Vocal buffering of the stress response: exposure to conspecific vocalizations moderates urinary cortisol excretion in isolated marmosets

Abstract

For many species, the presence of a significant social partner can lessen the behavioral and physiological responses to stressful stimuli. This study examined whether a single, individually specific, signature vocalization (phee call) could attenuate the physiological stress response that is induced in marmosets by housing them in short-term social isolation. Utilizing a repeated-measures design, adult marmosets (n = 10) were temporarily isolated from their long-term pair mate and exposed to three conditions: signature vocalizations from the pair mate, phee calls from an unfamiliar opposite sex individual, or no auditory stimuli. Levels of urinary cortisol were monitored as a physiological indicator of the stress response. Urinary cortisol levels were also monitored, while subjects remained undisturbed in their home cages to provide baseline levels. Temporarily isolated marmosets showed significantly higher levels of urinary cortisol than undisturbed marmosets. However, the nature of the acoustic stimulus experienced during isolation led to differences in the excretion of urinary cortisol. Isolated marmosets exposed to a familiar pair mate's vocalization showed significantly lower levels of urinary cortisol than when exposed to unfamiliar marmoset vocalizations (P < 0.04) or to no auditory stimuli (P < 0.03). Neither the duration of pairing nor the quality of relationship in the pair (indexed by spatial proximity scores) predicted the magnitude of reduction in cortisol in the familiar vocalization condition. The results presented here provide the first evidence that a single, individually specific communication signal can decrease the magnitude of a physiological stress response in a manner analogous to the physical presence of a social partner, a process we term “vocal buffering.”

Introduction

For members of a social species, one of the most profound moderators of function in the hypothalamic–pituitary–adrenal (HPA) axis during times of stress may be the presence of significant social partners (DeVries et al., 2003, Hennessy, 1997, Levine, 1993). This model of social support, sometimes known as buffering, postulates that during times of stress, the presence of significant social partners down-regulates activity in the HPA axis and hence serves to buffer the individual against the stressful stimulus (Cohen and Wills, 1985). Presumably, the beneficial effects associated with the physical presence of a significant social partner derive from the partner's ability to either modify the perceived intensity of the stressor, and/or to down-regulate the magnitude and duration of the HPA response during exposure to the stressful stimulus. This phenomenon has been well studied in the context of mother–offspring interactions. In some species, offspring that are exposed to stressful situations in the presence of the mother display significantly reduced behavioral and glucocorticoid responses to those stressors, relative to offspring exposed to the same stressors in the absence of the mother (see review in Hennessy, 1997). However, this effect of social support is not just limited to mother–infant dyads. Social buffering of behavioral and physiological responses to stress has also been reported in studies involving separation in adult nonhuman primates (e.g., heterosexual pairs: Mendoza and Mason, 1986, Smith et al., 1998; same-sex pairs: Gust et al., 1994).

The physical presence of a social partner is defined by an amalgamation of several distinct stimulus attributes (e.g., visual, olfactory, and auditory signals) as well as by the nature and pattern of social interactions. Presumably, the interactions of these signals function to provide a representation of one individual (the sender) and perhaps the relationship it represents to another individual (the receiver). In the absence of one or more of these signals, as often occurs in visually obscured environments, isolated individuals may attempt to maintain the positive effects of social support by using a single individually specific, or signature, signal rather than having to rely on a combination of several signals. Signature signals have been identified across a wide variety of birds and mammals (e.g., Charrier et al., 2003, Randall, 1989, Sayigh et al., 1998, Searby et al., 2004). Individually distinct communication signals are also prevalent among many nonhuman primates (e.g., Cleveland and Snowdon, 1982, Hammerschmidt and Todt, 1995, Jones et al., 1993, Jorgensen and French, 1998, Snowdon and Cleveland, 1980, Snowdon et al., 1983, Symmes et al., 1979). Therefore, if the physical presence of a social partner can moderate responses (i.e., reduce HPA axis activity) to stressful stimuli, it may be that, in addition to identity, these signature signals also communicate the beneficial effects of social support by communicating representations of individuals and the relationship they represent, in turn moderating the consequences of exposure to stressful events.

Marmosets, small tropical primates from the New World family Callitrichidae, are characterized by strong socioemotional attachments (“pair bonds”) between adult males and females, as well as cooperative infant care and prolonged residence of offspring in extended family groups (Rylands, 1993). Thus, social interactions are an important attribute of callitrichid life, and the resulting social relationships have profound impacts on marmoset physiology and behavior including reproduction (Ginther et al., 2001, Smith et al., 1997, Ziegler and Sousa, 2002), endocrine regulation (Shepherd and French, 1999, Smith and French, 1997, Smith et al., 1998), and communication (Elowson and Snowdon, 1994, Rukstalis et al., 2003, Snowdon and Elowson, 1999, Snowdon and Hodun, 1981, Vitale et al., 2003). In addition to these important social characteristics, marmosets also possess rich vocal repertoires that are used in a wide variety of contexts, including intragroup cohesion and maintenance of territories (Cleveland and Snowdon, 1982, Epple, 1968, Heymann, 1987, Hubrecht, 1985, Stevenson and Poole, 1976). These vocalizations are also known to have “signature” or individually distinct properties which can be used to distinguish individuals among group members (Jones et al., 1993, Jorgensen and French, 1998, Snowdon and Cleveland, 1980). In addition to strong attachments and signature vocalizations, marmosets have also been shown to experience social buffering of the stress response. Smith et al. (1998) demonstrated that marmosets removed from their home cage and exposed to a novel environment had significantly lower levels of urinary cortisol when their heterosexual pair mate was present versus when they were not. Therefore, the existence of complex social interactions, highly conserved, individually identifiable vocalizations, and evidence of social buffering make callitrichid primates ideally suited for examining the effects of individually specific signals on physiological responses to stressful events.

The purpose of the present study was to evaluate whether exposure to a single, individually specific, signal (i.e., phee call) could reduce the consequences of social separation and exposure to environmental novelty. Specifically, we predicted that exposure to the phee call of a significant social partner (i.e., pair mate) would reduce urinary cortisol excretion, a common indicator of psychosocial stress, in marmosets isolated and exposed to a novel environment. Additionally, it may be that exposure to vocalizations from individuals other than the focal animal's pair mate may also influence the physiological consequences of exposure to stressful events. However, given the individually specific nature of marmoset vocalizations and the long-term attachments formed between heterosexual pairs, we predicted that exposure to vocalizations from unfamiliar individuals would not reduce urinary cortisol production in socially isolated individuals exposed to novel environments. We also examined the effect of length of pairing and social proximity, if any, on urinary cortisol excretion upon exposure to a pair mate's vocalization.