Female counterstrategies against male infanticide
Extra-unit mating In total, 4,697 copulation events were observed. Only 317 of these copulation events were extra-unit matings (6.7%), and these involved almost exclusively males in other one-male units (309 events) rather than males in the all-male unit (8 events) (χ2 = 182.00, p < 0.001). Seven cases of extra-unit paternity, which involved resident males in other OMUs, were identified in 99 births (6.25%). All new males who took over OMUs came from AMUs, and accordingly were not the fathers of any offspring in the OMUs they entered.
Of the nine mothers with unweaned infants who left their original OMU and transferred to another OMU following male replacement, two mothers (Ln in Case #6, Yb in case #16) had previously been observed mating with the resident male of the destination OMU into which they transferred, but no extra-unit paternities were identified for these two cases.
Joint defense with other femalesIn cases of male attacks on unweaned infants, most females of the same OMU (83.1%, 59/71) were observed to joint with other females against the new male (Table S1 ). The lactating female whose infant was under attack was the initiator of 66.7% (10/15) of coalitions as she, after firstly seizing her infant, started shouting at the perpetrator. Then, other females of the same unit (and sometimes also juveniles) ran to the mother and subsequently started vocalizing loudly at the male. In the remaining instances the mother was still seizing her infant while other female already started shouting at the male before being joined by others (including the mother). In a few cases, non-lactating female collectively chased the perpetrator and keep it away. In response to females’ facing and shouting at the male, he usually retreated quickly and stayed away from the females for some time.
The number of adult females involved in collective defense was 3.38 ± 0.89 (mean ± SD). The defensive behavior of females resulted in 85.0% (17/20) of the initial male attacks being unsuccessful, and since in the majority of cases mothers whose infants were attacked subsequently transferred to another OMU, most of these infants (16/17, 94.1%) survived to weaning. The effectiveness of female joint defense appears to depend on the size of OMU (the number of individuals except infants), as OMU size was smaller in OMUs in which attacks had been observed than in OMUs in which no attacks were observed (Fig 3a , t = 4.57, df = 26, p < 0.001), and the size of OMUs in which infants died from infanticide was also significantly smaller than those in which males never killed infants (Fig 3b ,t = 2.53, df = 29, p < 0.05).
.
F emales’ transfer with infants Nearly a third of mothers (11/31) transferred to other OMUs or returned to stay with the old (ousted) male in response to the unsuccessful attacks on their offspring or another offspring in the OMU (Fig 1 ). The contexts of the nine mothers with dependent offspring who successfully immigrated into an already existing OMU included (i) five females who had at least one close female relative (mother-daughter, sister-sister) in the new unit they joined, (ii) two females who had been observed mating with the resident male of destination unit, and (iii) two sets of two females who collectively joined another OMU as a clique.