Selection shapes diverse animal minds

How animals perceive, process and respond to environmental cues is tightly tuned to the species-specific demands imposed by their ecology and life history. This specialisation is likely reflected in neural systems that support cognitive processes, as well as the behaviours expressed by those systems. In Heliconius butterflies the mushroom bodies - insect learning and memory centres - are significantly expanded relative to all other butterflies. Mushroom body expansion in Heliconius coincided with the evolution of a novel dietary shift towards active pollen feeding, and a spatial foraging behaviour, trap-lining, which is thought to require long-term spatial memory of visual scenes. I will discuss evidence that selection for trap-line foraging has reshaped Heliconius cognition along specific lines, reflected in both neuroanatomical specialisations and shifts in a restricted range of cognitive traits. By following the neural pathways that lead to and from the mushroom bodies, a mosaic pattern of neural adaptations is apparent, with shifts in cells and structures supporting visual and sparse coding within the mushroom body. Behavioural experiments closely mirror these changes, with improved performance in non-elemental learning and long-term memory in Heliconius, but specifically within a visual context. This provides a rare case where memory performance has been compared across species and sensory modalities, to identify a modality specific shift. These results are consistent with visual specialisation of the Heliconius mushroom body facilitating a specific enhancement of visual memory, likely due to the requirements of long-term foraging efficiency, and illustrate the precision with which selection can reshape animal cognition.

Dr Stephen Montgomery, University of Bristol, UK

Dr Stephen Montgomery, University of Bristol, UK

Dr Montgomery's primary interests are focused around how brains and behaviour evolve. He did his PhD on primate brain evolution with Dr Nick Mundy, at the University of Cambridge, and then held a series of fellowships at Oxford and UCL, with Professor Judith Mank, before returning to Cambridge to establish his own group. He is currently a Senior Research Fellow, funded by a NERC Fellowship and an ERC Starter Grant, and Proleptic Senior Lecturer at the School of Biological Sciences, University of Bristol. His group is interested in how brains adapt to different environments, how changes in brain structure produce behavioural differences, and how selection navigates developmental and functional constraints that may limit or channel the adaptive response. They take a comparative approach to tackling these questions, comparing molecular and phenotypic data across species. 

Work in comparative psychology and animal cognition assumes evolutionary—and hence psychological—continuity across the animal kingdom. The implicit assumption here is that it is possible to identify psychological phenomena in ways that are entirely species-neutral, and which allow us to map similarities and differences within and across different taxa. This assessment draws on psychological theories and capacities that have been, for the most part, derived from the study of our own species. If, however, human psychological capacities are “entangled”, in the philosopher, Alva Noë’s, sense, then we may be chasing a chimera. All is not lost, however, if we pare away the superficial and superfluous layers of human-flavoured intellectual coating that, ironically, are used to generate a sense of deep continuity. First and foremost, this entails recognising that evolution is not just about continuity but is also a diversity-generating process. Taking this on board allows us to generate a psychological landscape that can be explored very profitably by considering a more diverse range of theoretical frameworks in cognitive science. Using examples drawn from studies of memory and self-control, as well as recent work in neuroscience and brain evolution, I offer a view that may allow us to generate a better kind of continuity.

Professor Louise Barrett, University of Lethbridge, Canada

Professor Louise Barrett, University of Lethbridge, Canada

Louise Barrett was trained as an ecologist and anthropologist at University College London, UK. Following her PhD, Barrett took up a Postdoctoral Fellowship at the University of KwaZulu Natal, before returning to increasingly senior posts in the UK, and a Leverhulme Trust Research Fellowship. In 2007 she moved to the University of Lethbridge, where she is Professor of Psychology and Canada Research Chair (Tier I) in Evolution, Cognition and Behaviour. She has run two long-term field studies on baboons and vervet monkeys in South Africa, that have explored the complexities of social life and its potential links to the evolution of brain size and social intelligence. She studies the latter through the application of theories of embodied and inactive cognition. She is the author of Beyond the Brain: How Body and Environment Shape Animal and Human Minds. 

The challenges of social life have long been seen as key drivers of cognitive evolution. However, research has tended to focus on identifying human-like socio-cognitive traits in other animals while the fundamental question of whether and how animals benefit from making strategic social decisions has received relatively little attention. Here, we use automated field experiments to test whether wild jackdaws benefit from tracking information within dynamic social environments. For instance, while close proximity to other individuals can generate conflict, tolerating their presence may provide access to information or resources, so individuals may benefit from adjusting their tolerance in response to changing conditions. Accordingly, in an experiment where juveniles became valuable sources of information, we found that adults learned to increase their tolerance of juveniles and reduce aggressive displacements. This shows that jackdaws can adjust their information-use strategies to exploit new opportunities, in much the same way as the digital revolution has encouraged older people to learn from the youth. In a separate experiment, we find that jackdaws also modify their dyadic associations to interact preferentially with individuals that provide greater social foraging benefits. However, individuals in stable, long-term relationships maintained their associations regardless of the short-term benefits of strategic re-adjustment. This highlights a vital trade-off: investing in long-term bonds comes at the cost of missing out on benefits of wider social plasticity. We argue that understanding how animals manage trade-offs associated with different relationships is crucial to understand how sociality and cognition co-evolve.

Professor Alex Thornton, University of Exeter, UK

Professor Alex Thornton, University of Exeter, UK

Alex Thornton is a Professor of Cognitive Evolution at the University of Exeter (Cornwall Campus), where he runs the Wild Cognition Research Group. His work seeks to understand how the challenges faced by animals in their natural environments shape their mental processes, how the ability to learn from others affects the behaviour of individuals and groups, and how culture itself evolves. His research incorporates approaches from evolutionary biology, psychology and anthropology using a range of different study systems. Current research focuses primarily on cognition and behaviour in wild jackdaws, the cognitive requirements of human culture and the application of cognitive research in conservation. 

A recent development in understanding the evolution of cognition has been to adopt a within-species Behavioural Ecology Approach (BEA) in which individual differences in specific or general cognitive abilities are assayed, their role in shaping particular behaviours is assumed or demonstrated, and the fitness consequences of these differences are tracked. Cognitive abilities that correlate with fitness benefits are considered to be selected for and thus their evolution explained. Whilst intuitive, this approach raises questions about how different cognitive abilities are related to one another; how a cognitive ability, or suite of cognitive abilities, is (or are) linked to a behaviour, or a set of behaviours; and how selection feeds back to the cognitive and neural mechanisms underlying the behaviours. We propose a series of cartoon models that may depict these links. We then consider the implications for each model for our understanding of how cognitive abilities might evolve. This also allows us to derive a series of predictions about what sorts of relationships we might expect to see between measures of cognition and fitness if a particular model operates. We review existing studies that use the BEA to understand the evolution of cognition and examine which, if any, of the predictions are supported and thus which model(s) might operate. We recommend that future research adopting a BEA clearly specify their assumed model of linkage between particular cognitive abilities, the behaviour of interest and the fitness consequence, and interpret their findings in light of this assumed model.

Dr Joah Madden, University of Exeter, UK

Dr Joah Madden, University of Exeter, UK

Dr Madden works in the Centre for Research in Animal Behaviour at the University of Exeter. They started working on questions about sexual selection and mate choice, using spotted bowerbirds as a model system, before working on communication, conflict and cooperation in brood parasites and meerkats. They moved to the Department of Psychology, Exeter, in 2007 and since then they have been increasingly drawn to try to understand how selection may act on cognitive abilities that underpin natural behaviours. Dr Madden has used a within-species approach where they measure individual differences and explore resulting fitness consequences. They find this especially interesting when considering that the same (sets of) abilities may contribute to a disparate set of behaviours, raising the potential for conflicting selection pressures, and when considering why exaggeration of these abilities may not always be beneficial. Dr Madden has used a range of systems to ask these questions, with a recent focus on the genius of the bird world, the pheasant. 

The human brain is thought to be exceptionally adaptive. It has enabled us to control our environments and to thrive across the planet and beyond, with little indication that the human brain has exhausted its reach. The human brain's perceived adaptive nature and immense complexity have long made natural selection the only credible explanation for why it evolved. Despite this, I describe modelling results that find that the human brain size may not have evolved due to direct selection for it, but due to changes in developmental constraints. I explain how human brain expansion arises in this model as a side-effect of selection for increased ovarian follicle numbers and their genetic correlation with brain size, a correlation generated by development under a challenging ecology and seemingly cumulative culture. These results raise the intriguing possibility that the large human brain size is a spandrel rather than an adaptation and warn against persistent practices of considering adaptiveness or directional evolution as always caused by direct selection. I discuss challenges of establishing selection pressures with model-less empirical data, particularly regarding hominin brain size evolution, and how a modelling approach integrated with empirical data offers a way forward. I further discuss how modelling approaches may be used in the future not just to determine why the human brain could have evolved but to address why it actually did.

Dr Mauricio González-Forero, University of St Andrews, UK

Dr Mauricio González-Forero, University of St Andrews, UK

Dr González-Forero earned a BSc in Biology with a minor in mathematics at the University of Antioquia, Medellín, Colombia, in 2006. They then completed a PhD in Ecology and Evolutionary Biology in 2013 at the University of Tennessee, Knoxville, USA, supervised by Sergey Gavrilets. They subsequently did a 3-year postdoc at the University of Lausanne, supervised by Laurent Lehmann, where they formulated a mathematical model of brain life-history evolution. They then received a Marie Skłodowska-Curie Fellowship to work at the University of St Andrews, UK, supervised by Andy Gardner, to use the brain model to assess the ecological and social brain hypotheses for human brain size evolution. They have since remained at St Andrews formulating a mathematical synthesis of development and evolution to model the evolutionary and developmental dynamics of human brain size.

Despite the many challenges in measuring and quantifying cognition, it is nevertheless clear that cognitive abilities evolve, often due to natural selection. How does this process unfold? Is cognitive evolution rapid or slow? How strong is selection on cognition? Does it mostly depend on standing genetic variation or require populations to wait for beneficial mutations to arise? Answering these questions will provide novel insights into the mode and tempo of cognitive evolution. At stake is understanding whether species’ cognitive abilities are limited by their mutational potential or simply by the selective benefits of cognitive traits. Here, I provide an overview of our empirical work investigating the evolution of individual facial recognition in the paper wasp Polistes fuscatus. Population genomic analyses of paper wasps show that much of the strongest and most recent selective sweeps in these wasps have been targeted toward loci involved in learning, memory, and vision, all traits related to individual recognition. Critically, selection in other species and populations of paper wasps that lack individual recognition do not show a similar pattern of recent selection. Additionally, genes that are differentially expressed during social interactions are over-represented among those under strong selection. Selective sweeps are derived from a combination of both standing variation and de novo mutations in P. fuscatus, suggesting that the evolution of novel cognitive traits, such as individual recognition in P. fuscatus, may be at least partially mutation limited. Approaches for examining the mode and temp of cognitive evolution across taxa are discussed.

Dr Michael Sheehan, Cornell University, USA

Dr Michael Sheehan, Cornell University, USA

Michael Sheehan is an Associate Professor and Director of Graduate Studies in Neurobiology and Behavior at Cornell University, where he started as an Assistant Professor in 2015. His lab studies the evolution and consequences of social behaviour and cognition, especially as it relates to social recognition, using both paper wasps and house mice as empirical models. His work combines field studies, evolutionary analyses, physiology, neural recordings, as well as population and functional genomics to gain an integrated understanding of social behaviour and cognition. 

Reproductive success and its consequences for fitness are key to determining the role of natural selection in shaping cognitive abilities, and thus the all-important drivers for how and why cognition might change over evolutionary time. Demonstrating reproductive costs or benefits due to variation in cognitive ability, however, is hugely challenging. In some part this is because body size and body condition make the largest contribution to reproductive success, our ability to discriminate whether ‘being smart’ plays a role in fitness is already limited. Additionally, for many animals there is a multiplicity of ways to solve problems, only some of which will depend on their cognitive abilities. Escaping a predator, defending a food cache, or facing down a rival, for example, probably depend on energy levels, motivation, and morphological attributes. However, there may be specific instances in which key parts of animal’s life depend heavily, perhaps entirely on their cognitive abilities, and without them, the animal either won’t survive, or won’t get to reproduction.  Retrieving food caches seems to be one example, and evidence from nest building suggests that this might be another.  In the first, a specific cognitive ability (i.e. spatial memory) is key, while for nest building the pertinent ability is ‘experience’. Builders associate features of building, including material form and function, with consequent reproductive success. Until we can identify the specific cognitive abilities involved, perhaps ‘experience’ could also be used as a proxy in other contexts? If so, this might allow data collection in the field, and even in experiments that do not explicitly address cognition.  

Professor Sue Healy, University of St Andrews, UK

Professor Sue Healy, University of St Andrews, UK

From an undergraduate degree from Otago, New Zealand through a DPhil at Oxford on the behavioural and neural bases of food storing in birds, Sue has geographically moved steadily north in the UK, via Newcastle (Psychology) and Edinburgh (Evolutionary Biology) to St Andrews where she is now in the School of Biology. Her questions on animal cognition are addressed to testing spatial abilities in wild, free-living hummingbirds in the Rocky Mountains, and nest building in birds, in the lab and the field. With an interest in the evolution of the brain, she also examines the neural and hormonal bases of cognition where possible.