Intragroup social dynamics vary with the presence of neighbors in a cooperatively breeding fish

Intragroup social dynamics vary with the presence of neighbors in a cooperatively breeding fish Conflict is an inherent part of social life in group-living species. Group members may mediate con- flict through submissive and affiliative behaviors, which can reduce aggression, stabilize domin- ance hierarchies, and foster group cohesion. The frequency and resolution of within-group conflict may vary with the presence of neighboring groups. Neighbors can threaten the territory or resour- ces of the whole group, promoting behaviors that foster within-group cohesion. However, neighbors may also foster conflict of interests among group members: opportunities for subordin- ate dispersal may alter conflict among dominants and subordinates while opportunities for extra- pair reproduction may increase conflict between mates. To understand how neighbors mediate within-group conflict in the cooperatively breeding fish Neolamprologus pulcher, we measured be- havioral dynamics and social network structure in isolated groups, groups recently exposed to neighbors, and groups with established neighbors. Aggression and submission between the dom- inant male and female pair were high in isolated groups, but dominant aggression was directly pri- marily at subordinates when groups had neighbors. This suggests that neighbors attenuate conflict between mates and foster conflict between dominants and subordinates. Further, aggression and submission between similarly sized group members were most frequent when groups had neigh- bors, suggesting that neighbors induce rank-related conflict. We found relatively little change in within-group affiliative networks across treatments, suggesting that the presence of neighbors does not alter behaviors associated with promoting group cohesion. Collectively, these results pro- vide some of the first empirical insights into the extent to which intragroup behavioral networks are mediated by intergroup interactions and the broader social context. Key words: affiliation, colony, conflict, exponential random graph model, Neolamprologus pulcher, network. Conflict, which often manifests as aggression between group mem- subordinate individuals make use of both submissive and affiliative bers, is an inherent part of social life in group-living species and can displays to mitigate aggression from dominant group members and reduce the benefits of group living by increasing social stress, reduc- increase their likelihood of being tolerated within the group ing group productivity and leading to group dissolution if left unre- (Huntingford and Turner 1987; Bergmu ¨ ller and Taborsky 2005; solved (Aureli et al. 2002). In many group-living species, Bruintjes and Taborsky 2008). Submissive behavior can facilitate V C The Author(s) (2018). Published by Oxford University Press. 1 This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 2 Current Zoology, 2018, Vol. 0, No. 0 group stability by enforcing dominance hierarchies and settling con- understand how the presence of neighboring groups was associated flict among individuals within a social group (Huntingford and with changes in within-group social dynamics. To do this, we ex- Turner 1987; Hick et al. 2014), whereas affiliative behavior (e.g., perimentally created laboratory groups of Neolamprologus pulcher, grooming in primates) can reinforce friendships, encourage recon- a cooperatively breeding cichlid fish native to Lake Tanganyika in ciliation, and promote intragroup cooperation (Radford et al. East Africa. These fish form colonies of 2–200 permanent territorial 2016). groups (Stiver et al. 2007). Each group is composed of a dominant The frequency of aggressive, submissive, and affiliative behaviors male and female pair, with 0–20 subordinates that provide help in exchanged among group members often varies across groups the form of territory defense, territory maintenance, and alloparen- (Kutsukake and Clutton-Brock 2008; Madden et al. 2009; tal care (Wong and Balshine 2011a). Dominance is strictly size- Kutsukake and Clutton-Brock 2010), as well as across time within based, such that the dominant male and female are the largest indi- the same social group (Cantor et al. 2012; Godfrey et al. 2013; viduals and subordinates form sex-specific, size-based dominance Bierbach et al. 2014). Given that social stability and the nature of hierarches (Wong and Balshine 2011a). Male group members main- social interactions within a group influence individual fitness (Silk tain consistent differences in size, which likely reduces conflict that et al. 2003; Barocas et al. 2011; Archie et al. 2014), there have been would otherwise arise among similarly sized individuals (Heg et al. efforts to first quantify social structure within groups and then to 2004; Hamilton et al. 2005; Hamilton and Heg 2008). Submission understand factors that modulate social dynamics within and across is an effective behavior that reduces aggression in N. pulcher groups. Mounting evidence demonstrates that social interactions (Bergmu ¨ ller and Taborsky 2005; Bruintjes and Taborsky 2008), within a group are correlated with group-level attributes, such as while affiliative behavior is used to reinforce participation in terri- group size (Fischer et al. 2014), the relative size of group members tory defense (Bruintjes et al. 2016) and is associated with reduced (Hamilton et al. 2005), and the sex of group members (Kutsukake cortisol levels (Ligocki et al. 2015b). Neighboring groups are not and Clutton-Brock 2008). However, much less is known about how direct competitors for food: individuals in this species feed in the the social environment beyond the level of the group, specifically the water column and territories are used as protection from predators presence of other conspecific groups, influences social dynamics and breeding substrate (Gashagaza 1988; Wong and Balshine among group members. 2011a). Further, whole-group takeovers of neighboring territories There are at least 3 ways in which neighboring groups can alter are relatively rare in this species; instead, outside threats traditional- within-group dynamics by either reducing or promoting conflict ly come from single individuals seeking out reproductive opportuni- among a given subset of group members. First, neighboring groups ties, as male and female dominants lose reproduction to can threaten the territory or resources of an established group and subordinates within their group as well as to individuals in neigh- may incentivize group members to quickly resolve or reduce conflict boring groups (Dierkes et al. 1999; Stiver et al. 2009; Hellmann within their own group in order to facilitate cooperation in et al. 2015a). Further, outside individuals may threaten the stability between-group conflict (Radford 2008a; Radford and Fawcett of group composition. Subordinates disperse between groups 2014; Bruintjes et al. 2016). Studies examining territorial intrusions (Bergmu ¨ ller et al. 2005a; Jungwirth et al. 2015) and while domi- and conflict have found that affiliation between dominant and sub- nants generally do not move between territories, dominant males ordinate group members is greater when intergroup conflict is are polygynous and are at least capable of usurping a neighboring higher (Radford 2008a; Bruintjes et al. 2016). In these cases, affilia- male’s territory (O’Connor et al. 2015). Consequently, these outside tive behavior could be exchanged for continued subordinate partici- individuals may only threaten a subset of group members, while pation in out-group conflicts (Seyfarth and Cheney 1984; Radford having neutral or even beneficial effects on other group members. 2011). Second, neighboring groups can offer opportunities for egg After isolated groups stabilized following group formation, we dumping (Arnold and Owens 2002) or extra-pair fertilizations exposed groups to neighbors and observed behavioral interactions (Griffith et al. 2002; Hellmann et al. 2015a). This may result in within and between groups when groups were isolated, immediately increased aggression between mates, as the presence of neighboring after groups were introduced to neighbors, and 30 days after the males and females may provide fitness benefits to one mate by introduction of neighbors. Because changes in 1 individual’s behav- increasing the number of offspring sired, but may be costly to the ior are not independent of those of other group members, we used other mate who may care for offspring that are not their own. exponential random graph models (ERGMs), in addition to trad- Finally, neighboring groups offer opportunities for partner choice by itional regression models, to evaluate changes in social dynamics allowing subordinates to potentially move between groups. If neigh- (Wasserman and Pattison 1996). ERGMs control for the inter- bors provide an opportunity for subordinates to negotiate based on dependency of behavioral relationships within a group and evaluate outside options offered by neighboring groups, dominants should be the extent to which social dynamics among group members are more tolerant (less aggressive and demand less help) of current sub- influenced by individual attributes (e.g., sex and size) and structural ordinates (Bergmu ¨ ller et al. 2005b; Grinsted and Field 2017; dependency in social ties (e.g., reciprocity). By using these models, Hellmann and Hamilton forthcoming). Conversely, if neighbors we can test hypotheses related to understanding how social structure provide dominants with an opportunity to replace current subordi- varies across different contexts and how emergent properties of the nates, neighbors may allow dominants to demand more help and be social system may contribute to these changes (Wasserman and more aggressive to subordinates when neighbors are present Pattison 1996; Robins et al. 2007; Lusher et al. 2013; Silk and (Bruintjes and Taborsky 2008; Hellmann and Hamilton 2014; Fisher 2017). Hellmann et al. 2015b). Consequently, the way in which partner We had 3 separate predictions regarding the impact of neighbor- choice impacts behavioral dynamics between dominants and subor- ing groups on within-group social dynamics. If within-group dynam- dinates should depend on who is “choosing” partners in this system ics change in response to the threat that neighboring groups pose to (Noe ¨ and Hammerstein 1995; Cant and Johnstone 2009). the focal group, then we predicted that affiliative behavior would be We compared within-group interactions before neighboring higher when groups had neighbors compared with when groups groups were visible and while neighboring groups were visible to were isolated, as affiliative behavior can be used to promote Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 Hellmann and Hamilton  Neighbors and within-group conflict 3 cooperation among current group members when the potential for barrier. Neolamprologus pulcher can distinguish familiarity, iden- intergroup conflict is high (Radford 2008a, 2011). If within-group tity, and sex using only visual cues (Balshine-Earn and Lotem 1998; dynamics change in response to the reproductive opportunities Frostman and Sherman 2004; Hellmann and Hamilton 2014; offered by neighboring groups, then we predicted that aggression be- Kohda et al. 2015; Sabol et al. 2017), suggesting that chemical cues tween the dominant male and female pair would be higher when are not necessary for individuals to accurately identify the sex and neighbors were present compared with when they were absent, as re- familiarity of neighboring group members. productive conflict between mates is expected to be higher when All eggs that were laid during the experiment were removed im- there are additional opportunities for extra-pair fertilizations mediately from the groups to remove any confounding influence of (Eggert and Sakaluk 1995; Valera et al. 2003; Goetz et al. 2008). If parental care on intragroup and intergroup social interactions, al- within-group dynamics change in response to opportunities for sub- though a previous study in this species found that the structure of ordinates to move between groups, then we predicted that aggres- interaction networks does not strongly vary with reproductive sion and submission between dominants and subordinates would be events (Dey et al. 2015). Nevertheless, any behavioral observations altered when neighbors were present compared with when neigh- conducted within 24 h of egg-laying and removal were removed bors were absent. from the dataset. Group formation and behavioral observations Materials and Methods From October 2014 to November 2015, we formed 31 social groups of unrelated N. pulcher, each composed of a dominant male and fe- Study organism and housing conditions male breeding pair and 2–3 subordinates with at least 1 subordinate All experimental fish were wild caught or F1 offspring of wild- male and 1 subordinate female per group (n¼27 groups with 3 sub- caught fish from the Kambwimba region of Lake Tanganyika (8 32 ordinates, n¼4 groups with 2 subordinates). All individuals were S, 31 9 E). All wild-caught fish had been in captivity in our lab for likely to be reproductively mature (SL> 35 mm: Taborsky 1985) at least 3 months prior to the beginning of the experiment. We used and sexed by examination of external genital papillae. In these a total of n ¼ 106 individual fish for this experiment. Prior to the ex- groups, the dominant male was the largest (and therefore, most periment, all fish were marked uniquely with elastomer dye and dominant) fish in the group and the dominant female was the second given a dorsal fin clip to indicate sex. Fish recover from this proced- largest fish in the group. All subordinates were at least 5 mm shorter ure rapidly and receiving these markers has no apparent effect on in SL than the dominant male and female. We assembled up to 8 subsequent behavior (Stiver et al. 2004; Dey et al. 2015). On the day groups at a time; we formed new groups by reshuffling former group that groups were formed, all group members were weighed to the members or using new individuals. Some individuals were therefore nearest 0.001 g (Ohaus Adventurer Pro AV213C) and measured for members of 2 social groups throughout the course of the 13-month standard length (SL) to the nearest 0.01 mm (Fisher Scientific experiment. We controlled for pseudoreplication in 2 ways. First, in Traceable calipers). To mirror natural conditions, a 12:12 h our regression models, we used random effects of individual and light:dark schedule was maintained for the duration of the experi- group identity to control for variation in behavior due to individual ment and water temperatures were kept constant at 2761 C. Fish identity. Although we used n¼106 total fish in this experiment, we were fed daily and ad libitum with either TetraMin flakes (5 days a had a total of n¼76 unique fish that were still present in the groups week) or frozen Daphnia or Artemia (twice per week). by day 30 (see below). Second, we ran our network analysis (see below for “Materials and Methods”) removing the minimum num- Experimental setup ber of groups necessary to produce a network in which each individ- During the experiment, 2 groups were placed on opposite sides of ual was a member of only 1 social group. a barrier in a 208-L (122 cm long  32 cm wide  53 cm high) All subordinates were unrelated to the dominants within their aquaria lined with 3 cm of black sand substrate. Each group had 2 group, and all group members had never interacted prior to group inverted terracotta flowerpot halves that served as shelter and po- formation. Groups were given 30 days after group formation to sta- tential breeding substrate, as well as 2 PVC tubes near the top of bilize, as aggression can be high while dominance hierarchies are the tank that served as hiding spots for subordinate fish. The 2 being established. In the initial days after group formation, some groups in each aquarium were separated by 2 clear plexiglass bar- subordinates received high levels of aggression from the dominants riers that were flush against the walls and floor of the aquarium and were subsequently removed from the group to prevent further and extended above the top of the water line. This largely pre- injury. These subordinates were not replaced; instead, we controlled vented water flow between the 2 groups, although some water for changes in group composition throughout the experiment (due flow between the groups was possible and therefore, chemical to death or removal of group members) by only analyzing changes communication between the groups may have occurred. in network structure using groups that did not change in compos- Consequently, differences in interactions when neighbors were ition across time periods. blocked from view versus visible may not be due so much to the At 30 days post-group formation, we had 11 social units that complete absence of neighbors or different perceived densities of were composed only of a male and female pair (i.e., social mates) conspecifics, but rather due to the absence of interactions with and 20 social units that were composed of a dominant male and fe- neighboring groups and differences in how much groups know male pair with subordinates (n¼5 groups with 1 subordinate, n¼9 about their neighbors (e.g., group composition, size of group groups with 2 subordinates, and n¼6 groups with 3 subordinates). members). At 30-days post-formation, each pair or group was observed for During the first 40 days after group formation, there was an opa- 30 min daily for 10 days (days 30–39). After 10 days, the opaque que barrier between the plexiglass barriers that prevented groups barrier isolating the groups was removed, such that groups were from seeing or interacting with each other. After 40 days, the opaque now visible to their neighbor on the other side of the tank. Groups barrier was removed and groups were able to interact across a clear were observed for 30 min daily for the first 10 days after removal of Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 4 Current Zoology, 2018, Vol. 0, No. 0 the opaque barrier to understand the impact of novel neighbors on was the difference in the number of interactions between 2 time peri- intragroup dynamics (days 40–49). Thirty days after the removal of ods. Because all edges must be positive values, we added the abso- the opaque barrier, we observed groups again for 30 min daily for lute value of the minimum edge weight to all within-group edge 10 days (days 70–79). weights to make all values positive. Therefore, higher values denote All behavioral observations were recorded and videos were that behaviors were more common in the later period than the ear- scored by the same observer (J.K.H.) using species-specific etho- lier period. grams (Ligocki et al. 2015a; Reddon et al. 2015; Sopinka et al. We analyzed network structure using ERGMs (Wasserman and 2009). Behaviors were categorized as overt aggressive attacks (ram, Pattison 1996). As in traditional regression models, these models bite, mouth fight), restrained aggressive displays (fin raise, fast ap- test how independent variables predict the weight of the edges. proach, operculum spread, head jerk, head down display), submis- Because ERGMs control for the interdependency of social ties when sive displays (tail quivers, hook, submissive posture), and affiliative evaluating behavioral data, they improve upon traditional statistical behaviors (parallel swim, bump, join). Because aggression toward methods (Whitehead 2008). ERGMs allow for simultaneous estima- neighboring groups was across a barrier, we scored any behaviors as tions of substructures (e.g., transitivity of relationships) and individ- overt aggression when the actor made contact with the barrier (e.g., ual attributes (e.g., size) or pairwise attributes (e.g., sex homophily) rammed the barrier in an attempt to ram a neighboring fish). that contribute to network data. Similar to Dey et al. (2015),we Within-group overt aggression was relatively rare, so we combined assembled a supernetwork composed of all social groups and counts of overt and restrained aggression when analyzing within- restricted all possible edges to those occurring among group mem- group dynamics. bers (Figure 1). To examine factors that contribute significantly to determining network structure, we tested the independent variables of: (1) the “sum” term, which is similar to the intercept term in a re- Statistical and network analysis gression model, (2) status homophily, which tests whether interac- tions are more likely to occur between individuals of the same status To test how affiliative, submissive, and aggressive interactions be- (dominant or subordinate), (3) sexual homophily, which tests if tween the dominant male and female pair (i.e., between social there is an increased chance of interactions between same-sex dyads, mates) changed with the presence of neighbors, we used generalized (4) actor effect of sex, which tests if 1 sex is more likely to initiate linear mixed models [GLMMs; R packages lmerTest (Kuznetsova behavioral interactions, and (5) dyadic differences in size (SL), et al. 2017) and glmmADMB (Fournier et al. 2012; Skaug et al. which tests if interactions are more likely to occur between individu- 2016)] with a negative binomial distribution because count data als close in size. Further, we tested for structural dependence be- were overdispersed. Dependent variables for each model were the tween edges by evaluating the tendency for (6) reciprocity, which sum of all affiliative, submissive, or aggressive behaviors, respective- tests if the weight of an edge from 1 group member to another pre- ly, observed across the 10 observations. In each model, we included dicts the weight of the reciprocal edge, and (7) cyclical triads, or the fixed effects of period (pre-exposure to neighbors, days 30–39; im- tendency of individuals to form cyclical triads, which are markers of mediate post-exposure to neighbors, days 40–49; delayed post- unstable dominance hierarchies. These variables were chosen prior exposure to neighbors, days 70–79) and the presence of subordinate to examining the data based on our predictions outlined in the males or females in the group (binomial) to control for variation in behavioral interactions between dominant males and females due to group composition. For models of aggression and submission, we also included a fixed effect of the amount of aggression received from their social mate, to account for variation in dominant male or female behavior due to their social mate’s activity. For models examining affiliative behavior, we included the fixed effect of the amount of affiliation received from their social mate. We used Tukey’s honest significant difference tests (R package multcomp) as a post hoc analysis to control for multiple comparisons and analyze where the differences lie among the treatment groups. We also included random factors of individual ID (nested within group iden- tity) in all models, because many individuals were a member of 2 so- cial groups. Dominant death occurred in 2 groups; therefore, n ¼ 29 social pairs/groups were used to evaluate social dynamics between the dominant male and female pair. We analyzed social network structure in the groups with a dom- inant pair and subordinates [R version packages: statnet (Handcock et al. 2008), ergm (Hunter et al. 2008), and ergm.count (Krivitsky 2013)]. We built weighted, directed networks of aggression, submis- sion, and affiliation for the isolated time period. The weight of the edges in these networks was determined by the total number of interactions directed from 1 individual to another across the 10 observations in the time period. Then, to quantify how within-group social dynamics changed between pre-exposure (days 30–39) and Figure 1. Aggressive supernetwork structure for N. pulcher groups prior to ex- immediate post-exposure (days 40–49) periods and between imme- posure to neighbors (days 30–39). Larger nodes represent dominant individu- diate and delayed post-exposure (days 70–79) periods, we assembled als, red nodes represent females, and blue nodes represent males. Thicker edges indicate that more aggression was exchanged between a given dyad. difference networks. In the difference network, each edge weight Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 Hellmann and Hamilton  Neighbors and within-group conflict 5 introduction and based on characteristics that define sex-specific, 40–44). We also evaluated changes within the delayed post- size-based dominance hierarchies (Wong and Balshine 2011a; exposure period (days 70–74 vs. days 75–79). If within-group dy- McDonald and Shizuka 2012; Dey et al. 2013, 2015; Dey and namics merely change according to time, then we expected to see Quinn 2014). changes in social dynamics changing similarly within and across ERGMs use a Markov-chain Monte Carlo estimation technique time periods. These methods and results are included in the to approximate the maximum likelihood. We specified a sampling Supplementary Material. interval of 5000 and a burn-in of 50,000 proposals, and used Finally, we used GLMMs with a negative binomial distribution Poisson reference graphs for each model (Dey et al. 2015). Models to test for differences in aggression toward the neighbor group due were checked for degeneracy and goodness of fit using the to status (dominant or subordinate), sex (male or female), treatment mcmc.diagnostics function. Examinations of model diagnostics did period (days 40–49 or days 70–79), and neighbor group size. To not indicate a high correlation between status homophily and size understand how group dynamics impacted aggression to neighbor- difference so we included both terms in our ERGMs; further, the ing groups, we ran 2 GLMMs. The first had dominant aggression to effects of status homophily and size difference did not change when neighbors as the dependent variable, with fixed effects of dominant the only 1 or both variables were included in the models (Hellmann male aggression to dominant females, dominant female submission JK, Hamilton IM, unpublished data). To eliminate potential issues to the dominant male, and the presence of subordinates in the group with pseudoreplication in the ERGM models, which do not allow (binary). The second had subordinate aggression to neighbors as the for random effects of identity, we only used a subset of observed dependent variable, with fixed effects of aggression and affiliation groups such that each individual was only represented in the net- received from the dominants. For all models, we checked independ- work once. Therefore, we analyzed a total of 17 groups in the net- ent variables for collinearity. Individual ID nested within group was works examining social dynamics in isolated groups (n ¼ 5 groups included as a random effect for all models. of 3 individuals, n ¼ 6 groups of 4 individuals, n ¼ 6 groups of 5 individuals). Because difference networks contained only groups Results that did not change in composition between time periods, we ana- Are interactions between social pairs influenced by the lyzed 13 groups in the difference network comparing pre-exposure groups and immediate post-exposure groups with neighbors (n ¼ 5 presence of neighbors? groups of 3 individuals, n ¼ 2 groups of 4 individuals, n ¼ 6 groups Dominant males were less aggressive (i.e., had a lower frequency of of 5 individuals) and 11 groups in the difference network comparing aggressive acts) to dominant females when neighbors were recently immediate and delayed post-exposure groups (n ¼ 6 groups of 3 present compared with the pre-exposure period when groups were individuals, n ¼ 2 groups of 4 individuals, n ¼ 3 groups of 5 isolated (GLMM with Tukey’s HSD: Z ¼2.36, 0.048; Figure 2A); individuals). this trend weakened when neighbors were present for 30 days, such Because potential changes in group dynamics may be due to that there was no significant difference in dominant male aggression changes in time rather than changes in the visibility of neighbors, we between the pre-exposure period and the delayed post-exposure analyzed if network structure changed within the first 10 days of ob- period (Z ¼1.87, P ¼ 0.14; Figure 2A). Dominant females were servation (days 30–34 vs. days 35–39) and if the magnitude of that similarly aggressive to dominant males in the pre-exposure period change was less than the change in network structure between when compared with the post-exposure periods (immediate: Z ¼0.15, groups were isolated (days 35–39) and neighbors were visible (days P ¼ 0.99; delayed: Z ¼0.71, P ¼ 0.75), but were significantly Figure 2. (A) Dominant male aggression toward the dominant female was higher prior to exposure to neighbors (white: days 30–39) compared with the immedi- ate post-exposure period when neighbors were recently visible (light blue: days 40–49), but this trend did not persist into the delayed post-exposure period when neighbors had been present for 30 days (darker blue: days 70–79). (B) Dominant female submission toward the dominant male was higher prior to exposure to neighbors (white: days 30–39) compared with both periods when neighbors were present. Data presented are the residuals of the regression model without the fixed effect of treatment period, plotted against treatment period. Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 6 Current Zoology, 2018, Vol. 0, No. 0 neighbors had been present for 30 days (non-significant effect in im- mediate vs. delayed post-exposure difference network; Table 2). During the pre-exposure period when groups were isolated, there were similar frequencies of aggression between same-sex and oppos- ite sex group members, while submission was more frequently exchanged between opposite-sex group members relative to same- sex group members (sexual homophily: Table 1). Both aggression and submission became significantly more sexually homophilic when groups were initially exposed to neighbors compared with the pre-exposure period (positive effect of sexual homophily in the pre- vs. post-exposure difference network: Table 2); this effect persisted into the delayed post-exposure period for submission (non-signifi- cant effect of immediate vs. delayed post-exposure difference net- work; Table 2), but not for aggressive interactions (negative effect in difference network; Table 2). In the pre-exposure period, submission and aggression were most commonly exchanged between group members with a large size difference (positive effect of SL difference: Table 1 and Figure 4). However, groups in the immediate post-exposure Figure 3. Dominant females were more aggressive to dominant males in their period had significantly more aggression and submission group when there were subordinate females present in their group. Data pre- exchanged between similarly sized individuals compared with iso- sented are the residuals of the regression model without the fixed effect of lated groups (negative effect of size difference: Table 2 and subordinate female presence, plotted against the binomial variable of subor- Figure 4). When groups were isolated, dominant aggression and dinate female presence in the group. submission were more frequently exchanged between the domin- ant pair rather than directed toward subordinates (dominant sta- more submissive to dominant males during the pre-exposure period tus homophily: Table 1 and Figure 4); however, in the immediate compared with the post-exposure periods, even after controlling for post-exposure period, dominant aggression and submission were variation in the amount of aggression received from the dominant more frequently exchanged with subordinates than with their so- male (immediate: Z ¼7.19, P< 0.001; delayed: Z ¼6.65, 74 74 cial mate (Table 2 and Figure 4). All of these effects persisted into P< 0.001; Figure 2B). Dominant male aggression toward the dom- the delayed post-exposure period for aggressive interactions; inant female and dominant female submission were not significantly however, these changes in the submission network were stronger altered by the presence of subordinate males (DM: Z ¼0.43, in the immediate post-exposure period compared with the delayed P ¼ 0.67; DF: Z ¼1.42, P ¼ 0.16) or subordinate females (DM: post-exposure period (positive effects for both parameters; Z ¼0.93, P ¼ 0.35; DF: Z ¼0.12, P ¼ 0.90) in the group. 74 74 Table 2). However, dominant females were significantly more aggressive to Across all time periods, males and females initiated similar levels of dominant males when subordinate females were present in the group affiliative behaviors (no effect of actor sex in any network: Tables 1 (Z ¼ 2.81, P ¼ 0.005; Figure 3), although there was no significant and 2). Affiliation was exchanged primarily between males and impact of subordinate male presence (Z ¼ 0.25, P ¼ 0.80). femalesratherthan between individualsof the same sex in both the Dominant male or female affiliative behavior toward their mate was pre-exposure period and immediate post-exposure period (negative ef- not significantly different in the pre-exposure period compared with the fect of sexual homophily: Table 1; no effect of either parameter in the immediate post-exposure (DF: Z ¼0.63, P¼ 0.80; DM: Z ¼1.94, 74 74 difference network: Table 2); however, affiliation was significantly P¼ 0.13) period or the delayed post-exposure period (DF: Z ¼ 0.01, more sexually homophilic when groups had neighbors for 30 days P¼ 1.00; DM: Z ¼1.40, P¼ 0.34). It also did not vary with the pres- compared with recent neighbors (positive effect of sexual homophily ence of subordinate males (DF: Z ¼1.63, P¼ 0.10; DM: Z ¼ 0.40, 74 74 in thedifferencenetwork: Table 2). In the pre-exposure period when P¼ 0.69) or subordinate females (DF: Z ¼0.70, P¼ 0.48; DM: groups were isolated, affiliative behavior was more frequently Z ¼0.25, P¼ 0.80) in the group. exchanged between individuals with a large size difference relative to similarly sized individuals (SL, negative effect of size homophily; Are social interactions within a group influenced by the Table 1), but affiliative behavior between similarly sized individuals presence of neighbors? was more common in the immediate post-exposure period than the Because some individuals were used across more than 1 group and pre-exposure period (negative effect of size differences: Table 2). This ERGMs do not allow for random effects of individual identity, we effect continued to strengthen in the delayed post-exposure period analyzed network structure in a subset of groups to avoid potential (Table 2). Across both the pre-exposure and immediate post-exposure issues with pseudoreplication. Across all time periods, males and time period, there was a higher frequency of affiliation exchanged be- females initiated similar levels of aggression (non-significant effect of tween individuals of the same rank than between individuals of differ- actor sex in all networks: Tables 1 and 2). Females were more submis- ent ranks (i.e., dominant–dominant affiliation was more common than sive to other group members than males (negative effect of actor sex: dominant–subordinate affiliation; positive effect of status homophily, Table 1), although a significant effect of actor sex in the difference Table 1; no effect of status homophily in difference network, Table 2). network indicates that this trend was stronger prior to exposure to However, subordinates exchanged more affiliation with dominants neighbors compared with the immediate post-exposure period when neighbors were present for 30 days compared with when neigh- (Table 2). This difference between pre-exposure and post-exposure bors were recent (negative effect of subordinate homophily in the dif- groups persisted into the delayed post-exposure period where ference network: Table 2). Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 Hellmann and Hamilton  Neighbors and within-group conflict 7 Table 1. Results of ERGM fit for behavioral networks for groups prior to exposure to neighbors (days 30–39) Pre-exposure period Aggression Submission Affiliation Estimate SE P Estimate SE P Estimate SE P Sum 2.15 0.06 <0.001 0.20 0.19 0.30 2.12 0.06 <0.001 Difference in SL 0.06 0.003 <0.001 0.06 0.007 <0.001 0.05 0.003 <0.001 Sexual homophily 0.02 0.03 0.52 0.30 0.09 0.001 0.34 0.05 <0.001 Actor sex (male) 0.03 0.03 0.35 0.14 0.05 0.005 0.007 0.02 0.73 Status homophily Dominant–dominant 0.16 0.05 0.001 1.33 0.10 <0.001 2.36 0.05 <0.001 Subordinate–subordinate 1.09 0.07 <0.001 1.04 0.18 <0.001 0.42 0.06 <0.001 Cyclical triads 0.73 0.05 <0.001 0.94 0.32 <0.001 0.93 0.04 <0.001 Reciprocity 1.07 0.04 <0.001 1.59 0.15 <0.001 0.37 0.02 <0.001 Bold values indicate significance (P<0.05). Notes: we tested the effects of dyadic differences in SL (positive values signify that individuals with bigger size differences interact more frequently), sexual and status homophily (positive values signify that individuals of the same sex/status interact most frequently), actor effects of sex (positive values signify that males ini- tiate behaviors more frequently than females), cyclical triads, and reciprocity. Table 2. Results of ERGM fit for difference networks comparing network dynamics of pre-exposure groups that were isolated (days 30–39) and post-exposure groups recently exposed to neighbors (days 40–49), as well as comparing post-exposure groups recently exposed to neighbors (days 40–49) to those with neighbors that had been present for 30 days (days 70–79) Pre- vs. immediate post-exposure period Aggression Submission Affiliation Estimate SE P Estimate SE P Estimate SE P Sum 5.88 0.01 <0.001 4.39 0.04 <0.001 5.38 0.02 <0.001 Difference in SL 0.007 0.001 <0.001 0.004 0.002 0.04 0.003 0.001 0.02 Sexual homophily 0.02 0.01 0.01 0.09 0.02 <0.001 0.006 0.009 0.52 Actor sex (male) 0.003 0.007 0.63 0.13 0.03 <0.001 0.008 0.009 0.37 Status homophily Dominant–dominant 0.05 0.02 0.002 0.11 0.04 0.001 0.02 0.02 0.17 Subordinate–subordinate 0.02 0.01 0.14 0.01 0.03 0.72 0.03 0.02 0.07 Cyclical triads 0.004 0.006 0.51 0.07 0.03 <0.001 0.04 0.006 <0.001 Reciprocity 0.03 0.006 <0.001 0.10 0.02 <0.001 0.001 0.008 0.87 Immediate vs. delayed post-exposure period Aggression Submission Affiliation Estimate SE P Estimate SE P Estimate SE P Sum 5.23 0.02 <0.001 0.02 0.005 <0.001 5.29 0.02 <0.001 Difference in SL 0.001 0.002 0.66 0.02 0.005 <0.001 0.005 0.002 0.002 Sexual homophily 0.03 0.01 0.03 0.05 0.05 0.33 0.06 0.01 <0.001 Actor sex (male) 0.006 0.009 0.53 0.01 0.04 0.78 0.005 0.007 0.45 Status homophily Dominant–dominant 0.01 0.02 0.59 0.17 0.06 0.007 0.02 0.02 0.36 Subordinate–subordinate 0.02 0.02 0.32 0.09 0.06 0.13 0.05 0.02 0.02 Cyclical triads 0.03 0.008 0.002 0.04 0.03 0.21 0.04 0.008 <0.001 Reciprocity 0.04 0.01 <0.001 0.02 0.02 0.47 0.02 0.01 0.01 Bold values indicate significance (P<0.05). Notes: Negative estimates indicate that a given variable had a stronger influence in the earlier time period. Local network substructures we are unsure if changes in cyclical triads and reciprocity across Aggressive, submissive, and affiliative networks were characterized time periods (isolated: days 30–39, new neighbors: days 40–49) are by a strong, negative effect of reciprocity, which means that individ- due to time or experimental treatment. uals who received high levels of aggression, submission, and affili- ation were unlikely to reciprocate those behaviors. Similarly, networks were characterized by strong negative effects of cyclical Between-group aggression and neighbor triads (Table 1), which means that cyclical triads were less frequent characteristics on within-group interactions than expected by chance. Negative effects of both reciprocity and In general, there was a higher frequency of aggression toward neigh- cyclical triads indicate that dominance hierarchies are stable. As the bors during the initial post-exposure period (days 40–49) compared influence of cyclical triads and reciprocity on network structure was with the delayed post-exposure period (days 70–79; GLMM: significantly different between the beginning (days 30–34) and end Z ¼3.03, P ¼ 0.003). Across both time periods, dominants (days 35–39) of the isolated period (see Supplementary Material), were more aggressive (i.e., showed a higher number of aggressive Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 8 Current Zoology, 2018, Vol. 0, No. 0 Figure 4. ERGM parameter estimates (with standard error bars) for aggressive social networks due to dominant homophily (A) and size differences among group members (B) for the pre-exposure (days 30–39), immediate post-exposure (days 40–49), and delayed post-exposure (days 70–79) period. Positive estimates of dominant homophily indicate that dominants are aggressive primarily to their mate, whereas negative estimates of dominant homophily indicate that dominants are aggressive primarily toward subordinates. Positive estimates of size differences indicate that aggressive interactions are exchanged primarily between group members with a large size difference, whereas negative estimates of size differences indicate that aggressive interactions are exchanged primarily between group members with a small size difference. acts) to neighbors than subordinates (Z ¼10.36, P< 0.001) contributed to network structure varied across periods. Specifically, and males were more aggressive to neighbors than females aggression and submission between the dominant male and female (Z ¼ 2.39, P ¼ 0.02). Aggression toward neighbors was not sig- were both more frequent when groups were isolated than when nificantly altered by neighbor group size (Z ¼0.08, P ¼ 0.93). groups had recently been exposed to neighbors. In contrast, aggres- We also found that aggression toward neighbors was impacted by sion and submission between similarly sized group members and group-level dynamics. Dominant aggression toward neighbors was dominant aggression toward subordinates were relatively more fre- positively correlated with dominant male aggression toward the dom- quent in the immediate post-exposure period compared with the inant female (Z ¼ 2.03, P ¼ 0.04), but negatively correlated with pre-exposure period. dominant female submission to the dominant male (Z ¼2.98, An increase in aggression between dominant and subordinate P ¼ 0.003). Further, dominants tended to be more aggressive to neigh- individuals when neighbors were present is consistent with the hy- bors when there were no subordinates present in their group (i.e., pothesis that within-group dynamics are altered by the outside when the social unit was composed of only a dominant male and fe- options offered by neighboring groups. These findings are also con- male pair) compared with when there were subordinates in their social sistent with the results of several previous studies in this species that unit (Z ¼1.87, P ¼ 0.06). We found no evidence that subordinate suggest that subordinates “pay” more (in terms of an increase in aggression to neighbors was significantly impacted by affiliation received aggression and more help provided) to remain on the terri- (Z ¼0.10, P ¼ 0.92) or aggression (Z ¼0.04, P ¼ 0.97) tory when there are neighbors present (Bruintjes and Taborsky 53 53 received from the dominants. 2008; Hellmann and Hamilton 2014; Hellmann et al. 2015b). However, these results are in contrast to the results of a theoretical model (Hellmann and Hamilton forthcoming) and empirical studies Discussion in cichlid fish (Bergmu ¨ ller et al. 2005b) and paper wasps (Tibbetts and Reeve 2008; Grinsted and Field 2017) which show that subor- Previous studies have found that the presence and density of neigh- dinate help decreases as outside options increase. Biological market boring groups are correlated with increased subordinate cooper- theory predicts that outside options should benefit the partner that ation, increased subordinate eviction, and altered reproductive “chooses” (Noe ¨ and Hammerstein 1995; Bshary and Grutter 2002; sharing in N. pulcher (Hellmann and Hamilton 2014; Hellmann Bshary and Noe ¨ 2003), suggesting that the presence of neighbors et al. 2015a, 2015b). However, less was known about how neigh- increased partner choice for dominants in this study. However, it is bors impact group-level behavioral dynamics. By manipulating the likely that the “choosy” partner may vary among species as well as presence of neighbors, we can evaluate how neighbor groups change within the same species depending on the relative leverage that an both the magnitude and target of conflict within a group. This can individual has in a given social situation (Lewis 2002). For example, lend insight into who benefits most from the opportunities offered subordinates may have more leverage to negotiate based on outside by neighboring groups, and can better elucidate if groups perceive options in situations in which dominants gain large fitness benefits neighbors as reproductive competitors, future group members, or from subordinate help, relative to situations in which subordinate threats to the territory as a whole. help is less needed (Taborsky 1985; Zo ¨ ttl et al. 2013). Further, out- We found that group-level social dynamics were characterized side options may be relatively unimportant in informing social dy- by strong negative effects of reciprocity and cyclical triads across all namics in systems without pay-to-stay cooperation (Hellmann and time periods, demonstrating that network ties are self-organizing Hamilton forthcoming) or in groups where dominants and subordi- (i.e., the existence of certain ties promotes other ties to come into ex- nates are related (Cant and Johnstone 2009; Quinones et al. 2016). istence) and that dominance hierarchies are likely stable across all Consequently, greater exploration into how neighboring groups in- time periods (McDonald and Shizuka 2012). Further, we found that fluence cooperation and the stability of current social relationships network structure was influenced by the status, sex, and size of the would be highly beneficial to understanding when and to what group members, although the ways in which individual attributes Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 Hellmann and Hamilton  Neighbors and within-group conflict 9 extent group members can negotiate to improve their current social aggression and submission networks between pre-exposure groups situations based on outside options. (days 30–39) and immediate post-exposure groups (days 40–49) An increase in conflict among similarly sized individuals (i.e., were greater than changes between days 40–49 and days 70–79 rank-related conflict) when neighbors were present is also consistent (delayed post-exposure groups), despite greater differences in time with the hypothesis that the presence of neighbors alters within- between the latter 2 time periods. Finally, the changes in our net- group conflict by introducing the potential for group composition to work structure are consistent with what we predicted given that pre- change. In species with size-based dominance hierarchies such as N. vious studies found that the presence of neighbors increases conflict pulcher, conflict tends to be greatest between similarly sized group between dominants and subordinates (Hellmann and Hamilton members (Wong et al. 2007; Hamilton and Heg 2008; Heg and 2014; Hellmann et al. 2015b). Nevertheless, further research com- Hamilton 2008; Ang and Manica 2010) because relative fighting paring network structure in groups that form in the presence of ability is less certain between group members close in size (Reddon neighbors versus those that form while isolated and are later et al. 2011). When neighbors are present, current group members exposed to neighbors would help distinguish which, if any, of the may be in conflict over the joining of a new subordinate, which reported patterns are due to time rather than the presence of could benefit high-ranking group members who would gain benefits neighbors. from the additional help and protection associated with a greater In conclusion, for our network analysis, we used ERGMs to number of subordinates, but would be costly to low-ranking subor- evaluate social dynamics within groups (Wasserman and Pattison dinates who would descend in the dominance hierarchy if a larger 1996). These models control for the dependency among social rela- subordinate joined the group (Heg et al. 2005; Ligocki et al. 2015a). tionships, allowing us to not only ask questions about how individ- Conversely, the potential for subordinates to leave the group may ual attributes (e.g., actor sex) affect social dynamics, but to expand also disrupt the dominance hierarchy, as subordinate removal from the scope of our analysis to assess how structural dependency among the group induces temporary aggression between group members of social ties (e.g., status homophily) and emergent group level charac- adjacent rank as group members re-establish their rank position teristics of the social network itself (e.g., reciprocity: Silk and Fisher (Wong and Balshine 2011b). Collectively, these results suggest that 2017) influence group-level social dynamics. Here, we demonstrate providing opportunities for group composition to change may re- that the target of within-group conflict shifts in the presence of duce the stability of dominance hierarchies. neighbors: isolated groups are characterized by higher conflict be- We predicted that if neighbors represent reproductive opportuni- tween the dominant breeding pair while groups with neighbors are ties, aggression between the dominant male and female pair would characterized by rank-related conflict and conflict between domi- be higher in groups with neighbors compared with isolated groups. nants and subordinates. This suggests that, rather than promote Instead, we found that the presence of neighbors was associated group cohesion or reproductive conflict, neighbors may foster con- with reduced conflict between the dominant male and female pair, flict by introducing opportunities for group composition to change. lending little support to the hypothesis that neighbors promote re- However, further studies independently manipulating the availabil- productive conflict between social mates. Similarly, we observed lit- ity of outside opportunities for dominants and subordinates would tle change in the frequency and target of affiliation before and after elucidate when and to what extent partner choice impacts the struc- exposure to neighbors, providing little support for the hypothesis ture and stability of dominance interactions in animal societies. that out-group threats increase affiliation among group members by Further, measuring changes in hormone levels (e.g., cortisol) and threatening the group as a whole. This is in contrast to Bruintjes gene expression in these manipulative experiments may help us et al. (2016) and Radford (2008a), who found that post-conflict af- understand the ways in which subsets of group members are filiation increased following experimental territorial intrusions by impacted by the presence of these outside options. neighboring and non-neighboring conspecifics. However, territorial intrusions are a more intense form of out-group conflict than the constant nearby presence of neighboring groups where territory Ethics Statement boundaries were never crossed. Consequently, the intergroup con- All methods were approved by The Ohio State University IACUC flict present in our study may have never presented a great enough (protocol ID 2008A0095). As stated above, efforts were made to threat to the focal group to promote higher within-group affiliation. minimize injuries resulting from high levels of aggression by moni- Consistent with this hypothesis, Polizzi di Sorrentino et al. (2012) toring groups daily and removing subordinate group members that found that visual exposure to neighbors in tufted capuchins was not received high levels of aggression. sufficient to produce changes in within-group affiliation. Further, in green woodhoopoes, changes in within-group affiliative behavior were only seen when groups faced more intense and longer out- Author Contributions group threats (Radford 2008b). Consequently, further work is J.K.H. and I.M.H. conceived the study. J.K.H. conducted the experi- needed to understand how within-group affiliation varies with the ment, scored all behavior, performed the statistical analyses, and type and severity of out-group threat. wrote the initial manuscript. Both J.K.H. and I.M.H. revised the Changes among experimental periods could be due to time or manuscript. the presence of neighbors, as observations during the pre-exposure period occurred sooner to group formation than observations during the post-exposure periods. Our results are largely inconsistent with what would be expected under the time hypothesis; our supplemen- Acknowledgments tary analysis demonstrates that the influence of individual attributes We owe immense thanks to Dr Cody Dey for aid with ERGM network ana- (sex, status, size) on network structure had stabilized by day 40. lysis. This research was supported by The Ohio State University. J.K.H. is cur- Changes in the effect of these parameters on networks were greater rently supported by the National Institute of General Medical Sciences of the between time periods than within time periods. Further, changes in National Institutes of Health under Award Number F32GM121033. Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 10 Current Zoology, 2018, Vol. 0, No. 0 Fischer S, Zo ¨ ttl M, Groenewoud F, Taborsky B, 2014. Group-size-dependent Supplementary Material punishment of idle subordinates in a cooperative breeder where helpers pay Supplementary material can be found at https://academic.oup.com/cz. to stay. Proc R Soc B 281:20140184. Fournier D, Skaug H, Ancheta J, Ianelli J, Magnusson A et al., 2012. AD model builder: using automatic differentiation for statistical inference of References highly parameterized complex nonlinear models. Optim Method Softw 27: Ang TZ, Manica A, 2010. Aggression, segregation and stability in a domin- 233–249. ance hierarchy. Proc R Soc B 277:1337–1343. Frostman P, Sherman PT, 2004. Behavioral response to familiar and unfamil- Archie EA, Tung J, Clark M, Altmann J, Alberts SC, 2014. Social affiliation iar neighbors in a territorial cichlid Neolamprologus pulcher. Ichthyol Res matters: both same-sex and opposite-sex relationships predict survival in 51:283–285. wild female baboons. Proc R Soc B 281:pii: 20141261. Gashagaza M, 1988. Feeding activity of a Tanganyikan cichlid fish Arnold KE, Owens IPF, 2002. Extra-pair paternity and egg dumping in birds: Lamprologus brichardi. Afr Study Mongr 9:1–9. life history, parental care and the risk of retaliation. Proc R Soc B 269: Godfrey SS, Sih A, Bull CM, 2013. The response of a sleepy lizard social net- 1263–1269. work to altered ecological conditions. Anim Behav 86:763–772. Aureli F, Cords M, van Schaik CP, 2002. Conflict resolution following aggres- Goetz AT, Shackelford TK, Romero GA, Kaighobadi F, Miner EJ, 2008. sion in gregarious animals: a predictive framework. Anim Behav 64: Punishment, proprietariness, and paternity: men’s violence against 325–343. womenfrom anevolutionary perspective. Aggress Violent Behav 13: Balshine-Earn S, Lotem A, 1998. Individual recognition in a cooperatively 481–489. breeding cichlid: evidence from video playback experiments. Behaviour Griffith SC, Owens IPF, Thuman KA, 2002. Extra pair paternity in birds: a re- 135:369–386. view of interspecific variation and adaptive function. Mol Ecol 11: Barocas A, Ilany A, Koren L, Kam M, Geffen E, 2011. Variance in centrality 2195–2212. within rock hyrax social networks predicts adult longevity. PLoS One 6: Grinsted L, Field J, 2017. Market forces influence helping behaviour in co- e22375. operatively breeding paper wasps. Nat Commun 8:13750. Bergmu ¨ ller R, Heg D, Peer K, Taborsky M, 2005a. Extended safe havens and Hamilton IM, Heg D, 2008. Sex differences in the effect of social status on the between-group dispersal of helpers in a cooperatively breeding cichlid. growth of subordinates in a co-operatively breeding cichlid. J Fish Biol 72: Behaviour 142:1643–1667. 1079–1088. Bergmu ¨ ller R, Heg D, Taborsky M, 2005b. Helpers in a cooperatively breed- Hamilton IM, Heg D, Bender N, 2005. Size differences within a dominance ing cichlid stay and pay or disperse and breed, depending on ecological con- hierarchy influence conflict and help in a cooperatively breeding cichlid. straints. Proc R Soc B 272:325–331. Behaviour 142:1591–1613. Bergmu ¨ ller R, Taborsky M, 2005. Experimental manipulation of helping in a Handcock MS, Hunter DR, Butts CT, Goodreau SM, Morris M, 2008. cooperative breeder: helpers ‘pay to stay’ by pre-emptive appeasement. Statnet: software tools for the representation, visualization, analysis and Anim Behav 69:19–28. simulation of network data. J Stat Softw 24:11. Bierbach D, Oster S, Jourdan J, Arias-Rodriguez L, Krause et al., 2014. Social Heg D, Bender N, Hamilton I, 2004. Strategic growth decisions in helper cichl- network analysis resolves temporal dynamics of male dominance relation- ids. Proc R Soc B 271(Suppl. 6):S505–S508. ships. Behav Ecol Sociobiol 68:935–945. Heg D, Brouwer L, Bachar Z, Taborsky M, 2005. Large group size yields Bruintjes R, Lynton-Jenkins J, Jones JW, Radford AN, 2016. Out-group threat group stability in the cooperatively breeding cichlid Neolamprologus promotes within-group affiliation in a cooperative fish. Am Nat 187: pulcher. Behaviour 142:1615–1641. 274–282. Heg D, Hamilton IM, 2008. Tug-of-war over reproduction in a cooperatively Bruintjes R, Taborsky M, 2008. Helpers in a cooperative breeder pay a high breeding cichlid. Behav Ecol Sociobiol 62:1249–1257. price to stay: effects of demand, helper size and sex. Anim Behav 75: Hellmann JK, Hamilton IM, 2014. The presence of neighbors influences de- 1843–1850. fense against predators in a cooperatively breeding cichlid. Behav Ecol 25: Bshary R, Grutter AS, 2002. Experimental evidence that partner choice is a 386–391. driving force in the payoff distribution among cooperators or mutualists: Hellmann JK, Hamilton IM, forthcoming. Dominant and subordinate outside the cleaner fish case. Ecol Lett 5:130–136. options alter help and eviction in a pay-to-stay negotiation model. Behav Bshary R, Noe ¨ R, 2003. Biological markets: the ubiquitous influence of part- Ecol, https://doi.org/10.1093/beheco/ary036. ner choice on the dynamics of cleaner fish—client reef fish interactions. In: Hellmann JK, Ligocki IY, O’Connor CM, Reddon AR, Garvy KA et al., Hammerstein P, editor. Genetic and Cultural Evolution of Cooperation. 2015a. Reproductive sharing in relation to group and colony-level attributes Cambridge: MIT Press. 167–184. in a cooperative breeding fish. Proc R Soc B 282:20150954. Cant M, Johnstone R, 2009. How threats influence the evolutionary resolution Hellmann JK, Reddon AR, Ligocki IY, O’Connor CM, Garvy KA et al., of within-group conflict. Am Nat 173:759–771. 2015b. Group response to social perturbation: impacts of isotocin and the Cantor M, Wedekin LL, Guimar~ aes PR, Daura-Jorge FG, Rossi-Santos MR social landscape. Anim Behav 105:55–62. et al., 2012. Disentangling social networks from spatiotemporal dynamics: Hick K, Reddon Adam R, O’connor Constance M, Balshine S, 2014. Strategic the temporal structure of a dolphin society. Anim Behav 84:641–651. and tactical fighting decisions in cichlid fishes with divergent social systems. Dey CJ, Quinn JS, 2014. Individual attributes and self-organizational proc- Behaviour 151:47–71. esses affect dominance network structure in pukeko. Behav Ecol 25: Hunter DR, Handcock MS, Butts CT, Goodreau SM, Morris M, 2008. ergm: 1402–1408. a package to fit, simulate and diagnose exponential-family models for net- Dey CJ, Reddon AR, O’Connor CM, Balshine S, 2013. Network structure is works. J Stat Softw 24:1–29. related to social conflict in a cooperatively breeding fish. Anim Behav 85: Huntingford FA, Turner AK, 1987. Animal Conflict. London: Chapman & 395–402. Hill. Dey CJ, Tan QYJ, O’Connor CM, Reddon AR, Caldwell JR et al., Jungwirth A, Walker J, Taborsky M, 2015. Prospecting precedes dispersal and 2015. Dominance network structure across reproductive contexts in the co- increases survival chances in cooperatively breeding cichlids. Anim Behav operatively breeding cichlid fish Neolamprologus pulcher. Curr Zool 61: 106:107–114. 45–54. Kohda M, Jordan LA, Hotta T, Kosaka N, Karino K et al., 2015. Facial recog- Dierkes P, Taborsky M, Kohler U, 1999. Reproductive parasitism of brood- nition in a group-living cichlid fish. PLoS One 10:e0142552. care helpers in a cooperatively breeding fish. Behav Ecol 10:510–515. Krivitsky P, 2013. ergm.count: fit, simulate and diagnose exponential-family Eggert A-K, Sakaluk SK, 1995. Female-coerced monogamy in burying beetles. models for networks with count edges. R package version 311. Available Behav Ecol Sociobiol 37:147–153. from http://wwwstatnetorg/. Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 Hellmann and Hamilton  Neighbors and within-group conflict 11 Kutsukake N, Clutton-Brock TH, 2008. The number of subordinates moder- Reddon AR, Voisin MR, Menon N, Marsh-Rollo SE, Wong MYL et al., 2011. ates intrasexual competition among males in cooperatively breeding meer- Rules of engagement for resource contests in a social fish. Anim Behav 82: kats. Proc R Soc B 275:209–216. 93–99. Kutsukake N, Clutton-Brock TH, 2010. Grooming and the value of social Robins G, Pattison P, Kalish Y, Lusher D, 2007. An introduction to exponen- relationships in cooperatively breeding meerkats. Anim Behav 79:271–279. tial random graph (p*) models for social networks. Soc Networks 29: Kuznetsova A, Brockhoff P, Christensen R, 2017. lmerTest package: tests in 173–191. linear mixed effects models. J Stat Softw 82:1–26. Sabol AC, Hellmann JK, Gray SM, Hamilton IM, 2017. The role of ultraviolet col- Lewis RJ, 2002. Beyond dominance: the importance of leverage. Quart Rev oration in intrasexual interactions in a colonial fish. Anim Behav 131:99–106. Biol 77:149–164. Seyfarth RM, Cheney DL, 1984. Grooming, alliances and reciprocal altruism Ligocki IY, Balshine S, Hamilton Ian M, Reddon AR, Hellmann JK et al., in vervet monkeys. Nature 308:541–543. 2015a. Social status influences responses to unfamiliar conspecifics in a co- Silk JB, Alberts SC, Altmann J, 2003. Social bonds of female baboons enhance operatively breeding fish. Behaviour 152:1821–1839. infant survival. Science 302:1231–1234. Ligocki IY, Earley RL, Hellmann JK, Hamilton IM, 2015b. Variation in gluco- Silk MJ, Fisher DN, 2017. Understanding animal social structure: exponential ran- corticoid levels in relation to direct and third-party interactions in a social dom graph models in animal behaviour research. Anim Behav 132:137–146. cichlid fish. Physiol Behav 151:386–394. Skaug H, Fournier D, Bolker B, Magnusson A, Nielsen A, 2016. Lusher D, Koskinen J, Robins G, 2013. Exponential Random Graph Models Generalized linear mixed models using “AD Model Builder.” R package for Social Networks: Theory, Methods, and Applications. Cambridge: version 0833. Cambridge University Press. Sopinka NM, Fitzpatrick JL, Desjardins JK, Stiver KA, Marsh-Rollo SE et al., Madden JR, Drewe JA, Pearce GP, Clutton-Brock TH, 2009. The social net- 2009. Liver size reveals social status in the African cichlid Neolamprologus work structure of a wild meerkat population: 2. Intragroup interactions. pulcher. J Fish Biol 75:1–16. Behav Ecol Sociobiol 64:81–95. Stiver KA, Desjardins JK, Fitzpatrick JL, Neff B, Quinn JS et al., 2007. McDonald DB, Shizuka D, 2012. Comparative transitive and temporal order- Evidence for size and sex-specific dispersal in a cooperatively breeding cich- liness in dominance networks. Behav Ecol 24:511–520. lid fish. Mol Ecol 16:2974–2984. Noe ¨ R, Hammerstein P, 1995. Biological markets. Trends Ecol Evol 10: Stiver KA, Dierkes P, Taborsky M, Balshine S, 2004. Dispersal patterns and 336–339. status change in a co-operatively breeding cichlid Neolamprologus pulcher: O’Connor CM, Reddon AR, Ligocki IY, Hellmann JK, Garvy KA et al., 2015. evidence from microsatellite analyses and behavioural observations. J Fish Motivation but not body size influences territorial contest dynamics in a Biol 65:91–105. wild cichlid fish. Anim Behav 107:19–29. Stiver KA, Fitzpatrick JL, Desjardins JK, Balshine S, 2009. Mixed parentage in Polizzi di Sorrentino E, Schino G, Massaro L, Visalberghi E, Aureli F, 2012. Neolamprologus pulcher groups. J Fish Biol 74:1129–1135. Between-group hostility affects within-group interactions in tufted capuchin Taborsky M, 1985. Breeder–helper conflict in a cichlid fish with broodcare monkeys. Anim Behav 83:445–451. helpers: an experimental analysis. Behaviour 95:45–75. Quinones AE, van Doorn GS, Pen I, Weissing FJ, Taborsky M, 2016. Tibbetts EA, Reeve HK, 2008. Two experimental tests of the relationship be- Negotiation and appeasement can be more effective drivers of sociality than tween group stability and aggressive conflict in Polistes wasps. kin selection. Phil Trans R Soc Lond B Biol Sci 371:20150089. Naturwissenschaften 95:383–389. Radford AN, 2008a. Duration and outcome of intergroup conflict influences Valera F, Hoi H, Kri stı´n A, 2003. Male shrikes punish unfaithful females. intragroup affiliative behaviour. Proc R Soc B 275:2787–2791. Behav Ecol 14:403–408. Radford AN, 2008b. Type of threat influences postconflict allopreening in a Wasserman S, Pattison P, 1996. Logit models and logistic regressions for social social bird. Curr Biol 18:R114–R115. networks: i. An introduction to Markov graphs andp. Psychometrika 61: Radford AN, 2011. Preparing for battle? Potential intergroup conflict pro- 401–425. motes current intragroup affiliation. Biol Lett 7:26–29. Whitehead H, 2008. Analyzing Animal Societies: Quantitative Methods for RadfordAN, FawcettTW, 2014. Conflict between groups promotes later de- Vertebrate Social Analysis. Chicago: Chicago University Press. fense of a critical resource in a cooperatively breeding bird. Curr Biol 24: Wong M, Balshine S, 2011a. The evolution of cooperative breeding in the African 2935–2939. cichlid fish Neolamprologus pulcher. Biol Rev Camb Philos Soc 86:511–530. Radford AN, Majolo B, Aureli F, 2016. Within-group behavioural consequen- Wong M, Balshine S, 2011b. Fight for your breeding right: hierarchy re-establishment ces of between-group conflict: a prospective review. Proc R Soc B 283: predicts aggression in a social queue. Biol Lett 7:190–193. 20161567. Wong MYL, Buston PM, Munday PL, Jones GP, 2007. The threat of punish- Reddon AR, O’Connor CM, Marsh-Rollo SE, Balshine S, Gozdowska M ment enforces peaceful cooperation and stabilizes queues in a coral-reef fish. et al., 2015. Brain nonapeptide levels are related to social status and affilia- Proc R Soc B Biol Sci 274:1093–1099. tive behaviour in a cooperatively breeding cichlid fish. R Soc Open Sci 2: Zo ¨ ttl M, Frommen JG, Taborsky M, 2013. Group size adjustment 140072. to ecological demand in a cooperative breeder. Proc R Soc B 280:20122772. Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Current Zoology Oxford University Press

Intragroup social dynamics vary with the presence of neighbors in a cooperatively breeding fish

Free
11 pages

Loading next page...
 
/lp/ou_press/intragroup-social-dynamics-vary-with-the-presence-of-neighbors-in-a-Nzc5aey1Mv
Publisher
Editorial Office
Copyright
© The Author(s) (2018). Published by Oxford University Press.
ISSN
1674-5507
eISSN
2396-9814
D.O.I.
10.1093/cz/zoy025
Publisher site
See Article on Publisher Site

Abstract

Conflict is an inherent part of social life in group-living species. Group members may mediate con- flict through submissive and affiliative behaviors, which can reduce aggression, stabilize domin- ance hierarchies, and foster group cohesion. The frequency and resolution of within-group conflict may vary with the presence of neighboring groups. Neighbors can threaten the territory or resour- ces of the whole group, promoting behaviors that foster within-group cohesion. However, neighbors may also foster conflict of interests among group members: opportunities for subordin- ate dispersal may alter conflict among dominants and subordinates while opportunities for extra- pair reproduction may increase conflict between mates. To understand how neighbors mediate within-group conflict in the cooperatively breeding fish Neolamprologus pulcher, we measured be- havioral dynamics and social network structure in isolated groups, groups recently exposed to neighbors, and groups with established neighbors. Aggression and submission between the dom- inant male and female pair were high in isolated groups, but dominant aggression was directly pri- marily at subordinates when groups had neighbors. This suggests that neighbors attenuate conflict between mates and foster conflict between dominants and subordinates. Further, aggression and submission between similarly sized group members were most frequent when groups had neigh- bors, suggesting that neighbors induce rank-related conflict. We found relatively little change in within-group affiliative networks across treatments, suggesting that the presence of neighbors does not alter behaviors associated with promoting group cohesion. Collectively, these results pro- vide some of the first empirical insights into the extent to which intragroup behavioral networks are mediated by intergroup interactions and the broader social context. Key words: affiliation, colony, conflict, exponential random graph model, Neolamprologus pulcher, network. Conflict, which often manifests as aggression between group mem- subordinate individuals make use of both submissive and affiliative bers, is an inherent part of social life in group-living species and can displays to mitigate aggression from dominant group members and reduce the benefits of group living by increasing social stress, reduc- increase their likelihood of being tolerated within the group ing group productivity and leading to group dissolution if left unre- (Huntingford and Turner 1987; Bergmu ¨ ller and Taborsky 2005; solved (Aureli et al. 2002). In many group-living species, Bruintjes and Taborsky 2008). Submissive behavior can facilitate V C The Author(s) (2018). Published by Oxford University Press. 1 This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 2 Current Zoology, 2018, Vol. 0, No. 0 group stability by enforcing dominance hierarchies and settling con- understand how the presence of neighboring groups was associated flict among individuals within a social group (Huntingford and with changes in within-group social dynamics. To do this, we ex- Turner 1987; Hick et al. 2014), whereas affiliative behavior (e.g., perimentally created laboratory groups of Neolamprologus pulcher, grooming in primates) can reinforce friendships, encourage recon- a cooperatively breeding cichlid fish native to Lake Tanganyika in ciliation, and promote intragroup cooperation (Radford et al. East Africa. These fish form colonies of 2–200 permanent territorial 2016). groups (Stiver et al. 2007). Each group is composed of a dominant The frequency of aggressive, submissive, and affiliative behaviors male and female pair, with 0–20 subordinates that provide help in exchanged among group members often varies across groups the form of territory defense, territory maintenance, and alloparen- (Kutsukake and Clutton-Brock 2008; Madden et al. 2009; tal care (Wong and Balshine 2011a). Dominance is strictly size- Kutsukake and Clutton-Brock 2010), as well as across time within based, such that the dominant male and female are the largest indi- the same social group (Cantor et al. 2012; Godfrey et al. 2013; viduals and subordinates form sex-specific, size-based dominance Bierbach et al. 2014). Given that social stability and the nature of hierarches (Wong and Balshine 2011a). Male group members main- social interactions within a group influence individual fitness (Silk tain consistent differences in size, which likely reduces conflict that et al. 2003; Barocas et al. 2011; Archie et al. 2014), there have been would otherwise arise among similarly sized individuals (Heg et al. efforts to first quantify social structure within groups and then to 2004; Hamilton et al. 2005; Hamilton and Heg 2008). Submission understand factors that modulate social dynamics within and across is an effective behavior that reduces aggression in N. pulcher groups. Mounting evidence demonstrates that social interactions (Bergmu ¨ ller and Taborsky 2005; Bruintjes and Taborsky 2008), within a group are correlated with group-level attributes, such as while affiliative behavior is used to reinforce participation in terri- group size (Fischer et al. 2014), the relative size of group members tory defense (Bruintjes et al. 2016) and is associated with reduced (Hamilton et al. 2005), and the sex of group members (Kutsukake cortisol levels (Ligocki et al. 2015b). Neighboring groups are not and Clutton-Brock 2008). However, much less is known about how direct competitors for food: individuals in this species feed in the the social environment beyond the level of the group, specifically the water column and territories are used as protection from predators presence of other conspecific groups, influences social dynamics and breeding substrate (Gashagaza 1988; Wong and Balshine among group members. 2011a). Further, whole-group takeovers of neighboring territories There are at least 3 ways in which neighboring groups can alter are relatively rare in this species; instead, outside threats traditional- within-group dynamics by either reducing or promoting conflict ly come from single individuals seeking out reproductive opportuni- among a given subset of group members. First, neighboring groups ties, as male and female dominants lose reproduction to can threaten the territory or resources of an established group and subordinates within their group as well as to individuals in neigh- may incentivize group members to quickly resolve or reduce conflict boring groups (Dierkes et al. 1999; Stiver et al. 2009; Hellmann within their own group in order to facilitate cooperation in et al. 2015a). Further, outside individuals may threaten the stability between-group conflict (Radford 2008a; Radford and Fawcett of group composition. Subordinates disperse between groups 2014; Bruintjes et al. 2016). Studies examining territorial intrusions (Bergmu ¨ ller et al. 2005a; Jungwirth et al. 2015) and while domi- and conflict have found that affiliation between dominant and sub- nants generally do not move between territories, dominant males ordinate group members is greater when intergroup conflict is are polygynous and are at least capable of usurping a neighboring higher (Radford 2008a; Bruintjes et al. 2016). In these cases, affilia- male’s territory (O’Connor et al. 2015). Consequently, these outside tive behavior could be exchanged for continued subordinate partici- individuals may only threaten a subset of group members, while pation in out-group conflicts (Seyfarth and Cheney 1984; Radford having neutral or even beneficial effects on other group members. 2011). Second, neighboring groups can offer opportunities for egg After isolated groups stabilized following group formation, we dumping (Arnold and Owens 2002) or extra-pair fertilizations exposed groups to neighbors and observed behavioral interactions (Griffith et al. 2002; Hellmann et al. 2015a). This may result in within and between groups when groups were isolated, immediately increased aggression between mates, as the presence of neighboring after groups were introduced to neighbors, and 30 days after the males and females may provide fitness benefits to one mate by introduction of neighbors. Because changes in 1 individual’s behav- increasing the number of offspring sired, but may be costly to the ior are not independent of those of other group members, we used other mate who may care for offspring that are not their own. exponential random graph models (ERGMs), in addition to trad- Finally, neighboring groups offer opportunities for partner choice by itional regression models, to evaluate changes in social dynamics allowing subordinates to potentially move between groups. If neigh- (Wasserman and Pattison 1996). ERGMs control for the inter- bors provide an opportunity for subordinates to negotiate based on dependency of behavioral relationships within a group and evaluate outside options offered by neighboring groups, dominants should be the extent to which social dynamics among group members are more tolerant (less aggressive and demand less help) of current sub- influenced by individual attributes (e.g., sex and size) and structural ordinates (Bergmu ¨ ller et al. 2005b; Grinsted and Field 2017; dependency in social ties (e.g., reciprocity). By using these models, Hellmann and Hamilton forthcoming). Conversely, if neighbors we can test hypotheses related to understanding how social structure provide dominants with an opportunity to replace current subordi- varies across different contexts and how emergent properties of the nates, neighbors may allow dominants to demand more help and be social system may contribute to these changes (Wasserman and more aggressive to subordinates when neighbors are present Pattison 1996; Robins et al. 2007; Lusher et al. 2013; Silk and (Bruintjes and Taborsky 2008; Hellmann and Hamilton 2014; Fisher 2017). Hellmann et al. 2015b). Consequently, the way in which partner We had 3 separate predictions regarding the impact of neighbor- choice impacts behavioral dynamics between dominants and subor- ing groups on within-group social dynamics. If within-group dynam- dinates should depend on who is “choosing” partners in this system ics change in response to the threat that neighboring groups pose to (Noe ¨ and Hammerstein 1995; Cant and Johnstone 2009). the focal group, then we predicted that affiliative behavior would be We compared within-group interactions before neighboring higher when groups had neighbors compared with when groups groups were visible and while neighboring groups were visible to were isolated, as affiliative behavior can be used to promote Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 Hellmann and Hamilton  Neighbors and within-group conflict 3 cooperation among current group members when the potential for barrier. Neolamprologus pulcher can distinguish familiarity, iden- intergroup conflict is high (Radford 2008a, 2011). If within-group tity, and sex using only visual cues (Balshine-Earn and Lotem 1998; dynamics change in response to the reproductive opportunities Frostman and Sherman 2004; Hellmann and Hamilton 2014; offered by neighboring groups, then we predicted that aggression be- Kohda et al. 2015; Sabol et al. 2017), suggesting that chemical cues tween the dominant male and female pair would be higher when are not necessary for individuals to accurately identify the sex and neighbors were present compared with when they were absent, as re- familiarity of neighboring group members. productive conflict between mates is expected to be higher when All eggs that were laid during the experiment were removed im- there are additional opportunities for extra-pair fertilizations mediately from the groups to remove any confounding influence of (Eggert and Sakaluk 1995; Valera et al. 2003; Goetz et al. 2008). If parental care on intragroup and intergroup social interactions, al- within-group dynamics change in response to opportunities for sub- though a previous study in this species found that the structure of ordinates to move between groups, then we predicted that aggres- interaction networks does not strongly vary with reproductive sion and submission between dominants and subordinates would be events (Dey et al. 2015). Nevertheless, any behavioral observations altered when neighbors were present compared with when neigh- conducted within 24 h of egg-laying and removal were removed bors were absent. from the dataset. Group formation and behavioral observations Materials and Methods From October 2014 to November 2015, we formed 31 social groups of unrelated N. pulcher, each composed of a dominant male and fe- Study organism and housing conditions male breeding pair and 2–3 subordinates with at least 1 subordinate All experimental fish were wild caught or F1 offspring of wild- male and 1 subordinate female per group (n¼27 groups with 3 sub- caught fish from the Kambwimba region of Lake Tanganyika (8 32 ordinates, n¼4 groups with 2 subordinates). All individuals were S, 31 9 E). All wild-caught fish had been in captivity in our lab for likely to be reproductively mature (SL> 35 mm: Taborsky 1985) at least 3 months prior to the beginning of the experiment. We used and sexed by examination of external genital papillae. In these a total of n ¼ 106 individual fish for this experiment. Prior to the ex- groups, the dominant male was the largest (and therefore, most periment, all fish were marked uniquely with elastomer dye and dominant) fish in the group and the dominant female was the second given a dorsal fin clip to indicate sex. Fish recover from this proced- largest fish in the group. All subordinates were at least 5 mm shorter ure rapidly and receiving these markers has no apparent effect on in SL than the dominant male and female. We assembled up to 8 subsequent behavior (Stiver et al. 2004; Dey et al. 2015). On the day groups at a time; we formed new groups by reshuffling former group that groups were formed, all group members were weighed to the members or using new individuals. Some individuals were therefore nearest 0.001 g (Ohaus Adventurer Pro AV213C) and measured for members of 2 social groups throughout the course of the 13-month standard length (SL) to the nearest 0.01 mm (Fisher Scientific experiment. We controlled for pseudoreplication in 2 ways. First, in Traceable calipers). To mirror natural conditions, a 12:12 h our regression models, we used random effects of individual and light:dark schedule was maintained for the duration of the experi- group identity to control for variation in behavior due to individual ment and water temperatures were kept constant at 2761 C. Fish identity. Although we used n¼106 total fish in this experiment, we were fed daily and ad libitum with either TetraMin flakes (5 days a had a total of n¼76 unique fish that were still present in the groups week) or frozen Daphnia or Artemia (twice per week). by day 30 (see below). Second, we ran our network analysis (see below for “Materials and Methods”) removing the minimum num- Experimental setup ber of groups necessary to produce a network in which each individ- During the experiment, 2 groups were placed on opposite sides of ual was a member of only 1 social group. a barrier in a 208-L (122 cm long  32 cm wide  53 cm high) All subordinates were unrelated to the dominants within their aquaria lined with 3 cm of black sand substrate. Each group had 2 group, and all group members had never interacted prior to group inverted terracotta flowerpot halves that served as shelter and po- formation. Groups were given 30 days after group formation to sta- tential breeding substrate, as well as 2 PVC tubes near the top of bilize, as aggression can be high while dominance hierarchies are the tank that served as hiding spots for subordinate fish. The 2 being established. In the initial days after group formation, some groups in each aquarium were separated by 2 clear plexiglass bar- subordinates received high levels of aggression from the dominants riers that were flush against the walls and floor of the aquarium and were subsequently removed from the group to prevent further and extended above the top of the water line. This largely pre- injury. These subordinates were not replaced; instead, we controlled vented water flow between the 2 groups, although some water for changes in group composition throughout the experiment (due flow between the groups was possible and therefore, chemical to death or removal of group members) by only analyzing changes communication between the groups may have occurred. in network structure using groups that did not change in compos- Consequently, differences in interactions when neighbors were ition across time periods. blocked from view versus visible may not be due so much to the At 30 days post-group formation, we had 11 social units that complete absence of neighbors or different perceived densities of were composed only of a male and female pair (i.e., social mates) conspecifics, but rather due to the absence of interactions with and 20 social units that were composed of a dominant male and fe- neighboring groups and differences in how much groups know male pair with subordinates (n¼5 groups with 1 subordinate, n¼9 about their neighbors (e.g., group composition, size of group groups with 2 subordinates, and n¼6 groups with 3 subordinates). members). At 30-days post-formation, each pair or group was observed for During the first 40 days after group formation, there was an opa- 30 min daily for 10 days (days 30–39). After 10 days, the opaque que barrier between the plexiglass barriers that prevented groups barrier isolating the groups was removed, such that groups were from seeing or interacting with each other. After 40 days, the opaque now visible to their neighbor on the other side of the tank. Groups barrier was removed and groups were able to interact across a clear were observed for 30 min daily for the first 10 days after removal of Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 4 Current Zoology, 2018, Vol. 0, No. 0 the opaque barrier to understand the impact of novel neighbors on was the difference in the number of interactions between 2 time peri- intragroup dynamics (days 40–49). Thirty days after the removal of ods. Because all edges must be positive values, we added the abso- the opaque barrier, we observed groups again for 30 min daily for lute value of the minimum edge weight to all within-group edge 10 days (days 70–79). weights to make all values positive. Therefore, higher values denote All behavioral observations were recorded and videos were that behaviors were more common in the later period than the ear- scored by the same observer (J.K.H.) using species-specific etho- lier period. grams (Ligocki et al. 2015a; Reddon et al. 2015; Sopinka et al. We analyzed network structure using ERGMs (Wasserman and 2009). Behaviors were categorized as overt aggressive attacks (ram, Pattison 1996). As in traditional regression models, these models bite, mouth fight), restrained aggressive displays (fin raise, fast ap- test how independent variables predict the weight of the edges. proach, operculum spread, head jerk, head down display), submis- Because ERGMs control for the interdependency of social ties when sive displays (tail quivers, hook, submissive posture), and affiliative evaluating behavioral data, they improve upon traditional statistical behaviors (parallel swim, bump, join). Because aggression toward methods (Whitehead 2008). ERGMs allow for simultaneous estima- neighboring groups was across a barrier, we scored any behaviors as tions of substructures (e.g., transitivity of relationships) and individ- overt aggression when the actor made contact with the barrier (e.g., ual attributes (e.g., size) or pairwise attributes (e.g., sex homophily) rammed the barrier in an attempt to ram a neighboring fish). that contribute to network data. Similar to Dey et al. (2015),we Within-group overt aggression was relatively rare, so we combined assembled a supernetwork composed of all social groups and counts of overt and restrained aggression when analyzing within- restricted all possible edges to those occurring among group mem- group dynamics. bers (Figure 1). To examine factors that contribute significantly to determining network structure, we tested the independent variables of: (1) the “sum” term, which is similar to the intercept term in a re- Statistical and network analysis gression model, (2) status homophily, which tests whether interac- tions are more likely to occur between individuals of the same status To test how affiliative, submissive, and aggressive interactions be- (dominant or subordinate), (3) sexual homophily, which tests if tween the dominant male and female pair (i.e., between social there is an increased chance of interactions between same-sex dyads, mates) changed with the presence of neighbors, we used generalized (4) actor effect of sex, which tests if 1 sex is more likely to initiate linear mixed models [GLMMs; R packages lmerTest (Kuznetsova behavioral interactions, and (5) dyadic differences in size (SL), et al. 2017) and glmmADMB (Fournier et al. 2012; Skaug et al. which tests if interactions are more likely to occur between individu- 2016)] with a negative binomial distribution because count data als close in size. Further, we tested for structural dependence be- were overdispersed. Dependent variables for each model were the tween edges by evaluating the tendency for (6) reciprocity, which sum of all affiliative, submissive, or aggressive behaviors, respective- tests if the weight of an edge from 1 group member to another pre- ly, observed across the 10 observations. In each model, we included dicts the weight of the reciprocal edge, and (7) cyclical triads, or the fixed effects of period (pre-exposure to neighbors, days 30–39; im- tendency of individuals to form cyclical triads, which are markers of mediate post-exposure to neighbors, days 40–49; delayed post- unstable dominance hierarchies. These variables were chosen prior exposure to neighbors, days 70–79) and the presence of subordinate to examining the data based on our predictions outlined in the males or females in the group (binomial) to control for variation in behavioral interactions between dominant males and females due to group composition. For models of aggression and submission, we also included a fixed effect of the amount of aggression received from their social mate, to account for variation in dominant male or female behavior due to their social mate’s activity. For models examining affiliative behavior, we included the fixed effect of the amount of affiliation received from their social mate. We used Tukey’s honest significant difference tests (R package multcomp) as a post hoc analysis to control for multiple comparisons and analyze where the differences lie among the treatment groups. We also included random factors of individual ID (nested within group iden- tity) in all models, because many individuals were a member of 2 so- cial groups. Dominant death occurred in 2 groups; therefore, n ¼ 29 social pairs/groups were used to evaluate social dynamics between the dominant male and female pair. We analyzed social network structure in the groups with a dom- inant pair and subordinates [R version packages: statnet (Handcock et al. 2008), ergm (Hunter et al. 2008), and ergm.count (Krivitsky 2013)]. We built weighted, directed networks of aggression, submis- sion, and affiliation for the isolated time period. The weight of the edges in these networks was determined by the total number of interactions directed from 1 individual to another across the 10 observations in the time period. Then, to quantify how within-group social dynamics changed between pre-exposure (days 30–39) and Figure 1. Aggressive supernetwork structure for N. pulcher groups prior to ex- immediate post-exposure (days 40–49) periods and between imme- posure to neighbors (days 30–39). Larger nodes represent dominant individu- diate and delayed post-exposure (days 70–79) periods, we assembled als, red nodes represent females, and blue nodes represent males. Thicker edges indicate that more aggression was exchanged between a given dyad. difference networks. In the difference network, each edge weight Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 Hellmann and Hamilton  Neighbors and within-group conflict 5 introduction and based on characteristics that define sex-specific, 40–44). We also evaluated changes within the delayed post- size-based dominance hierarchies (Wong and Balshine 2011a; exposure period (days 70–74 vs. days 75–79). If within-group dy- McDonald and Shizuka 2012; Dey et al. 2013, 2015; Dey and namics merely change according to time, then we expected to see Quinn 2014). changes in social dynamics changing similarly within and across ERGMs use a Markov-chain Monte Carlo estimation technique time periods. These methods and results are included in the to approximate the maximum likelihood. We specified a sampling Supplementary Material. interval of 5000 and a burn-in of 50,000 proposals, and used Finally, we used GLMMs with a negative binomial distribution Poisson reference graphs for each model (Dey et al. 2015). Models to test for differences in aggression toward the neighbor group due were checked for degeneracy and goodness of fit using the to status (dominant or subordinate), sex (male or female), treatment mcmc.diagnostics function. Examinations of model diagnostics did period (days 40–49 or days 70–79), and neighbor group size. To not indicate a high correlation between status homophily and size understand how group dynamics impacted aggression to neighbor- difference so we included both terms in our ERGMs; further, the ing groups, we ran 2 GLMMs. The first had dominant aggression to effects of status homophily and size difference did not change when neighbors as the dependent variable, with fixed effects of dominant the only 1 or both variables were included in the models (Hellmann male aggression to dominant females, dominant female submission JK, Hamilton IM, unpublished data). To eliminate potential issues to the dominant male, and the presence of subordinates in the group with pseudoreplication in the ERGM models, which do not allow (binary). The second had subordinate aggression to neighbors as the for random effects of identity, we only used a subset of observed dependent variable, with fixed effects of aggression and affiliation groups such that each individual was only represented in the net- received from the dominants. For all models, we checked independ- work once. Therefore, we analyzed a total of 17 groups in the net- ent variables for collinearity. Individual ID nested within group was works examining social dynamics in isolated groups (n ¼ 5 groups included as a random effect for all models. of 3 individuals, n ¼ 6 groups of 4 individuals, n ¼ 6 groups of 5 individuals). Because difference networks contained only groups Results that did not change in composition between time periods, we ana- Are interactions between social pairs influenced by the lyzed 13 groups in the difference network comparing pre-exposure groups and immediate post-exposure groups with neighbors (n ¼ 5 presence of neighbors? groups of 3 individuals, n ¼ 2 groups of 4 individuals, n ¼ 6 groups Dominant males were less aggressive (i.e., had a lower frequency of of 5 individuals) and 11 groups in the difference network comparing aggressive acts) to dominant females when neighbors were recently immediate and delayed post-exposure groups (n ¼ 6 groups of 3 present compared with the pre-exposure period when groups were individuals, n ¼ 2 groups of 4 individuals, n ¼ 3 groups of 5 isolated (GLMM with Tukey’s HSD: Z ¼2.36, 0.048; Figure 2A); individuals). this trend weakened when neighbors were present for 30 days, such Because potential changes in group dynamics may be due to that there was no significant difference in dominant male aggression changes in time rather than changes in the visibility of neighbors, we between the pre-exposure period and the delayed post-exposure analyzed if network structure changed within the first 10 days of ob- period (Z ¼1.87, P ¼ 0.14; Figure 2A). Dominant females were servation (days 30–34 vs. days 35–39) and if the magnitude of that similarly aggressive to dominant males in the pre-exposure period change was less than the change in network structure between when compared with the post-exposure periods (immediate: Z ¼0.15, groups were isolated (days 35–39) and neighbors were visible (days P ¼ 0.99; delayed: Z ¼0.71, P ¼ 0.75), but were significantly Figure 2. (A) Dominant male aggression toward the dominant female was higher prior to exposure to neighbors (white: days 30–39) compared with the immedi- ate post-exposure period when neighbors were recently visible (light blue: days 40–49), but this trend did not persist into the delayed post-exposure period when neighbors had been present for 30 days (darker blue: days 70–79). (B) Dominant female submission toward the dominant male was higher prior to exposure to neighbors (white: days 30–39) compared with both periods when neighbors were present. Data presented are the residuals of the regression model without the fixed effect of treatment period, plotted against treatment period. Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 6 Current Zoology, 2018, Vol. 0, No. 0 neighbors had been present for 30 days (non-significant effect in im- mediate vs. delayed post-exposure difference network; Table 2). During the pre-exposure period when groups were isolated, there were similar frequencies of aggression between same-sex and oppos- ite sex group members, while submission was more frequently exchanged between opposite-sex group members relative to same- sex group members (sexual homophily: Table 1). Both aggression and submission became significantly more sexually homophilic when groups were initially exposed to neighbors compared with the pre-exposure period (positive effect of sexual homophily in the pre- vs. post-exposure difference network: Table 2); this effect persisted into the delayed post-exposure period for submission (non-signifi- cant effect of immediate vs. delayed post-exposure difference net- work; Table 2), but not for aggressive interactions (negative effect in difference network; Table 2). In the pre-exposure period, submission and aggression were most commonly exchanged between group members with a large size difference (positive effect of SL difference: Table 1 and Figure 4). However, groups in the immediate post-exposure Figure 3. Dominant females were more aggressive to dominant males in their period had significantly more aggression and submission group when there were subordinate females present in their group. Data pre- exchanged between similarly sized individuals compared with iso- sented are the residuals of the regression model without the fixed effect of lated groups (negative effect of size difference: Table 2 and subordinate female presence, plotted against the binomial variable of subor- Figure 4). When groups were isolated, dominant aggression and dinate female presence in the group. submission were more frequently exchanged between the domin- ant pair rather than directed toward subordinates (dominant sta- more submissive to dominant males during the pre-exposure period tus homophily: Table 1 and Figure 4); however, in the immediate compared with the post-exposure periods, even after controlling for post-exposure period, dominant aggression and submission were variation in the amount of aggression received from the dominant more frequently exchanged with subordinates than with their so- male (immediate: Z ¼7.19, P< 0.001; delayed: Z ¼6.65, 74 74 cial mate (Table 2 and Figure 4). All of these effects persisted into P< 0.001; Figure 2B). Dominant male aggression toward the dom- the delayed post-exposure period for aggressive interactions; inant female and dominant female submission were not significantly however, these changes in the submission network were stronger altered by the presence of subordinate males (DM: Z ¼0.43, in the immediate post-exposure period compared with the delayed P ¼ 0.67; DF: Z ¼1.42, P ¼ 0.16) or subordinate females (DM: post-exposure period (positive effects for both parameters; Z ¼0.93, P ¼ 0.35; DF: Z ¼0.12, P ¼ 0.90) in the group. 74 74 Table 2). However, dominant females were significantly more aggressive to Across all time periods, males and females initiated similar levels of dominant males when subordinate females were present in the group affiliative behaviors (no effect of actor sex in any network: Tables 1 (Z ¼ 2.81, P ¼ 0.005; Figure 3), although there was no significant and 2). Affiliation was exchanged primarily between males and impact of subordinate male presence (Z ¼ 0.25, P ¼ 0.80). femalesratherthan between individualsof the same sex in both the Dominant male or female affiliative behavior toward their mate was pre-exposure period and immediate post-exposure period (negative ef- not significantly different in the pre-exposure period compared with the fect of sexual homophily: Table 1; no effect of either parameter in the immediate post-exposure (DF: Z ¼0.63, P¼ 0.80; DM: Z ¼1.94, 74 74 difference network: Table 2); however, affiliation was significantly P¼ 0.13) period or the delayed post-exposure period (DF: Z ¼ 0.01, more sexually homophilic when groups had neighbors for 30 days P¼ 1.00; DM: Z ¼1.40, P¼ 0.34). It also did not vary with the pres- compared with recent neighbors (positive effect of sexual homophily ence of subordinate males (DF: Z ¼1.63, P¼ 0.10; DM: Z ¼ 0.40, 74 74 in thedifferencenetwork: Table 2). In the pre-exposure period when P¼ 0.69) or subordinate females (DF: Z ¼0.70, P¼ 0.48; DM: groups were isolated, affiliative behavior was more frequently Z ¼0.25, P¼ 0.80) in the group. exchanged between individuals with a large size difference relative to similarly sized individuals (SL, negative effect of size homophily; Are social interactions within a group influenced by the Table 1), but affiliative behavior between similarly sized individuals presence of neighbors? was more common in the immediate post-exposure period than the Because some individuals were used across more than 1 group and pre-exposure period (negative effect of size differences: Table 2). This ERGMs do not allow for random effects of individual identity, we effect continued to strengthen in the delayed post-exposure period analyzed network structure in a subset of groups to avoid potential (Table 2). Across both the pre-exposure and immediate post-exposure issues with pseudoreplication. Across all time periods, males and time period, there was a higher frequency of affiliation exchanged be- females initiated similar levels of aggression (non-significant effect of tween individuals of the same rank than between individuals of differ- actor sex in all networks: Tables 1 and 2). Females were more submis- ent ranks (i.e., dominant–dominant affiliation was more common than sive to other group members than males (negative effect of actor sex: dominant–subordinate affiliation; positive effect of status homophily, Table 1), although a significant effect of actor sex in the difference Table 1; no effect of status homophily in difference network, Table 2). network indicates that this trend was stronger prior to exposure to However, subordinates exchanged more affiliation with dominants neighbors compared with the immediate post-exposure period when neighbors were present for 30 days compared with when neigh- (Table 2). This difference between pre-exposure and post-exposure bors were recent (negative effect of subordinate homophily in the dif- groups persisted into the delayed post-exposure period where ference network: Table 2). Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 Hellmann and Hamilton  Neighbors and within-group conflict 7 Table 1. Results of ERGM fit for behavioral networks for groups prior to exposure to neighbors (days 30–39) Pre-exposure period Aggression Submission Affiliation Estimate SE P Estimate SE P Estimate SE P Sum 2.15 0.06 <0.001 0.20 0.19 0.30 2.12 0.06 <0.001 Difference in SL 0.06 0.003 <0.001 0.06 0.007 <0.001 0.05 0.003 <0.001 Sexual homophily 0.02 0.03 0.52 0.30 0.09 0.001 0.34 0.05 <0.001 Actor sex (male) 0.03 0.03 0.35 0.14 0.05 0.005 0.007 0.02 0.73 Status homophily Dominant–dominant 0.16 0.05 0.001 1.33 0.10 <0.001 2.36 0.05 <0.001 Subordinate–subordinate 1.09 0.07 <0.001 1.04 0.18 <0.001 0.42 0.06 <0.001 Cyclical triads 0.73 0.05 <0.001 0.94 0.32 <0.001 0.93 0.04 <0.001 Reciprocity 1.07 0.04 <0.001 1.59 0.15 <0.001 0.37 0.02 <0.001 Bold values indicate significance (P<0.05). Notes: we tested the effects of dyadic differences in SL (positive values signify that individuals with bigger size differences interact more frequently), sexual and status homophily (positive values signify that individuals of the same sex/status interact most frequently), actor effects of sex (positive values signify that males ini- tiate behaviors more frequently than females), cyclical triads, and reciprocity. Table 2. Results of ERGM fit for difference networks comparing network dynamics of pre-exposure groups that were isolated (days 30–39) and post-exposure groups recently exposed to neighbors (days 40–49), as well as comparing post-exposure groups recently exposed to neighbors (days 40–49) to those with neighbors that had been present for 30 days (days 70–79) Pre- vs. immediate post-exposure period Aggression Submission Affiliation Estimate SE P Estimate SE P Estimate SE P Sum 5.88 0.01 <0.001 4.39 0.04 <0.001 5.38 0.02 <0.001 Difference in SL 0.007 0.001 <0.001 0.004 0.002 0.04 0.003 0.001 0.02 Sexual homophily 0.02 0.01 0.01 0.09 0.02 <0.001 0.006 0.009 0.52 Actor sex (male) 0.003 0.007 0.63 0.13 0.03 <0.001 0.008 0.009 0.37 Status homophily Dominant–dominant 0.05 0.02 0.002 0.11 0.04 0.001 0.02 0.02 0.17 Subordinate–subordinate 0.02 0.01 0.14 0.01 0.03 0.72 0.03 0.02 0.07 Cyclical triads 0.004 0.006 0.51 0.07 0.03 <0.001 0.04 0.006 <0.001 Reciprocity 0.03 0.006 <0.001 0.10 0.02 <0.001 0.001 0.008 0.87 Immediate vs. delayed post-exposure period Aggression Submission Affiliation Estimate SE P Estimate SE P Estimate SE P Sum 5.23 0.02 <0.001 0.02 0.005 <0.001 5.29 0.02 <0.001 Difference in SL 0.001 0.002 0.66 0.02 0.005 <0.001 0.005 0.002 0.002 Sexual homophily 0.03 0.01 0.03 0.05 0.05 0.33 0.06 0.01 <0.001 Actor sex (male) 0.006 0.009 0.53 0.01 0.04 0.78 0.005 0.007 0.45 Status homophily Dominant–dominant 0.01 0.02 0.59 0.17 0.06 0.007 0.02 0.02 0.36 Subordinate–subordinate 0.02 0.02 0.32 0.09 0.06 0.13 0.05 0.02 0.02 Cyclical triads 0.03 0.008 0.002 0.04 0.03 0.21 0.04 0.008 <0.001 Reciprocity 0.04 0.01 <0.001 0.02 0.02 0.47 0.02 0.01 0.01 Bold values indicate significance (P<0.05). Notes: Negative estimates indicate that a given variable had a stronger influence in the earlier time period. Local network substructures we are unsure if changes in cyclical triads and reciprocity across Aggressive, submissive, and affiliative networks were characterized time periods (isolated: days 30–39, new neighbors: days 40–49) are by a strong, negative effect of reciprocity, which means that individ- due to time or experimental treatment. uals who received high levels of aggression, submission, and affili- ation were unlikely to reciprocate those behaviors. Similarly, networks were characterized by strong negative effects of cyclical Between-group aggression and neighbor triads (Table 1), which means that cyclical triads were less frequent characteristics on within-group interactions than expected by chance. Negative effects of both reciprocity and In general, there was a higher frequency of aggression toward neigh- cyclical triads indicate that dominance hierarchies are stable. As the bors during the initial post-exposure period (days 40–49) compared influence of cyclical triads and reciprocity on network structure was with the delayed post-exposure period (days 70–79; GLMM: significantly different between the beginning (days 30–34) and end Z ¼3.03, P ¼ 0.003). Across both time periods, dominants (days 35–39) of the isolated period (see Supplementary Material), were more aggressive (i.e., showed a higher number of aggressive Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 8 Current Zoology, 2018, Vol. 0, No. 0 Figure 4. ERGM parameter estimates (with standard error bars) for aggressive social networks due to dominant homophily (A) and size differences among group members (B) for the pre-exposure (days 30–39), immediate post-exposure (days 40–49), and delayed post-exposure (days 70–79) period. Positive estimates of dominant homophily indicate that dominants are aggressive primarily to their mate, whereas negative estimates of dominant homophily indicate that dominants are aggressive primarily toward subordinates. Positive estimates of size differences indicate that aggressive interactions are exchanged primarily between group members with a large size difference, whereas negative estimates of size differences indicate that aggressive interactions are exchanged primarily between group members with a small size difference. acts) to neighbors than subordinates (Z ¼10.36, P< 0.001) contributed to network structure varied across periods. Specifically, and males were more aggressive to neighbors than females aggression and submission between the dominant male and female (Z ¼ 2.39, P ¼ 0.02). Aggression toward neighbors was not sig- were both more frequent when groups were isolated than when nificantly altered by neighbor group size (Z ¼0.08, P ¼ 0.93). groups had recently been exposed to neighbors. In contrast, aggres- We also found that aggression toward neighbors was impacted by sion and submission between similarly sized group members and group-level dynamics. Dominant aggression toward neighbors was dominant aggression toward subordinates were relatively more fre- positively correlated with dominant male aggression toward the dom- quent in the immediate post-exposure period compared with the inant female (Z ¼ 2.03, P ¼ 0.04), but negatively correlated with pre-exposure period. dominant female submission to the dominant male (Z ¼2.98, An increase in aggression between dominant and subordinate P ¼ 0.003). Further, dominants tended to be more aggressive to neigh- individuals when neighbors were present is consistent with the hy- bors when there were no subordinates present in their group (i.e., pothesis that within-group dynamics are altered by the outside when the social unit was composed of only a dominant male and fe- options offered by neighboring groups. These findings are also con- male pair) compared with when there were subordinates in their social sistent with the results of several previous studies in this species that unit (Z ¼1.87, P ¼ 0.06). We found no evidence that subordinate suggest that subordinates “pay” more (in terms of an increase in aggression to neighbors was significantly impacted by affiliation received aggression and more help provided) to remain on the terri- (Z ¼0.10, P ¼ 0.92) or aggression (Z ¼0.04, P ¼ 0.97) tory when there are neighbors present (Bruintjes and Taborsky 53 53 received from the dominants. 2008; Hellmann and Hamilton 2014; Hellmann et al. 2015b). However, these results are in contrast to the results of a theoretical model (Hellmann and Hamilton forthcoming) and empirical studies Discussion in cichlid fish (Bergmu ¨ ller et al. 2005b) and paper wasps (Tibbetts and Reeve 2008; Grinsted and Field 2017) which show that subor- Previous studies have found that the presence and density of neigh- dinate help decreases as outside options increase. Biological market boring groups are correlated with increased subordinate cooper- theory predicts that outside options should benefit the partner that ation, increased subordinate eviction, and altered reproductive “chooses” (Noe ¨ and Hammerstein 1995; Bshary and Grutter 2002; sharing in N. pulcher (Hellmann and Hamilton 2014; Hellmann Bshary and Noe ¨ 2003), suggesting that the presence of neighbors et al. 2015a, 2015b). However, less was known about how neigh- increased partner choice for dominants in this study. However, it is bors impact group-level behavioral dynamics. By manipulating the likely that the “choosy” partner may vary among species as well as presence of neighbors, we can evaluate how neighbor groups change within the same species depending on the relative leverage that an both the magnitude and target of conflict within a group. This can individual has in a given social situation (Lewis 2002). For example, lend insight into who benefits most from the opportunities offered subordinates may have more leverage to negotiate based on outside by neighboring groups, and can better elucidate if groups perceive options in situations in which dominants gain large fitness benefits neighbors as reproductive competitors, future group members, or from subordinate help, relative to situations in which subordinate threats to the territory as a whole. help is less needed (Taborsky 1985; Zo ¨ ttl et al. 2013). Further, out- We found that group-level social dynamics were characterized side options may be relatively unimportant in informing social dy- by strong negative effects of reciprocity and cyclical triads across all namics in systems without pay-to-stay cooperation (Hellmann and time periods, demonstrating that network ties are self-organizing Hamilton forthcoming) or in groups where dominants and subordi- (i.e., the existence of certain ties promotes other ties to come into ex- nates are related (Cant and Johnstone 2009; Quinones et al. 2016). istence) and that dominance hierarchies are likely stable across all Consequently, greater exploration into how neighboring groups in- time periods (McDonald and Shizuka 2012). Further, we found that fluence cooperation and the stability of current social relationships network structure was influenced by the status, sex, and size of the would be highly beneficial to understanding when and to what group members, although the ways in which individual attributes Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 Hellmann and Hamilton  Neighbors and within-group conflict 9 extent group members can negotiate to improve their current social aggression and submission networks between pre-exposure groups situations based on outside options. (days 30–39) and immediate post-exposure groups (days 40–49) An increase in conflict among similarly sized individuals (i.e., were greater than changes between days 40–49 and days 70–79 rank-related conflict) when neighbors were present is also consistent (delayed post-exposure groups), despite greater differences in time with the hypothesis that the presence of neighbors alters within- between the latter 2 time periods. Finally, the changes in our net- group conflict by introducing the potential for group composition to work structure are consistent with what we predicted given that pre- change. In species with size-based dominance hierarchies such as N. vious studies found that the presence of neighbors increases conflict pulcher, conflict tends to be greatest between similarly sized group between dominants and subordinates (Hellmann and Hamilton members (Wong et al. 2007; Hamilton and Heg 2008; Heg and 2014; Hellmann et al. 2015b). Nevertheless, further research com- Hamilton 2008; Ang and Manica 2010) because relative fighting paring network structure in groups that form in the presence of ability is less certain between group members close in size (Reddon neighbors versus those that form while isolated and are later et al. 2011). When neighbors are present, current group members exposed to neighbors would help distinguish which, if any, of the may be in conflict over the joining of a new subordinate, which reported patterns are due to time rather than the presence of could benefit high-ranking group members who would gain benefits neighbors. from the additional help and protection associated with a greater In conclusion, for our network analysis, we used ERGMs to number of subordinates, but would be costly to low-ranking subor- evaluate social dynamics within groups (Wasserman and Pattison dinates who would descend in the dominance hierarchy if a larger 1996). These models control for the dependency among social rela- subordinate joined the group (Heg et al. 2005; Ligocki et al. 2015a). tionships, allowing us to not only ask questions about how individ- Conversely, the potential for subordinates to leave the group may ual attributes (e.g., actor sex) affect social dynamics, but to expand also disrupt the dominance hierarchy, as subordinate removal from the scope of our analysis to assess how structural dependency among the group induces temporary aggression between group members of social ties (e.g., status homophily) and emergent group level charac- adjacent rank as group members re-establish their rank position teristics of the social network itself (e.g., reciprocity: Silk and Fisher (Wong and Balshine 2011b). Collectively, these results suggest that 2017) influence group-level social dynamics. Here, we demonstrate providing opportunities for group composition to change may re- that the target of within-group conflict shifts in the presence of duce the stability of dominance hierarchies. neighbors: isolated groups are characterized by higher conflict be- We predicted that if neighbors represent reproductive opportuni- tween the dominant breeding pair while groups with neighbors are ties, aggression between the dominant male and female pair would characterized by rank-related conflict and conflict between domi- be higher in groups with neighbors compared with isolated groups. nants and subordinates. This suggests that, rather than promote Instead, we found that the presence of neighbors was associated group cohesion or reproductive conflict, neighbors may foster con- with reduced conflict between the dominant male and female pair, flict by introducing opportunities for group composition to change. lending little support to the hypothesis that neighbors promote re- However, further studies independently manipulating the availabil- productive conflict between social mates. Similarly, we observed lit- ity of outside opportunities for dominants and subordinates would tle change in the frequency and target of affiliation before and after elucidate when and to what extent partner choice impacts the struc- exposure to neighbors, providing little support for the hypothesis ture and stability of dominance interactions in animal societies. that out-group threats increase affiliation among group members by Further, measuring changes in hormone levels (e.g., cortisol) and threatening the group as a whole. This is in contrast to Bruintjes gene expression in these manipulative experiments may help us et al. (2016) and Radford (2008a), who found that post-conflict af- understand the ways in which subsets of group members are filiation increased following experimental territorial intrusions by impacted by the presence of these outside options. neighboring and non-neighboring conspecifics. However, territorial intrusions are a more intense form of out-group conflict than the constant nearby presence of neighboring groups where territory Ethics Statement boundaries were never crossed. Consequently, the intergroup con- All methods were approved by The Ohio State University IACUC flict present in our study may have never presented a great enough (protocol ID 2008A0095). As stated above, efforts were made to threat to the focal group to promote higher within-group affiliation. minimize injuries resulting from high levels of aggression by moni- Consistent with this hypothesis, Polizzi di Sorrentino et al. (2012) toring groups daily and removing subordinate group members that found that visual exposure to neighbors in tufted capuchins was not received high levels of aggression. sufficient to produce changes in within-group affiliation. Further, in green woodhoopoes, changes in within-group affiliative behavior were only seen when groups faced more intense and longer out- Author Contributions group threats (Radford 2008b). Consequently, further work is J.K.H. and I.M.H. conceived the study. J.K.H. conducted the experi- needed to understand how within-group affiliation varies with the ment, scored all behavior, performed the statistical analyses, and type and severity of out-group threat. wrote the initial manuscript. Both J.K.H. and I.M.H. revised the Changes among experimental periods could be due to time or manuscript. the presence of neighbors, as observations during the pre-exposure period occurred sooner to group formation than observations during the post-exposure periods. Our results are largely inconsistent with what would be expected under the time hypothesis; our supplemen- Acknowledgments tary analysis demonstrates that the influence of individual attributes We owe immense thanks to Dr Cody Dey for aid with ERGM network ana- (sex, status, size) on network structure had stabilized by day 40. lysis. This research was supported by The Ohio State University. J.K.H. is cur- Changes in the effect of these parameters on networks were greater rently supported by the National Institute of General Medical Sciences of the between time periods than within time periods. Further, changes in National Institutes of Health under Award Number F32GM121033. Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 10 Current Zoology, 2018, Vol. 0, No. 0 Fischer S, Zo ¨ ttl M, Groenewoud F, Taborsky B, 2014. Group-size-dependent Supplementary Material punishment of idle subordinates in a cooperative breeder where helpers pay Supplementary material can be found at https://academic.oup.com/cz. to stay. Proc R Soc B 281:20140184. Fournier D, Skaug H, Ancheta J, Ianelli J, Magnusson A et al., 2012. AD model builder: using automatic differentiation for statistical inference of References highly parameterized complex nonlinear models. Optim Method Softw 27: Ang TZ, Manica A, 2010. Aggression, segregation and stability in a domin- 233–249. ance hierarchy. Proc R Soc B 277:1337–1343. Frostman P, Sherman PT, 2004. Behavioral response to familiar and unfamil- Archie EA, Tung J, Clark M, Altmann J, Alberts SC, 2014. Social affiliation iar neighbors in a territorial cichlid Neolamprologus pulcher. Ichthyol Res matters: both same-sex and opposite-sex relationships predict survival in 51:283–285. wild female baboons. Proc R Soc B 281:pii: 20141261. Gashagaza M, 1988. Feeding activity of a Tanganyikan cichlid fish Arnold KE, Owens IPF, 2002. Extra-pair paternity and egg dumping in birds: Lamprologus brichardi. Afr Study Mongr 9:1–9. life history, parental care and the risk of retaliation. Proc R Soc B 269: Godfrey SS, Sih A, Bull CM, 2013. The response of a sleepy lizard social net- 1263–1269. work to altered ecological conditions. Anim Behav 86:763–772. Aureli F, Cords M, van Schaik CP, 2002. Conflict resolution following aggres- Goetz AT, Shackelford TK, Romero GA, Kaighobadi F, Miner EJ, 2008. sion in gregarious animals: a predictive framework. Anim Behav 64: Punishment, proprietariness, and paternity: men’s violence against 325–343. womenfrom anevolutionary perspective. Aggress Violent Behav 13: Balshine-Earn S, Lotem A, 1998. Individual recognition in a cooperatively 481–489. breeding cichlid: evidence from video playback experiments. Behaviour Griffith SC, Owens IPF, Thuman KA, 2002. Extra pair paternity in birds: a re- 135:369–386. view of interspecific variation and adaptive function. Mol Ecol 11: Barocas A, Ilany A, Koren L, Kam M, Geffen E, 2011. Variance in centrality 2195–2212. within rock hyrax social networks predicts adult longevity. PLoS One 6: Grinsted L, Field J, 2017. Market forces influence helping behaviour in co- e22375. operatively breeding paper wasps. Nat Commun 8:13750. Bergmu ¨ ller R, Heg D, Peer K, Taborsky M, 2005a. Extended safe havens and Hamilton IM, Heg D, 2008. Sex differences in the effect of social status on the between-group dispersal of helpers in a cooperatively breeding cichlid. growth of subordinates in a co-operatively breeding cichlid. J Fish Biol 72: Behaviour 142:1643–1667. 1079–1088. Bergmu ¨ ller R, Heg D, Taborsky M, 2005b. Helpers in a cooperatively breed- Hamilton IM, Heg D, Bender N, 2005. Size differences within a dominance ing cichlid stay and pay or disperse and breed, depending on ecological con- hierarchy influence conflict and help in a cooperatively breeding cichlid. straints. Proc R Soc B 272:325–331. Behaviour 142:1591–1613. Bergmu ¨ ller R, Taborsky M, 2005. Experimental manipulation of helping in a Handcock MS, Hunter DR, Butts CT, Goodreau SM, Morris M, 2008. cooperative breeder: helpers ‘pay to stay’ by pre-emptive appeasement. Statnet: software tools for the representation, visualization, analysis and Anim Behav 69:19–28. simulation of network data. J Stat Softw 24:11. Bierbach D, Oster S, Jourdan J, Arias-Rodriguez L, Krause et al., 2014. Social Heg D, Bender N, Hamilton I, 2004. Strategic growth decisions in helper cichl- network analysis resolves temporal dynamics of male dominance relation- ids. Proc R Soc B 271(Suppl. 6):S505–S508. ships. Behav Ecol Sociobiol 68:935–945. Heg D, Brouwer L, Bachar Z, Taborsky M, 2005. Large group size yields Bruintjes R, Lynton-Jenkins J, Jones JW, Radford AN, 2016. Out-group threat group stability in the cooperatively breeding cichlid Neolamprologus promotes within-group affiliation in a cooperative fish. Am Nat 187: pulcher. Behaviour 142:1615–1641. 274–282. Heg D, Hamilton IM, 2008. Tug-of-war over reproduction in a cooperatively Bruintjes R, Taborsky M, 2008. Helpers in a cooperative breeder pay a high breeding cichlid. Behav Ecol Sociobiol 62:1249–1257. price to stay: effects of demand, helper size and sex. Anim Behav 75: Hellmann JK, Hamilton IM, 2014. The presence of neighbors influences de- 1843–1850. fense against predators in a cooperatively breeding cichlid. Behav Ecol 25: Bshary R, Grutter AS, 2002. Experimental evidence that partner choice is a 386–391. driving force in the payoff distribution among cooperators or mutualists: Hellmann JK, Hamilton IM, forthcoming. Dominant and subordinate outside the cleaner fish case. Ecol Lett 5:130–136. options alter help and eviction in a pay-to-stay negotiation model. Behav Bshary R, Noe ¨ R, 2003. Biological markets: the ubiquitous influence of part- Ecol, https://doi.org/10.1093/beheco/ary036. ner choice on the dynamics of cleaner fish—client reef fish interactions. In: Hellmann JK, Ligocki IY, O’Connor CM, Reddon AR, Garvy KA et al., Hammerstein P, editor. Genetic and Cultural Evolution of Cooperation. 2015a. Reproductive sharing in relation to group and colony-level attributes Cambridge: MIT Press. 167–184. in a cooperative breeding fish. Proc R Soc B 282:20150954. Cant M, Johnstone R, 2009. How threats influence the evolutionary resolution Hellmann JK, Reddon AR, Ligocki IY, O’Connor CM, Garvy KA et al., of within-group conflict. Am Nat 173:759–771. 2015b. Group response to social perturbation: impacts of isotocin and the Cantor M, Wedekin LL, Guimar~ aes PR, Daura-Jorge FG, Rossi-Santos MR social landscape. Anim Behav 105:55–62. et al., 2012. Disentangling social networks from spatiotemporal dynamics: Hick K, Reddon Adam R, O’connor Constance M, Balshine S, 2014. Strategic the temporal structure of a dolphin society. Anim Behav 84:641–651. and tactical fighting decisions in cichlid fishes with divergent social systems. Dey CJ, Quinn JS, 2014. Individual attributes and self-organizational proc- Behaviour 151:47–71. esses affect dominance network structure in pukeko. Behav Ecol 25: Hunter DR, Handcock MS, Butts CT, Goodreau SM, Morris M, 2008. ergm: 1402–1408. a package to fit, simulate and diagnose exponential-family models for net- Dey CJ, Reddon AR, O’Connor CM, Balshine S, 2013. Network structure is works. J Stat Softw 24:1–29. related to social conflict in a cooperatively breeding fish. Anim Behav 85: Huntingford FA, Turner AK, 1987. Animal Conflict. London: Chapman & 395–402. Hill. Dey CJ, Tan QYJ, O’Connor CM, Reddon AR, Caldwell JR et al., Jungwirth A, Walker J, Taborsky M, 2015. Prospecting precedes dispersal and 2015. Dominance network structure across reproductive contexts in the co- increases survival chances in cooperatively breeding cichlids. Anim Behav operatively breeding cichlid fish Neolamprologus pulcher. Curr Zool 61: 106:107–114. 45–54. Kohda M, Jordan LA, Hotta T, Kosaka N, Karino K et al., 2015. Facial recog- Dierkes P, Taborsky M, Kohler U, 1999. Reproductive parasitism of brood- nition in a group-living cichlid fish. PLoS One 10:e0142552. care helpers in a cooperatively breeding fish. Behav Ecol 10:510–515. Krivitsky P, 2013. ergm.count: fit, simulate and diagnose exponential-family Eggert A-K, Sakaluk SK, 1995. Female-coerced monogamy in burying beetles. models for networks with count edges. R package version 311. Available Behav Ecol Sociobiol 37:147–153. from http://wwwstatnetorg/. Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018 Hellmann and Hamilton  Neighbors and within-group conflict 11 Kutsukake N, Clutton-Brock TH, 2008. The number of subordinates moder- Reddon AR, Voisin MR, Menon N, Marsh-Rollo SE, Wong MYL et al., 2011. ates intrasexual competition among males in cooperatively breeding meer- Rules of engagement for resource contests in a social fish. Anim Behav 82: kats. Proc R Soc B 275:209–216. 93–99. Kutsukake N, Clutton-Brock TH, 2010. Grooming and the value of social Robins G, Pattison P, Kalish Y, Lusher D, 2007. An introduction to exponen- relationships in cooperatively breeding meerkats. Anim Behav 79:271–279. tial random graph (p*) models for social networks. Soc Networks 29: Kuznetsova A, Brockhoff P, Christensen R, 2017. lmerTest package: tests in 173–191. linear mixed effects models. J Stat Softw 82:1–26. Sabol AC, Hellmann JK, Gray SM, Hamilton IM, 2017. The role of ultraviolet col- Lewis RJ, 2002. Beyond dominance: the importance of leverage. Quart Rev oration in intrasexual interactions in a colonial fish. Anim Behav 131:99–106. Biol 77:149–164. Seyfarth RM, Cheney DL, 1984. Grooming, alliances and reciprocal altruism Ligocki IY, Balshine S, Hamilton Ian M, Reddon AR, Hellmann JK et al., in vervet monkeys. Nature 308:541–543. 2015a. Social status influences responses to unfamiliar conspecifics in a co- Silk JB, Alberts SC, Altmann J, 2003. Social bonds of female baboons enhance operatively breeding fish. Behaviour 152:1821–1839. infant survival. Science 302:1231–1234. Ligocki IY, Earley RL, Hellmann JK, Hamilton IM, 2015b. Variation in gluco- Silk MJ, Fisher DN, 2017. Understanding animal social structure: exponential ran- corticoid levels in relation to direct and third-party interactions in a social dom graph models in animal behaviour research. Anim Behav 132:137–146. cichlid fish. Physiol Behav 151:386–394. Skaug H, Fournier D, Bolker B, Magnusson A, Nielsen A, 2016. Lusher D, Koskinen J, Robins G, 2013. Exponential Random Graph Models Generalized linear mixed models using “AD Model Builder.” R package for Social Networks: Theory, Methods, and Applications. Cambridge: version 0833. Cambridge University Press. Sopinka NM, Fitzpatrick JL, Desjardins JK, Stiver KA, Marsh-Rollo SE et al., Madden JR, Drewe JA, Pearce GP, Clutton-Brock TH, 2009. The social net- 2009. Liver size reveals social status in the African cichlid Neolamprologus work structure of a wild meerkat population: 2. Intragroup interactions. pulcher. J Fish Biol 75:1–16. Behav Ecol Sociobiol 64:81–95. Stiver KA, Desjardins JK, Fitzpatrick JL, Neff B, Quinn JS et al., 2007. McDonald DB, Shizuka D, 2012. Comparative transitive and temporal order- Evidence for size and sex-specific dispersal in a cooperatively breeding cich- liness in dominance networks. Behav Ecol 24:511–520. lid fish. Mol Ecol 16:2974–2984. Noe ¨ R, Hammerstein P, 1995. Biological markets. Trends Ecol Evol 10: Stiver KA, Dierkes P, Taborsky M, Balshine S, 2004. Dispersal patterns and 336–339. status change in a co-operatively breeding cichlid Neolamprologus pulcher: O’Connor CM, Reddon AR, Ligocki IY, Hellmann JK, Garvy KA et al., 2015. evidence from microsatellite analyses and behavioural observations. J Fish Motivation but not body size influences territorial contest dynamics in a Biol 65:91–105. wild cichlid fish. Anim Behav 107:19–29. Stiver KA, Fitzpatrick JL, Desjardins JK, Balshine S, 2009. Mixed parentage in Polizzi di Sorrentino E, Schino G, Massaro L, Visalberghi E, Aureli F, 2012. Neolamprologus pulcher groups. J Fish Biol 74:1129–1135. Between-group hostility affects within-group interactions in tufted capuchin Taborsky M, 1985. Breeder–helper conflict in a cichlid fish with broodcare monkeys. Anim Behav 83:445–451. helpers: an experimental analysis. Behaviour 95:45–75. Quinones AE, van Doorn GS, Pen I, Weissing FJ, Taborsky M, 2016. Tibbetts EA, Reeve HK, 2008. Two experimental tests of the relationship be- Negotiation and appeasement can be more effective drivers of sociality than tween group stability and aggressive conflict in Polistes wasps. kin selection. Phil Trans R Soc Lond B Biol Sci 371:20150089. Naturwissenschaften 95:383–389. Radford AN, 2008a. Duration and outcome of intergroup conflict influences Valera F, Hoi H, Kri stı´n A, 2003. Male shrikes punish unfaithful females. intragroup affiliative behaviour. Proc R Soc B 275:2787–2791. Behav Ecol 14:403–408. Radford AN, 2008b. Type of threat influences postconflict allopreening in a Wasserman S, Pattison P, 1996. Logit models and logistic regressions for social social bird. Curr Biol 18:R114–R115. networks: i. An introduction to Markov graphs andp. Psychometrika 61: Radford AN, 2011. Preparing for battle? Potential intergroup conflict pro- 401–425. motes current intragroup affiliation. Biol Lett 7:26–29. Whitehead H, 2008. Analyzing Animal Societies: Quantitative Methods for RadfordAN, FawcettTW, 2014. Conflict between groups promotes later de- Vertebrate Social Analysis. Chicago: Chicago University Press. fense of a critical resource in a cooperatively breeding bird. Curr Biol 24: Wong M, Balshine S, 2011a. The evolution of cooperative breeding in the African 2935–2939. cichlid fish Neolamprologus pulcher. Biol Rev Camb Philos Soc 86:511–530. Radford AN, Majolo B, Aureli F, 2016. Within-group behavioural consequen- Wong M, Balshine S, 2011b. Fight for your breeding right: hierarchy re-establishment ces of between-group conflict: a prospective review. Proc R Soc B 283: predicts aggression in a social queue. Biol Lett 7:190–193. 20161567. Wong MYL, Buston PM, Munday PL, Jones GP, 2007. The threat of punish- Reddon AR, O’Connor CM, Marsh-Rollo SE, Balshine S, Gozdowska M ment enforces peaceful cooperation and stabilizes queues in a coral-reef fish. et al., 2015. Brain nonapeptide levels are related to social status and affilia- Proc R Soc B Biol Sci 274:1093–1099. tive behaviour in a cooperatively breeding cichlid fish. R Soc Open Sci 2: Zo ¨ ttl M, Frommen JG, Taborsky M, 2013. Group size adjustment 140072. to ecological demand in a cooperative breeder. Proc R Soc B 280:20122772. Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy025/4961448 by Ed 'DeepDyve' Gillespie user on 08 June 2018

Journal

Current ZoologyOxford University Press

Published: Apr 5, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off