Colours influence numerous aspects of an animal’s ecology and the adaptive significance of colour variation has been intensively studied in diverse taxonomic groups. This study was motivated by the question: Why do garter snakes vary in colour? To answer this question, I focused on Western Terrestrial Garter Snakes, Thamnophis elegans, which exhibit geographic variation in colouration (dark morph vs. light morph), and two different species of garter snake that occur in the same geographic region but vary in colour –light T. elegans and the Common Garter Snake, T. sirtalis. My work provides an objective quantification and analysis of snake colour and evaluates the influence of colour variation on ecological processes such as thermoregulation, crypsis, and antipredator behaviour. I compared body temperatures (Tbs) with available thermal opportunities, both in wild snakes and in a field experiment, to determine how snakes with contrasting colours differ in thermoregulation and temperature-dependent behaviours. Gravid females of the light and dark colour morphs of T. elegans exhibited comparable thermoregulatory behaviour at high temperatures; however, dark T. elegans maintained elevated Tbs when available temperatures dropped. In the field, dark-coloured snakes were more likely to be moving when first detected when Tbs were high, but this trend was reversed in light T. elegans. I quantified crypsis of snakes, in terms of colour and brightness, by measuring the spectral reflectance of snakes and the surrounding habitat. These data were visually modeled from the perspective of potential snake predators and human researchers. Overall, snakes selected basking sites that maximized crypsis and both colour morphs of T. elegans were equally cryptic. There was evidence suggesting that T. sirtalis was more cryptic than light T. elegans to snake predators. I collected a series of behavioural measurements for snakes pre- and post-capture. Light T. elegans were more likely to be moving when originally detected in the field than dark snakes. Distance to cover and injuries were important factors in explaining the antipredator behaviour of snakes in the field. Snakes became generally faster with increasing Tbs, but differences attributable to colour morph were not straightforward. A higher proportion of T. elegans of both colour morphs exhibited some type of movement when exposed to a simulated predatory attack. Thamnophis sirtalis, on the other hand, hung limp and motionless in the air. The less cryptic light T. elegans had a higher probability of having an injury than T. sirtalis but injury patterns between the equally cryptic light and dark T. elegans differed by sex. The relationships between colour and these various traits were complex, but, taken together, they highlighted how thermal ecology, crypsis, and anti-predator behaviours were related to a snake’s visual appearance. These results therefore provide an ecological underpinning for future genetic studies to identify potential candidate genes that may be responsible for the control of colour pattern in garter snakes. / Graduate / 0329 / 0306 / 0472

With reference to anatomical studies it appears that the aerpents are indeed a neglected group ot reptiles. Although some early writers recorded a few fragmentary obser-vations on the anatomy of snakes, no attempt was made to undertake a complete comprehensive study. Cole (1944) has rightfully referred to this section of vertebrate anatomy as "almost virgin field." The serpent area of reptilian anatomy is deserving ot a more intensive investigation because of the considerable number ot adaptations demonatrated by this group of reptiles in relation to their peculiar types ot locomotion and methods ot feeding. It is important to understand the anatomical specializations of this group of animals if we are to understand more completely the evolutionary trends among the vertebrates. It is also quite evident that compara-tive descriptions of certain serpents, which have been based mainly on external characteristics, need to be supplemented by accounts of their internal anatomy.

Environmental cues such as day length and temperature contribute to timing of biological rhythms in seasonal breeders. Life-history transitions such as spring emergence from hibernation, migration, or mating must be coordinated with environmental conditions or survival is compromised. Therefore, there must be chemical signaling pathways in the brain that transduce seasonally-changing sensory inputs into signals that initiate a hormonal cascade, culminating in reproductive behavior. The relative importance of environmental cues to reproductive timing varies with species, time of year, and sex, and the mechanisms driving these differences remain unknown. The role of photoperiod in regulating reproductive behavior has been explored in birds and mammals, but much less is known about the role of so-called supplementary cues such as temperature, which is crucial in the timing of ectotherm reproduction. This is a critical gap in our knowledge, because shifts in seasonal temperatures due to climate change could create a mismatch between peak reproductive behavior and resources necessary for gestation and offspring survival. Deiodinase 2 (DIO2) enzyme is a critical component of the pathway that mediates reproduction in photoperiod-activated seasonal breeders, but whether deiodinase 2 is sensitive to seasonal changes in environmental temperature is unknown. In this study, we used an ectothermic vertebrate known to be a temperature-activated seasonal breeder, the red-sided garter snake (Thamnophis sirtalis parietalis), to investigate changes in hypothalamic DIO2 in response to hibernation at 4°C and 12°C. We captured male and female snakes in Manitoba, Canada as they returned to their winter den site from summer feeding grounds. Snakes were hibernated in complete darkness at either 4°C or 12°C for up to 16 weeks. A subset of each sex and temperature group were euthanized at intervals, and the brains collected and processed for DIO2 immunohistochemistry. DIO2-specific staining was found in the anterior hypothalamus, in the periventricular hypothalamic nucleus and ventral pre-optic area, along the longitudinally central region of the olfactory tract, in the bed nucleus of the stria terminalis, caudally in the cortex and optic tectum, and in the lateral septal nucleus. DIO2-stained area in the anterior hypothalamus was quantified. Male T. sirtalis in both the 4°C and 12°C groups were found to have an increase in DIO2-specific staining in the anterior hypothalamus after 8 weeks in hibernation. Female T. sirtalis were found to have an increase in DIO2-specific staining in the anterior hypothalamus after 8 weeks in the 12°C group only. These findings shed light on the neuroendocrine pathway through which environmental cues other than photoperiod influence the timing of seasonal reproduction, and support the hypothesis that at least some components of this pathway are conserved across seasonal breeders.

Resource availability follows seasonal cycles in environmental conditions. To align physiology and behavior with prevailing environmental conditions, seasonal animals integrate cues from the environment with their internal state. One of the systems animals use to integrate those cues is the hypothalamus-pituitary-adrenal (HPA) axis and its primary effector, glucocorticoid hormones. The HPA axis has wide-ranging effects on physiology and behavior and, in the context of a glucocorticoid stress response, is known to mediate tradeoffs between immediate survival and future fitness. The HPA axis also plays an important role in facilitating predictable life-history events. Variation in HPA axis activity has been reported in all vertebrates, often coordinating seasonal reproduction and possibly also transitions between life-history stages. My dissertation research used red-sided garter snakes (Thamnophis sirtalis parietalis) to examine the role of the HPA axis in regulating seasonal life-history transitions, especially in females. In Chapter 2, I hypothesized that seasonal plasticity in stress responses is regulated, in part, by changes in the responsiveness of the adrenal glands to adrenocorticotropic hormone (ACTH). I found that glucocorticoid responses to ACTH challenge were smaller in males than in females during the spring, suggesting that reports of reduced stress responsiveness in males may reflect lower adrenal responsiveness to ACTH. The sex difference in mating season duration and consequently also in the timing of migration led me to hypothesize that sex differences in HPA axis activity could explain sex differences in the timing of migration. Furthermore, adrenal responsiveness to ACTH also varied seasonally in males, but not females, suggesting that female stress responses, which have not been studied, may not vary seasonally. In Chapter 3, I investigated potential seasonal variation in female stress responses, which have not previously been examined. In males, baseline glucocorticoids decrease over the course of the mating season resulting in significantly lower baseline levels in males that have begun to migrate. I hypothesized that a change in HPA axis activity occurs during spring and fall migration. Peak stress-induced glucocorticoid concentration occurred at an earlier sampling time in females during the spring compared to the fall. Peak stress-induced glucocorticoid concentrations also occurred at a later sampling time in migrating females than in pre-migratory females during the spring, suggesting that negative feedback regulation of the HPA axis changes as soon as females begin to migrate during the spring. Female red-sided garter snakes are biennial breeders that give birth approximately every other year implying that a female's recent reproductive history can influence whether or not she will reproduce in a given year. Body condition can be used as a proxy for recent reproductive history and can be related to baseline and stress-induced glucocorticoid concentrations. In Chapter 4, I hypothesized that hormonal and behavioral stress responses vary with body condition. Baseline glucocorticoids did not vary with body condition, but females in low body condition showed a significantly larger increase in plasma glucocorticoids in response to capture stress. Body condition, but not capture stress, influenced latency to copulate, suggesting that females are resistant to the behavioral effects of capture stress during the spring mating season. Only females in low body condition increased latency to copulate in response to injection of a physiological (15 µg) dose of exogenous CORT, while all females responded to a pharmacological (60 µg) dose, indicating that behavioral responses to exogenous glucocorticoids vary with female body condition. These data suggest that variation in body condition may be associated with differences in HPA axis sensitivity and/or glucocorticoid receptor (GR) density in the brain. I directly tested if there is a relationship among body condition, reproductive history and HPA axis activity in Chapter 5. I found that glucocorticoid stress responses and mating behavior did not vary with body condition, nor was body condition related to brain GR or reproductive condition (parturient vs post-parturient females). Only unreceptive females showed a significant stress-induced increase in glucocorticoids, suggesting that reduced stress responsiveness is associated with receptivity. Parturient females mated faster (were more proceptive) than post-parturient females. These data suggest that HPA axis activity influences reproductive "decisions" by modulating receptivity, while proceptivity is related primarily to recent reproductive history. Together, these chapters help characterize how HPA axis activity varies with season, sex, reproductive history and migration status. By systematically probing the HPA axis in a single, tractable system, I have gained insight into how changes in the HPA axis support and modulate transitions between life-history stages. These results highlight the HPA axis' important function in mediating the critical trade-offs all animals must navigate to be successful in a changing world.

Graduation date: 2006 / Most vertebrates exhibit seasonality in many life history traits. Such seasonal rhythms are temporally organized via the transduction of environmental cues (e.g., photoperiod, temperature) into appropriate endocrine signals. However, among ectothermic vertebrates that undergo continuous winter dormancy, temperature is the only environmental cue available for synchronizing seasonal rhythms. Most intriguing is that in species where reproduction occurs immediately following spring emergence, the associated changes in neurophysiology and behavior that accompany reproduction likely occur during winter dormancy. The purpose of this dissertation research was to explore the mechanisms by which temperature cues affect the chronobiology and seasonal reproduction of red-sided garter snakes (Thamnophis sirtalis parietalis). Because of their roles in circadian organization and energy balance, melatonin and corticosterone are likely hormonal components of these time-keeping systems. I first characterized the interactions between melatonin and corticosterone to better understand the hormonal mechanisms facilitating temperature-induced reproduction. Melatonin and corticosterone additively inhibit reproductive behavior during the spring mating season. Experimental manipulations with a serotonin receptor antagonist suggest the mechanism underlying these effects involves a serotonin-regulated system. Although melatonin does not influence corticosterone responses to capture stress, capture stress significantly increases melatonin concentrations. To investigate the functional significance of these interactions in regulating temperature-induced reproduction, I measured body temperatures of snakes as well as circadian melatonin and corticosterone cycles during winter dormancy and spring emergence using a combination of field and laboratory experiments. Surprisingly, an increase in body temperature is not necessary for emergence from winter dormancy. Rather, critically low temperatures may serve as a zeitgeber entraining an endogenous circannual cycle that regulates emergence. Decreased environmental temperatures, in the absence of changing photoperiod cues, modulate circadian melatonin and corticosterone rhythms during hibernation. Such temperature-induced changes in hormone rhythms may facilitate seasonal reproductive behavior following spring emergence. Furthermore, a phase-shift in corticosterone rhythms during the mating season may regulate the seasonal transition between reproductive and non-reproductive states in red-sided garter snakes. Such studies investigating the environmental and hormonal mechanisms underlying time-keeping systems may provide valuable insight into the potential impact of environmental perturbations (e.g., climate change) on seasonal rhythms in physiology and behavior.

Movements of animals presumably reflect their changing needs and the changing availability of necessary resources. In cold climates, snakes often make long seasonal migrations between hibernacula and summer habitats, Finding suitably deep hibernacula with minimal delay could be critical. I hypothesized that such animals should have highly developed navigational ability. By contrast, snakes living in mild climates can hibernate in shallow sites and probably do not migrate; if so, they might be expected to show poorly developed orientation mechanisms!. The objectives of this study were to determine movement patterns and navigational ability of garter snakes (Thamnophis) living in a mild climate and compare them with a congeneric population known to be migratory. From 1986 - 1988 I examined, using mark-recapture, movement behaviour of two populations of garter snakes at Spectacle Lake Provincial Park (SLPP) on Vancouver Island, British Columbia, focusing on several components typically associated with migratory behaviour - distances travelled, population directionality, fidelity to seasonal sites, distinctness of seasonal habitats, and route directness. Thamnophis sirtalis, the common garter snake, is the most widely distributed North American snake species and high latitude populations are migratory. Thamnophis ordinoides, the northwestern garter snake, is restricted to the Pacific northwest and migratory behaviour has never been reported. Both species displayed combinations of traits clearly suggesting nonmigratory behaviour. These included short-distance (< 500 m), random movements, a lack of den fidelity, and variation in the maintenance of specific home ranges between successive years. Home ranges overlapped between individuals, averaged less than 0.3 ha measured over a single active season, and were not clearly distinct from denning areas. Although some directionality of movement was evident, it was likely related to foraging strategy and unlike the typical unidirectional movements undertaken by migrating snakes, Sexual and reproductive differences in any of these traits generally were not significant. The navigational abilities of a migratory population of T. sirtalis from Wood Buffalo National Park (WBNP) in northern Alberta were examined as were those of the nonmigratory populations of snakes from SLPP. Displacement studies were carried out during the active seasons of 1986 - 1988 to determine the level of orientational abilities present in each population and to examine potential orientation cues. Snakes were displaced from their home range and tested in an arena under a variety of conditions, The results demonstrated that T. sirtalis from both SLPP and WBNP possessed advanced navigational abilities. Advanced skills may be absent in T. ordinoides. Thamnophis sirtalis at both study sites demonstrated time-compensated solar orientation as determined by 6 hr phase-delayed tests. Pheromone trails produced by recently copulated females (but not unmated females) also provided an orientation guide for displaced WBNP males, but results from SLPP were less conclusive. Thamnophis ordinoides did not respond in a discernible way to either cue. Navigational skills thus vary relatively little between migrating and nonmigrating populations of the same species but may be poorly developed in completely nonmigratory species. / Graduate

Natural history is a cornerstone of both theoretical and applied ecology, and provides the context for comparative and experimental studies. I studied the natural history of the Plains Garter Snake (Thamnophis radix) near the northern limit of its range in Miquelon Lake Provincial Park. Alberta, Canada. As is typical of garter snakes, T. radix at Miquelon were sexually dimorphic as adults - females grew faster, reached larger sizes and had relatively shorter tails than males. Mating occurred in spring and females gave birth in August; litter size ranged from 8-50 and was correlated with size of female. These snakes were generalist predators, but fed mainly on anurans when small, adding mammals and other prey as they grew. Despite sexual size dimorphism, I found no difference in diet between adult males and females. Feeding frequency was low overall (22%). especially in gravid females. In summer, snakes were found mainly near cover on south-facing field edges or in marshes, but not in forests, and moved relatively little. However, movement distances increased in fall as snakes sought hibernating sites. Snakes hibernated for up to eight months, either individually or in pairs, in rodent burrows. This study also revealed the importance of radiotelemetry when studying movements of small, secretive animals, as there were significant differences between the results obtained through opportunistic captures versus telemetry: habitat descriptions from opportunistic captures therefore were biased, although they still captured some elements of microhabitat use by T. radix. The data from this study will be important for comparative studies of the ecology of garter snakes and for the management of this species on the human-altered landscape surrounding Miquelon Lake.

Mechanisms of how red-sided garter snakes (Thamnophis sirtalis parietalis) travel up to 18 km from summer feeding sites to hibernation dens are not understood. In this study, monthly and seasonal trailing behavior were investigated to determine whether red-sided garter snakes prefer to follow trails of snakes from the same den (den-mates) versus trails made by other conspecifics (non-den-mates). Snakes from five different hibernacula in Manitoba, Canada, were involved in the study. Eighteen were adults and 15 were subadults. Subadult red-sided garter snakes do not return to hibernacula until their second year of life, so their trailing behavior was of interest in learning how they might first find hibernation sites.

Many animals exhibit seasonal changes in life-history stages, and these seasonal transitions are often accompanied by dramatic switches in behavior. While the neuroendocrine mechanisms that regulate such behavioral transitions are poorly understood, arginine vasotocin (AVT) and neuropeptide Y (NPY) are excellent candidates because they regulate reproductive and feeding behavior, respectively. In this study, I asked if seasonal changes in AVT and/or NPY are concomitant with spring migration away from the breeding grounds, as male and female red-sided garter snakes (Thamnophis sirtalis parietalis) are transitioning from reproductive to non-reproductive behavior during this time. To address this question, I collected male and female snakes in different migratory stages during the spring and fall. Brains were processed for AVT and NPY immunohistochemistry and the total number of immunoreactive (-ir) cells quantified for each individual. As predicted, males had significantly more AVT-ir cells in the preoptic area and bed nucleus of the stria terminalis, brain regions important for courtship behavior, during the spring mating season compared to the fall. Females had significantly more AVT-ir cells in the preoptic area during the spring compared to the fall and, surprisingly, did not exhibit seasonal changes in NPY. In contrast, males had significantly more NPY-ir cells in the cortex, a region important for spatial memory, and in the posterior hypothalamus during the fall compared to the spring, which likely reflects increased feeding behavior during the summer foraging period. Neither AVT- nor NPY-ir cell number varied significantly with migratory status, indicating that seasonal changes in these neuropeptides are not directly related to migration. I then asked if the observed seasonal changes in AVT and NPY in males and females are related to the transition from reproductive to non-reproductive states. Compared to courting males, non-courting males had significantly more AVT-ir cells in the supraoptic nucleus and more NPY-ir cells in the cortex. AVT- and NPY-ir cells did not differ between unmated and mated females. Collectively, my results suggest that AVT and NPY play a role in regulating seasonal transitions in male reproductive behavior, rather than regulating migration per se. Further, these data indicate that both AVT and NPY are regulating reproductive behavior differently in males versus females. These data provide the framework for future studies examining the mechanisms regulating transitions between reproductive, migratory and foraging behaviors.

While nocturnal basal snakes have rod-dominant retinae, diurnal garter snakes have all-cone retinae. Previous work from the Chang lab identified three visual pigments expressed in the photoreceptors of Thamnophis proximus: SWS1, LWS and RH1. I further characterized T. proximus photoreceptors using electron microscopy, immunohistochemistry, and in vitro protein expression. T. proximus have four types of morphological cones: double cones, large single cones, small single cones, and very small single cones. Some small single cones have rod-like features, such as rod-like outer-segment membranes and a lack of micro-droplets. Immunohistochemistry showed that rod-specific transducin is expressed in some T. proximus photoreceptors. In vitro expression of T. proximus RH1 produced a functional rhodopsin with λmax at 485nm, which corresponds to microspectrophotometry measurement from some small single cones. Current results suggest that small single cones of T. proximus may have evolved from ancestral rods, and secondarily acquired a cone-like morphology as adaptation to diurnality.