Presence of an audience influences habitat use in male Poecilia reticulate

Meghan C. Pierse1,3, Kady L. Loubier1,3, and Teresa L. Dzieweczynski1,2

1. Department of Psychology, University of New England Biddeford, ME 04005 2. Corresponding author: Teresa L. Dzieweczynski, assistant professor, Department of Psychology, University of New England, 11 Hills Beach Rd, Biddeford, ME 04005, email: [email protected] 3. These authors contributed equally to every aspect of this study.

Volume 20, Issue 2 August 2010

Presence of an audience influences habitat use in male guppies Poecilia reticulate

Meghan C. Pierse1,3, Kady L. Loubier1,3, and Teresa L. Dzieweczynski1,2

1. Department of Psychology, University of New England Biddeford, ME 04005

2. Corresponding author: Teresa L. Dzieweczynski, assistant professor, Department of Psychology, University of New England, 11 Hills Beach Rd, Biddeford, ME 04005, email: [email protected]

3. These authors contributed equally to every aspect of this study.

ABSTRACT Although much work has focused on how the presence of an audience may influence and mate choice, little is known about how this may impact other behaviors such as resource defense or habitat use. While the social behavior of guppies, Poecilia reticulata, has been extensively studied in the contexts of predator inspection and mate choice copying, the existence of true audience effects has not been investigated in this . P. reticulata do not have a resource-based mating system, but males have been observed to form dominance hierarchies and compete for access to females. Male P. reticulata behavior may be altered when an audience is present to maintain access to females. To test this, pairs of male P. reticulata (n = 60) were allowed to interact in the presence of another male, a female, or no audience. Our findings show that male P. reticulata spent more time in the resource zone when an audience was present (p < 0.05). Males spent the most time in the audience zone and chase one another the most when a female audience was present (p < 0.05). These findings provide an important addition to the existing literature on audience effects by showing that the presence of other individuals can affect more than courtship and aggressive behavior.

INTRODUCTION influences the behavior of the interactants (audience effects: Evans & In recent years, there has been Marler 1994; Le Roux et al. 2008) or growing interest in how conspecific how that individual alters its behavior signals are generated and perceived after observing the interaction amongst groups of individuals beyond (eavesdropping: Naguib & Todt 1997; that of the well-studied, signaler- Oliveira et al. 2001). The presence of receiver dyad models. Understanding an audience has been shown to these conspecific signals can provide a markedly affect both aggressive and more accurate depiction of signaling courtship behavior (see Matos & interaction systems that occur in social Schlupp 2005 for review), and . The study of communication eavesdropping individuals obtain networks has predominately focused on valuable information about the physical either how the presence of an individual and social environment with little effort not directly involved in an interaction or cost (e.g. food patch location: Ryer &

2 Journal of Young Investigators August 2010

Olla 1991; Reader et al. 2003; mate associate in mixed sex shoals, creating quality: Doutrelant & McGregor 2000; a social network in which individuals Mennill et al. 2003). can transmit and receive information to and from multiple individuals. P. Moreover, the sex of the reticulata have been found to engage in audience influences the nature of the tit for tat (i.e. turn-taking during predator audience effect. For example, in inspection; Dugatkin & Alfieri 1991), Siamese fighting fish, males are less mate choice copying (i.e. females aggressive in male-male interactions choose mates by watching other when a female audience is present; matings; Godin et al. 2005), and social however, aggression increases if a (Laland & Williams 1997), male has a nest (Matos & McGregor which allow them to gain 2002; Dzieweczynski et al. 2005). environmentally-relevant information by Likewise, male Atlantic mollies change watching conspecifics. Given the their mate preference when a male influence that social environment has audience is present (Plath et al. 2008). been found to have on behavior in this These sex-specific audience effects species, we hypothesized that the have rarely been examined outside the presence of an audience influences context of courtship and aggressive behavior as well. behavior (but see Evans & Marler Even non-territorial species like 1994), and the effects of an audience the P. reticulata may compete for on resource defense have not been access to resources, especially when explored. food or access to females is scarce (Basquill & Grant 1998). Resource Poecilia reticulata, familiarly defense theory predicts that aggression known as guppies, are a live-bearing should increase as the value of a fish native to Trinidad and parts of resource increases (Enquist & Lemar South America. They are sexually 1987). The nature of aggressive dimorphic, with females being larger contests changes significantly when the than males, and have a non-resource value of the resource in question based, promiscuous mating system changes or if asymmetries in ownership (Liley & Seghers 1975). In such a arise. In addition to resource availability mating system, both males and females (Kolluru & Grether 2005), other factors will mate multiple times and females such as population density (Cole & and males choose one another based Noakes 1980), operational sex ratio on attributes of the partner rather than (Grant et al. 2000), and habitat some external resource, such as food, complexity (Blanchet et al. 2006) can that the partner might provide. Males affect the amount and intensity of can either court females through the aggression. use of ‘S’ shaped sigmoid displays or This study examines whether the force copulate and all males can use presence of an audience as well as the either of these strategies at any time sex of that audience influences the time (Farr 1976). Males vie against one that male P. reticulata spend near a another for physical proximity to resource and the aggressiveness of receptive females to increase their their behavior. We hypothesized that chances of mating (Bruce & White males would differ in their use of space 1995; Pilastro et al. 2007). P. reticulata

2 Journal of Young Investigators August 2010

based on the presence of an audience 10 audience males, and 10 audience and that male P. reticulata would spend females. A power analysis (P = 0.934) more time in a “preference zone” indicated that this sample size was containing resources when an audience adequate. All fish were purchased from is present. While P. reticulata do not Petco® in Biddeford, Maine. Tanks appear to defend areas of a stream or were kept at 78oF, and all animals were form dominance hierarchies in the wild fed Top flakes twice daily. (Farr 1975; Houde 1997), they have Fish were given two weeks to acclimate been found to act aggressively to to the room before testing. P. reticulata defend food patches (Magurran & were sexed as described above (Liley Seghers 1991). Moreover, linear & Seghers 1975) and housed in two dominance hierarchies that reduce separate 10 gallon tanks (31 x 25 x 51 aggression and provide access to cm) that each had a filter and a heater. resources have been found in P. Females were separated from the reticulata under lab conditions (Bruce & males for two weeks to eliminate sperm White 1995; Morrell et al. 2007). Thus, storage. males may spend more time near the A 10-gallon tank with brown resources to defend them from the gravel on the bottom was used as the other P. reticulata. experimental tank. The tank was We also hypothesized that the divided into three zones (resource, presence of a female audience would neutral, and audience) of equal size by cause male P. reticulata to spend more placing tape on the outside walls of the time near the resources and exhibit tank (see Figure 1). The resource zone more aggressive behavior in contained a real plant, a fake brown comparison to when a male audience is tree stump with an opening that acted present. Farr (1976) demonstrated that as refuge, and larger rocks were also males increased their courtship display placed throughout this zone. The frequency as the number of males and presence of plants and other forms of females present increased and that shelter located in the resource zone in males showed increased chasing this study were meant to mimic the behavior and sneak copulation structurally diverse habitats P. attempts in the presence of another reticulata experience in the wild. The male (Magellan et al. 2005). Thus, if a neutral zone was located in the middle male perceives the female as a of the tank, between the resource and potential mate, he may invest his time audience zones, which were empty in interacting with the other male or except for the gravel covering the attempting to lure the female into the bottom of the tank. The audience zone resource area where he might use the consisted of a 1L plastic soda bottle materials located there as refuge to with the top cut off in which the court uninterrupted (Hibler & Houde audience fish were placed during 2006). experiments.

METHODS AND MATERIALS

A total of 80 feeder P. reticulata, were used, including 60 male subjects,

3 Journal of Young Investigators August 2010

Tukey tests were conducted to examine differences in the time spent in the different zones as well as in the aggressive behavior between the treatments. Pearson correlations were conducted to determine if the location of Male A influenced the location of Male B in a pair, and vice versa, for each zone and treatment tested. All

Figure 1. Experimental set-up. The three statistics were conducted using zones, from left to right, are the resource zone, SigmaStat 3.0. neutral zone, and audience zone. In the resource zone, the brown circle represents a RESULTS fake tree stump and the small white circles depict stones. The hexagon with a fish inside represents the audience in the soda bottle The amount of time the located in the audience zone. interacting subjects spent in each zone during the female, male, and no Ten trials were conducted for audience treatments were examined. each of the three treatments: female Differences were found among the audience, male audience, and no treatment groups for time spent in the audience present. At the start of each resource zone (F2,59 = 10.46, P < 0.05) trial, two male P. reticulata of similar as well as in the audience zone (F2,59 = size were removed from the male 6.62, P < 0.05). As expected, males subject holding tank and placed in the spent more time in the resource zone neutral zone in a clear bottle for a when a male or female audience was three-minute acclimation period. A present, than when no audience was white partition was placed between the present (Tukey: q ≥ 3.58, P < 0.05; Fig. clear plastic bottle and the rest of the 2). However, the sex of the audience tank where the experimental fish were did not affect the time that males spent kept to prevent males from viewing the in the resource zone (Tukey: q = 2.12, audience zone prior to testing. P > 0.05), although males spent more Following the acclimation period, the time in the audience zone when no partition was removed and the time audience or a female audience was each subject spent in each zone present compared to when a male (resource, neutral, and audience), as audience was present (Tukey: q ≥ 3.48, well as any aggressive behaviors, such P < 0.05; Fig. 2). No significant as biting and chasing, were recorded differences were found for time spent in for a period of 10 minutes. The clear the neutral zone (F2,59 = 2.47, P > 0.05; plastic bottle was kept in the tank in the Fig. 2). no audience trials, so that everything except for the presence of an audience would be identical in the treatments. Males that were used for a given treatment were never used again. Kruskal-Wallis one-way analyses of variance (ANOVA) with post-hoc

4 Journal of Young Investigators August 2010

Figure 2. Mean length of time that males spent in the resource zone (black bars), neutral zone Figure 3. Mean number of chases males in a (light gray bars), and audience zone (dark gray pair performed to one another when a male, bars) during the control, male audience and female, or no audience was present. Error bars female audience groups. Error bars are ± 1 represent ± 1 SEM. SEM.

Generally, the time one male Males also performed more spent in a given area of the tank did not chases when a female audience was affect the amount of time that the other present compared to when a male male spent in that same area. There audience was present (Tukey: q = 4.19, was no significant difference in the P < 0.05). Chasing occurred the most amount of time paired males spent in in the female audience treatment and the resource zone in the presence of a the least in the no audience treatment. female audience (r = -0.03, P > 0.05), a Biting was observed once in the trials male audience (r = 0.39, P > 0.05), or where no audience was present, not at when no audience was present (N = all when a male audience was present, 20, r = -0.60, P > 0.05). The presence and three times in the presence of a of another male also did not affect the female audience. time that males spent in the audience DISCUSSION zone during any of the treatments (r ≤ -

0.14, P > 0.05). The presence of one This study explored whether the male in the neutral zone was correlated presence of a conspecific audience and with the presence of the other male the sex of that audience affects when there was no audience (N = 20, r resource defense behavior in male P. = 0.74, P < 0.05), but not when an reticulata. Males spent significantly audience was present (r ≥ 0.32, P > more time in the resource zone when 0.05). an audience was present compared to Aggression was rare and the control trials where no audience consisted predominantly of chasing. was present. These results are Males chased one another more when comparable to previous studies in an audience was present than when it Siamese fighting fish and Atlantic was absent (Tukey: q 3.62, P < 0.05; ≥ mollies, whose behavior changes in Fig. 3). the presence of an audience (e.g.

5 Journal of Young Investigators August 2010

Doutrelant et al. 2001; Plath et al. audience was present. Previous work 2008). has shown that males devote more We found that males spent less time to pursuing females when a time near the audience when a male competitor is present compared to audience was present and performed when they are alone with the female more chases when a female audience (Magellan et al. 2005). However, in our was present. Audience effect studies in study, the presence of a female fish in prior studies focused on how audience did not affect the amount of courtship or aggression changes when time males spent in this area. This a third party is present. Here we suggests that males are not spending demonstrated that the presence of an more time in this area when a female audience can also influence resource audience is present, but rather that they defense and habitat use, even in a non- are decreasing time spent here when a territorial species like P. reticulata. male audience is present. This could We also found that interacting suggest that males are avoiding or are males spent more time in the resource uninterested in the male audience, zone when an audience was present. which seems surprising in a shoaling There are a number of potential fish like P. reticulata which prefers to explanations for why this occurred. associate with other individuals rather While P. reticulata do not have a than swim alone. Male P. reticulata do resource based mating system, the spend more time shoaling with females formation of dominance hierarchies and than with other males (Lindstrom & aggression during foraging have been Ranta 1993) and shoal less than observed in this species (Bruce & females (Croft et al. 2003), which may White 1995). Thus, it could be explain why they preferred to associate beneficial for males to defend with the female audience. resources when an audience is present Alternatively, it is possible that as a means of preserving shelter that the amount of time males spent in this could be used to retreat from predators area decreased when the male or aggressive males. Seeking shelter audience was present because males from predators is an unlikely are spending more time in the resource explanation, however, as we would zone in this treatment. With the current expect males to spent equal amounts methodology, it is difficult to determine of time in this area across treatments whether spending time in the resource and this is not the case. Male P. zone is actively chosen by the males or reticulata experience less courtship that it results from males avoiding the interference and perform fewer chases audience zone. Regardless, the when visual barriers are present (Hibler presence of an audience does clearly & Houde 2006). Therefore, it is also impact where males are found in the possible that males were spending time tank. in the resource zone to find hidden Wild populations of P. reticulata areas to court a female without vary in the degree of aggression they interference from other males. exhibit in foraging situations and likely Males spent more time in the in the degree of male-male competition audience zone when a female audience as well (Magurran & Seghers 1991). was present than when a male Males do not typically show overt

6 Journal of Young Investigators August 2010

aggression but have been found to were housed in a stock tank with other compete for positions near females, males. Thus, they could have which affects their ability to inseminate established relationships with other females (Matthews et al. 1997). This males, allowing the formation of a lack of overt aggression observed in stable hierarchy and decreased many wild populations may reflect aggression and time spent near the familiarity-based social stability rather male audience. Ideally, individuals than a true lack of aggression in this would have no prior contact with one species. In support of this, unfamiliar another prior to experimentation, and males do engage in more aggressive this issue should be addressed in the behavior than familiar males (Price & future. Rodd 2006). Males rarely delivered In conclusion, these results nips to one another in our study and support the idea that audience effects nipping behavior tended to occur play a role in resource defense in male primarily when a female audience was P. reticulata. While other studies have present. Aggression, in the form of examined how male P. reticulata alter chases, was more common in the their courtship and aggressive behavior presence of a female audience than in in relation to intensity of male the presence of no audience or a male competition (Price & Rodd 2006) and audience, suggesting that males act predation risk (Evans et al. 2002), no aggresively to compete for a position study has examined audience effects in near the female. Males also performed this species. Furthermore, our findings more chases when a male audience showed that the sex of the audience was present than when there was no influenced resource defense in male P. audience, suggesting that the presence reticulata, where males spend more of any audience leads to increased time near the resource when a male male aggression. audience is present. This study While the time each male in a provides a novel addition to the study of pair spend in the different areas as not communication networks by examining correlated, they did spend time resource defense. swimming together, which could influence where males are found in the tank. Individuals of a number of fish ACKNOWLEDGEMENTS species, including P. reticulata, bluegill sunfish, and threespine stickleback, We would like to acknowledge B. preferentially associate with familiar Heather Bleakley, Maryann Corsello, individuals (e.g. Krause et al. 2000 for and Jennifer Wieselquist for their review; Croft et al. 2004), which might comments on an earlier version of this enhance predator escape responses or manuscript. This study was conducted even increase foraging efficiency. as an independent research component Familiarity may also help stabilize of the Advanced Methods and dominance hierarchies, hence reducing Techniques in Psychobiology course. unnecessary aggression amongst This project was funded through the individuals in a population. department of psychology at the While males did not have University of New England and met contact with females in the lab, they IACUC compliance.

7 Journal of Young Investigators August 2010

Siamese fighting fish, splendens. REFERENCES Behavioral Ecology 12, 283-286.

Basquill SP, Grant JWA. (1998) An Dugatkin LA, Alfieri M. (1991) Tit-for-tat increase in habitat complexity reduces in P. reticulata (Poecilia reticulata): the aggression and monopolization of food relative nature of cooperation and by zebra fish ( rerio). Canadian defection during predator inspection. Journal Zoology 76, 770-772. Evolutionary Biology 5, 300-309.

Blanchet S, Dodson JJ, et al. (2006) Dzieweczynski TL, Earley RL, et al. Influence of habitat structure and fish (2005) Audience effect is context density on Atlantic salmon Salmo salar dependent in siamese fighting fish, L. territorial behavior. Journal of Fish Betta splendens. Behavioral Ecology Biology 68, 951-957. 16, 1025-1030.

Bruce KE, White WG. (1995) Agonistic Enquist M, Lemar O. (1987) Evolution relationships and sexual behaviour of fighting behavior: the effect of patterns in male P. reticulata, Poecilia variation in resource value. Journal reticulata. Behaviour 50, 1009- Theoretical Biology 127, 187-205. 1021. Evans CS, Marler P. (1994) Food Cole KS, Noakes DLG. (1980) calling and audience effects in male Development of early social behavior of chickens, Gallus gallus: their rainbow trout, Salmo gairdneri (Pisces, relationships to food availability, Salmonidae). Behavioural Processes 5, courtship and social facilitation. Animal 97-112. Behaviour 47, 1159-1170.

Croft DP, Arrowsmith B J, et al. (2003) Mechanisms underlying shoal Evans JP, Kelley JL, et al. (2002) composition in the Trinidadian P. Female behavior mediates male reticulata (Poecilia reticulata). Oikos courtship under predation risk in the P. 100, 429-438. reticulata (Poecilia reticulata). Behavioral Ecology Sociobiology 52, Croft DP, Arrowsmith BJ, et al. (2004) 496-502. Intra-sexual preferences for familiar fish in male P. reticulata. Journal Fish Farr JA. (1975) The role of predation in Biology 64, 279-283. the evolution of social behavior of natural populations of the P. reticulata, Doutrelant C, McGregor PK. (2000) Poecilia reticulata (Pisces: Poeciliidae). Eavesdropping and mate choice in Evolution 29, 151-158. female fighting fish. Behaviour 137, 1655-1669. Farr JA. (1976) Social facilitation of male sexual behavior, intrasexual Doutrelant C, McGregor PK, et al. competition and sexual selection in the (2001) The effect of an audience on P. reticulata, Poecilia reticulata. intrasexual communication in male Evolution 30, 707-717.

8 Journal of Young Investigators August 2010

Godin J-G J, Herdman EJE, et al. (2005) Social influences on female Liley NR, Seghers BH. (1975) Factors mate choice in the P. reticulata, affecting the morphology and behavior Poecilia reticulata: generalized and of the P. reticulata (Poecilia reticulata) repeatable trait-copying behavior. in Trinidad. In Function and Behavior in Animal Behaviour 69, 999-1005. Evolution (Manning, A., ed.), pp. 92- 118. Oxford: Oxford University Press. Grant JWA, Gaboury CL, et al. (2000) Competitor-to-resource ratio, a general Lindstrom K, Ranta E. (1993) Social formulation of operational sex ratio, as preferences by male P. reticulata, a predictor of competitive aggression in Poecilia reticulata, based on shoal size Japanese medaka (Pisces: Oryziidae). and sex. Animal Behaviour 46, 1029- Behavioral Ecology 11, 670-675. 1031.

Hibler TL, Houde AE. (2006) The effect Magellan K, Petterson LB, et al. (2005) of visual obstructions on the sexual Quantifying male attractiveness and behavior of P. reticulata: the mating behavior through phenotypic importance of privacy. Animal size manipulation in the Trinidadian P. Behaviour 72, 959-964. reticulata, Poecilia reticulata. Behavioral Ecology Sociobiology 58, Houde AE. (1997) Sex, color, and mate 366-374. choice in P. reticulata. Princeton, NJ: Princeton University Press. Magurran AE, Seghers BH. (1991) Variation in schooling and aggression Kolluru GB, Grether GF. (2005) The amongst P. reticulata (Poecilia effects of resource availability on reticulata) populations in Trinidad. alternative mating tactics in P. reticulata Behaviour 118, 214-234. (Poecilia reticulata). Behavioral Ecology 16, 294-300. Matthews IM, Evans JP, et al. (1997) Male display rate reveals ejaculate Krause J, Butlin RK, et al. (2000) The characteristics in the Trinidadian P. social organization of fish shoals: a test reticulata Poecilia reticulata. of the predictive power of laboratory Proceedings Royal Society London B experiments for the field. Biological 264, 695-700. Reviews 75, 477-501. Matos RJ, McGregor PK. (2002) The Laland KN, Williams K. (1997) Shoaling effect of the sex of an audience on generates social learning of foraging male-male displays of Siamese fighting information in P. reticulata. Animal fish (Betta splendens). Behaviour 139, Behaviour 53, 1161-1169. 1211-1221.

Le Roux A, Cherry MI, et al. (2008) The Matos RJ, Schlupp I. (2005) Performing audience effect in a facultatively social in front of an audience: signalers and mammal, the yellow mongoose, the social environment. In Animal Cynictis penicillata. Animal Behaviour Communication Networks (McGregor, 75, 943-949.

9 Journal of Young Investigators August 2010

P. K., ed.), pp. 63-83. Cambridge: Ryer CH, Olla BL. (1991) Information Cambridge University Press. transfer and the facilitation and inhibition of feeding in a shoaling fish. Mennill DJ, Boag PT, et al. (2003) The Environmental Biology of Fishes 30, reproductive choices of eavesdropping 317-323. female black-capped chickadees, Poecilie atricapillus. Naturwissenschaften 90, 577-582.

Morrell LJ, Hunt KL, et al. (2007) Diet, familiarity and shoaling decisions in P. reticulata. Animal Behaviour 74, 311- 319.

Naguib M, Todt D. (1997) Effects of dyadic interactions on other conspecific receivers in nightingales. Animal Behaviour 54, 15351-15543.

Oliveira RF, Lopes M, et al. (2001) Watching fights raises fish levels. Nature 409, 475.

Pilastro A, Mandelli M, et al. (2007) Copulation duration, insemination efficiency and male attractiveness in P. reticulata. Animal Behaviour 74, 321- 328.

Plath M, Blum D, et al. (2008) Audience effect alters mating preferences in a livebearing fish, the Atlantic molly, Poecilia mexicana. Animal Behaviour 75, 21-29.

Price AC, Rodd FH. (2006) The effect of social environment on male-male competition in P. reticulata (Poecilia reticulata). 112, 22-32.

Reader SM, Kendall JR, et al. (2003) Social learning of foraging sites and escape route in wild Trinidadian P. reticulata. Animal Behaviour 66, 729- 739.

10 Journal of Young Investigators August 2010