Robustness of dispersal-mediated intergroup cooperation: extending the Rodrigues–Barker–Robinson multi-group model to realistic spatial networks, conflict costs, discrimination mechanisms, resource landscapes, and cultural institutions

Active cooperation between members of separate social groups is rare in nature, occurring mainly in some ants and primates. Rodrigues, Barker & Robinson (2023) built an inclusive-fitness model in which each patch holds three breeding groups, and showed that localized dispersal builds relatedness both within and between resident groups, so that two resident groups can evolve mutual tolerance and cooperatively disrupt a third (typically incomer) group; however, the feedback between intergroup behaviour and population structure can make this cooperation evolutionarily unstable. The authors themselves flag that the result rests on simplified spatial structure, extreme conflict costs, unmodelled discrimination mechanisms, a survival-versus-fecundity resource abstraction, and the absence of cultural institutions, and that its stability under realistic ecological trajectories is unknown. No published extension of the framework along these directions is known to date. The open problem is to determine how the evolution and evolutionary stability of intergroup cooperation changes when the model's simplifying assumptions are replaced by realistic structure.

Background

In most social animals, interactions within groups are cooperative while interactions between groups involve conflict or, at best, tolerance; active cooperation between members of distinct, separate groups — such as jointly defending or attacking other groups — is very rare, documented mainly in some ants and primates (Robinson & Barker 2017; De Dreu & Triki 2022). Explaining this rarity is a problem for social evolution theory. The standard tool is inclusive-fitness (kin selection) analysis, which weights the fitness effects of a behaviour on social partners by their genetic relatedness to the actor; a classic result (Taylor 1992) is that in a viscous population — one with limited dispersal — relatedness among neighbours builds up, but the same population viscosity intensifies competition among kin, so the two effects often cancel. Whether and how this cancellation plays out when cooperation is organized between groups rather than within them has been far less explored. Rodrigues, Barker & Robinson (2023) presented a mathematical model addressing exactly this. The population is subdivided into patches, each containing three breeding sites occupied by groups of $n$ pre-reproductive individuals; groups are resident (born in the local patch) or incomer (born elsewhere). With probability $y$, two dominant-acting groups attack a third, succeeding with probability $\sigma$; failed attacks cost the attackers their breeding prospects, and eliminating a group frees resources that raise either the survival or the fecundity of survivors' offspring. A fraction $d$ of offspring disperses long distance at cost $c$, a fraction $m$ of those remaining moves between the patch's groups, and grouping rules are strict: philopatric offspring (those remaining on their natal patch) group only with other philopatric offspring, and incomers only with other incomers. The central results are that localized dispersal builds relatedness between resident groups, favouring the evolution of resident–resident tolerance combined with cooperative participation in attacks that preferentially eliminate incomer groups — even without any ability to recognize or target incomers — and that the resulting intergroup cooperation sits in an ecological feedback that may undermine its own evolutionary stability. The authors explicitly present their model as a starting point whose robustness is unestablished: real populations occupy heterogeneous spatial networks of groups rather than isolated three-group patches; the cost of a failed conflict is assumed to be total breeding failure rather than milder or ecologically side-effecting costs; the behavioural mechanisms that could produce preferential targeting of incomers (recognition cues, group-size assessment, active coalition formation) are not modelled; resources are abstracted into a single survival-or-fecundity benefit rather than multiple, spatiotemporally heterogeneous, complementary resources; and for humans, cultural institutions and norms that alter payoffs and recognition are outside the framework entirely, as is any quantitative account of transient versus long-term stability under changing ecological conditions.

Problem Statement

Determine, within the Rodrigues–Barker–Robinson patch-and-group framework, how the evolution and the evolutionary stability of intergroup cooperation — resident–resident tolerance combined with cooperative participation in the disruption of a third group — change when the model's simplifying assumptions are replaced by realistic structure. The open directions flagged by the authors, all bearing on whether the published result is robust or an artifact of the simplifications, include: (i) heterogeneous spatial networks of groups with explicit geographic proximity, in place of isolated patches with three breeding areas and strict philopatric-versus-incomer grouping rules — specifically how such networks alter encounter rates, within- and between-group relatedness, and selection on intergroup behaviour; (ii) conflict-cost functions of less-steep severity and with alternative ecological effects than total breeding failure, beyond the authors' single supplementary first pass, and whether the dispersal–cooperation relationship is preserved across systematic variation of cost functional forms; (iii) explicit mechanistic discrimination processes — recognition or familiarity cues, assessment of group size or quality, and active coalition formation — and how such mechanisms influence the resident–resident tolerance plus joint-attack-on-incomers pattern that the original model obtained without any targeting mechanism; (iv) multi-resource landscapes with spatially and temporally heterogeneous, complementary resources, in place of the survival-versus-fecundity abstraction, including when intergroup tolerance opens the way to active resource exchange; (v) for humans, integration of cultural evolution and institution-building — norms and institutions that alter payoffs and recognition — into the framework, so as to explain persistent large-scale intergroup cooperation; and (vi) the transient versus long-term evolutionary stability of intergroup cooperation under realistic ecological parameter trajectories such as resource depletion or changes in dispersal regime, characterizing when and how the model's ecology–structure feedback may undermine the cooperation it initially favoured. These six directions are independent openings onto the single question, and resolving any one of them addresses the robustness question.

References

1. Rodrigues AMM, Barker JL, Robinson EJH (2023) The evolution of intergroup cooperation. Philosophical Transactions of the Royal Society B 378, 20220074. DOI 10.1098/rstb.2022.0074. https://doi.org/10.1098/rstb.2022.0074 (full text also at PMC9939261, https://pmc.ncbi.nlm.nih.gov/articles/PMC9939261/)
2. Rodrigues AMM, Barker JL, Robinson EJH (2022) From inter-group conflict to inter-group cooperation: insights from social insects. Philosophical Transactions of the Royal Society B 377, 20210466. DOI 10.1098/rstb.2021.0466. https://doi.org/10.1098/rstb.2021.0466
3. Robinson EJH, Barker JL (2017) Inter-group cooperation in humans and other animals. Biology Letters 13, 20160793. DOI 10.1098/rsbl.2016.0793. https://doi.org/10.1098/rsbl.2016.0793
4. De Dreu CKW, Triki Z (2022) Intergroup conflict: origins, dynamics and consequences across taxa. Philosophical Transactions of the Royal Society B 377, 20210134. DOI 10.1098/rstb.2021.0134. https://doi.org/10.1098/rstb.2021.0134
5. Taylor PD (1992) Altruism in viscous populations — an inclusive fitness model. Evolutionary Ecology 6, 352–356. DOI 10.1007/BF02270971. https://doi.org/10.1007/BF02270971
6. Inaba M, Akiyama E (2025) Environmental variability promotes the evolution of cooperation among geographically dispersed groups on dynamic networks. PLOS Complex Systems 2(4), e0000038. DOI 10.1371/journal.pcsy.0000038 (preprint arXiv:2411.12348). https://doi.org/10.1371/journal.pcsy.0000038
7. Grueter CC (2026) How resource heterogeneity and social threat shape intergroup tolerance: insights from a spatial agent-based model. Preprint. DOI 10.64898/2026.02.21.707157. https://doi.org/10.64898/2026.02.21.707157
8. Grueter CC, Pozzi L (2025) Close encounters of the friendly kind: pacific between-group interactions in primates. Biological Reviews 100, 2257–2284. DOI 10.1111/brv.70046. https://doi.org/10.1111/brv.70046
9. Glowacki L (2024) The evolution of peace. Behavioral and Brain Sciences 47, e1. DOI 10.1017/s0140525x22002862. https://doi.org/10.1017/s0140525x22002862
10. Arbon JJ, Radford AN (2026) Pre-emptive behaviour in a landscape of intergroup conflict. Trends in Ecology & Evolution, in press (published online July 2026). DOI 10.1016/j.tree.2026.06.002. https://doi.org/10.1016/j.tree.2026.06.002