Factors influencing within-group conflict over defence against conspecific outsiders seeking breeding positions

Abstract

In social species, groups face a variety of threats from conspecific outsiders. Defensive actions are therefore common, but there is considerable variation in which individuals contribute and to what extent. There has been some theoretical exploration of this variation when the defence is of shared resources, but the relative contributions when a single intruder threatens a particular breeding position have received less attention. Defensive actions are costly, both for the individual and dependent young, and contributions are likely to differ depending on individual sex, rank and size, current breeding stage, infanticide risk and relatedness levels. Here, we model analytically the relative fitness benefits of different group members to engaging in defence against individual intruders and determine when within-group conflicts of interest might arise over these defensive contributions. Conflicts of interest between the challenged breeder and other group members depend on relatedness to the brood and the potential relatedness reduction if an intruder acquires breeding status. Conflicts are more likely to occur when there is a low chance of winning the contest, low infanticide rates, inefficient defence from helpers, a long remaining brood-dependency period and high external (non-contest-related) mortality. Our work can help explain variation in defensive actions against out-group threats.

Keywords: animal groups; breeding position; defence; inclusive fitness; intrusions; subordinate contributions.

Figure 1.

Group set-up in the example of a fish species. BS, the same-sex breeder as the intruder; BO, the opposite-sex breeder to the intruder; H, the helper and I, the intruder. BS and BO are equally sized. H's size can range from half the size to the same size as the breeders. I's size can range from the same size to 25% larger than the breeders. Silhouettes are based on photographs of Neolamprologus pulcher taken by Ines Braga Goncalves.

Figure 2.

The helper's inclusive fitness when defending (becoming involved in the contest with the intruder) and defeating the intruder (black) and when not defending and the intruder takes over BS's position (red). Panels show inclusive fitness as functions of the (a) size difference between intruder and breeders (s_I); (b) steepness of the threshold curve for involvement that is required to successfully defend against the intruder (k₁); (c) probability of infanticide (μ); (d) duration of remaining brood dependence (k₃); (e) helper inefficiency (k₂); and (f) external mortality (1 − p_surv). Solid lines refer to H_SO, dashed lines to H_O, and dotted lines to H_S. In the case of defending and defeating the intruder, H_O and H_S can expect equal fitness returns (black dashes). Fixed parameters are: s_I = 0.13, μ = 0.05, k₁ = 20, k₂ = 2, k₃ = 10, p_surv = 0.4. (Online version in colour.)

Figure 3.

Schematic summary of positioning of switch points for each group member, depending on variation in the different parameter values.

Figure 4.

Effects of combinations of two parameters with external mortality (high and low) on the likelihood of within-group conflicts of interest. Area of parameter space where the interests of group members are aligned (yellow and dark blue area, denoted ‘no conflict’), where a conflict between BS and H_SO occurs (light blue), and where a conflict between BS and H_O occurs (green and light blue). Yellow area is where becoming involved in a contest and defeating the intruder is favourable for all group members; dark blue area is where not becoming involved in a contest and the intruder replaces BS is favourable for all group members. (a,b) Intruder's size advantage s_I versus steepness of threshold curve k₁; (c,d) infanticide survival 1 − μ versus duration of brood dependence k₃; (e,f), helper inefficiency k₂ versus steepness of threshold curve k₁. Fixed parameters are: s_I = 0.13, μ = 0.05, k₁ = 20, k₂ = 2, k₃ = 10, (a,c,e) p_surv = 0.4 and (b,d,f) p_surv = 0.8. (Online version in colour.)