Stereotypy and perseveration in captive European starlings: consequences for decision-making

Repetitive and unvarying behaviour, such as the pacing of a polar bear in a zoo enclosure or the circling of a mouse in a laboratory cage, is common in captive animals. These mysterious behaviour patterns with no obvious function are collectively known as stereotypies, and their explanation has been a topic of much debate. Do they have a function in allowing animals to cope with their inadequate artificial environments, or are they behavioural symptoms of underlying psychiatric disease? Evidence supporting the latter hypothesis is beginning to accumulate. Experiments with caged birds have played an important role in linking stereotypies in captive animals with similar behaviour patterns in humans with psychiatric illness such as schizophrenia. Stereotyping birds and humans both show very similar behaviour on a computer game, known as the gambling task, in which they have to choose which of two buttons will provide reward. In comparison to non-stereotyping individuals, stereotypers show a tendency to repeatedly make the same choice, a phenomenon known as perseveration. Both stereotypy and perseveration are likely to be symptoms of dysfunction in the brain systems responsible for the sequencing of behaviour patterns. Specifically, it is believed that humans and animals with these symptoms have damage to the pathways involved in stopping the current response and allowing the organism to switch to a different one. This hypothesis has worrying implications for the widespread use of laboratory animals as models of normal human behaviour. Many of the standard behavioural tasks performed on laboratory animals are likely to be affected by perseveration, however, at this point we have very little idea how big this problem is likely to be. The aim of our research is therefore to explore the links between stereotypic behaviour and performance on some standard behavioural tasks used to assess normal, adaptive behaviour. We have chosen to study captive European starlings for two reasons: first because the existing evidence for a link between cage stereotypies and perseveration comes mainly from birds, and second because wild-caught birds are important for the study of adaptive behaviour in the lab. It is therefore particularly important to know how much their behaviour might be altered when they develop stereotypies. We will create groups of birds differing in the degree of stereotypy they exhibit by using manipulations of both developmental history and current caging. Birds that are caught from the wild as adults and housed in large cages should show the lowest levels of stereotypy, whereas birds taken from their nests as chicks, hand raised and then housed in small cages should show the highest levels. We will measure stereotypies by analysing videotapes of the birds' behaviour, both by hand, and by using computer software that automatically tracks the position of a bird in a cage and looks for patterns in its behaviour. In addition, we will test birds on a number of computer-controlled tasks: a gambling task similar the one mentioned above, a timing task, and a task that tests sensitivity to risk. We have chosen these latter two tasks because they are commonly used in studies of behaviour. Finally, we will look for links between the degree of stereotypy an individual bird exhibits and its behaviour on the tasks. Our study will provide the first data exploring how stereotypic behaviour in laboratory animals affects normal behaviour on a range of common tasks. These data will allow us to assess the potential scientific costs of studying animals with stereotypies. If these costs are significant, our findings will provide strong support for further improving the cage environments of laboratory animals in order to prevent them from developing stereotypies.

Stereotypies, defined as behaviour patterns that are repetitive, unvarying and with no apparent function, are common in laboratory animals. A recent hypothesis suggests that animals performing stereotypies may have pathological changes to the brain mechanisms responsible for sequencing normal motor behaviour. If this is correct, it places major doubt over the use of such animals as models of normal behaviour. The aim of this proposal is therefore to explore the consequences of stereotypic behaviour for performance on some operant behavioural tasks typically used to explore adaptive behaviour. We hypothesise that animals with high levels of stereotypy will show increased perseveration on operant tasks, with clear predictions for the results obtained. We have chosen to study captive European starlings because wild-caught birds are important model systems for the functional and mechanistic study of adaptive behaviour. Birds differing in the degree of stereotypy will be created via manipulations of both developmental history and current caging. Behaviour will be recorded using standard manual scoring techniques and a state-of-the-art automated system. Stereotypies will be quantified using standard counting, and two novel techniques capable of detecting subtle differences in behaviour: Markov chains analysis of behavioural sequences and T-pattern analysis. Birds will be tested on three operant tasks: a gambling task, a fixed-interval timing task, and a risk-sensitive foraging task. Results obtained from these tasks will be correlated with data on stereotypies. We will provide the first data exploring how stereotypic behaviour in laboratory animals affects adaptive behaviour. We will also provide further evidence on the details of the changes in behavioural control present in stereotypic animals. In addition, the project will provide data on the relative contributions of developmental history and current caging to the development of stereotypies.