ACKNOWLEDGMENTS 16 We thank Yvonne Cämmerer and Bettina Rinjes for help in maintaining the crickets, 17 Nedim Tüzün and Silvana Meli for help in performing the experiments and scoring the 18 videos, Anne Rutten for building the database, Jon Brommer for feedback on initial 19 data analyses, and Petri T. Niemelä for feedback on the study design. We thank two 20 anonymous reviewers for constructive feedback. 21 TITLE 22 Interacting with the enemy: indirect effects of personality on conspecific

more »

... ression in 23 crickets 24 RUNNING TITLE 25 Indirect effects of personality on aggression 26 ABSTRACT 27 In animal contests, individuals respond plastically to the phenotypes of the opponents 28 that they confront. These 'opponent' -or 'indirect' -effects are often repeatable, e.g., 29 certain opponents consistently elicit more or less aggressiveness in others. 30 'Personality' (repeatable among-individual variance in behavior) has been proposed as 31 an important source of indirect effects. Here, we repeatedly assayed aggressiveness of 32 wild-caught adult male field crickets Gryllus campestris in staged dyadic fights, 33 measuring aggressiveness of both contestants. Measurements of their personality in 34 non-social contexts (activity and exploration behavior) enabled us to ask whether 35 personality caused indirect effects on aggressiveness. Activity, exploration, and 36 aggressiveness were positively associated into a behavioral syndrome eliciting 37 aggressiveness in conspecifics, providing direct evidence for the role of personality in 38 causing indirect effects. Our findings imply that a multivariate view of phenotypes that 39 includes indirect effects greatly improves our ability to understand the ecology and 40 evolution of behavior. 41 KEY WORDS: animal personality, behavioral syndrome, aggression, indirect effects, 42 social behavior, crickets 43 65 to opponent body size (Moore et al. 1997). 66 Indirect effects can have major consequences for ecological and evolutionary 67 processes especially when caused by repeatable differences in phenotypes of social 68 partners that are underpinned by genetic variation. In such cases, the genes of an 69 individual influence the (behavioral) expression of a trait expressed by an interacting 70 individual; such effects are termed 'indirect genetic effects' (IGEs) (Wolf et al. 1998, 71 1999; McGlothlin et al. 2010). IGEs can greatly influence evolutionary processes 72 particularly when they are correlated with direct genetic effects (DGEs) of an 73 individual's genes on its own phenotype. In mice, for example, aggressiveness is 74 heritable (a DGE), aggressiveness elicited in conspecifics is heritable too (an IGE), and 75 aggressive genotypes elicit greater levels of aggressiveness in conspecifics. Such 76 positive genetic correlations between DGEs and IGEs are predicted to speed up the 77 response to directional selection compared to predictions derived from classic 78 quantitative genetics theory (e.g., McGlothlin et al. 2010). By contrast, negative 79 genetic correlations between DGEs and IGEs might instead impose evolutionary 80 constraints. In gulls, for example, genes expressed in females that contribute to early 81 egg laying (a DGE) actually delay the egg laying date of other females when expressed 82 in their male partners (an IGE) (Brommer and Rattiste 2008). Obviously, phenotypes 83 that are relatively repeatable (e.g., body size) and affect fitness in interaction partners 84 are most likely to induce indirect effects in nature. 85 The consequences of indirect (genetic) effects for evolutionary processes are 86 even more profound when we fully acknowledge the multivariate nature of most 87 phenotypes (Moore et al. 1997; McGlothlin and Brodie 2009; Araya-Ajoy and 88 Dingemanse 2014; Bijma 2014). Behavioral traits are often structured into syndromes 89 (Garamszegi et al. 2012), i.e., whole suites of behaviors are correlated at the among-90 individual level; such syndromes may impose constrains on micro-evolution of 91 behavior (Dochtermann and Dingemanse 2013). This would, for example, occur when 92 selection favors multivariate behavioral phenotypes that are not present in the 93 population due to strong syndrome structure. To date, studies of syndromes have 94 considered only correlations among 'direct' sources of behavioral variation. However, 95 it is likely that indirect effects are also integrated as parts of syndromes. This would 96 occur if, for instance, opponent personality is itself an important source of indirect 97 effects on focal behavior. Indirect effects are expected for behaviors such as 98 aggressiveness and dominance (Moore et al. 1997, 2002), which are themselves often 99 correlated with other behaviors such as boldness, exploration and activity (Garamszegi 100 et al. 2012). Social partner effects should therefore be incorporated as part of 101 behavioral syndromes to better understand the evolution of correlated traits. 102 Aggressiveness represents a good example of an 'interacting phenotype' 103 studied both in quantitative genetics and behavioral ecology (Wilson et al. 2009; Briffa 104 et al. 2015). Aggressiveness is interesting because of its reciprocal nature and 105