Dietary cognition in educated predators: implications for the evolution of prey defence strategies

Many insects defend themselves against predation with toxic chemicals. They often advertise their chemical defences to potential attackers (generally birds attempting to eat them) using conspicuous warning colouration, as seen in ladybirds, wasps and brightly-coloured butterflies. Typically, scientists have assumed that birds learn to associate the warning colouration with the effects of the toxins, and as a result totally avoid individuals with similar colour patterns in the future. However, this assumption has recently been shown to be wrong. Birds actually continue to eat toxic insects at low levels, even when they know that they are defended. They carefully control the number of toxic insects that they eat in order to gain the nutrients contained in the toxic prey, but ensuring they do not eat enough insects to cause them any serious health problems. The proposed experiments aim to investigate what factors influence birds' choices to eat toxic prey. We would expect that birds will eat fewer toxic insects when they can gain the nutrients they need from other sources i.e. when there are more alternative non-toxic prey in the environment, or when non-toxic prey have higher levels of nutrients. We would also expect the presence of other toxic species to influence birds' foraging decisions. However, we do not currently know what predators learn about their prey, or how they decide how 'valuable' toxic prey are as a food source. We will investigate this by giving wild-caught starlings, which are housed in a laboratory, sequences of toxic and non-toxic insects, and measuring which insects they choose to eat. The toxicity of the insects will be manipulated by injecting them with quinine solution (a mild toxin), and their nutrient content can be manipulated by injecting them with a protein solution. The results will allow us to understand what birds learn about the food they eat, how they use this information when deciding what to eat, and how the nutrition of a prey species influences the benefit of it being toxic. We will also produce mathematical simulations that investigate what foraging strategies predators' should use to maximize their nutrient intake whilst keeping their toxin intake as low as possible. These simulations will then be extended to investigate how predators' foraging strategies determine under what circumstances prey will benefit from being toxic, and when such prey will benefit from advertising their toxicity with bright warning colouration. These simulations will also be used to help us to understand how spraying crops with toxic chemicals (to stop birds eating them) influences birds decisions to eat toxic insects like ladybirds and bees. This might also allow us to decide how best to stop these insect species declining in numbers.

Dating back to Wallace and Darwin, anti-predator strategies have been a fundamental test-case for refining our understanding of evolutionary processes. None more so than for aposematism, where species have conspicuous warning coloration to advertise their toxins to predators. Predator cognition has been crucial to understanding the evolution of aposematism and mimicry (where species share the same warning pattern), and the process of avoidance learning by naïve predators has underpinned theories in this field. Our research takes a necessary and alternative view by considering the role of 'educated' predators in the evolution of defensive strategies. Predators do not just learn to avoid toxic prey, but learn to balance toxin intake with nutritional value - even toxic prey can be nutritionally profitable. Predators are not naive for long, but soon become knowledgeable educated predators. By studying the dietary cognition of educated predators, we will understand how they integrate information about nutrients and toxins in their dietary choices, and the effects that this has on prey defence strategies. We predict that our empirical data will significantly challenge current evolutionary theories, and lead to a re-evaluation of the study of aposematism and mimicry. We will also use computer simulations to predict optimal predator strategies in complex prey environments, and provide the first models to explore the co-evolution between cognitive strategies of predators and the nutrients and defensive strategies in prey. Whilst our main aim is to develop a new theoretical framework for the study of aposematism and mimicry, our empirical work can be extended and applied across a broad spectrum of different research fields, such as neuroscience, experimental psychology, evolutionary ecology and animal nutrition and welfare. Our work may also contribute to developing more effective avian crop repellents, and conservation strategies for declining aposematic species.