Power struggles, dominance testing, and reproductive skew

Models of reproductive skew are concerned with the partitioning of reproduc tion between dominant and subordinate members of a group. In an interesting extension of these models, Reeve and Ratnieks briefly considered whether i t might benefit subordinates to engage in aggressive behavior to test the f ighting ability of a dominant. Their analysis suggested that such testing s hould be more probable in groups that feature high skew and, hence, perhaps among closer relatives (because high relatedness favors high skew). Here w e explore in more detail the possibility of dominance testing. Three models that differ in the outcome of fights over dominance are presented: in the first model, the loser of the challenge is killed; in the second model, the loser is evicted from the nest; and, in the third model, the loser becomes (or remains) subordinate. In each case we consider the independent effects of the parameters that determine skew (namely, relatedness, group producti vity, and ecological constraints) on the predicted level of dominance testi ng. We then construct an amalgamated model to examine situations where tigh ts may lead to any one of the three outcomes. Our analysis reveals that. in the majority of casts, higher relatedness will in fact lead to lower level s of aggression. Moreover, dominance testing need not be associated with hi gh skew. Rather, the relationship between skew and dominance testing will d epend on which factor (relatedness, group productivity or Level of ecologic al constraints) is principally responsible for variation in the distributio n of reproduction.