Non-lethal fungal infection could reduce aggression towards strangers in ants

The study investigates how a non-lethal fungal ectoparasite, Rickia wasmannii, alters social interactions in ants by modifying cuticular hydrocarbon (CHC) cues that mediate nestmate recognition. In social insects, CHC profiles are key signals used to discriminate nestmates from non-nestmates; perturbations to these profiles can alter aggression and acceptance dynamics within and between colonies. Previous work showed that fungal infection broadens the host’s nestmate recognition template and increases the proportion of the linear alkane tricosane (n-C23) on infected Myrmica workers’ cuticles. The central hypotheses were: (a) elevated n-C23 in infected individuals reduces their own aggression toward others, and/or (b) increased n-C23 on an opponent acts as a pacifying signal, reducing aggression directed at that opponent regardless of its infection status. The study experimentally manipulated CHC profiles to test these hypotheses and to clarify the role of n-C23 in mediating aggression toward strangers.

The introduction outlines extensive literature on chemical recognition in social insects, emphasizing CHCs as primary cues for nestmate discrimination. CHC profiles can include over 100 hydrocarbons and convey information on colony identity as well as individual traits such as age, task, and sex. Prior studies have shown that adding single CHC components can modulate aggression responses in social Hymenoptera. Parasites and pathogens are known to alter host CHC profiles, potentially affecting nestmate recognition and social parasitism dynamics. Against this background, the fungus Rickia wasmannii is highlighted for its capacity to alter Myrmica ants’ CHC profiles, notably increasing n-C23 and n-C24, and to open infected societies to outsiders, suggesting a mechanistic link between pathogen-induced CHC shifts and reduced aggression.

Study species and colonies: Colonies of the ant Myrmica sabuleti, both infected (N = 9) with the ectoparasitic fungus Rickia wasmannii and uninfected (N = 6), were collected near Cluj-Napoca, Romania, and maintained under controlled laboratory conditions (approx. 22 °C, 12:12 light cycle) with ad libitum diet. Species identity and infection status were confirmed morphologically and by DNA barcoding. Experimental design: Five types of behavioral assays were conducted using live workers (infected or uninfected) interacting with standardized dummies (freshly killed conspecific corpses) under controlled conditions: 1) Hexane-washed dummies: Corpses washed five times with hexane to remove CHCs, serving as odourless controls. 2) Unmanipulated infected vs. uninfected corpses: Baseline aggression toward infected and uninfected dummies. 3) CHC reconstitution: Hexane-washed dummies treated with cuticular hydrocarbon (CHC) extracts from either uninfected or infected workers, to test whether transferred CHC profiles reproduce aggression patterns seen with natural corpses. 4) n-C23 supplementation: Uninfected corpses treated topically with synthetic n-C23 (dose ~80 ng) to mimic the elevated n-C23 characteristic of infected ants. 5) CHC extract + n-C23: Hexane-washed dummies treated with CHC extract from uninfected ants supplemented with synthetic n-C23. Behavioral assays and scoring: Interactions were observed after first contact between a live worker and a presented dummy. Behaviors were categorized as socio-positive (e.g., grooming), neutral (e.g., anointment), or aggressive (e.g., mandible gaping, biting, dragging, stinging). An aggression index (AI) was calculated for each encounter as the number of aggressive responses divided by the total number of interactions. Observers were blind to the infection status of the live worker. Chemical manipulations and verification: CHC removal was achieved by five sequential hexane washes; control extracts from washed dummies showed only trace residual CHCs relative to unwashed ants, confirming effective removal. CHC extracts for reconstitution were prepared from donor ants, evaporated, and applied to dummies. Synthetic n-C23 was dissolved in hexane and applied topically to achieve a target dose (~80 ng) approximating the relative abundance found on infected ants. Gas chromatography–mass spectrometry (GC-MS) verified: (i) effective CHC removal from washed dummies, and (ii) that n-C23 supplementation on uninfected dummies yielded a relative abundance (mean ± SD: 22.4 ± 6.0%) comparable to infected ants and significantly higher than untreated uninfected ants (13.8 ± 1.3%). Statistics: Aggression indices were analyzed using generalized linear models/mixed models (GLM/GLMM) with appropriate post hoc comparisons (e.g., Tukey’s tests). Where applicable, colony identity was included; analyses of deviance and model diagnostics were applied. Sample sizes varied across treatments (see results and supplementary tables).

- Baseline controls: Virtually no aggression occurred toward hexane-washed (odourless) dummies (GLM, P = 0.99, N = 36), confirming the role of CHCs in recognition and the efficacy of CHC removal. - CHC extract reconstitution: Applying CHC extracts from infected or uninfected ants to washed dummies generally reproduced aggression patterns observed with unmanipulated corpses, validating the CHC-transfer approach (e.g., U–I corpses vs U–I extract: GLMM z = 0.54; U–I corpses vs U–I extract: GLM z = −0.03). One comparison differed significantly (I–I corpses vs I–I extract, P = 0.03), indicating that transfers may not always perfectly replicate natural cues. - Infection-linked acceptance: Infected workers elicited less aggression when dummies bore CHC extracts from infected individuals compared to uninfected combinations (e.g., U–U extract vs I–I extract, P = 0.01), mirroring reduced aggression towards infected non-nestmates. - n-C23 as a pacifying signal: Adding synthetic n-C23 to uninfected dummies significantly reduced aggression from uninfected ants (U–U + n-C23 vs U–U: GLM, P = 0.001). Supplementing uninfected CHC extracts with n-C23 similarly reduced aggression in both uninfected and infected ants (U–U extract vs U–U extract + n-C23: P = 0.001, N = 21; I–U extract vs I–U extract + n-C23: P = 0.01, N = 15), underscoring the pacifying effect of n-C23. - Asymmetry in responses: Infected ants appeared more peaceful toward infected non-nestmates but not toward uninfected non-nestmates, suggesting condition-dependent social tolerance. - Chemical confirmation: n-C23 enrichment on treated uninfected dummies (22.4 ± 6.0% relative abundance) matched infected ants and exceeded untreated uninfected ants (13.8 ± 1.3%; LM t = 4.46, P < 0.001).

The experiments support the hypothesis that elevated n-C23 acts primarily on the opponent by reducing aggression directed toward individuals bearing this compound, rather than broadly reducing the intrinsic aggressiveness of infected ants. This mechanistic link explains previous observations that Rickia wasmannii–infected Myrmica ants show increased acceptance of infected non-nestmates. By widening the recognition template through a specific CHC component, the parasite can modulate colony-level social dynamics. Such parasite-induced changes in chemical signaling may facilitate the entry of infected non-nestmates, foreign queens, or other social parasites and can, over time, increase genetic heterogeneity and alter colony structure. While some variation existed between responses to n-C23-supplemented corpses versus CHC-extract-treated dummies, the convergent outcome across manipulations indicates that n-C23 is a key pacifying cue linked to fungal infection.

This study demonstrates that a single CHC component associated with fungal infection, n-C23, reduces aggression toward conspecifics in Myrmica ants, functioning as a pacifying signal. Parasitic modification of host chemical cues can thus relax nestmate discrimination and reshape social interactions, potentially altering colony composition and genetic structure. Future work should test the generality of n-C23–mediated pacification across Myrmica species and ecological contexts, dissect synergistic effects with other CHCs, examine the neural and sensory mechanisms underpinning reduced aggression, and assess long-term colony-level consequences under natural conditions.

- Chemical reconstructions did not always perfectly recapitulate natural interactions, as indicated by a significant deviation in one comparison (I–I corpses vs I–I extract, P = 0.03), suggesting that CHC-transfer methods may incompletely reproduce natural cue context. - Some treatments had modest sample sizes (e.g., N = 5–21 in specific assays), which may limit statistical power for certain comparisons. - Laboratory conditions and the use of corpses/dummies, rather than live opponents, may not fully capture the dynamics of natural encounters. - The study focused on n-C23; potential interactions with other CHC components (e.g., n-C24) were not exhaustively tested, leaving the full cue blend effects unresolved.