dc.description.abstracteng | Many group living mammals minimize the costs of sociality by forming stable affiliative
relationships, termed social bonds, with other group members. These bonds are highly
adaptive as they increase fitness, and strongly bonded partners often provide each other with
social support. However, the dynamics of hormones and behavior regulating social bonds
between adults in non-reproductive contexts under natural conditions remain elusive. Three
major neuroendocrine systems regulate animal behavior – the oxytocinergic system, the
hypothalamic-pituitary-gonadal (HPG) axis, and the hypothalamic-pituitary-adrenal (HPA)
axis. Oxytocin is the primary hormone implicated in bond formation in mammals, while
the HPG axis is involved in regulating reproductive behaviors, and the HPA axis regulates
metabolism and helps individuals cope with social challenges. Each system is able to interact
with the other in ways that can fine-tune animal behavior and bonding. The steroid peptide
theory of social bonds (S/P theory) proposes that testosterone mediates a trade-off between
competition and nurturing/bonding. Thus, the S/P theory predicts that the ideal physiological
conditions to induce bond formation is high oxytocin coupled with low testosterone levels.
Additionally, the HPA axis is highly responsive to the state of an individual’s social bonds and
support network. Investigating the interplay between physiology and bond formation under
natural conditions is an important element of linking sociality to fitness. While methodologies
for measuring hormone concentration non-invasively opens up many opportunities to study
this interplay under natural conditions, these methods require validation to ensure that their
measurements are biologically meaningful.
The overarching aim of this thesis was to use the S/P theory as a framework to study the
relationship between oxytocin and testosterone in relation to social bonding between adults
and nurturing male infant care under in natural conditions in male Barbary macaques (Macaca
sylvanus). As previous studies have suggested that oxytocin release after bonding behaviors is
potentially partner specific, I tested for specificity of oxytocin release after two bonding behaviors
– grooming and male-infant-male interactions. Furthermore, I predicted low testosterone to
occur during male bonding and nurturing infant care as the S/P theory predicts that testosterone
mediates a trade-off between competition and nurturing. Simultaneously, I predicted that
cortisol levels would be negatively correlated with male-infant-male interactions, a behavior
that strengthens male bonds and is predictive of coalitionary support. Over the course of my
studies I collected over 4000 hours of behavioral data and 650 urine samples from 14 adult males
Barbary macaques from a semi-free ranging population. Within this dataset, urine samples were
time-matched to social interactions of interest. These events resulted in 152 urine samples for
oxytocin and 650 urine samples for testosterone and cortisol each. I additionally collected 32
urine samples from eight immature males for hormone measurement validation purposes. I conducted a validation of commonly used testosterone and cortisol enzyme immunoassays
(EIA) to assess gonadal and adrenocortical activity from the urine of male Barbary macaques.
Measurements from EIA were compared to those measured by liquid chromatography–mass
spectrometry (LC–MS) to determine whether EIA measurements accurately reflected the
concentration of the target hormones in urine samples. I found that cortisol EIA measurements
from unprocessed urine had a strong positive correlation with measurements from LC–MS
in both adult and immature males. This strong correlation suggests that the cortisol EIA
antibody does not substantially cross-react with other metabolites and is suitable to assess
adrenocortical activity in male Barbary macaques of both age classes. EIA measurements of
testosterone had a strong positive correlation with LC–MS, but only if steroids in the urine were
deconjugated via an enzymatic hydrolysis prior to measurement by EIA in adult males. By
contrast, testosterone EIA and LC–MS measurements of samples from immature males did not
correlate significantly. As expected, testosterone in adult males was significantly higher than
those of immature males when measured by LC–MS but not by EIA, biologically validating
testosterone measurements of LC–MS but not EIA. Collectively, these results suggest that the
testosterone EIAs are suitable to assess gonadal activity in adult but not immature males, and
only if an enzymatic hydrolysis of steroids is performed prior to analysis.
I investigated oxytocin in relation to two bonding behaviors, specifically grooming and
male-infant-male interactions. Oxytocin was higher after grooming with any partner. I
could not formally test whether oxytocin release was partner specific after grooming as most
grooming occurred between bond partners. Nevertheless, this finding is in line with the S/P
theory as well as previous studies that found elevated oxytocin following non-sexual affiliative
behaviors in other species. Naturally elevated oxytocin levels following grooming interactions
did not increase the probability to accept a solicitation for help (from any partner) during
an agonistic conflict. It may be that oxytocin does not universally increase the propensity to
cooperate, although I cannot exclude the possibility that oxytocin may promote cooperation
with bonded partners in within-group conflicts. Following male-infant-male interactions,
oxytocin was only elevated after interactions with non-bond partners but not after interactions
with bond partners. Furthermore, male-infant-male interactions were negatively associated
with testosterone and cortisol. These results support the S/P theory proposing that low
testosterone is beneficial for bonding. Collectively, these results suggest that male-infant-male
interactions serve to create new bonds while not physiologically affecting existing ones. Lastly,
testosterone and cortisol levels were both positively correlated with male infant care, suggesting
that this behavior is conducted under competitive rather than nurturing contexts. As males
preferentially interact with infants based on their past mating success with the mother, infant
care could take a protective role, as with most other Old World primates. Alternatively, infant
care could be considered mating effort as it increases future copulations with the mother.
Over the course of these studies, I have touched upon the core aspects that define the study of
behavioral endocrinology. First, I corroborated previous calls for the need to validate steroid
EIAs for each new species or sample matrix, something that needs to be remembered now that
these methods are more common. Then, collectively, I was the first to investigate the joint levels of oxytocin, testosterone, and cortisol in relation to bonding behavior between same-sex
adults under natural conditions. Overall, my thesis provides general support for the S/P theory
and demonstrates the benefit of investigating several major hormonal axes regulating animal
social behavior, particularly with regards to social bonding. Further studies involving multiple
theoretical frameworks and hormonal axes will help elucidate the dynamics of bond formation
and maintenance between adults. | de |