A neuroethological approach to memory and cognition in monkeys

Why Study Social Memory and Cognition in Monkeys?
Monkeys are the best model systems in which to study the neurobiology of complex cognition and memory. Brain organization and function have fundamental similarities across species of primates allowing us to learn a great deal about human brain function through the study of other primates. For example, we know that in both humans and monkeys, the medial temporal lobe brain region is essential for normal memory function. In humans, declarative memory (memory for facts and events) is typically assessed by asking people to remember highly relevant human-specific stimuli such as spoken or written words, stories, or common pictures. A common strategy used to study the neurobiology of memory in monkeys has been to train them to perform human-inspired memory tasks in which they are required to remember pictures or common objects. However, monkeys are not just little people in fur suits and the use of human-centric stimuli disregards the species-specific learning and memory problems monkeys have evolved to face in the wild. For example, many monkey species live in complex social groups involving dynamic supportive and antagonistic relationships. Moreover, it is widely thought that much of monkey mental energy is devoted to managing their social life' - 5. Millions of years of complex social life has produced primate brains that are especially well suited for perceiving, remembering, and acting on social information including social rank, kinship, and alliances within the troop. Despite this evidence. systematic tests of social memory in monkeys have been largely ignored in neuroscience research.

Our goal is to develop new behavioral tasks that test monkey social cognition. The development of these more naturalistic tests will offer two fundamental advantages over the typical memory tests presently used with monkeys in neuroscience research. First, memory tests using highly relevant and easily learned social information will provide closer parallels to the experimental approaches typically used to test declarative memory in humans. Second, this approach will allow experimental study of the brain basis of social cognition in primates. While this is an area of great interest in human psychology, nonhuman primate models of social cognition are not well-established. The development of these novel cognitive tasks for monkeys is significant because it will allow powerful experimental study of the centrally important human capacities for social learning, memory, and cognition.

The Social World of Rhesus Monkeys

Social groups of rhesus monkeys consist of a stable linear dominance hierarchy such that all female members of one matriline (i.e., a group of closely related females) outrank all the female members of a lower matriline6 -9 . A dominance hierarchy also exists between the animals within each matriline. Moreover, these dominance relationships are "transitive' such that if monkey A is dominant to monkey B and monkey B is dominant to monkey C, then monkey A is always dominant to monkey C. In addition to the stable linear structure of social dominance, there are also nonlinear transient alliances and friendships between monkeys that form and dissolve over shorter time periods. Indeed, accurate memory for both long-term and transient social relationships is critical for social and reproductive success in monkeys. Thus a "social stimulus set" provides fertile ground for testing a wide range of complex forms of memory and cognition in monkeys. One form of complex cognition that is relevant to this social stimulus set is transitive inference.

Transitive inference refers to the ability to flexibly manipulate knowledge of the relationships between items in a linear hierarchy (e.g., A>B>C>D>E). Previous studies have concluded that both monkeys 10-12 and rats 13-16 can solve tasks of transitive inference and that this function depends on the hippocampal region, part of the medial temporal lobe. However, these previous studies used non-naturalistic stimuli and artificially constructed hierarchies that took long periods of training for animals to learn. In contrast, an obvious stimulus set for a test of transitive inference in monkeys is the natural, and presumably easy to learn, social dominance hierarchy. Here we propose to develop such a test of social transitive inference for monkeys. We will teach animals a linear social dominance hierarchy using video clips of animals interacting in dominance situations. We will then determine if monkeys have learned the dominance hierarchy by asking them to indicate the relative dominance of two monkeys that were not previously shown together, thus assessing the ability of monkeys to apply transitive inference across the hierarchy.

Learning Who's the Boss: The Social Transitive Inference Task

In the initial training phase of our social transitive inference task, monkeys will be shown short videos of pairs of monkeys engaged in dominance interactions (e.g.. one monkey eating while another waits, one monkey displacing another, or one monkey threatening while the other one cowers). Monkeys will view films of the following dominance interactions: A+B-, B+C-, C+D-, D+E-, and E+F-, where "+" indicates the animal that wins the competition. These interactions correspond to the social dominance hierarchy: A>B>C>D>E>F. Immediately after each film, subject monkeys will be shown still pictures of the two animals in the film and will be rewarded for touching the `winner' (i.e., the "boss'") of the competition.

During the subsequent test phase of the task, we will test the monkey's memory of interactions by showing pictures of pairs of monkeys, but without a video showing one monkey being dominant to the other. Subject monkeys will be required to indicate which monkey of the pair is dominant. Most interesting will be critical probe tests in which monkeys will be challenged with images of monkeys B and D, whom they have never seen interact directly. Although B and D have never been seen together before, the transitive property of rhesus social dominance requires that B is dominant to D by virtue of their relationships to other monkeys in the hierarchy, A>B>C>D>E. Monkeys that can correctly determine this indirect relationship between monkey B and D must be applying transitive inference to do so. Based on data from field studies, we predict that monkeys will not only learn the initial dominance discriminations easily but that they will express transitive knowledge of this relationship during the test phase of the task.

Summary and Future Directions

The development of this social transitive inference task has both immediate and long­term applications for neuroscience research. The immediate application is in experimental studies designed to define the contribution of particular medial temporal lobe brain areas to social transitive inference. Hampton's lab will define the role of different medial temporal lobe brain areas in transitive inference by examining the effects of selective brain lesions. Suzuki's lab will use the task to examine the patterns of transitive-inference-related brain activity seen in the structures of the medial temporal lobe using neuronal recording techniques. It will also be of great interest to develop parallel tasks for social memory in humans whose brain activity can be studied using non-invasive brain imaging techniques. This proposal also has long-term implications for the development of a whole new family of naturalistic tasks for the study of learning, memory, and cognition using social information in monkeys. This new family of naturalistic tasks will provide a powerful set of tools with which to explore the neurobiology of a wide range of higher cognitive functions in monkeys.

Reference List

1. M. Tomasello and J. Call, Primate Cognition ( Oxford University Press, New York, 1997).

2.L. Barrett, P. Henzi, R. Dunbar, Trends Cogn Sci. 7, 494-497 (2003).

3. R. W. Byrne and N. Corp, Proc.Biol.Sci. 271, 1693-1699 (2004).

4. L. Marino, Brain Behav. Evol. 59, 21-32 (2002).

5. A. Jolly, Science 153, 501-506 (1966).

6. .T. B. Silk, R. M. Seyfarth, D. L. Cheney, Behaviour 136, 679-703 (1999). 7. D. L. Cheney and R. M. Seyfarth, Anim.Behav. 28, 362-367 (1980).

7. D. L. Cheney and R. M. Seyfarth, Anim.Behav. 28, 362-367 (1980).

8. D. L. Cheney and R. M. Seyfarth, How Monkeys See the World (University of Chicago Press, Chicago and London, 1990).

9. T. J. Bergman, J. C. Beehner. D. L. Cheney, R. M. Seyfarth, Sc ie nce 302. 1234­ 1236 (2003).

10. F. R. Treichler, M.A. Raghanti, D. N. Van Tilburg, J.Exp.Psychol.Anim Behav.Process 29, 211-221 (2003).

11. F. R. Treichler and D. Van Tilburg, Anim Cogn 5, 97-105 (2002).

12. C. A. Buckmaster, H. Eichenbaum, D. G. Amaral, W. A. Suzuki, P. R. Rapp, J . Neurosci, in press.

13. K. L. Agster, N. J. Fortin, H. Eichenbaum, J.Neurosci. 22, 5760-5768 (2002).