Contextual use of male-male social information by Trinidadian guppies Authors: Alix J P Brusseau, Félixe Dumaresq-Synnott, Jade Morris, Abigail C Nagl, Indar W Ramnarine, Adam L Crane, Grant E Brown Published: 2025 ABSTRACT AbstractDecision-making among prey often involves balancing fitness-related activities, such as foraging and mating, with the need to avoid predation. These trade-offs may be influenced by sex, especially among sexually dimorphic species where males and females face different selection pressures. Consistent with the ‘Distracted Male Hypothesis’, male Trinidadian guppies (Poecilia reticulata), may show a reduced response to publicly available personal risk assessment cues (i.e. chemosensory predation risk cues) relative to females due to the relatively high costs associated with lost courtship and mating opportunities. Male guppies may compensate for a reduced response to personal information by increasing their use of conspecifics as a source of social information. Here, we tested this hypothesis using wild-caught male Trinidadian guppies, examining their use of visual social cues from alarmed conspecific males in the presence vs. absence of females. We found that when stimulus females were absent, focal males increased their inspection of an alarmed male stimulus shoal, suggesting the use of social information. However, when a stimulus female was present, males did not increase inspection of the male stimulus shoal. Rather, they exhibited high rates of inspection towards the female stimulus. These suggest that male guppies may adjust their antipredator behaviour depending on social context, likely reflecting underlying reproductive trade-offs.Supplementary InformationThe online version contains supplementary material available at 10.1007/s10071-025-02027-z.Keywords: Antipredator behaviours, Inspection, Mating opportunities, Predation risk, Sex differences Introduction Prey individuals are continually faced with the conflicting demands of detecting and avoiding predators and the need to engage in fitness related activities such as foraging and mating (Lima and Dill 1990; Brown and Godin 2023). The costs associated with making these decisions, however, increase when prey cannot reliably recognize risks (Dill 1987; Donaldson-Matasci et al. 2010; Crane et al. 2024). Thus, the ability to make these decisions depends on both the quality and reliability of the information available (Dall et al. 2005; Crane et al. 2024). To assess predation risk and adjust behavioural decisions, prey often integrate both personal and social information, reducing reliance on a single source of information (Brown and Laland 2003; Dall et al. 2005; Brown et al. 2011; Feyten et al. 2021). Personal information, gathered through direct interaction with the environment (Dall et al. 2005), may provide accurate and reliable cues regarding predation threats (Brown and Godin 1999; Wagner and Danchin 2010; Carthey and Blumstein 2018) but may be costly due to proximity to a predator. To mitigate these costs, prey may use the behaviours of conspecifics or heterospecific guild members as a source of social information (Brown and Laland 2003; Dall et al. 2005; Valone 2007; Schmidt et al. 2010). However, while social information can be seen as less risky than personal information, it may also be less reliable (Giraldeau et al. 2002; Feyten et al. 2021) as it may be prone to copying errors (Giraldeau et al. 2002; Koops 2004; Rieucau and Giraldeau 2011). Using a combination of personal and social cues would allow prey to optimize predator avoidance, foraging, and mating decisions (Dall et al. 2005; Schmidt et al. 2010). The ability to make decisions can also be modulated by other factors such as sex. In sexually dimorphic species, males and females may face vastly different trade-offs (Magurran and Nowak 1991; Andersson 1994; Schuett et al. 2010). According to the Asset Protection Principle (Clark 1994), individuals with higher reproductive investments should exhibit greater degrees of risk aversion (Reinhardt and Healey 1999; Katwaroo-Andersen et al. 2016). Consistent with this model, females should exhibit stronger responses towards predation cues (relative to males) as a result of greater investments in reproductive tissues (Katwaroo-Andersen et al. 2016; Brusseau et al. 2023). In contrast, strong sexual selection may result in males prioritizing mating opportunities over actively avoiding predation threats, as males are forced to compete for access to reproductively active females (Magurran and Nowak 1991; Andersson 1994). The Distracted Male Hypothesis (Dumaresq Synnott et al. 2025) posits that in sexually dimorphic species, males may be less responsive to acute personal risk assessment information indicating risk due to the high costs associated with lost mating opportunities. For example, Brusseau et al. (2023), found that while adult female Trinidadian guppies (Poecilia reticulata) showed strong antipredator responses to conspecific alarm cues (as a form of personal information), reproductively active males exhibited a weaker or absent response. Supporting the hypothesis that these differences are due to differential reproductive investment, Dumaresq Synnott et al. (2025) demonstrated that non-reproductively active male and female guppies responded similarly to acute chemosensory indicators of risk. In a related study, Mirza et al. (2001) demonstrated that while male and female red swordtails (Xiphophorus helleri) both respond to conspecific alarm cues, the response pattern (vertical space use) differed with sex. Interestingly, the live bearing lace brachy (Brachyrhaphis rhabdophora) has a lower degrees of sexual dimorphism (based on size and colouration) compared to other poecilids (Reznick et al. 2021) and (consistent with the Distracted Male Hypothesis) do not show sex differences in response to conspecific alarm cues (Duffy and Johnson 2024). Male prey may be able to compensate for their reduced responses towards personal information by prioritizing social information (Brown and Laland 2003; Schmidt et al. 2010; Feyten et al. 2021). Social information is less risky to obtain but may be less reliable, as it can be manipulated or misinterpreted by receivers (Giraldeau et al. 2002; Dall et al. 2005). The use of social information to detect predation risk has also been shown in male threespine sticklebacks (Gasterosteus aculeatus), as they decreased courtship in response to visual and olfactory cues from females that have been exposed to predation risk (Dellinger et al. 2018). More recently, Brusseau et al. (2025a, b) demonstrated that when female guppies exhibited antipredator responses, observing males increased their fright responses, and decreased their courtship of females, regardless of the personal information (i.e. chemosensory cues) detected by males. This suggests that male guppies rely on social risk assessment information from females as a reliable cue to, at least partially, compensate for not responding acute personal information. This likely allows males to manage the trade-offs between reproduction and avoiding predators. Given that female guppies are more risk averse compared to males in response to personal risk assessment information, they can be considered a reliable source of social information. However, it is unknown if conspecific males would also function as an equally reliable source of social information. Under natural conditions, Trinidadian guppies often form sexually segregated shoals, with females avoiding males to reduce predation risk (Croft et al. 2004, 2006). As a result, male guppies would be exposed to predation threats under varying social contexts (i.e. in the presence vs. absence of conspecific females). We would predict that males use same-sex shoal mates as a potential source of social risk assessment information when females are absent. However, the presence of conspecific females would reduce the response of males to risk assessment information from conspecific males. Here, we tested this hypothesis using wild-caught adult male Trinidadian guppies. Specifically, we exposed focal males to a shoal of conspecific males in the presence vs. absence of conspecific females. Using a repeated measures approach, we quantified focal male behaviour before and after the male stimulus shoal was exposed to a realistic predator model (Fig. 1), recording the time focal males spent inspecting stimulus males and stimulus females (when present). We predict that in the presence of calm females, focal males should ignore the alarmed males and show no increase in antipredator behaviour. In contrast, in the absence of stimulus females, focal males should display risk-averse behaviours in response to social cues from an alarmed male shoal. Photo of model predator used to elicit response in male stimulus shoal Methods Ethics This experiment was approved by Concordia University’s Animal Research Ethics Board (Protocol #30000255) and by the University of the West Indies, St. Augustine (Protocol CREC.SA.1581). Test fish and model predator We caught male and female Trinidadian guppies with a 3 mm mesh seine net in the Lopinot River in the Northern Range Mountains of Trinidad in May 2024. The Lopinot River is a high predation stream with a diverse predator guild including pike cichlids (Crenicichla sp.), blue acara (Andinoacara pulcher), brown coscarob (Cichlasoma tenia), and wolf fish (Hoplias malabaricus) (Deacon et al. 2015). We immediately transported guppies to the University of the West Indies, St. Augustine, Trinidad, and housed them in 275 L glass aquaria. We kept the holding tanks at ~ 23 °C, under a 12 L:12D cycle, and they contained two bio-sponge filters. We fed guppies twice daily with commercial flake food, at least an hour before and after observations. In total, we used 288 males and 108 females. For the visual predation stimulus, we employed a model predator. We used 4 identical model predators that mimic shiner fish (Animated Lure ©), that were 13.3 cm in total length (Fig. 1). Equipped with a small motor and a propeller (positioned on the anterior tip), the model predator could move independently once placed in water, effectively mimicking natural predator attacks. We opted to use a moving visual predator model, as our previous work shows that male guppies do not exhibit overt responses to acute chemical indicators of predation risk (Brusseau et al. 2023). Experimental set up Test tanks consisted of a series of 24 L glass aquaria (dimensions 21.5 cm × 28 cm × 40.5 cm; width × length × height), filled with ~ 18 L of dechlorinated water. Each setup included two test tanks (focal male tank and male stimulus tank) positioned side by side, with a smaller tank (predator tank, dimensions 21.5 cm × 14 cm × 40.5 cm; width × length × height) situated behind the left test tank (male stimulus tank; Fig. 2). We placed opaque black plastic around each pair of tanks to prevent guppy visual communication across tank pairs. In the right test tank (focal male tank), we placed a 1 L (8 × 8 × 20 cm) transparent plastic chamber (‘female chamber’ hereafter) that had opaque black plastic on one half to visually isolate it from the male stimulus tank (Fig. 2). Additionally, an opaque plastic barrier covered the back quarter of the focal male tank, blocking the focal fish’s view of the predator tank. To confirm this, we placed a GoPro inside the focal male tank and verified that no fish were visible in the male stimulus tank. We placed a shoal of three males in the male stimulus tank and one focal male in the focal male tank (Fig. 2). For half the trials, we placed a shoal of three females in the female chamber. The model predator, when used, was placed in the predator tank. This arrangement ensured that the shoal of males could not see the females, and the focal male could not see the model predator. All fish had 15 min to acclimate before testing. Schematic of the experimental set-up used in the experiment. Dotted lines represent opaque barriers that blocked visual cues. The setup included a shoal of three male guppies in the left tank (male stimulus tank), a focal male in the right tank (focal male tank), and a model predator in a smaller tank behind the male stimulus tank (predator tank). Depending on the treatment, a shoal of female guppies was placed in the female chamber or was absent Observations consisted of a 4-minute pre-exposure period and a 4-minute exposure period to the predator model. During both observation periods, we simultaneously recorded the behaviour of the focal male and the stimulus males. For the focal male, we recorded whether it was swimming (defined as swimming at a normal pace of less than 5 cm per second as in Brusseau et al. 2023) every 10 s; hereafter ‘swimming activity’. We also recorded the total number of inspection attempts toward stimulus males (defined as actively approaching the tank containing the stimulus shoal, head directed toward the glass between stimulus and focal tanks) and inspection attempts toward females (defined as whether the focal male approached the female chamber with its head up pointing at one female, and within one body length of the chamber, Brusseau et al. 2024). Finally, we recorded shoaling index for the shoal of males, where scores ranged between 1 (no fish within one body length of another) and 3 (all fish within one body length of each other) (Brusseau et al. 2023). The score was recorded every 10 s during each 4-minute period (i.e., 24 times each), and a mean shoaling index was calculated. Increased shoaling index and decreased movement are well established antipredator behaviours in fishes (Brown and Godin 1999; Brusseau et al. 2023). Inspection behaviour was used as a proxy for courtship (female inspection) and potential sexual competition (male inspection) (Magurran and Nowak 1991; Magurran 2005). After the observations, we returned fish to their housing tanks, and they were not tested again. Our experimental design allowed us to test the response to a calm vs. alarmed male stimulus shoal (i.e. before vs. after presentation of the model predator) in the presence vs. absence of a shoal of calm stimulus females. We conducted 18 replicates for each treatment combination (i.e., female absent or present and predator model absent or present). Statistical analyses We conducted linear mixed effects models, using the package lme4 in R (RStudio v1.2.1335) to test the effects of the predator model presence (i.e., pre-exposure vs. exposure period) and female presence (i.e., no females in the chamber vs. females present in the chamber) on the shoaling index of the male stimulus shoal and the occurrence of focal male swimming, the occurrence of inspection attempts toward females, and the occurrence of inspection attempts toward stimulus males. Because each focal fish was tested twice (pre-exposure and exposure period), individual identity (fish ID) was included as a random factor to account for repeated measures. We examined the assumptions of normality and homoscedasticity by generating QQ plots of the residuals from each linear model and inspecting the residuals vs. fitted values plots. No datasets needed to be transformed. For all analyses, the significance level was set at α = 0.05. Results Stimulus shoal When exposed to the predator model, the male stimulus shoal significantly increased shoaling index (LRT = 6.17; P = 0.012; Fig. 3; Table 1), confirming that the shoal perceived the model predator as a predation risk. We also confirmed that the presence vs. absence of a female stimulus shoal had no significant effect on the shoaling response of the stimulus males (LRT = 1.31; P = 0.251; Fig. 3; Table 1), as their visual cues were blocked by the opaque barrier. Mean ± SE of shoaling index for the male stimulus shoals before (“absent”) and during (“present”) presentation of the model predator in the predator tank Table 1Output of the repeated measures ANOVAS testing the effects of predator presence and female presence on shoaling index, swimming activity, inspection attempts toward females and inspection attempts toward malesSEELRTP valuea) Male stimulus shoalShoaling indexPredator presence × female presence2.570.0110.912Predator presence1.276.174 0.012 Female presence2.591.3130.251b) Focal male Swimming activity Predator presence × female presence1.790.0620.801Predator presence0.894.283 0.038 Female presence2.041.0360.308 Inspection toward females Predator presence × female presence1.261.2810.257Predator presence0.630.3040.581Female presence1.1928.729 < 0.0001 Inspection toward males Predator presence × female presence1.903.1730.074Predator presence0.9613.242 0.0002 Female presence1.7313.508 0.0002 Bold P-values indicate statistical significance. SEE = standard error of the estimate; LRT = likelihood ratio test Output of the repeated measures ANOVAS testing the effects of predator presence and female presence on shoaling index, swimming activity, inspection attempts toward females and inspection attempts toward males a) Male stimulus shoal Shoaling index b) Focal male Swimming activity Bold P-values indicate statistical significance. SEE = standard error of the estimate; LRT = likelihood ratio test Focal males Our analysis of focal male responses revealed no significant interactions between female presence and predator presence for any of the behavioural metrics (swimming activity LRT = 0.062; inspection towards male shoal LRT = 3.17; inspection towards female shoal LRT = 1.28, P > 0.05 for all; Table 1). Focal males inspected the female chamber when the female stimulus shoal was present (LRT = 28.73; P < 0.0001; Fig. 5; Table 1), confirming that the males could perceive and react to the females in the chamber. The presence of the model predator did not influence the rate of inspection attempts toward females (LRT = 0.30; P = 0.581; Fig. 4; Table 1). However, focal males engaged in more swimming activity in the presence of an alarmed male stimulus shoal (LRT = 4.28; P = 0.038; Fig. 3; Table 1), and there was no effect of female shoal presence on focal male swimming activity (LRT = 1.04; P = 0.308; Fig. 3; Table 1). Mean ± SE of occurrence of swimming activity for the focal male before (“absent”) and during (“present”) presentation of the model predator in the predator tank. The focal male was either exposed to no females (blue circle) or a shoal of 3 females in the female chamber (orange triangle) Mean ± SE of occurrence of inspection attempts towards stimulus females for the focal male before (“absent”) and during (“present”) presentation of the model predator in the predator tank. The focal male was either exposed to no females (blue circle) or a shoal of 3 females in the female chamber (orange triangle) For the level of inspection attempts toward stimulus males, there were significant main effects of both the presence of females (LRT = 13.51; P = 0.0002; Fig. 5; Table 1) and the presence of the model predator (LRT = 13.24; P = 0.0002; Fig. 5; Table 1). As the interaction between female presence and predator presence was marginal (LRT = 3.17; P = 0.07), we performed Tukey tests (using the emmeans package in RStudio v1.2.1335) to investigate the significant main effects further. In the absence of the female stimulus shoal, the focal male significantly increased inspection toward stimulus males (Tukey estimate = −5.36 ± 1.35, P = 0.001; Fig. 5). However, when a female shoal was present, the focal male engaged in lower rates of inspection toward stimulus males (Tukey estimate = −1.95 ± 1.35, P = 0.46; Fig. 5). Rather, the focal male increased its rate of inspection toward females (see above) Fig. 6. Mean ± SE of occurrence of male inspection for the focal male before (“absent”) and during (“present”) presentation of the model predator in the predator tank. The focal male was either exposed to no females (blue circle) or a shoal of 3 females in the female chamber (orange triangle) Discussion Our results suggest that the use of same-sex social information cues by male Trinidadian guppies is dependent upon the social context (i.e. presence vs. absence of females). When females were absent, focal males showed increased inspection of the alarmed male stimulus shoal. However, the presence of females led to fewer inspections of the alarmed male shoal in favour of increased inspection of the female shoal. Brusseau et al. (2025a, b) have previously demonstrated that while male Trinidadian guppies do not exhibit increased predator avoidance responses to acute chemosensory risk cues (i.e. conspecific alarm cues, disturbance cues), they can modulate their predator avoidance based on a female stimulus shoal’s response to risky information. Our results extend these findings, demonstrating that the social context (presence vs. absence of females) alters the response of male guppies to same-sex social information. Curiously, we found that regardless of female presence, focal males also increased their swimming activity when the stimulus shoal displayed heightened antipredator responses. Given that the focal males could not see the model predator, they may have perceived the stimulus male shoal as an uncertain or ambiguous risk (Crane et al. 2024). While a reduction in movement towards a known indicator of predation risk is a well-used metric for antipredator responses, it may also be indicative of increased information acquisition (Brusseau et al. 2024, 2025a, b). For example, guppies faced with uncertain or ambiguous predation risk exhibited increased inspection behaviour towards a novel predator model (Brusseau et al. 2024). In the current experiment, since focal males were not exposed to any form of acute personal information (they could not see the model predator, nor were they exposed to a chemosensory cue), it is possible that this increase in swimming activity may be an attempt to increase information acquisition (Crane et al. 2021; Brusseau et al. 2024, 2025a, b). While additional studies are required to test this supposition, we observed that this increased activity was directed towards the alarmed male stimulus shoal only when a female stimulus shoal was absent. In sexually dimorphic species such as guppies, the trade-off between allocating energy to courtship vs. predator avoidance is strongly skewed toward reproduction for males, compared to females (Magurran and Nowak 1991; Andersson 1994). However, males also face higher predation risk due to their colourful patterns and conspicuous courtship display (Godin and McDonough 2003). Therefore, it may be beneficial for males to use social information from both male and female conspecifics to reduce predation risk and make informed behavioural decisions. However, since the costs of lost mating opportunities would disproportionately disadvantage males in a highly sexually dimorphic species, males use social information from sexual competitors (i.e., other males) only when females are not available. This reliance on “cheaper” social information regarding predation risk, should provide males with reliable risk assessment while maintaining courtship opportunities. Alternatively, males may have reduced their response towards male social information to increase their perceived attractiveness towards females. Dugatkin and Godin (1992) suggested that males that actively inspect potential predators may be perceived as bolder and more attractive mates compared to those that do not inspect. An intriguing possibility is that focal males that ignore male social information may be perceived as bolder, more attractive mates (Godin and Dugatkin 1996). We are unable to rule out this alternative mechanism as we did not record data on the activity of the female shoal. However, it is likely that this alternative mechanism would have an additive effect on the focal males ‘discounting’ of same-sex social information observed here. Future studies are required to disentangle these mechanisms. While previous studies have shown the use of same-sex social information among male fishes, these studies have typically been studied in the context of competition or mating. For example, male three-spined sticklebacks (Gasterosteus aculeatus) and sailfin mollies (Poecilia latipinna) copy mating choices of conspecific males (Witte and Ryan 2002; Frommen et al. 2009). Male guppy aggression increases under conditions of male-biased sex ratios (Chuard et al. 2022). Fewer studies, however, have focused on the context of predation risk. Kelly and Godin (2001) have shown that though male guppies reduce mating competition and aggressive behaviours in the presence of a predator, they still attempt mating with females, regardless of local predation risk. This is consistent with our current study, as we observed that males significantly increased inspection of conspecific males only when no females were present. Under natural conditions, many prey fishes engage in sexual segregation, as females seek to reduce predation risks by avoiding shoaling with males (Croft et al. 2004, 2006; Griffiths et al. 2014; Rystrom et al. 2018). Thus, males may benefit from same-sex social information when in male biased sex ratios (Chuard et al. 2022) and shift to female social information when in female biased sex ratios. It has been previously shown that when presented with personal cues indicating risk and social cues indicating safety, guppies prioritize social information when making decisions regarding foraging within novel patches (Feyten et al. 2021). The use of conflicting information has also been demonstrated in fathead minnows (Pimephales promelas), where individuals that had conditioned to recognize an odour as safe were able to immediately relearn it as risky when the odour was paired with alarmed conspecifics rather than with alarm cues alone (i.e., personal information) (Crane and Ferrari 2015). In our study, the shoal of females may have acted as a safety cue, as they were displaying calm behaviours (i.e., did not receive any cues indicating risk), potentially conflicting with the alarmed male shoal. Additionally, adding the female shoal increased the total group size, potentially altering the focal male’s perceived level of risk. Our results could suggest that focal males tended to ignore risk-related social cues from the male shoal and instead relied on the social safety cues from the female shoal. It has been hypothesized that individuals might rely more on safety cues because they provide more precise information than risky cues (Luttbeg et al. 2020). Thus, focal males may have been using the female shoal as an additional source of social information regarding safety, rather than solely as potential mating opportunities. It is worth noting that the current study was conducted on a population of guppies collected from a high predation risk population, the Lopinot River (Deacon et al. 2018). Future studies should explore the impacts of background predation risk on the use of male vs. females as social information sources, particularly under fully natural conditions. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements We thank Kharran Deonarinesingh for his valuable assistance in the laboratory. A.L.C. and G.E.B. contributed equally to this project and should be considered as ‘joint senior authors’. Author contributions G.E.B, A.J.P.B, A.L.C. and F.D-S. contributed to the study conception and design. All authors participated in data collection. Statistical analyses and figures were done by A.J.P.B. I.W.R. secured collection and animal care permits for field work in Trinidad. The first draft of the manuscript was written by A.J.P.B and F.D.S and all authors commented on previous versions of the manuscript. All authors contributed to and approved the final manuscript. Funding This work was financially supported by the Natural Sciences and Engineering Research Council of Canada Discovery Grants program and Concordia University to G.E.B. Data availability Data are available in supplemental material. Declarations Conflict of interest The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement Data are available in supplemental material.