Details of Award
NERC Reference : NE/D010772/1
Predator-prey interactions and the evolution of prey aggregation
Grant Award
- Principal Investigator:
- Professor J Krause, University of Glasgow, Environmental and Evolutionary Biology
- Co-Investigator:
- Professor GD Ruxton, University of St Andrews, Biology
- Grant held at:
- University of Glasgow, Environmental and Evolutionary Biology
- Science Area:
- Terrestrial
- Marine
- Freshwater
- Overall Classification:
- Marine
- ENRIs:
- Natural Resource Management
- Global Change
- Biodiversity
- Science Topics:
- Population Genetics/Evolution
- Community Ecology
- Population Ecology
- Behavioural Ecology
- Abstract:
- Predation is a key factor in the structuring of ecological populations and communities. The behavioural adaptations of predators and prey to each other can have a fundamental effect on population processes by affecting how individuals in a population interact. This project considers the behavioural adaptations and counter-adaptations of predator and prey and how these translate into larger scale phenomena at the group and population levels. Specifically, we consider how the sensory systems of predators can become confused by large groups of prey, how this can be made worse by prey behaviour and appearance, how it can be ameliorated by predator behaviour, and how the behaviours investigated translate into observed natural variation in the form and composition of animal groups at the single-group and population scale. We will use the powerful computational modelling techniques of artificial neural networks and genetic algorithms, fully validated with experiments on predation by sticklebacks and humans. The predatory sensory system is represented by a multilayer artificial neural network with a static, previously obtained, mapping unit that generates the aforementioned 'confusion effect', and a trainable decision making network that interprets the cognitive map and chooses prey items in a way that ameliorates predator confusion. This decision-making unit is trained using a process similar to natural selection (a genetic algorithm) while the whole network is presented with numerous images of prey groups. Previous modelling and experimental studies indicate that the prey individuals chosen will be from the edge of the group, of odd appearance, from small groups etc. Here we will fully investigate key questions addresses by the project: How does prey grouping ecology affect predator success? How might natural population-level variation in animal groups (group size, composition etc) affect predator success? How does complexity of the prey group affect predator success and under what circumstances might a predator choose to specialise on a particular prey type? These simulations will generate multiple predictions that will be subject to full experimental validation with experiments on sticklebacks predating real and computer generated swarms of Daphnia, and humans predating virtual groups of prey on a computer screen. Images of the real and virtual prey groups immediately prior to strike by the study organism will be fed into the predatory neural network and the choice of network and real organism compared. Having validated the neural network model, and investigated the influence of prey group behaviour on predator success, we will consider the counter case of how predator prey choice may effect group formation of prey. Individual-based models of the prey shoal, herd, and swarm will be developed from previous publications and parameters given a genetic basis so that virtual prey groups can evolve in form and composition in response to predation by the neural network predator. The predator will remove individuals from the prey group, and evolution of prey group form examined. How does the optimal predator strike strategy affect the ultimate form and composition of a prey group? Can between-group variation in the form and composition of prey groups be explained by variation in predator characteristic and/or initial form of the groups? Other fundamental questions will be addressed. Ultimately, the whole predator-prey modelling framework will be integrated into a full coevolutionary system in which predator and prey evolve simultaneously. These sophisticated simulations will be used to probe fundamental population-level questions such as: Can natural variation between animals groups in form and composition be explained by predation? What types of predator might be associated with different types of prey structure in natural populations? How might predator strategy and prey group form change through evolutionary time?
- Period of Award:
- 1 Apr 2007 - 31 Mar 2010
- Value:
- £70,784 Split Award
Authorised funds only
- NERC Reference:
- NE/D010772/1
- Grant Stage:
- Completed
- Scheme:
- Standard Grant (FEC)
- Grant Status:
- Closed
- Programme:
- Standard Grant
This grant award has a total value of £70,784
FDAB - Financial Details (Award breakdown by headings)
| DI - Other Costs | Indirect - Indirect Costs | DA - Investigators | DA - Estate Costs | DI - T&S |
|---|---|---|---|---|
| £5,406 | £25,660 | £33,684 | £4,525 | £1,509 |
If you need further help, please read the user guide.