Differences in neural activity, but not behavior, across social contexts in guppies, Poecilia reticulata

Animals are continually faced with the challenge of producing context-appropriate social behaviors. In many instances, appropriate behaviors differ by social situation. However, in some instances, the same behaviors are employed across different social contexts, albeit in response to distinct stimuli and with distinct purposes. We took advantage of behavioral similarities across mating and aggression contexts in guppies, Poecilia reticulata, to understand how patterns of neural activity differ across social contexts when behaviors are nonetheless shared. While there is growing interest in understanding behavioral mechanisms in guppies, resources are sparse. As part of this study, we developed a neuroanatomical atlas of the guppy brain as a research community resource. Using this atlas, we found that neural activity in the preoptic area reflected social context, whereas individual differences in behavioral motivation paralleled activity in the posterior tuberculum and ventral telencephalon (teleost homologs of the mammalian ventral tegmental area and lateral septum, respectively). Our findings suggest independent coding of social salience versus behavioral motivation when behavioral repertoires are shared across social contexts. Choosing behaviors appropriate to the current social situation is of central importance to animals. Interactions with different social partners (e.g., mates, competitors, or offspring) generally require distinct behavioral repertories. However, in some cases, similar behaviors are used across social contexts. The neural mechanisms underlying social behavior are particularly intriguing in these situations, where the same behaviors are produced in response to distinct social stimuli and for distinct purposes. We took advantage of behavioral similarities across mating and aggression interactions in Trinidadian guppies to explore how social information is reflected in the brain when fish perform a common set of behaviors across contexts. We found that activity in distinct brain regions reflects social context versus behavioral motivation, suggesting a means by which social inputs and behavioral outputs can be coded independently of one another.


Introduction
Producing behaviors appropriate to the current social situation is a central challenge for animals, requiring the integration of both internal and external cues. Integrating cues from conspecifics is particularly critical, as interactions with potential mates, competitors, and offspring generally require distinct behavioral repertoires and physiological states (e.g., Davies et al. 2012). However, in some instances, similar behaviors are deployed across social contexts, albeit with distinct purposes and complex consequences for evolutionary trajectories (Hebets et al. 2016). For example, frogs, birds, rodents, and humans produce similar vocalizations in for mate attraction and territory defense (e.g., Portfors 2007;Wells 2007; Communicated by A. Pilastro Catchpole 2008). While behaviors are the same across contexts in these cases, they are produced in response to distinct stimuli, indicating that divergent sensory inputs are converted into similar behavioral outputs. The neural underpinnings promoting context-dependent behaviors are particularly intriguing in instances where seemingly identical behaviors are employed during functionally distinct social interactions, as the degree to which these contexts are dissociable has important implications for the evolution of communication signals (Hebets et al. 2016).
Trinidadian guppies, Poecilia reticulata, perform similar, highly stereotyped behaviors during mating and aggressive interactions, and provide an excellent system in which to examine how social salience is reflected in the brain. Owing to extensive work in both the wild and the lab, much is known about the environmental cues influencing behaviors and their ultimate adaptive significance in this species (e.g., Kolluru and Grether 2005;Botham et al. 2008;Chapman et al. 2009;Huizinga et al. 2009;Mariette et al. 2010;Heintz et al. 2015;Burns et al. 2016 for recent examples ;Houde 1997;Magurran 2005 for review). Guppies are live-bearing fish with internal fertilization, and males spend the majority of their time in pursuit of females. Male guppies display a stereotyped courtship behavior known as a sigmoid display, during which they orient themselves perpendicularly to a female, assume the characteristic S-shape that gives the display its name, and quiver their bodies. As an alternative to these overt courtship displays, male guppies also attempt to gain fertilizations by forced/sneaky copulation. In this case, males approach females from behind and below and thrust their gonopodium (intromittent organ) forward towards the female's genital pore. In addition, males will bite, head-butt, and body-slam females, presumably to get their attention when they are unresponsive (reviewed in Houde 1997).
Despite obvious functional differences, male guppies direct a strikingly similar set of behaviors towards other males. While male-male aggression is uncommon in wild populations (Farr 1976(Farr , 1980, it can be readily be elicited by shifts in social environment, including the introduction of unfamiliar males, short-term changes in population density, altered rearing conditions, and resource availability (Farr and Herrnkind 1974;Bruce and White 1995;Jirotkul 1999;Field and Waite 2004;Kolluru and Grether 2005;Price and Rodd 2006;Fischer et al. 2016). Although male guppies respond both directly and indirectly to the behaviors of other males (Farr 1976;Bruce and White 1995;Price and Rodd 2006;Nuffer and Alburn 2010), the precise functional nature of male-male displays remains unresolved because: (1) the non-intuitive nature of male-male courtship displays means they are often not quantified (Gorlick 1976;Bruce and White 1995;Price and Rodd 2006); (2) these displays are uncommon in natural settings (Farr 1976(Farr , 1980, perhaps due to familiarity and established hierarchies among males (Bruce and White 1995;Price and Rodd 2006);and (3) studies conflict as to the consequences of these interactions for mating success (increased mating success : Gorlick 1976;Kodric-Brown 1992;Bruce and White 1995;no effect: Farr 1976;Houde 1988;Price and Rodd 2006). Nonetheless, the same courtship displays, forced copulation attempts, and physical contacts that achieve successful copulations are routinely employed in aggressive contexts when fish are establishing dominance hierarchies (Houde 1988(Houde , 1997Nuffer and Alburn 2010;Fischer et al. 2016). Given overlap at the behavioral level, how are male guppies nonetheless attuned to the obvious and functionally important contextual differences between mating and aggressive encounters? In the present study, we examine the neural mechanisms mediating behavior across social contexts in guppies. We begin by characterizing behavior in mating versus aggressive contexts. Using an atlas of the guppy brain we constructed, we next describe patterns of neural activation associated with mating versus aggressive contexts in 13 brain regions. We focus our analysis on brain regions mediating social behaviors that are evolutionarily ancient and functionally conserved across vertebrates (the Social Decision Making Network, SDMN; O'Connell and Hofmann 2011a, b). Finally, we combine behavioral measures with neural activity data to understand associations between neural induction and behavioral output. The results of this study build resources for future work examining neural mechanisms of behavior in this increasingly popular study system and provide insights into how distinct brain regions code social context versus social motivation.

Animals
All fish used in this study were sexually mature males from a single lab-reared population derived from the Marianne River Drainage in the Northern Range Mountains of Trinidad. Fish were housed in mixed-sex group tanks of approximately equal densities (~1:1 male to female; 40-50 fish per 60-L tank). All tanks contained gravel substrate, plants that acted as a refuge, and filters to clean and aerate the water. Fish were kept on a 12:12 h light cycle (lights on 7:00 am to 7:00 pm) and fed a measured food diet once daily. Fish received Tetramin™ tropical flake paste and hatched Artemia cysts on an alternating basis.

Behavior
Males were assigned to one of three experimental groups: aggression, mating, or isolation (n = 10 per group). Fish were assayed concurrently in sets of three per day, with one representative from each experimental condition. All fish run on the same day were selected from the same holding tank and fish from different holding tanks were balanced across days. Fish were placed in individual 2.5-L tanks on the afternoon preceding behavioral trials. Behavioral trials were conducted the following morning, 2 h after lights-on and lasted 60 min thereafter. In the aggression condition, two unfamiliar males were introduced into the focal male's tank at the start of the trial. In the mating condition, two unfamiliar females were introduced. In the isolation condition males remained alone in their tanks throughout the trial. Behaviors were continuously recorded by two independent observers using JWatcher™ software. Each observer watched either the aggression or the mating condition at alternating 15-min intervals, such that behaviors recorded for each social condition were evenly distributed among observers. It was not possible to use blind observers, as experimental group (i.e., the sex of conspecifics) was obvious during observation. Tanks were isolated from one another by opaque barriers so fish could not see one another and behavioral trials were conducted behind a blind with tanks lit from above to reduce visibility of the observers to the fish.
We followed protocols previously established in our lab to define and record behaviors (Fischer et al. 2016). Previous work in our lab demonstrates that guppies perform similar behaviors in aggression and mating trials and so the same behaviors were scored in both contexts. These included the number and duration of sigmoid displays, the number of forced copulation attempts, the number of gonopodial swings, the number of physical contacts not related to forced copulation (i.e., nipping, body slamming, tail slapping), and the number of posturing incidents (when fish line up nose to nose).
Female guppies are sexually receptive only as virgins and in a brief time window following parturition (reviewed in Houde 1997;Magurran 2005). To mitigate the confounding effects of successful copulations on our results, we minimized the potential for male copulatory success by using only unreceptive, non-virgin females. Indeed, only a single male successfully copulated with a focal female and subsequently ceased interacting with that female. We excluded this male from all further analyses due to the confounding effects of behavioral and physiological responses following successful copulation.

Tissue collection
Whole brains were collected immediately following behavioral trials. Guppies were anesthetized by rapid cooling, followed by decapitation. Whole heads were fixed in 4% paraformaldehyde at 4°C overnight and then transferred to 30% sucrose for dehydration. Following dehydration, whole heads were embedded in mounting media (Tissue-Tek® O.C.T. Compound, Electron Microscopy Sciences, Hatfield, PA, USA), rapidly frozen, and stored at − 80°C until cryosectioning. Heads were sectioned in the coronal plane at 14 μm and thaw-mounted into two replicate series onto charged slides (Superfrost Plus, VWR, Randor, PA, USA). We used alternate series to prevent double-counting of neurons that span adjacent sections. Tissue sections were stored at − 20°C until immunohistochemical staining.

Immunohistochemistry
We used a phospho-S6 antibody that targets phosphorylated ribosomes (pS6; Life Technologies, Carlsbad, CA, USA) to assay neural activity. Ribosomes become phosphorylated following changes in electrical activity in neurons and the pS6 antibody therefore acts as a general marker of neural activation, akin to immediate early genes (Knight et al. 2012). As time course is critical for experiments involving immediate early genes and can vary across species, we assessed staining intensity at three time points (30, 60, and 90 min) in sample guppy tissue prior to the experiment and chose the 60-min time point based on these preliminary results (data not shown).
We followed standard immunohistochemical procedures for antibody staining to label pS6-positive neurons. Briefly, we quenched endogenous peroxidases using a 3% H 2 O 2 solution, blocked slides in 5% normal goat serum diluted in 1X phosphate-buffered saline (PBS) and 0.03% Triton-X for 1 h, and then incubated slides in the anti-pS6 primary antibody (Life Technologies, Waltham, MA, USA) at a concentration of 1:500 in blocking solution overnight at 4°C. The following day, we incubated slides in secondary antibody (Jackson ImmunoResearch, West Grove, PA, USA) at a concentration of 1:200 in blocking solution for 2 h, incubated slides in an avidin-biotin complex (ABC) solution (Vector Laboratories, Burlingame, CA, USA) for 2 h, and treated slides with 3,3′diaminobenzidine (DAB; Vector Laboratories, Burlingame, CA, USA) for 5 min to produce a color reaction. Slides were rinsed in 1X PBS prior to and following all the above steps. Finally, slides were rinsed in water, dehydrated in increasing concentrations of ethanol (50, 75, 95, 100, 100%), and coverslipped with Permount (Fisher Scientific, Hampton, NH, USA).

Microscopy and cell counting
To reliably quantify neural activity across candidate brain regions, we created a guppy brain atlas (Supplemental Material S1). We examined coronal brain sections of multiple male and female guppies stained using cresyl violet to assess morphology. We identified brain regions using neuroanatomical information from other fishes (Anken and Rahmann 1994;Wullimann et al. 1996;Munchrath and Hofmann 2010) and have made the atlas freely available online.
We photographed brain tissue at × 20 magnification on a light microscope (Zeiss AxioZoom, Zeiss, Oberkochen, Germany) attached to a camera (ORCA-ER, Hamamatsu, San Jose, CA, USA) and analyzed cell counts from photographs using FIJI software (Schindelin et al. 2012). We outlined and measured focal brain regions (Fig. 1) and counted all stained cells within a given region. All regions extended across multiple sections, and we quantified cell number for each region in all possible sections. We counted cells in only a single hemisphere per section. We counted cells in the left hemisphere unless there was substantial unilateral tissue damage, in which case we used the right hemisphere in the damaged section.

Statistical analysis
We initially compared negative binomial and normal distributions, and selected the former based on qq-plots, the distribution of residuals, and metrics of overall model fit. We tested for the influence of social context (aggression versus mating) on behavior using generalized linear mixed models with a negative binomial distribution appropriate for count data with unequal variances. We tested for differences in the number of times each behavior was performed during the 60-min trial. In addition, we summed the counts of all behaviors (excluding sigmoid duration which was measured in seconds) into a single total behavioral metric to assess overall behavioral activity. We chose this approach as (1) summing preserves the count nature of the original data, and (2) we have previously shown (Fischer et al. 2016)-and confirmed with exploratory analyses here-that the correlations between behaviors are not consistent across contexts, and therefore, the same principal components or factors cannot accurately summarize behavioral variation across experimental groups.
We used linear mixed models with a negative binomial distribution to test for differences in neural activation based on social context (aggression versus mating versus isolation). The model included social context and brain region as independent variables and the number of pS6-positive cells per section as the dependent variable. We included fish identity as a random effect to control for repeated sampling within and among regions. As we expected that only some regions would show activation differences based on social context, we used Tukey-corrected post hoc tests to test which regions differed in activation across contexts.
Finally, we tested whether activation in some regions predicted individual differences in behavior. To do this, we ran separate models for each brain region, in which the number of pS6-positive cells per region, social context, and their interaction predicted the total number of behaviors during the trial. We excluded fish from the isolation treatment for this analysis because we had no behavioral data for these fish. We chose to use our total behavioral metric because (1) we wanted to use a metric that reflected general behavioral motivation, and (2) this approach increased the power of the statistical analysis. All statistical analyses were performed in SAS 9.4 (SAS Institute, Cary, NC, USA).
Data availability The datasets generated during and analyzed during the current study are available in the Electronic Supplemental Material (Supplemental Material S2).

Results
Fish performed similar behaviors during aggressive and mating interactions (Fig. 2), and we observed no statistically significant differences (p > 0.05) in the number of single or overall behaviors performed across mating and aggressive contexts (Table 1).
We assessed differences in neural activation among social contexts (aggression, mating, isolation) by quantifying regionspecific pS6 immunoreactivity. Social context influenced neural induction of pS6 in a brain region-specific manner (context*region: F 24,1778 = 3.36, p < 0.0001). Post hoc analyses (Table 2) revealed differential activation in the preoptic area (POA; F 2,1778 = 6.58, p = 0.001): fish in the mating context had higher pS6 activation in the POA compared to fish in the aggression (t = 2.86, adjusted p = 0.019) or isolation (t = 3.39, adj p = 0.002) contexts, which did not significantly differ from one another in their extent of neural activation (t = 0.53, Tukey adj p = 0.8569) (Fig. 3). No other brain regions differed in activity among experimental groups after correction for multiple hypothesis testing (Fig. 3).
Finally, we tested for both context-dependent and contextindependent relationships between pS6 immunoreactivity and behavior. pS6 induction in the posterior tuberculum (TPp: Although behaviors varied among individuals, males did not differ in the number of behaviors performed between contexts. All behaviors are counts except sigmoid duration, which was measured in seconds. The total behavior measure used in additional analyses represents the sum of these behaviors (except sigmoid duration). Median, first and third quartiles, whiskers (± 1.5 interquartile range), and outliers are plotted F 1,15 = 5.02, p = 0.041, r = 0.46) and the lateral part of the ventral telencephalon (Vl: F 1,15 = 4.69, p = 0.047, r = 0.41) were positively associated with behavior in a contextindependent manner (Fig. 4). The relationship between pS6 induction and behavior did not differ between mating and aggression contexts in any brain region (Table 3).

Discussion
In the present study, we sought to shed light on the neural mechanisms mediating behavioral interactions across social contexts and to understand how context is reflected at the neural level when animals perform similar behaviors. To do so, we characterized neural activity patterns associated with shared behavioral repertoires across mating and aggression contexts in adult male guppies. Adult male guppies performed similar behaviors during aggression and mating interactions. Extensive work has demonstrated that male guppies use alternative behavioral strategies under differing environmental conditions (e.g., Kolluru and Grether 2005;Botham et al. 2008;Chapman et al. 2009;Huizinga et al. 2009;Mariette et al. 2010;Heintz et al. 2015;Burns et al. 2016 for recent examples; Houde 1997; Magurran 2005 for review), yet competitive interactions between males have rarely been considered, as aggressive interactions are relatively uncommon in natural populations (Farr 1976(Farr , 1980 and the functional consequences of male-male interactions remain unresolved (Farr 1976;Gorlick 1976;Houde 1988;Kodric-Brown 1992;Bruce and White 1995;Price and Rodd 2006;Nuffer and Alburn 2010). Our data suggest that, under the same environmental conditions, male guppies direct similar behaviors with similar frequencies towards male competitors and female potential mates (see also Fischer et al. 2016), and we therefore suggest that other environmental factors are more important in modifying the frequency and type of behaviors males perform. Given the similarities we observe at the behavioral level, how is social context encoded in the brain?

Differential neural activity in the POA across social contexts
Despite the lack of behavioral differences between contexts, mating and aggression present distinct challenges and opportunities for males, and we therefore expected social context to be encoded at the neural level. Indeed, we observed that neural activation increased in the preoptic area (POA) of males following interactions with female-but not male-conspecifics. The POA is an evolutionarily ancient brain region that is largely hodologically, molecularly, and functionally conserved across vertebrates. The identification of neural  Table 2 Post hoc analyses of regional differences in neural induction. Abbreviations for brain regions are defined in Fig. 1  In male fish, electrical stimulation of the POA has been shown to increase reproductive behaviors (Demski and Knigge 1971;Satou et al. 1984;Wong 2000) and decrease aggression (Demski and Knigge 1971), while ablation of the POA eliminates spawning and other sexual behaviors (Macey et al. 1974;Koyama et al. 1984). Although male guppies perform similar behaviors across mating and aggression contexts, copulation is only possible in interactions with females. The differences we describe in POA activation may be related to this critical distinction. Indeed, increased POA immediate early gene activity has been observed in response to sexualbut not aggressive-interactions in male voles (Wang et al. 1997) and male birds (Riters and Ball 1999 Alger et al. 2009). Moreover, neural activity in the POA has been shown to be important for ejaculation in rodents and monkeys (Malsbury 1971;Slimp et al. 1978), as well as some fish (Demski 1973;Satou et al. 1984), although we note that none of the fish included in this study successfully copulated. As internal fertilization is rare in fish, guppies provide a unique system in which to more explicitly examine mechanistic convergence of ejaculation behavior in future.
The differential activation of POA neurons in response to social stimuli raises the question as to how specific subpopulations of POA neurons function differently across social contexts. Previous work across a wide range of fishes demonstrates that nonapeptide signaling in the POA plays a central role in regulating aggressive, affiliative, and sexual behaviors (Foran and Bass 1999;Semsar et al. 2001;Larson et al. 2006;Greenwood et al. 2008;Almeida et al. 2012;Godwin and Thompson 2012;O'Connell et al. 2012;Ramallo et al. 2012). Given the small fraction of POA neurons that release nonapeptides, a lack of pS6 induction does not preclude higher nonapeptide release during aggressive encounters and characterizing the expression patterns of specific neuromodulatory cell groups during different social interactions in guppies will complement our analysis of broad activation patterns throughout the POA.
The only other study exploring neural mechanisms of behavior in guppies of which we are aware examined neural activity patterns following exposure to same-sex conspecifics in females (Cabrera-Álvarez et al. 2017). The authors found that POA activation (measured using egr-1 immunolabeling) in female guppies varied depending on the number of female social partners (Cabrera-Álvarez et al. 2017). Coupled with evidence for sex differences in affiliative propensity (Magurran et al. 1992;Griffiths and Magurran 1998;Magurran 1998;Croft et al. 2003) and possibly social cognitive ability (Kotrschal et al. 2012;Lucon-Xiccato et al. 2016), the contrast between our results and the patterns of egr-1 induction in female guppies suggests the intriguing possibility that sex differences in POA function may reflect sex differences in the responses to same-sex social partners, i.e., affiliation among females vs. aggression among males. As activity measures using different immediate early genes are broadly, but not entirely, concordant (Velho et al. 2005;Burmeister et  Relationship between neural activity and behavior. Increased neural induction in a the TPp and b Vl was associated with an increased number of total behaviors independent of social context (gray regression line). Context-specific trends are also shown (aggression = light orange circles and line, mating = dark orange circles and line), though the relationship (i.e., the slope of the line) did not differ across contexts in either brain region Model-estimated numbers of pS6-positive cells per section for each individual, rather than raw cell counts, are plotted Table 3 Test of social context dependent and independent associations between neural induction and behavior by brain region. Abbreviations for brain regions are defined in the Fig. 1 Knight et al. 2012), the possible sex differences in POA function warrant direct experimental comparison. In sum, our observations add to the body of work documenting widespread functional conservation of the POA in regulating social behavior across vertebrates. Differential POA activity across social contexts, even when the behaviors performed across those contexts are shared, suggests a role for the POA in integrating sensory information to establish context-specific decision-making strategies.

Neural activity in the TPp and Vl predict behavioral activity
In addition to identifying brain regions reflecting social context differences, we asked whether neural activation in some regions predicted behavior, either in a context-dependent or context-independent manner. We did not identify any regions with context-dependent associations but did identify two brain regions in which levels of activation were positively associated with the number of behaviors performed in both aggressive and mating contexts: the posterior tuberculum (TPp) and the lateral part of the ventral telencephalon (Vl). Homologies of the TPp and Vl to mammalian brain regions remain somewhat contentious (Northcutt 2008;Vargas et al. 2009). In fish, the TPp is generally proposed to be homologous to part of the mammalian ventral tegmental area (VTA), and Vl, together with the ventral part of the ventral telencephalon (Vv), is the putative homolog of the mammalian lateral septum (LS) Wullimann 2001, 2004). Acknowledging ongoing controversy concerning these homologies, we interpret our findings in view of what is currently known about these regions from fish and other vertebrates. The VTA is a central component of the dopaminergic reward system, which plays a critical role in evaluating stimulus salience and motivating behavior. The LS receives projections from the VTA as well as the hippocampus, the hypothalamus, and the midbrain (Meibach and Siegel 1977;Swanson and Cowan 1979;Staiger and Nürnberger 1989). It plays a role in both sexual and social behavior, in particular in the context of social memory, social recognition, and evaluating stimulus novelty (Maeda and Mogenson 1981;Dantzer et al. 1988;Landgraf et al. 1995;Liebsch et al. 1996;Bielsky et al. 2005). In short, both regions are critical in evaluating, responding to, and retaining social information. Moreover, the connections between these two regions are thought to be particularly important in modulating goal-directed behavior (reviewed in O'Connell and Hofmann 2011a, b). Given that all behaviors we assayed are typically directed at other individuals, and that we found group-independent associations between neural activity and behavior in these particular regions, we propose that increased neural induction in these regions reflects increased social behavioral motivation across contexts.
Though the majority of functional studies on behavioral motivation come from mammals, substantial work in songbirds provides a particularly interesting comparison, as birds perform very similar singing behaviors across social contexts, analogous to the similarities in behavior across contexts we observed in guppies. Moreover, songbirds' motivation to sing is clearly distinct from their ability to sing, as the latter is controlled by a well-defined set of song nuclei in the brain, while the former is controlled largely by the POA, the VTA, and the LS (reviewed in Riters 2012). Increased singing behavior is associated with increased neural induction in the VTA during both sexual and agonistic encounters (Maney and Ball 2003;Heimovics and Riters 2005;Bharati and Goodson 2006). Lesion and electrophysiological studies link activity in the VTA to proper production of sexual song (Yanagihara and Hessler 2006;Hara et al. 2007;Huang and Hessler 2008), and LS lesions disrupt aggressive responses to territory intrusion (Ramirez et al. 1988).
As in birds, the context-independent association we observed between activation in VTA and LS homologs and behavior may reflect evaluation of social stimuli and motivation to respond to these stimuli generally. Parallel patterns of activity and association with behavior in TPp and Vl are consistent with the hypothesis that connections between these brain regions play a role in mediating their activity and, by extension, behavioral output (Maeda and Mogenson 1981). Importantly, our observation that TPp and Vl activation are positively associated with behavioral activity regardless of context suggests that individual behavioral differences may stem largely from differences in motivation that are independent of social context. Thus, integration of social, other external, and internal cues, may be able to modulate TPp and Vl activity to modify behavioral motivation across social as well as other behavioral contexts.

Conclusions
Although male guppies performed similar behaviors in aggression versus mating contexts, we observed distinct patterns of activation associated with social context and behavioral output at the neural level. Induction of pS6 in the POA differed based on the social context male guppies experienced, while pS6 activation in TPp and Vl was similarly positively associated with behavioral output across contexts. Taken together, the patterns we observe support a model in which distinct aspects of behavior are mediated by a balance of activity across distributed nodes of the SDMN (Teles et al. 2015), such that activity in distinct brain regions reflects behavioral context versus social motivation in a species in which behavioral repertoires are shared across different types of social interactions.