Social bonds in birds are associated with brain size and contingent on the correlated evolution of life-history and increased parental investment

Biological Journal of the Linnean Society, 2010, 100, 111–123. With 4 figures Social bonds in birds are associated with brain size and contingent on the correlated evolution of life-history and increased parental investment SUSANNE SHULTZ* and ROBIN I. M. DUNBAR Received 10 September 2009; accepted for publication 22 December 2009 bij_1427 111..123 In birds, large brains are associated with a series of population-level phenomena, including invasion success, species richness, and resilience to population decline. Thus, they appear to open up adaptive opportunities through flexibility in foraging and anti-predator behaviour. The evolutionary pathway leading to large brain size has received less attention than behavioural and ecological correlates. Using a comparative approach, we show that, independent of previously recognized associations with developmental constraints, relative brain size in birds is strongly related to biparental care, pair-bonding, and stable social relationships. We also demonstrate correlated evolution between large relative brain size and altricial development, and that the evolution of both traits is contingent on biparental care. Thus, biparental care facilitates altricial development, which permits the evolution of large relative brain size. Finally, we show that large relative brain size is associated with pair-bond strength, itself a likely consequence of cooperation and negotiation between partners under high levels of parental investment. These analyses provide an evolutionary model for the evolution of and prevalence of biparental care, altricial development, and pair-bonding in birds. © 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 111–123. ADDITIONAL KEYWORDS: biparental care – DISCRETE – pair-bonding. The considerable variation in relative brain size both between and within bird taxa (Portmann, 1947) has variously been explained by either constraints (developmental or life-history strategy) or by adaptive benefits. Recent studies provide correlative evidence of adaptive benefits by demonstrating associations between relative brain size and architecture in birds and invasion success (Sol & Lefebvre, 2000; Sol et al., 2005a, b), resilience to population decline (Shultz et al., 2005), behavioural innovations (Lefebvre et al., 1997; Nicolakakis & Lefebvre, 2000; Sol, Timmermans & Lefebvre, 2002; Lefebvre, Reader & Sol, 2004), tool use (Lefebvre, Nicolakakis & Boire, 2002), and various other indices of behavioural flexibility. These characteristics suggest that large relative brain *Corresponding author. E-mail: size may confer a benefit through behavioural flexibility, which allows individuals to cope better with environmental unpredictability. Such adaptive benefits must outweigh the costs because there has been a consistent increase in relative brain size over evolutionary time observed across birds and other vertebrate taxa (Jerison, 1970, 1973; Finarelli, 2008). The most obvious costs incurred by growing and supporting large brains are slow lifehistory traits and high metabolic costs. Precocial developmental state at hatching is strongly associated with small relative brain size, whereas large brains are strongly associated with an altricial developmental state at hatching (Bennett & Harvey, 1985; Iwaniuk & Nelson, 2003). Additionally, relatively large-brained species have longer developmental periods and length of parental care (Iwaniuk & Nelson, 2003; Winkler, Leisler & Bernroider, 2004), © 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 111–123 111 British Academy Centenary Research Project, Institute of Cognitive & Evolutionary Anthropology, University of Oxford, 64 Banbury Road, Oxford OX2 6PN, UK 112 S. SHULTZ and R. I. M. DUNBAR groups, rather than the sheer number of associates. A better test of the social brain hypothesis is to look at the degree of bondedness and stability in relationships between individuals (Dunbar & Shultz, 2007; Shultz & Dunbar, 2007). Emery (2004) and Emery et al. (2007) identified relationships between relative brain size and social cohesion in birds; species found in small groups or pairs tend to have larger relative brains than those occurring outside of small, cohesive groups. In the present study, we apply the social brain hypothesis to birds by evaluating associations between relative brain size and the degree of cohesion in mating and social grouping patterns. We test several hypotheses about relationships between brain and telencephalon size and behavioural characteristics, development patterns, and ecological trends for 135 bird species. First, we predict that large relative brain size will be associated with stable and cohesive social grouping and reproductive partnerships. Second, we predict that models including both behavioural (bonding patterns) and development will best predict relative brain size across species. Third, we predict that, over evolutionary time development, biparental care and pair-bonding will be tightly correlated rather than evolving independently. MATERIAL AND METHODS DATA We compiled a database containing body and brain sizes from 135 species of birds from Portmann (1947). The brain size data in this sample are based on direct measurement of anatomical brains rather than endocranial volumes. Although there is a good correlation between intracranial volume and brain size (Iwaniuk & Nelson, 2002) and the sample of cranial volumes available is larger, we wanted to compare model performance between measures of forebrain (telencephalon) volume, total brain size, and relative brain size. A comparison is best carried out using a database with the same species composition. Additionally, using a dataset from a single source avoids introducing additional error caused aggregating datasets with different measurement protocols (Healy & Rowe, 2007). We also collated information on behaviour and ecology from the literature. Except where noted, behavioural and ecological characteristics for European and Northern African species were collated from species profiles in Cramp & Simmons (1977–1983), Cramp (1985–1992), and Cramp & Perrins (1993– 1994); data for ratites were taken from Davies (2002) and Madge & McGowan (2002) and those for Psittaciformes from Juniper & Parr (1997). We included all species from Portmann where we were able to find an explicit description for all measured characteristics. © 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 111–123 suggesting that the investment required per offspring is higher for parents of relatively large brained offspring. By contrast, young hatching at a precocial developmental stage reach independence at an early age and require less investment and protection by parents. Thus, large brained species are associated with a suite of life-history characteristics including altricial development at ha (...truncated)