Extended and cumulative effects of experimentally induced intergroup conflict in a cooperatively breeding mammal

Abstract

Conflict between rival groups is rife in nature. While recent work has begun exploring the behavioural consequences of this intergroup conflict, studies have primarily considered just the 1-2 h immediately after single interactions with rivals or their cues. Using a habituated population of wild dwarf mongooses (Helogale parvula), we conducted week-long manipulations to investigate longer-term impacts of intergroup conflict. Compared to a single presentation of control herbivore faeces, one rival-group faecal presentation (simulating a territorial intrusion) resulted in more within-group grooming the following day, beyond the likely period of conflict-induced stress. Repeated presentations of outsider cues led to further changes in baseline behaviour by the end of the week: compared to control weeks, mongooses spent less time foraging and foraged closer to their groupmates, even when there had been no recent simulated intrusion. Moreover, there was more baseline territorial scent-marking and a higher likelihood of group fissioning in intrusion weeks. Consequently, individuals gained less body mass at the end of weeks with repeated simulated intrusions. Our experimental findings provide evidence for longer-term, extended and cumulative, effects of an elevated intergroup threat, which may lead to fitness consequences and underpin this powerful selective pressure.

Keywords: cooperation; dwarf mongooses; field experiment; intergroup conflict; territorial intrusion.

Figure 1.

Illustration of a typical treatment week. In the experiment, each group (n = 7) received two week-long treatments: an Intrusion week where the faecal presentations and call playback simulated the presence of a rival group, and a Control week where herbivore faeces and calls were used on an equivalent schedule. Data collected on Day 2 were used to investigate the extended effects after the first simulated intrusion on Day 1, while data collected on Day 6 were used to investigate the cumulative effects after experiencing repeated simulated intrusions. Rare group-level behaviours were also collected over the course of the week. (Online version in colour.)

Figure 2.

Effect of treatment (Control: herbivore; Intrusion: rival group) on dwarf mongoose grooming behaviour. (a) On Day 2 of Intrusion weeks (the day after the first simulated intrusion), grooming bouts were longer than those in Control weeks, but the effect was more pronounced for males than females. On Day 6 of Intrusion weeks (after repeated simulated intrusions), the grooming-bout duration was longer than in Control weeks, but the effect was more pronounced both (b) for males than females and (c) for subordinates than dominants. Shown are back-transformed predicted means (square points) ± s.e. (within the bounds of the square point in some cases) from the mixed models presented in electronic supplementary material, table S1 (for a) and S2 (for b,c), with the raw data (circular points) for each individual. Dashed lines connect data from the same individuals; orphan points are instances where data were available from an individual in only one treatment. In (a), n = 67 mean bout durations from 44 individuals in seven groups; in (b,c), n = 73 mean bout durations from 47 individuals in seven groups.

Figure 3.

Effect of treatment (Control: herbivore; Intrusion: rival group) on dwarf mongoose foraging behaviour. On Day 6 of Intrusion weeks (after repeated simulated intrusions), there was (a) less foraging activity (proportion of scan-samples where the group was recorded as foraging), and (b) individuals foraged closer to one another (nearest-neighbour distance) compared to Control weeks; these effects were not apparent on Day 2. Shown in (a) are values for each group (circular points; n = 7), with dashed lines connecting data from the same groups; boxplots indicate the median and quartiles; whiskers represent data within quartiles ± 1.5 times the interquartile range. Shown in (b) are back-transformed predicted means (square points) ± s.e. (within the bounds of the square point) from the mixed models presented in electronic supplementary material, table S4, alongside raw data (circular points; jitter function applied to spread the points horizontally). n = 451 nearest-neighbour distances from 53 individuals in seven groups on Day 2 and n = 490 nearest-neighbour distances from 54 individuals in seven groups on Day 6.

Figure 4.

Effect of treatment (Control: herbivore; Intrusion: rival group) on dwarf mongoose body-mass changes. On Day 6 of Intrusion weeks (after repeated simulated intrusions), (a) adults and (b) independently foraging pups gained less body mass compared to Control weeks; these effects were not apparent on Day 2. Shown are predicted means (square points) ± s.e. (within the bounds of the square point in some cases) from the mixed models presented in electronic supplementary material, table S5 (back-transformed for pups on Day 6), with the raw data (circular points) for each individual. Dashed lines connect data from the same individuals; orphan points are instances where data were available from an individual in only one treatment. In (a), n = 62 body-mass changes from 39 individuals in seven groups on Day 2 and n = 68 body-mass changes from 39 individuals in seven groups on Day 6. In (b), n = 38 body-mass changes from 25 individuals in seven groups on Day 2 and n = 43 body-mass changes from 26 individuals in seven groups on Day 6.