ABSTRACT
Urban wildlife often suffer poorer health than their counterparts living in more pristine environments due to exposure to anthropogenic stressors such as habitat degradation and environmental contamination. As a result, the health of urban versus nonurban snakes might be assessed by differences in their plasma biochemistries. We compared the plasma profiles of western tiger snakes (Notechis scutatus occidentalis) from a heavily urbanized wetland and a natural, nonurbanized wetland. Despite the urbanized snakes having lower body mass index, we found no significant difference between the plasma profiles of the two populations. We collected snakes from each population and kept them in captivity for 6 mo, providing them with stable conditions, uncontaminated (exempt from heavy metals and pesticides) food and water, and lowered parasite intensity in an attempt to promote better health through depuration. After captivity, snakes experienced a significant improvement in body mass index and significant changes in their plasma profiles. Snakes from the natural wetland initially had more variation of DNA damage; mean concentration of DNA damage in all snakes slightly decreased, but not significantly, after captivity. We present the plasma biochemistry profiles from western tiger snakes both before and after captivity and suggest a period of removal from natural stressors via captivity may offer a more reliable result of how plasma profiles of healthy animals might appear.
INTRODUCTION
Assessing and monitoring health and disease in reptiles is a complex and often difficult process, particularly if there are no health parameter baseline references for healthy individuals. Measurable signs of unhealthy individuals such as low body condition scores, wounds, infections, and high parasite burdens can be obvious; however, as reptiles often do not show physiological changes until very late stages of many ailments (Selleri and Hernandez-Divers 2006), a normal-appearing reptile may not be reflecting its current health condition. As an alternative, plasma biochemistry profiles can provide an important snapshot of the physiologic health of free-living individuals (Eatwell et al. 2014) because analytes and biomarker measurements can be linked to disease (Jacobson et al. 1991), tissue damage (Campbell 2006), and contaminant toxicity (Komoroske et al. 2011; Koch and Hill 2017; Villa et al. 2017) before physical signs can be detected. However, the plasma profile reference profiles for healthy individuals must first be determined for each species and, for reptiles, there are few baseline measurements available compared to other taxa (Campbell 2006). In addition, developing a reference interval is more difficult for reptiles than for other taxa because reptilian plasma analytes can be influenced by physiologic traits such as age, sex, feeding, reproductive status (Coz-Rakovac et al. 2011), and environmental conditions; for example, season, and toxin exposure (Arthur et al. 2008).
Many ecosystems around the world are increasingly affected by the rapid expansion of urbanization. Nevertheless, many wildlife populations persevere or thrive in urban environments (Luniak 2004; Bateman and Fleming 2012). Generalist and opportunistic species usually adjust to or benefit from disturbed environments, while species with more specific ecologic requirements merely persist within suitable remnant fragmented habitats (French et al. 2018); however, urbanization is generally regarded as detrimental to wildlife health (Murray et al. 2019). Common stressors to which urban wildlife are exposed include noise, human and machine disturbance, light pollution, and toxic contaminants (French et al. 2018). Exposure to any or all these stressors can have a significant impact on the health of urban wildlife (Liker et al. 2008; Murray et al. 2019; Winchell et al. 2019). Identifying and assessing stressors and understanding their impacts on wildlife health can be a difficult task, yet this is crucial for conservation and management of urban wildlife.
The Australian tiger snake (Notechis scutatus) is a 1-m elapid commonly found occupying wetlands and wet forests in cool climate areas with high rainfall. Populations currently persist in many urban wetlands, including those in Perth, Western Australia (Lettoof et al. 2020a) and Melbourne, Victoria (Butler et al. 2005). Western tiger snakes (Notechis scutatus coccidentalis) have received considerable research attention focusing on evolutionary plasticity and behavior (Bonnet et al. 2002; Aubret et al. 2011; Aubret 2015) and several studies alluding to signs of Perth urban tiger snakes suffering poor health. For example, snakes in a highly urbanized wetland (Herdsman Lake, central Perth) have a high proportion and degree of tail loss and injury (Aubret 2005; Aubret et al. 2005). Urban tiger snakes in Perth are accumulating a suite of heavy metals and anticoagulant rodenticide (Lettoof et al. 2020a, b). We hypothesized that tiger snakes in urban wetlands suffer poorer health than do tiger snakes in natural wetlands and predicted that this would be reflected in their plasma biochemistry. By removing snakes from these wetlands, holding them over a period of captivity with fresh water, uncontaminated food, and worming treatment, and measuring their plasma profiles before and after captivity, we aimed to determine the plasma profiles of healthy tiger snakes. Our objectives were to compare the differences in plasma biochemistry profiles between tiger snakes from a highly urbanized wetland and from a natural wetland to measure the change in plasma profiles after a period of captivity and to explore the plasma biochemistry profile changes that may reflect healthier tiger snakes.
MATERIALS AND METHODS
Field collection and sites
In October 2018, 20 adult (i.e., >530 mm snout-vent length; Shine 1987), tiger snakes were hand collected from two sites in the greater Perth region of Western Australia; 10 (seven males and three females) from the heavily urbanized wetland Herdsman Lake (HL; 31°55′12″S, 115°4819″E) and 10 (six males and four females) from Loch McNess (YC; 31°32′44″S, 115°40′50″E), a wetland beyond the fringe of urban Perth and located within Yanchep National Park. These sites were selected based on their similarity of structure and proximity to each other, yet are characterized by very different degrees of urbanization. The sites also differ in sediment contamination of metals. Lettoof et al. (2020a) report sediment concentrations of arsenic, copper, lead, and zinc in HL exceeding trigger values of the Australian and New Zealand guidelines and the revised Australia and New Zealand Environment and Conservation Council and Agriculture and Resource Management Council of Australia and New Zealand (ANZECC and ARMCANZ 2000) and the ANZECC and ARMCANZ Sediment Quality Guidelines (Simpson et al. 2013); mercury and selenium levels from YC were also found to exceed trigger values. Tiger snakes from HL have higher liver concentrations of antimony, arsenic, barium, chromium, cobalt, molybdenum, and silver, whereas YC snakes had higher cadmium, copper, lead, mercury, and selenium (Lettoof et al. 2020a).
Data collection
On the day of capture, body weight, snout-vent length (SVL), and tail length were recorded for each snake. An external examination was conducted for four types of parasites (ticks, skin worms, gastric nematodes, and oral trematodes) and for injuries and wounds, and the stomach was palpated for detection of food. On the day of capture, a blood sample was taken from the heart using a lithium heparin-coated 23-ga needle, placed in a heparinized tube, and centrifuged within 30 min of extraction at 2,000 × G for 10 min at room temperature. Plasma was separated into three vials per snake, rapidly frozen in liquid nitrogen, and stored at –80 C until analyzed. The first blood sample was taken within 6 h of snake capture to best represent wild plasma profiles, and the second sample was taken after 6 mo of captivity. To reduce influences of feeding on plasma analyte concentrations, snakes were fasted for 2 wk before the second blood sampling. Body weight was recorded monthly at 2 wk after feeding.
Snake husbandry
Snakes were housed individually in locked plastic tubs (70×50×40 cm) and provided with a water bowl, plastic hide, and shredded hemp as bedding. Water and cleaning were provided ad libitum but a maximum of once a week. Smaller snakes were offered a thawed adult mouse once every 2 wk and once a week for larger snakes. The majority of snakes ate every feeding day. The holding laboratory was temperature controlled, ranging from 16–20 C, and a looped strip of 32 C heat cord passed beneath one end of each tub, allowing for thermoregulation. The room was lit with natural light to allow natural periods of activity. Snakes were collected and held from October to April reflecting their peak seasonal activity. Animal care and sample collection was approved by the Animal Ethics Committee at Curtin University (no. ARE2018-23) and the Department of Biodiversity, Conservation and Attractions (no. 08-002624-02), Kensington, Western Australia.
Deworming treatment
We treated all snakes to a deworming regime because western tiger snakes are frequently infected with the gastric nematode Ophidascaris pyrrhus (Lettoof et al. 2020c). Fenbendazole (Panacur 100, Intervet Australia, Bendigo East, Victoria, Australia) was administered orally for three monthly treatments increasing in dosage (10, 20, 40 mg/kg) and each treatment consisted of four dosages, one dose every 4 d. We collected feces before and after treatment and conducted standardized fecal floats, but no eggs were detected.
Plasma biochemistry profile, oxidative DNA damage, and HSP70
The plasma concentration of aspartate aminotransferase (AST), bile acids (BA), creatine kinase (CK), uric acid (UA), glucose (GLU), total calcium (CA), phosphorus (PHOS), total protein (TP), albumin (ALB), globulin (GLOB), potassium (K), and sodium (NA) were measured using VetScan Avian/Reptilian Profile Plus reagent rotor in the VetScan Whole Blood Analyzer (Abaxis, Inc., Union City, California, USA). The coefficients of variation for each analyte using the VetScan Avian/Reptilian Profile Plus reagent rotor are: AST (3.6–4.3%), BA (4.5–4.9%), CK (3.6–6.0%), UA (3.9–4.8%), GLU (1.4–1.6%), CA (2.9–3.4%), PHOS (2.6–4.9%), TP (1.2–1.9%), ALB (3.6–4.3%), GLOB (3.5–4.4%), K (5.7–6.3%), and NA (1.6–1.8%).
Measures of DNA damage have been frequently used as a biomarker of metal-induced toxicity and carcinogenesis in organisms (Kasprzak 2002; Simonyan et al. 2018; Finlayson et al. 2019). The by-product, 8-hydroxy-20-deoxyguanosine (8-oxo-dG), is the most studied and detected indicator of oxidative DNA damage (Olsson et al. 2012) and is formed during DNA replication (Haghdoost et al. 2005). We measured the concentration of 8-oxo-dG in tiger snake plasma before and after captivity using the HT 8-oxo-dg ELISA Kit II (Trevigen, Gaithersburg, Maryland, USA). The analysis was performed as outlined by the manufacturer.
The stress protein HSP70 is a commonly used biomarker of environmentally induced stress (Tsan and Gao 2004). As there are no commercial HSP70 enzyme-linked immunosorbent assay (ELISA) kits available specifically for reptiles, we tested an avian HSP70 ELISA for measurement of HSP70 in tiger snake plasma (Chicken Heat Shock 70 KDa Protein ELISA Kit, Abbexa, Milton, Cambridge, UK). This trial was unsuccessful, however, with only nonspecific binding evident.
Statistical analysis
We compared changes in body condition calculated as body mass index (BMI=[body mass (g)/SVL2 (cm)]×100) of snakes over the period of captivity using a repeated measures analysis of variance (ANOVA) with individual snakes as the random factor (Zipkin et al. 2020). We assigned each snake a single SVL measurement (the mean of their before and after captivity lengths) due to the margin of error when measuring snakes (Rivas et al. 2008).
A principle component analysis (PCA) was undertaken to identify plasma profile differences between site, sex, and captivity status (Lê et al. 2008). The PCA analysis does not allow for missing data and so, for the purposes of statistical analysis, analytes below the limit of detection were assumed to be the value of half the limit of detection, analytes above the detection range were assumed to be the upper detection limit, and individual analytes that were missing (due to instrument error or to death of an individual before completion of the trial) were assumed to be the mean of that analyte for snakes before and after captivity, respectively.
Multiple comparisons of analytes were performed by one-way ANOVA followed by a Tukey's test. Data were considered significantly different with alpha=0.05. To account for variation due to dehydration, 8-oxo-dG concentrations were normalized by total serum protein (Gagnon and Rawson 2016). All statistical analyses were conducted using R statistical software (R Studio version 4.0.2; R Development Core Team 2019).
Detection of outliers was done with Reference Value Advisor, version 2.1 (Geffre et al. 2011). Outliers were identified visually as well as statistically using Tukey methods and the Dixon-Reed test and removed if deemed an aberrant observation (e.g., high AST and CK values due to failed cardiac puncture attempts).
RESULTS
One individual from each site died prior to the conclusion of the experiment. The HL snake was in the poorest condition of all snakes upon capture and rarely ate; it died within 2 mo of captivity. The postmortem examination revealed no obvious pathologies, so we assumed that this snake may have been close to its natural death. The YC snake ate every feeding day and was in good condition for 4 mo. It regurgitated a semidigested mouse 6 d after feeding and died 2 d after that. The postmortem revealed no obvious abnormalities except it had a large nematode burden (n=42). It is possible that the snake's stomach may have been blocked by the dead nematodes after the deworming treatment. Plasma analytes and DNA damage concentrations from these snakes were included in the mean values for before captivity.
Body condition
Snakes from HL were of lower mean (±SD) body condition (BMI 3.82±0.33 SD) than were YC snakes (BMI 4.19±0.69 SD) upon capture. There was no significant difference between the mean body condition of both sexes from each site throughout the experiment (t-test, P=0.975). Snakes increased in body condition significantly over the 6 mo of captivity (ANOVA, P<0.001; Fig. 1), but the interaction between site and month was not significant (P=0.069). At the end of captivity, the mean body condition of HL snakes was similar (BMI 5.05±0.43 SD) to YC snakes (BMI 4.95±1.01 SD). The YC snakes had higher variation in body condition (sample variance 0.86) than did HL snakes (sample variance 0.54) throughout the entire captive period.
The mean body mass index (BMI) change of western tiger snake (Notechis scutatus occidentalis) from two populations over 6 mo of captivity. Month 0=body mass index upon capture. Error bars represent SD. HL=Herdsman Lake (urbanized wetland); YC=Loch McNess, Yanchep National Park (natural wetland).
The mean body mass index (BMI) change of western tiger snake (Notechis scutatus occidentalis) from two populations over 6 mo of captivity. Month 0=body mass index upon capture. Error bars represent SD. HL=Herdsman Lake (urbanized wetland); YC=Loch McNess, Yanchep National Park (natural wetland).
Plasma biochemical profiles
As the concentrations of BA were below the limit of detection for all but two snakes, we did not report them. The PCA revealed two principle component axes, Dim1 and Dim2 (dimension), which account for 53.6% of the total variability. Plasma biochemistry profiles were significantly different for snakes before and after captivity (Dim1 ANOVA, P<0.001), but not between snakes from sites (Dim1 ANOVA, P=0.625) nor between snakes of different sexes (Dim1 ANOVA, P=0.342; Fig. 2a, b, c). Specifically, UA, GLU, and PHOS were significantly lower (P<0.001) after captivity and TP, ALB, and GLOB were significantly higher (P<0.001) after captivity (Fig. 3). The DNA damage biomarker was not significantly different after captivity (ANOVA, P=0.551), with mean 8-oxo-dG concentrations decreasing from 0.83±0.33 SD nmol/g to 0.77±0.35 SD nmol/g precaptivity and post-captivity, respectively. Plasma analyte profiles are presented for both before and after captivity in Table 1.
Principle component analysis of western tiger snake (Notechis scutatus occidentalis) plasma biochemistry profiles grouped by (a) change over captivity, (b) site, and (c) sex. Before=wild snakes upon capture. After=the same snakes after 6 mo of captivity. Snakes are from two populations around Perth, Western Australia: HL=Herdsman Lake (urbanized wetland); YC=Yanchep National Park (natural wetland).
Principle component analysis of western tiger snake (Notechis scutatus occidentalis) plasma biochemistry profiles grouped by (a) change over captivity, (b) site, and (c) sex. Before=wild snakes upon capture. After=the same snakes after 6 mo of captivity. Snakes are from two populations around Perth, Western Australia: HL=Herdsman Lake (urbanized wetland); YC=Yanchep National Park (natural wetland).
Box-and-whisker plots of western tiger snake (Notechis scutatus occidentalis) plasma analytes from HL=Herdsman Lake (urbanized wetland) and YC=Yanchep National Park (natural wetland) before and after 6 mo of captivity. Black horizontal line=median. Analytes with an asterisk indicate significant (P<0.05) changes after captivity. Please defer to the Materials and Methods section for the abbreviated terms.
Box-and-whisker plots of western tiger snake (Notechis scutatus occidentalis) plasma analytes from HL=Herdsman Lake (urbanized wetland) and YC=Yanchep National Park (natural wetland) before and after 6 mo of captivity. Black horizontal line=median. Analytes with an asterisk indicate significant (P<0.05) changes after captivity. Please defer to the Materials and Methods section for the abbreviated terms.
Biochemistry profiles and descriptive statistics of western tiger snakes (Notechis scutatus occidentalis) before and after 6 mo captivity in 2019–20. Snakes are from Herdsman Lake (urbanized wetland) and Yanchep National Park (natural wetland) around Perth, Western Australia. Outliers were removed if n<20 for before captivity and n<18 after captivity.

DISCUSSION
We compared the plasma profiles of wild-caught (before captivity) adult tiger snakes from a heavily urbanized wetland (HL) to those from a natural wetland (YC), with the hypothesis that snakes from the former site would be in poorer health than those from the latter, and that this would be reflected by differences in their plasma profiles. Surprisingly, there was no significant difference in the plasma profiles of tiger snakes between the urbanized wetland (HL) and the natural wetland (YC) or between male and female snakes (Fig. 2b, c), despite snakes from HL being of lower body condition and generally exposed to higher concentrations of metals and anthropogenic disturbance compared to snakes from YC (Lettoof et al. 2020a). Other studies have found strong differences in plasma profiles of organisms exposed to different environmental stressors; for instance, in green turtles (Chelonia mydas) exposed to the toxic cyanobacterium Lyngbya majuscula (Arthur et al. 2008), Pigeon Guillemots (Cepphus columba) from oil-spill areas (Seiser et al. 2000) and Libyan jirds (Meriones libycus) from heavy metal–contaminated urban sites (Adham et al. 2011). Our results suggested either the stress or contaminants accumulated by tiger snakes at HL are not different enough to influence the plasma profiles, or tiger snakes from YC are experiencing similar stress and contaminant accumulation. However, snakes from these sites may suffer from other impacts to their health, such as immunosuppression and aberrant behavior, which are not detected in plasma profiles but warrant further investigation.
Our results also suggested that developing baseline plasma profiles based on wild populations may not accurately represent the plasma profiles of healthy individuals, or the analytes we measured may not be diagnostically sensitive for poor health in tiger snakes, as snakes from HL were noticeably in poorer condition upon capture yet their plasma profiles were similar to YC snakes. Following 6 mo of captivity with minimal stressors, the experimental depuration period resulted in a significant increase in body condition for all snakes (Fig. 1) and a significant change in plasma profiles (Fig. 2a).
By the end of captivity, snakes from both sites were of similar mean body condition. Yanchep snakes had much higher body condition variation throughout the experiment than those from HL, as YC snakes took food infrequently, possibly due to what appeared to be their more-defensive and nervous behavior. The increase in body condition for all snakes was likely influenced by the captive diet. Western tiger snakes primarily eat frogs; because frogs are not commercially available as pet food in Australia, we were restricted to feeding them mice, which provide more calories compared to frogs (Wiseman et al. 2019). Also, snakes from both sites may have been of lower than normal condition and returned to a healthier condition during captivity. Shedding nematodes by worming treatment may have also contributed to an increase in body condition.
Stress can negatively impact the health of snakes. Some stressors of wild snakes include fluctuations in food availability, habitat disturbance, and contamination from urbanization, predation, and conspecific competition (Van Waeyenberge et al. 2018). We removed snakes from these stressors, but captivity can present different stressors such as unfamiliar scents, reduced freedom of movement, and different food. Snakes exhibit acute stress responses from capture and short-term (3 d–8 wk) captivity (Mathies et al. 2001; Sykes and Klukowski 2009), but only one study measured stress in snakes over a longer period. Sparkman et al. (2014) found wild-caught gravid garter snakes (Thamnophis elegans) plasma corticosterone levels and heterophilto-lymphocyte ratios increased over 4 mo of captivity. We had no obviously pregnant females, so these results may not be comparable to our study.
Acute and chronic stress responses are also highly variable between species and individuals (Sparkman et al. 2014; Fischer and Romero 2019). We attempted to reduce acute stress in snakes by minimizing interaction (3 d/wk) and feeding them uncontaminated food and water over the duration of captivity. We consider any potential stress to be chronic. Measuring chronic stress in captive snakes is difficult, but obvious symptoms are behavior changes, increase in lesions and infections, and weight loss (Van Waeyenberge et al. 2018). We did not detect any increase in lesions or infections, and most tiger snakes ate almost immediately when presented with food and increased in weight. Although we attempted to measure stress (HSP70), there was no reactivity between the avian assay and tiger snake plasma; however, we consider the lack of chronic stress symptoms to be an indication that these snakes were less stressed than in the wild.
Baseline healthy plasma profiles are difficult to establish for reptiles due to their variable physiology. The concentration of plasma analytes can be influenced by the season (Machado et al. 2006; Bryant et al. 2012), sex and reproductive status (Christopher et al. 1999; Coz-Rakovac et al. 2011), diet, and time between feeding (Smeller et al. 1978; Stinner and Ely 1993; Moon et al. 1999). We kept tiger snakes in stable conditions for 6 mo to minimize these variables and consider the plasma analyte concentrations after captivity to better reflect those of healthier western tiger snakes freed of multiple environmental stressors.
By the end of this period of captivity, concentrations of TP, ALB, and GLOB in tiger snake plasma significantly increased and UA, GLU, and PHOS concentrations significantly decreased. Total protein, UA, and GLU can be influenced by the diet, hydration, and nutritional status in reptiles (Campbell 2006; Hamilton et al. 2016). We could not determine when the snakes had last eaten when they were sampled before capture, but they were fasted for 2 wk before they were sampled after being placed into captivity, so after captivity values were not influenced by recent feeding. We doubt that an increase in TP was caused from dehydration, as snakes had a constant supply of fresh water, and there was neither a corresponding increase in UA nor electrolytes (NA, K, and PHOS). Rather, TP concentrations may have increased in captivity from a more calorie-rich diet of mice as well as from a lowered parasite intensity. As ALB concentrations generated by the dye-binding methods are inaccurate compared to protein electrophoresis (Muller and Brunnberg 2010), caution must be taken when interpreting reptile ALB for health assessments. Furthermore, as GLOB concentration is calculated from the TP and ALB values, our globulin results were likely influenced by the change in diet and hydration during captivity.
Glucose and UA of tiger snakes before captivity were significantly higher and highly variable compared to those after captivity. As both of these analytes will increase from recent feeding (Smeller et al. 1978; Moon et al. 1999; Lam and Halán 2017) and GLU can increase due to stress (Skoczylas and Sidorkiewicz 1974; Stinner and Ely 1993; Laderberg 2015), it is possible that the high concentrations and variations of these analytes can be attributed to recent feeding and stress of capture. Concentrations of GLU also increase and UA concentration can decrease during the season of peak activity (Coz-Rakovac et al. 2011; Silva et al. 2011). However, a decrease in GLU concentration can also be influenced by high protein diets (Campbell 2006) and lower body temperatures (Laderberg 2015). Plasma PHOS concentrations are often used to test for renal disease and nutritional deficiency in reptiles (Knotek et al. 2003; Campbell 2006); however, the concentrations and the PHOS:CA ratios were not high enough to suggest that either of these conditions occurred in our snakes. The concentration of PHOS can also increase during periods of sexual activity (Coz-Rakovac et al. 2011); the lower PHOS levels after captivity probably reflect both the end of the breeding season and the fasting period.
We could not determine what parameters influenced the snake plasma profiles, but they were likely a combination of fasting, diet, and season. However, with the provision of an uncontaminated diet, deworming treatment, and reduction of stressors throughout captivity we consider the plasma profiles of snakes after captivity to be a better representation of healthier tiger snakes. We recommend plasma profiles be measured in wild western tiger snakes at our sites during similar seasons (October and April) to disentangle the influence of season on analytes. We also recommend measuring plasma profiles of snakes in October from sites with notably fewer stressors to see if plasma profiles are closer to our after-captivity results to establish if these profiles reflect healthier snakes. Our results can be used as baselines for future comparative studies to assess the health of other western tiger snake populations and for veterinary diagnosis of captive snakes.
Surprisingly, snakes from HL expressed lower mean 8-oxo-dG concentrations than did YC snakes upon capture. After captivity, snakes from HL had slightly higher variation in 8-oxo-dG and YC snakes less variation (Fig. 3). A slight, not statistically significant, reduction in mean 8-oxo-dG concentrations for all snakes after captivity suggested that their removal from contaminated environments for 6 mo may have instigated depuration, but the elapsed time was insufficient to see statistically significant changes due to the slow metabolism of reptiles. Measuring oxidative DNA damage is often used as a biomarker for metal toxicity (Collins et al. 1996; Kasprzak 2002), and high variation in biomarkers is often an indicator of contaminant accumulation (van der Oost et al. 2003; Gagnon and Rawson 2016). Despite being exposed to generally lower concentrations of contaminants, snakes from YC were found to have higher liver concentrations of cadmium, copper, lead, mercury, and selenium (Lettoof et al. 2020a), and perhaps these metals are more genotoxic than the metals (antimony, arsenic, barium, chromium, cobalt, molybdenum, and silver) that HL snakes are exposed to. The limited number of studies make it difficult to conclude whether changes in reptilian DNA are influenced by genotoxic agents or are a result of natural phenomena (Novillo et al. 2006), and more research is needed to determine if 8-oxo-dG is a reliable biomarker of contaminant toxicity in reptiles.
ACKNOWLEDGMENTS
We thank the Holsworth Wildlife Research Endowment for providing funding support for this study. We also thank Luke Allen from Venom Supplies for training D.C.L. to take blood via cardiocentesis and for husbandry advice, Kady Grosser, Serin Subaraj, and Ross McGibbon for the assistance with collecting snakes and restraining for blood sampling, and two anonymous reviewers for comments that improved the manuscript.