predictions of estimates and uncertainties for the trajectories of aphid species and overall vector pressure per disease type was presented at the RSS statistical conference. An agronomist who advises potato growers has provided guidance and feedback for disseminating the work.
• Modern approaches to time series data analysis: nonlinear Hidden Markov models and change points. Objectives are improving modelling of ecological time series, with a focus on animal movement data as a test case study. A review paper, Understanding and predicting animal movement in the Anthropocene, was published, which provides recommendations for improved modelling, data collection, and better linkages with policy. Simulation studies to test change-point detection in animal movement data were designed.
• Statistical methods for biodiversity modelling.  Objectives are developing computational tools to evaluate reliability, limitations, and scalability of JSDMs under realistic ecological conditions. jsdmSimR, an R package simulating biodiversity landscapes with controlled environmental structure, species correlations, and sampling biases is in development, enabling benchmarking JSDM implementations and identifying strengths and weaknesses of current approaches.

Year 2 progress highlights:

• Natural pest management: An initial model of two key aphid species, including one that spreads blueberry scorch virus, has been developed. Next steps will add the role of natural enemies such as parasitoids to help design more sustainable pest control strategies.

• Tree population dynamics: New software now allows researchers to model long-lived species such as trees, accounting for competition between seedlings and mature individuals. This will help explore whether natural adaptation or assisted migration is the best way to protect forests from climate change.

• Pest life cycle modelling: A general framework has been created to model stage-based organisms like potato cyst nematodes and cabbage stem flea beetles, major crop pests. Two scientific papers are nearly ready for publication.

• Predicting aphid outbreaks: By linking aphid arrival to daily temperatures, researchers are developing predictive models that could underpin a future farmer-facing app to forecast pest outbreaks.

• Analysing time series data): Work is advancing on more accurate statistical methods to detect changes in complex ecological time series, with new approaches being trialled.

• Biodiversity modelling: A powerful simulation package has been created to test how well statistical models cope with challenges like biased data and large-scale biodiversity problems.

This programme is developing advanced statistical and mathematical tools to better understand how natural populations, biodiversity, and ecosystems respond to change.

Year 1 progress highlights:

• Collaborations: New research groups have formed within the programme, including a dedicated team on agricultural pest modelling, while partnerships with the James Hutton Institute, Liverpool University, Harper Adams, and ADAS are linking models with real-world data and field experiments.

• Natural pest management: A spatial model has been built to test how pests and their natural enemies behave at field and landscape scales, with data now being collected from new field experiments.

• Tree biodiversity under climate change: Work is underway to model how tree populations respond through both genetic adaptation and plasticity (short-term flexibility), improving predictions of forest resilience.

• Crop pest life cycles: A stage-based model of the cabbage stem flea beetle, a major threat to oilseed rape, has been developed to guide more effective crop protection strategies.

• Aphid forecasting: UK-wide models of virus-spreading aphids have been built, with prototype apps and indicators now available to help potato growers and other end users.

• Analysing ecological change: Progress has been made on advanced statistical methods to detect changes in ecological time series data, with new code and simulations developed.

• Biodiversity modelling: A framework has been created to improve models that predict how multiple species respond to environmental change, laying the groundwork for better climate change impact assessments.