Studies of birds have greatly advanced our understanding of how testosterone modulates complex phenotypes, specifically its role in mediating male reproductive and associated behaviors. Yet most of the foundational studies have been limited to northern latitude breeding species despite the fact that they represent only a small fraction of worldwide avian diversity. In contrast, phylogenetic, life-history, and mating system diversity all reach their apex in neotropical avifauna and yet these birds, along with more southern latitude species, remain very poorly understood from an endocrine perspective. Despite the relatively limited previous work on taxa breeding in Central and South America, empirical findings have had a disproportionately large impact on our understanding of testosterone's role in everything from geographic variation to behavioral roles and neuroplasticity. Here, we synthesize how studies of neotropical breeding avifauna have advanced our understanding of how testosterone's actions can and are associated with the broad patterns of phenotypic diversity that we see in birds. In addition, we outline how these studies can be used individually or in a comparative context to address fundamental questions about the environmental endocrinology of testosterone and to understand the diversity of roles that testosterone plays in mediating behavioral variation, reproductive strategies, and associated life-history trade-offs.

Altricial young are dependent on adults for protection and food, and they display nutritional need by begging to elicit feeding from parents. Begging at high levels can be energetically expensive and attract predators; thus, an individual must balance its nutritional needs with these potential costs. Further, because a parent is limited in the amount of food it can provide, begging can contribute to both parent-offspring conflict and sibling-sibling competition. Many extrinsic and intrinsic factors may contribute to begging behavior. One intrinsic factor of interest is corticosterone (CORT), a metabolic hormone hypothesized to play a role in regulating a nestling's begging behavior. We investigated the hypothesis that increased exposure to CORT influences nestling begging behavior in an altricial species, the Florida scrub-jay (Aphelocoma coerulescens). We treated one nestling per treatment nest with a twice-daily dose of exogenous hormone via a CORT-injected waxworm, whereas a second individual received a vehicle-injected waxworm. We monitored individual nestling and adult behavior at all nests with the use of high-definition video cameras on several days during treatment. We found no difference in begging rate between CORT fed and vehicle fed nestlings within a treatment nest. Further, to determine whether CORT treatment had indirect effects on the entire brood, we monitored additional nests, in which nestlings were not manipulated. When treatment and controls were compared, overall begging rates of nestlings in treatment nests were greater than those in control nests. This result suggests that CORT treatment of an individual altered its behavior, as well as that of its siblings.

Many tropical birds have slow-paced life history strategies, exhibiting lower metabolic rates, reduced annual investment in reproduction, and longer lifespans relative to birds at higher latitudes. Life history strategies have been relatively well documented in adult individuals in the tropics, but we know comparatively little about the immature life history stage. Here we examine strategies of feather replacement (molt) and fattening in immature Rufous-collared Sparrows (Zonotrichia capensis) in a high elevation equatorial population, following a parallel, previous study on an arctic congener, the White-crowned Sparrow (Zonotrichia leucophrys gambelii). In captivity, Rufouscollared Sparrows incurred energetic costs of experimentally induced feather growth, similar to those previously described for Zonotrichia at higher latitudes. In contrast, free-ranging immature Rufous-collared Sparrows in natural molt had fat stores that declined over time, opposite to patterns evident in arctic Zonotrichia that fatten before migration. Equatorial birds in good condition molted more heavily (controlling for fat stores), suggesting that body condition limits the intensity of molt. Heavily molting equatorial sparrows also had lower amounts of fat (controlling for body condition), suggesting a trade-off between allocation of resources to fat stores versus feather growth. Molt progressed slowly in Rufous-collared Sparrows relative to previously described patterns in their arctic congener, which is concordant with a slower pace-of-life syndrome in tropical, as compared with high latitude, birds.

We studied Florida scrub jay (Aphelocoma coerulescens) nestlings to examine the relationship between parental feeding rates and levels of corticosterone (CORT), a metabolic and stress-related steroid hormone hypothesized to play a role in mediating begging behavior. It has been documented that nutritional deficiency results in increased glucocorticoid levels in nestling birds. Further, previous studies have found that CORT levels of Florida scrub jay nestlings are negatively correlated with parental nest attendance and provisioning rates; however, the behavioral observations were made several days before the collection of samples to assess CORT levels. Few studies have investigated whether experience immediately before sampling impacts nestling glucocorticoid levels, especially in a free-living species. By monitoring parental activity at the nest before sample collection, we found that nestling CORT levels varied as a function of parental provisioning rate and the time since their last feed. However, counter to our predictions, higher provisioning rates and more recent feedings were associated with higher CORT levels in nestlings rather than lower CORT levels. These results suggest that some aspect of parental provisioning results in increased CORT levels in nestling Florida scrub jays.

Plasma glucocorticoid (CORT) levels collected within 3min of capture are commonly believed to reflect pre-stressor, baseline CORT levels. Differences in these "baseline" values are often interpreted as reflecting differences in health, or the amount of social and environmental stress recently experienced by an individual. When interpreting "baseline" values it is generally assumed that any effect of capture-and-handling during the initial sampling period is small enough and consistent enough among individuals to not obscure pre-capture differences in CORT levels. However, plasma CORT increases in less than 3min post-capture in many free-living, endothermic species in which timing has been assessed. In addition, the rate of CORT secretion and the maximum level attained (i.e., the degree of stress-responsiveness) during a severe stressor often differs among individuals of the same species. In Florida scrub-jays (Aphelocoma coerulescens), an individual's stress-responsiveness during a 30min post-capture stressor is correlated with CORT levels in samples collected within 1.5min of capture, suggesting there is an intrinsic connection between stress-responsiveness and pre-capture CORT levels. Although differences in stress-responsiveness accounted for just 11% of the variance in these samples, on average, higher stress-responsive jays (top third of individuals) had baseline values twice that of lower stress-responsive jays (bottom third). Further, plasma CORT levels begin to increase around 2min post-capture in this species, but the rate of increase between 2 and 3min differs markedly with CORT increasing more rapidly in jays with higher stress-responsiveness. Together, these data indicate that baseline CORT values can be influenced by an individual's stress response phenotype and the differences due to stress-responsiveness can be exaggerated during sample collection. In some cases, the effects of differences in stress-responsiveness and the increase in CORT during sample collection could obscure, or supersede, differences in pre-capture plasma CORT levels that are caused by extrinsic factors.

There is increasing evidence that individual differences in the physiological stress response are persistent traits in many animals. To test the hypothesis that the stress-induced CORT (SI-CORT) response is repeatable over the adult life span of Florida scrub-jays (Aphelocoma coerulescens), we sampled 32 male and 25 female free-living scrub-jays (aged 2-13 years) during a 9-year period (2004-2012). Each individual was sampled two to five times and samples were collected one or more years apart during the pre-breeding season (Jan-March). In addition, individuals sampled over the greatest time period (6-8 years) were analyzed separately to more closely assess long-term repeatability. SI-CORT was repeatable in females, but not males, when values were not corrected for confounding variables (agreement repeatability). However, when the year and time of day of sample collection were controlled (adjusted repeatability), SI-CORT was repeatable in both sexes. SI-CORT was also repeatable in the males and females sampled 6-8 years apart. Finally, baseline CORT levels of males, but not females, exhibited low but significant repeatability when adjusted for year. The results of this study demonstrate that differences in SI-CORT levels were repeatable within adult scrub-jays sampled up to 8 years apart. Further, the female SI-CORT response was more consistent between pre-breeding seasons than males, which may have resulted from males having higher SI-CORT plasticity in response to environmental conditions. These data support the hypothesis that the SI-CORT response of Florida scrub-jays develops before adulthood and persists throughout much, if not all, of their natural adult life span.

Intraspecies differences are fundamental to natural selection, yet individual differences in cognition in free-living populations have received little attention. Proactive and reactive coping styles describe individual differences in personality and related stress physiology; however, the coping style model can be extended to include predictions regarding measures of cognition. We compared two measures of personality (neophobia and exploratory behaviour) included in the coping style model to cognitive performance on colour-based associative and reversal learning tests in adult Florida scrub-jays, Aphelocoma coerulescens. Also, as exogenous glucocorticoid treatment can affect cognitive performance, we examined whether an individual's naturally occurring physiological phenotype, reflected by corticosterone measures obtained during development and at the time of the learning tests, covaried with learning performance. Performance on associative and reversal learning tests were inversely related. Scrub-jays with low levels of corticosterone as 11-day-old nestlings performed better on an associative learning test as adults, and there was a marginally nonsignificant trend for nestlings with high levels of corticosterone to perform better on a reversal learning test. There was also a marginally nonsignificant trend for neophobic birds to perform better on reversal learning tests. There were no relationships either between adult stress-induced corticosterone levels and learning, or between exploratory behaviour and learning. Our findings provide evidence that variation in sensitivity to environmental conditions, as reflected by an individual's coping style, underlie the specific strategy by which individuals perform cognitive tasks (i.e. cognitive style). Florida scrub-jays experience a trade-off in performance between types of learning that covary with early corticosterone exposure.

In mid- to high-latitude songbirds, seasonal reproduction is stimulated by increasing day length accompanied by elevated plasma sex steroid levels, increased singing, and growth of the song control nuclei (SCN). Plasticity of the SCN and song behavior are primarily mediated by testosterone (T) and its metabolites in most species studied thus far. However, the majority of bird species are tropical and have less pronounced seasonal reproductive cycles. We have previously documented that equatorial rufous-collared sparrows (Zonotrichia capensis) exhibit seasonal neuroplasticity in the SCN. Manipulating T in these birds, however, did not alter singing behavior. In the current study, we investigated whether T mediates plasticity of the SCN in a similar manner to temperate songbirds. In the first experiment, we treated captive male birds with T or blank implants during the nonbreeding season. In a second experiment, we treated captive male birds with either blank implants, T-filled implants, T with flutamide (FLU; an androgen receptor antagonist) or T with FLU and 1,4,6-androstatriene-3,17-dione (ATD; an estrogen synthesis inhibitor) during the breeding season. In both experiments, the volumes of the brain areas high vocal center (HVC), Area X, and robust nucleus of the arcopallium (RA) were measured along with singing behavior. In summary, T stimulated growth of HVC and RA, and the combined effect of FLU and ATD reversed this effect in HVC. Area X was not affected by T treatment in either experiment. Neither T-treated birds nor controls sang in captivity during either experiment. Together, these data indicate that T mediates seasonal changes in the HVC and RA of both tropical and higher- latitude bird species even if the environmental signals differ. However, unlike most higher-latitude songbirds, we found no evidence that motivation to sing or growth of Area X are stimulated by T under captive conditions.