A trait-based approach to understand the evolution of complex coalitions in male mammals

Lucretia E. Olson 0 1 Daniel T. Blumstein 0 1 0 The Author 2009. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions , please 1 Department of Ecology and Evolutionary Biology, University of California , Los Angeles, CA 90095-1606, USA Coalitions occur when multiple individuals cooperate against a common opponent or for a common goal. Coalition formation is a complex behavior, typically described in highly social and cognitively complex species. Surprisingly, we know little about the social and environmental factors that may select for the evolution of coalitions. We studied the evolution of coalitionary behavior by first redefining it in a continuous way that acknowledges variation in the degree to which animals collaboratively work toward a common goal. We then examined the evolutionary association of coalition complexity with 3 social factors (estrous duration, group size, and presence of a dominance hierarchy) and 3 environmental factors (habitat type, diurnality, and diet type). We found that estrous duration, group size, and dominance hierarchy were significantly correlated with coalition complexity and thus conclude that social factors are relatively more important in the evolution of complex coalitions than are environmental factors. From these results, we infer that complex coalitions may be the product of social factors that reduce female monopolizability and encourage the aggregation of multiple males. Key words: coalitions, mammals, mutual tolerance, sociality. [Behav Ecol 20:624-632 (2009)] parative analysis to evaluate the extent to which specific social and environmental factors are responsible for the evolution of complex coalitions. - Btion between 2 or more individuals against a third during ehavioral ecologists typically define coalitions as cooperaan aggressive or competitive encounter (Harcourt 1992). Coalitions may also be formed for intergroup contests, where individuals of one group work together to defend a territory against outside males, as seen in cheetahs (Acinonyx jubatus), or to take over already established breeding groups, as seen in meerkats (Suricata suricatta) (Caro and Collins 1987; Doolan and Macdonald 1996). Although coalitions are formed by both males and females, the purpose of these coalitions usually differs. Female coalitions are generally formed to increase access to resources, as seen in coatis (Nasua narica), as well as more commonly in primates (Wrangham 1980; Gompper et al. 1997). Male groups may be formed for a variety of reasons. For instance, Cape ground squirrels (Xerus inauris) form all-male groups to increase their protection from predators (Waterman 1997), whereas coastal river otters (Lutra canadensis) form groups of mostly unrelated males to increase hunting success (Blundell et al. 2004). Because these types of cooperation are potentially explained by the mutualistic benefits of aggregation, we will restrict our focus here to coalitions formed to increase access to reproductive females. Reproductive coalition formation among males has been hypothesized by van Hooff and van Schaik (1992) to be difficult to explain because the main resource generally sought by males is successful fertilization, which is nondivisible; thus, it would appear to be most beneficial for a single male to attempt to monopolize as many females as possible and to be intolerant to the presence of competitor males. Therefore, the evolution of male coalitions is not readily explained (van Schaik 1996; Watts 1998). In this comparative review, we first propose a new metric to describe coalitionary behavior in a range of social species. Then, we use this metric in a comAs coalitions are now defined, their occurrence is largely limited to higher primates. An underlying assumption is that only primates are capable of assessing one anothers competitive ability and, using this information, are thus able to select coalition partners (Harcourt 1992). This definition, however, forces us to view coalitionary behavior as a trait that has sprung into being fully formed, with no intermediate steps along the way to explain its evolution. Yet, recent research has shown that males of less social species are also capable of aiding, or at least tolerating, each other. Feral horses (Equus caballus) jointly defend their harems against rival male intruders (Feh 1999), whereas striped hyenas (Hyaena hyaena) will mutually tolerate up to 2 other males on the same territory to maximize their access to females (Wagner et al. 2008). As coalitions are now defined, these species would not be termed coalitionary. However, these species seem to possess attributes of coalitionary behavior, and thus, there may be utility in a broader definition that recognizes these attributes. Moreover, because coalitions, as currently defined, either occur in a highly developed form or are not said to occur at all, we possess a limited understanding of the factors that have led to their evolution. A definition that acknowledges the graded nature of coalitionary behavior will allow us to conduct formal comparative analyses and determine the correlation of environmental and social traits with increased coalition complexity. We suggest that 3 key traits define complex coalitionary behavior: mutual tolerance, collaboration (against inter- or intragroup conspecifics), and preference for certain partners over others during intragroup competition. Generally, species such as baboons (Papio spp.) (Noe and Sluijter 1995) or bottlenose dolphins (Tursiops truncatus) (Connor et al. 1992) that possess all 3 traits are currently recognized as those that form coalitions. Species that have only a single trait can be viewed as having rudimentary coalitionary behavior. Although a temporal element may be associated with complex social interactions (e.g., reciprocal altruism requires actors to associate with each other over a period of timeTrivers 1971; Axelrod and Hamilton 1981), we did not include the temporal element on its own because it seems inherent to the other 3 traits. For instance, species with solely mutual tolerance usually interact for only a short amount of time, whereas those with more complex interactions tend to be longer term. We do not intend for these traits to be viewed as independent of one another, but rather as part of a continuum, with each trait building on those beneath it to define coalition complexity. Thus, species of low complexity will be characterized by mutual tolerance only; those of moderate complexity will have mutual tolerance and some form of collaboration; and coalitionarily complex species will possess mutual tolerance, collaboration, and within-group partner preference. The 3 traits will thus be viewed as a single continuous variable of coalition complexity. Mutual tolerance can be said to occur whenever a social group contains 2 or more males; thus, all rudimentary coalitions ar (...truncated)