Faculty of Life Sciences, University of Leipzig, 04103 Leipzig, GERMANY.
Competition for access to females is the major principle governing the differential reproductive success observed among males in most animal species. Dominance hierarchies arise from differences in competitive abilities between individual males and often translate into reproductive benefits for high-ranking males (i.e. rank-based reproductive skew). Frequently, dominance hierarchies correspond to relatively simple variables that describe general physical abilities of individuals, such as body size or development of weaponry. In many species, however, the determinants of dominance status are more complex than that, and represent interactions between a variety of individual attributes, which collectively determine the ability to become dominant. Primates represent an interesting taxon to study this phenomenon because they live in complex social systems and further display elaborate cognitive abilities. Thus, the makeup of traits that determine which individuals will achieve high status is expected to be particularly complex in primates.
For several reasons, crested macaques (Macaca nigra) are a well suited taxon to study the mechanisms that underlie differential dominance achievement and maintenance between individual males. First, high dominance rank is associated with high mating and reproductive success, highlighting the importance of becoming as high-ranking as possible with regard to fitness. Second, male dominance hierarchies in crested macaques can be described as very dynamic in both, the level of males commonly migrating in and out of groups, and rank changes within the group that frequently occur outside the context of migration.
The overall aim of this thesis was therefore to investigate mechanisms underlying individual dominance rank trajectories in male crested macaques and to highlight possible, individual and social, determinants of how males can achieve and maintain the highest rank possible. In study 1, I address the problem of how dominance hierarchies can be reliably estimated even when conditions such as frequent migration events and changes within the hierarchy make the application of traditional approaches difficult, if not impossible. Studies 2 and 3 describe how male personality as an example of an intrinsic property can contribute to rank trajectories. Finally, in study 4, I investigate how coalitions, as an example for influences of the social environment, impact rank dynamics.
The data for this study were collected between 2009 and 2011 in the Tangkoko- Batuangus Nature Reserve in the north of Sulawesi, Indonesia. Study subjects were the 37 adult males residing in two social groups during this time. During focal animal sampling on these males, continuous data were recorded on social, aggressive and self- directed behaviour. In addition, the identities of adult individuals in spatial proximity were noted at regular intervals, as well as the focal animals’ position with respect to the core of the social group. In total, more than 2,000 hours of focal animal data were collected (mean = 66.1h, range = 0.6 – 130.0h per male, total = 2447.2h). Finally, two playback experiments were conducted to supplement the observational study of personality. With the presentation of dog bark bouts, I tested whether males differ in boldness, while neophilia was measured as the response to donkey brays. Statistical methods employed during data analysis include non-parametric tests, factor analysis and linear mixed models.
In study 1, I validate Elo-rating as a useful method to quantify dominance hierarchies. Elo-rating is rooted in the rating of competitive chess players and has a range of hitherto overlooked advantages over more commonly used methods of measuring dominance hierarchies in animals. These advantages are particularly important with respect to obtaining dominance measures in the context of overly dynamic relations between individuals. The applicability of standard ranking algorithms is limited to situations in which group composition is relatively stable and in which the majority of relationships between any pair of individuals are known. Elo-rating, in contrast, uses a relatively simple algorithm, during which an individual’s Elo-rating is updated after each single interaction this individual was involved in. The underlying principle of how ratings change reflects the expected as compared to the observed outcome of single dominance interactions. In this way, Elo-rating allows the estimation of dominance status on a very fine-grained time scale, without the need to aggregate dominance data over substantial periods of time. In addition, Elo-rating results in dominance hierarchies that closely match those derived from commonly used methods – given the data allow the application of these methods. Elo-rating therefore provides the necessary tool for reliable assessment of dominance status in dynamic systems, such as male crested macaque hierarchies. Furthermore, it allows to address conveniently questions related to individual rank trajectories.
In study 2, I suggest a personality structure for crested macaque males that consists of five personality factors: connectedness, sociability, anxiety, aggressiveness, and boldness. Connected males spent their time with a high diversity of female and male group members in spatial proximity. Sociable males spent more time grooming and had more diverse grooming partners. Anxious males were inactive, approached females rarely and showed high frequencies of self-directed behaviors. Aggressive males exhibited high rates of aggressive and threat behaviors towards other group members, with the notable exception of other adult males. Finally, bold males showed consistently stronger responses to the playback of dogs. The general makeup of crested macaque personality resembles to a high degree that of other macaque species. Yet, a notable difference to other macaques is the presence of connectedness, which covers aspects of social network diversity. These results not only contribute to our understanding of the evolution of personality structure in primates, they further set the grounds to investigate the possible adaptive value of differential expression of specific personality factors with regards to dominance rank – the topic of the subsequent study.
In study 3, I tested the relationships between the five personality factors determined in study 2 and dominance rank as determined with Elo-ratings. For this, I divided the study period into two-month blocks (necessary to obtain repeated personality scores for each male), for which I gathered corresponding current rank (Elo-rating), past rank trajectory (i.e. the difference between current Elo-rating and the Elo-rating corresponding to the preceding time block), and future rank trajectory (i.e. the difference between Elo-rating of the time block after and current Elo-rating). The first set of results revealed that four personality factors (connectedness, aggressiveness, anxiety and boldness) co-varied with concurrent dominance rank. However, none of the personality factors was predicted by past rank trajectory. The second set of results indicated that future rank trajectory was predicted by connectedness and anxiety, independent of the co-variation between current rank and personality scores. More connected and less anxious males were more likely to rise in rank as compared to less connected and more anxious males. These results indicate that personality can affect social careers and not vice versa. On the one hand, connectedness might reflect the ability of males to optimize their access to social partners in a dynamic environment. Here, the ability to diversify, rather than to concentrate, relationships might be crucial given the frequent changes among males caused by migration and rank changes. On the other hand, anxiety as correlate of future success might relate to physiological adaptions in male ability to cope with environmental and social stress. In sum, the results of study 3 highlight the salience of personality as an individual feature with potential impact on male dominance careers.
In the final study 4, I investigate the consequences of coalitions with regard to future rank of males. Overall, coalitions were relatively infrequent (0.03 events per hour), and the majority (about 65% of events) was composed of only males, while about 35% of coalitions included at least one female participant. All coalitions observed were targeted at single adult males. Although rare, coalitions had pronounced effects on individual ranks in the future. As could have been expected, a male’s role in a given coalition (participant or target) had the most pronounced effect, i.e. participating males generally rose in rank while targeted males dropped. However, these effects interacted with other important coalition characteristics. For example, coalitions in all-up configuration (all participants rank below the target) resulted in greater changes in both participants and targets as compared to bridging and all-down coalitions (target ranks between and below all participants, respectively). Additional results indicated that coalitions acted in an additive way to age-predicted patterns of rank trajectories, for example, while old males generally dropped in rank, they dropped less if they participated in coalitions. Further characteristics that influenced rank consequences of coalitions include the degree of feasibility (difference in rank between target and the combination of participants) and whether coalitions were composed of males only, or included females. These results contribute to our understanding of how coalitions impact dominance rank trajectories by highlighting the importance of several coalition characteristics with respect to the consequences of coalitions. Coalitions can therefore be regarded as an effective strategy employed by males to maximize their dominance rank, which in turn is associated with advantages regarding access to females.
This thesis provides evidence for the complex interplay between factors and their collective impact on dominance rank trajectories of male crested macaques. Two general classes of such factors can be distinguished: individual attributes and social environment. This thesis suggests that both, individual attributes (e.g. personality) and social environment (e.g. coalitions), have important consequences regarding the rank a male can achieve and whether he can maintain it once he reached it. In addition, individual characteristics and social environment are likely to interact with each other, for example by personalities that facilitate the formation of bonds and/or coalitions. Ultimately, if we want to understand what determines rank-based reproductive skew, we need to consider the complexity of mechanisms that govern rank trajectories that individual males will follow and we further have to take into account the likely diversity and cross-species differences of these mechanisms.