The role of decision accuracy in the evolution of niche width

Why are some animals generalists and others specialists? Humans are the ultimate generalists, eating thousands of different types of food and making a living in a thousand different ways. Some species of fly are quite the opposite, however, and will spend their entire life living on one single species of plant. The advantages of generalisation seem obvious: an animal that can use everything should have no problems finding food or somewhere to live. The reasons why many millions of economically important animal species such as plant eating insects and many disease-causing parasites are so specialised, however, is less clear. These issues are important because the level of specialisation of an animal is a key factor in its ability to survive environmental change. Recently it has been suggested that specialisation may evolve to avoid confusion. Just as a person interested in sports searching a cluttered TV schedule might focus only on the word 'sport' and filter-out all other information, it is reasoned that an insect flying over a cluttered field of plants might try to focus on a smaller number of plants and become a specialist. Most experimental studies conducted so far do indeed indicate that specialists find it easier to locate and select suitable resources. I recently published a computer modelling study that suggests the conditions in nature in which this neural limitations hypothesis ('specialising to avoid confusion') could work are just the ones actually found in nature. The purpose of this fellowship is to establish just how important the neural limitations hypothesis is in the evolution of specialist animal lifestyles. My published computer model consists of a virtual world where animals search a cluttered environment for appropriate resources using eyes and a sensory system. I train the virtual animals using natural selection to become more specialised and then determine whether their nervous systems become less confused. I will reconstruct this model, this time including virtual noses instead of virtual eyes to look for resources. Modelling smell is important as most specialised animals actually sniff out food rather than looking for it. A 'smell' is also commonly a more complex and confusing signal than a 'sight' and so I predict that the neural limitations hypothesis is especially likely to be important in animals that use smell to locate resources. I will test this prediction with my model and then check to see that I can reproduce my model predictions in a real living system by recreating model simulations in a laboratory microcosm (a 'little world' in the laboratory) using the fruit fly, Drosophila, an animal that uses smell to find food. I will then take the insights I have gained from my computer model of smell and feed these into a more traditional class of model called an evolutionary genetic model that contains assumptions about the genetics of specialisation. This will tell me whether the neural limitations mechanism can drive the specialisation process to completion and even split populations into new species. The project will ultimately help us to better understand specialisation and how animals respond to environmental change. It could also help to protect crops against pests. Intercropping (planting more that one type of crop plant together in a field) is an increasingly popular agricultural method because it appears to lessen pest insect attacks. This reduction in attack could be due to the confusion pests experience in a more complex field environment. By studying how animals become confused we may be able to design intercropping strategies to even better confuse pests and so protect crops in a totally environmentally friendly way.