Species-specific variations in corticosterone concentrations in captive snake shed skins: implications for housing systems and welfare

Abstract

Background

Captive reptiles may experience stress related to environmental conditions, and housing design is considered a key factor influencing their welfare. This study assessed corticosterone concentrations in shed skins as a non-invasive biomarker of chronic stress in three commonly kept snake species—Boa imperator, Python regius, and Pantherophis guttatus—maintained either in low-stimulus rack systems or enriched terraria. Shed skins were collected from privately owned snakes and from a breeders, prepared following standardized protocols, and analyzed using ELISA assays. Non-parametric statistical methods (Mann–Whitney U tests) were applied due to non-normal data distribution.

Results

Variation in corticosterone concentrations was observed among the studied species, with higher values recorded in P. guttatus compared to B. imperator and P. regius. The effects of housing conditions differed between species. In P. regius, individuals kept in enriched terraria exhibited higher corticosterone concentrations than those housed in rack systems, whereas the opposite pattern was observed in P. guttatus, where snakes maintained in rack systems showed elevated corticosterone levels. No clear housing-related differences were observed in B. imperator. Species-specific relationships between enclosure volume and corticosterone concentration were identified, with positive associations in P. regius, negative associations in P. guttatus, and no apparent relationship in B. imperator.

Conclusions

The study indicates interspecific variation in corticosterone deposition in snake shed skin and demonstrates that the relationship between housing conditions and corticosterone levels is species-specific. More active species, such as P. guttatus, may be more sensitive to spatial restriction and limited environmental complexity, whereas less active species, such as P. regius and B. imperator, may respond differently to housing conditions. These findings underscore the importance of species-tailored welfare guidelines, suggesting that enclosure design and enrichment strategies should be adapted to the behavioural ecology and physiological characteristics of individual snake species rather than applied uniformly across taxa.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12917-026-05471-2.

Background

The maintenance of animals under captive conditions has been shown to influence physiological processes, including endocrine regulation associated with stress and energy metabolism. In vertebrates, chronic exposure to less than optimal environmental conditions may alter glucocorticoid secretion, may affect health, behaviour, and overall welfare [1–3]. As a result, physiological indicators are increasingly used to assess the effects of captivity and housing conditions in managed animal populations [1, 4–7]. Reptiles represent a notably difficult group in this context due as a result of their diverse ecological strategies, behavioural patterns, and metabolic adaptations. Compared to mammals and birds, physiological responses of reptiles to captivity remain less well understood, and available data are often limited to a small number of taxa [8–10]. Within reptiles, snakes are especially underrepresented in physiological welfare studies, despite being among the most commonly kept reptiles in captivity [8, 11, 12].

Corticosterone is the primary glucocorticoid hormone in reptiles and plays a central role not only in stress responses but also in the regulation of energy balance and activity [13–15]. Importantly, baseline corticosterone concentrations may vary substantially between species as a function of ecology, behaviour, and metabolic demands [14–16]. This complexity complicates the interpretation of corticosterone levels as direct indicators of stress, particularly when comparisons are made across species with contrasting life-history strategies [13, 17]. Among captive snakes, housing systems differ markedly in environmental complexity and spatial structure. Two systems are especially prevalent: enriched terraria that allow greater environmental heterogeneity and movement, and rack systems noted by reduced space and sensory stimulation. Although both systems are widely used, their physiological implications stay controversial, and existing studies have yielded inconsistent results [9, 18–21]. Moreover, species-specific responses to these housing conditions are rarely addressed [9, 10, 22–25].

Given these knowledge gaps, the present study aimed to investigate associations between housing system (rack systems versus enriched terraria) and corticosterone concentrations in the shed skin of three captive snake species differing in activity patterns and ecological strategies: Boa imperator, Python regius, and Pantherophis guttatus. By focusing on within-species comparisons and considering species-specific biological characteristics, this study tries to add to a more subtle understanding of how housing conditions relate to physiological variation in captive snakes.

Results

Corticosterone concentrations

Variation in corticosterone concentrations was observed among the studied species (Table 1). These values are presented descriptively, as interspecific differences in corticosterone deposition may reflect species-specific physiological features, including variation in baseline hormone levels, metabolic regulation, and sensitivity to captive conditions. Corticosterone concentrations measured in Pantherophis guttatus tended to be higher than those observed in Boa imperator and Python regius, suggesting that behavioural ecology and activity patterns may modulate hormone deposition in shed skin. When considered independently of housing conditions, the documented patterns indicate that corticosterone deposition in shed skin may differ between snake species, which could have implications for understanding species-specific endocrine functioning and physiological responses to environmental challenges.

Table 1.

Descriptive values of the corticosterone in shed skins in tested snakes [ng/g]

	N	Mean ± SE	Range	Median	Q25-Q75
Boa imperator	32	14.68 ± 2.65	0.30–58.62.30.62	8.73	3.03–24.88
Python regius	62	26.64 ± 5.55	0.10–256.45.10.45	8.73	3.17–32.42
Pantherophis guttatus	35	1112.88 ± 271.14	5.50–6427.81.50.81	391.94	135.19–1486.53.19.53

When examining corticosterone levels in the shed skins of the studied snakes in relation to their housing conditions (Fig. 1), it can be observed that, in the case of boas, these levels did not differ significantly; the mean values were comparable among different maintenance regimes (U = 88.00, p = 1.00). Conversely, a distinct pattern was observed in both the python and corn snakes. Studied pythons maintained in enriched environments exhibited significantly higher corticosterone concentrations in shed skins compared to those housed in low-stimulus rack systems (U = 338.00, p = 0.0479). In contrast, for corn snakes, the method of housing demonstrated a statistically significant effect on corticosterone levels, although the pattern was inverse. Specifically, corn snakes kept in rack systems defined by reduced environmental stimuli showed significantly higher corticosterone levels compared to those housed in enriched environments (U = 10.00, p < 0.0001). Effect sizes derived from the standardized test statistic (Z) indicated a negligible effect of housing system in Boa (r = 0.004), a small effect in Python regius (r = 0.25), and a large effect in Pantherophis guttatus (r = 0.80), according to Cohen’s guidelines (small: 0.10; medium: 0.30; large: ≥ 0.50).

Fig. 1.

Comparison of mean corticosterone levels in three studied snake species housed in terrarium and rack system environment (a,b– differs statistically at P < 0.05, A,B - differs statistically at P < 0.001)

Additionally, the correlation analysis between enclosure cubic capacity (measured in cubic meters) and corticosterone concentrations in shed skins revealed species-specific patterns. In pythons, a positive and statistically significant correlation was observed (R = 0.3309; p = 0.009), indicating that larger enclosure volumes are associated with higher corticosterone concentrations in this species. Conversely, for corn snakes, the correlation was negative and statistically significant (R = −0.3434; p = 0.043), suggesting that smaller enclosure sizes are associated with increased corticosterone levels in this species. There was no effect of cage size on corticosterone levels in the studied boa snakes (R=−0.03; p = 0.89).

Discussion

Recent years have seen a growing adoption of non-plasma biological matrices for assessing corticosterone levels [26–28]. Utilizing snake shed skins as research material allows for non-invasive sample collection. Additionally, corticosterone and other stress-related hormones present in keratinized samples reflect hormone deposition over time, offering valuable insights for studying chronic stress. This approach has been tested in reptiles [26, 28, 29], birds [4], fish [5] and mammals [6, 7]. Such methods are particularly relevant for evaluating the impact of husbandry conditions on animal welfare [13].

The present study reveals marked species-specific differences in corticosterone concentrations in snake shed skins. The difference between the corn snake and both the boa and the royal python is substantial, averaging over 2000 ng/g. Considering the taxonomic classification of the studied species and their differing activity levels, it may be inferred that they employ distinct hormonal strategies to cope with life in captivity. Boas and pythons are generally more sedentary animals, typically becoming active only in specific contexts such as foraging or reproduction [30]. Heavy-bodied snake species, which typically employ an ambush or sit-and-wait predatory strategy, appear to exhibit a higher tolerance for spatial restriction in captivity [31, 32]. In contrast, the corn snake, exhibits high levels of diurnal activity, actively foraging across large territories [8]. Importantly, corticosterone plays a central role not only in stress responses but also in regulating energy metabolism and daily activity patterns. Baseline levels of this hormone are closely linked to allostatic processes, including the mobilisation and allocation of energy resources required for routine physiological functioning [33, 34]. Species differing in foraging strategy and locomotor activity may therefore exhibit distinct baseline corticosterone profiles that are not necessarily indicative of stress exposure per se. Ambush predators such as boas and pythons typically display low routine activity levels and prolonged periods of inactivity, whereas actively foraging species such as Pantherophis guttatus maintain higher levels of movement and exploratory behaviour. Consequently, elevated corticosterone concentrations observed in more active species may partly reflect differences in metabolic regulation and energy expenditure rather than heightened stress responses [35, 36]. This underscores the need for caution when interpreting interspecific variation in corticosterone levels, particularly in captive settings. Given the lack of species-specific validation of shed skin corticosterone in the studied taxa, the results should be interpreted as relative differences rather than absolute indicators of stress. The magnitude of the observed effects varied between species. In our study in tested Boa, the negligible effect size suggests that housing system was not associated with meaningful differences in corticosterone concentrations under the conditions of this study. In Python regius, although the difference between housing systems reached statistical significance, the small effect size indicates that the practical relevance of this difference may be limited. In contrast, Pantherophis guttatus showed a large effect size, suggesting a more pronounced association between housing system and corticosterone levels. These findings indicate potential species-specific differences in physiological responsiveness to housing conditions. However, given the observational design and the focus on a single endocrine parameter, the results should be interpreted with caution. Further studies incorporating additional welfare indicators would be necessary to confirm the biological significance of these associations.

Considerable debate persists regarding the impact of limited space, which is typical for pet snakes, on their welfare status [2, 37–39]. From the perspective of enhancing captive animal welfare, increasing enclosure dimensions appears justified. Multiple studies have shown that snakes and other captive reptiles housed in larger enclosures utilize environmental enrichments more effectively and display greater behavioral diversity [8, 10, 20, 21, 40–42]. However, empirical evidence directly linking confined spaces, such as rack systems or small terrariums, to elevated stress levels in snakes remains limited. The present study contributes new insights to this issue and suggests a potential association that warrants further controlled investigation. The findings demonstrate pronounced variation in corticosterone deposition among the three commonly bred and kept snake species examined. Rack system housing was associated with elevated corticosterone levels in corn snakes, which are characterized by high activity levels. In contrast, this effect was not observed in the less active royal pythons or boas, where hormone levels remained low. These results prompt a critical reconsideration of whether the universal recommendation to increase enclosure size is appropriate for all snake species. It is plausible that limited space may be more suitable for reclusive species such as royal pythons than for larger, enriched enclosures.

A major limitation in interpreting the results of this study is the absence of established reference values for corticosterone concentrations in snakes. For instance, corticosterone levels in P. guttatus (range values for snakes from this study: 5.50–6427.81.50.81 ng/g) — even in enriched terraria — already exceed the upper range reported by Berkvens et al. [28], who validated a shed skin corticosterone enzyme immunoassay in the African house snake (Lamprophis fuliginosus) and found values from approximately 9.24 to 82.12 ng/g (mean 33.87 ng/g). This indicates that corn snakes tested in this study exhibit substantially higher corticosterone concentrations, particularly under rack system conditions. The physiology, skin growth rate, kinetics of corticosterone deposition in keratin, shedding frequency, and baseline corticosterone may differ between P. guttatus and L. fuliginosus. Thus, direct comparison is indicative but not definitive. The extremely elevated levels observed in rack-housed corn snakes suggest a pronounced physiological response, but in the absence of established species-specific reference values, it is not possible to determine whether these concentrations unequivocally reflect chronic stress or represent normal interspecific differences in hormone deposition patterns. The findings underscore the need for additional studies across multiple snake taxa to establish reliable baseline ranges for corticosterone in shed skin and to clarify the extent to which housing conditions drive the observed variation.

Several limitations should be considered when interpreting the results of this study. Shed skins were collected from privately owned snakes maintained under natural husbandry conditions. While this enhances ecological validity, it introduces variability in enclosure size, enrichment, heating, UVB provision, and handling frequency that could not be fully standardized. Housing information was obtained through standardized owner-completed questionnaires, which may introduce reporting variability. Future studies utilizing controlled experimental designs would enable more precise isolation of specific environmental factors. Another limitation is the absence of established species-specific reference ranges for corticosterone concentrations in snake shed skin. Consequently, findings should be interpreted primarily as relative differences between groups rather than definitive indicators of physiological stress. Additional validation studies are required to establish normative baseline values across taxa. Despite these limitations, the use of naturally shed skins offers a non-invasive method for assessing long-term glucocorticoid deposition, and this study provides novel comparative data across commonly kept snake species. These findings represent an important step toward developing species-specific welfare assessment tools for captive reptiles.

In contrast to mammals and birds, for which baseline and stress-induced glucocorticoid ranges are well documented, no standardized values exist to serve as benchmarks for different snake species. As a result, it is not possible to determine with certainty whether the corticosterone levels measured in this study indicate a physiological stress response or fall within the normal hormonal range for these taxa. This lack of comparative data complicates the interpretation of welfare implications associated with the observed hormone concentrations. Further research is required to establish species-specific baseline corticosterone values in snakes under both natural and controlled conditions, thereby providing a necessary framework for interpreting results such as those presented here.

In summary, the scientific literature highlights environmental enrichment and increased enclosure size as key factors in reducing stress among captive snakes. Within rack systems, incorporating elements that enhance environmental complexity is advisable to mitigate the adverse effects of monotonous settings, particularly for more active species such as colubrids. However, accurate welfare assessment requires consideration of potential interspecific differences in hormonal responses to captivity-related stress. Therefore, husbandry recommendations should be tailored to the specific needs of individual species rather than applying uniform guidelines to the entire suborder Serpentes. When evaluating the effects of environmental conditions on reptile behavior and welfare, conclusions should be drawn cautiously, as the studies cited demonstrate diverse impacts depending on methodology and species.

Conclusion

The findings indicate that corticosterone levels in snake shed skins vary significantly among species and housing systems, reflecting species-specific physiological responses to captive environments. Corn snakes, which are naturally more active, exhibited higher corticosterone concentrations, particularly when maintained in rack systems. This suggests that limited space and reduced environmental complexity may be associated with elevated corticosterone levels in more exploratory species. In contrast, this effect was not observed in boas or royal pythons, which are less active and did not show the same association with housing type under the conditions of this study. The results only partially support the initial hypothesis that snakes housed in rack systems would exhibit higher stress hormone levels than those kept in terrariums. While this pattern was evident in corn snakes, it was not observed in boas, and an inverse relationship was found in royal pythons. These findings demonstrate that the relationship between housing type and physiological stress is species-specific and influenced by behavioral ecology.

These results further emphasize that welfare assessments and husbandry recommendations for snakes should not be generalized across taxa. Instead, enclosure design, environmental enrichment, and space provision should be tailored to each species’ behavioral ecology and stress physiology. Implementing species-specific welfare guidelines is essential for reducing chronic stress and improving the overall well-being of captive snakes.

Materials and methods

Study design

The study used shed skins from snakes kept as pets by private owners. The snakes were not kept for the purposes of the experiment; they belonged to private owners who kept them in their home terrariums or racks. The experiment used snake shed skins from 3 species of snakes: Pantherophis guttatus (Corn snake), Python regius (Royal python/Ball python), Boa imperator (Central American Boa) (Table 2) in two different housing systems with different levels of environmental enrichment: enriched terrarium, rack system without environmental enrichment. All snakes participating in the experiment were females that had been kept in the current enclosure for a minimum of 6 months.

Table 2.

Characteristics of the studies snakes (N – number of animals)

	Rack system (N)	Terrarium (N)	Total (N)
Boa imperator	33	7	40
Python regius	68	16	84
Pantherophis guttatus	21	15	36
Total (N)	122	38	160

Shed skins were collected from snakes kept by private owners (in the case of snakes kept in terrariums), snakes bred in the rack system’s conditions were obtained from a reptile breeder - the environmental conditions were the same for all snakes in the rack system containers.

Prior to obtaining skin samples from individual owners, each participant was asked to complete a standardized questionnaire concerning the housing conditions of their snake [see Additional file 1]. The questionnaire included items regarding the dimensions of the enclosure, the presence and type of heating, the use of UVB lighting, and the frequency and manner of handling the snake both inside and outside the terrarium. Respondents were also asked to indicate whether their snake was housed in a terrarium with environmental enrichment — defined as the presence of additional structural elements beyond the substrate, water bowl, and hiding place (e.g., branches or other features promoting exploration) — or in a rack system container, defined as a flat enclosure equipped only with substrate and a water bowl, without environmental enrichment. Finally, each owner confirmed that the questionnaire had been completed truthfully.

Rationale for the use of shed skin as a biological matrix

Shed skin was selected as the biological matrix for corticosterone quantification due to its non-invasive nature and its suitability for assessing long-term glucocorticoid deposition. In contrast to plasma sampling, which reflects acute circulating hormone levels and may itself be influenced by handling or restraint, corticosterone incorporated into keratinized tissues represents hormone deposition occurring during skin formation. As a result, shed skin provides an integrated measure of glucocorticoid exposure over an extended period rather than a transient physiological response. This approach has been validated in reptiles and other vertebrates using keratinized matrices such as shed skin, feathers, hair, and wool, where hormone concentrations are considered indicative of longer-term endocrine activity. In the context of privately owned captive snakes, the use of naturally shed skin enables hormone assessment without additional disturbance to the animals and is therefore particularly appropriate for welfare-related investigations.

Snakes and enclosures collected data

The study used shed skins from female snakes with an average age of 2.82 ± 1.98 (years old ± SD; pythons), 3.05 ± 2.03 (boa), and 2.64 ± 1.97 (corn snake) for snakes kept in terrariums. Females kept in the rack system were 2 years old (100%, all species). Information on lighting and heating equipment was recorded for all terrarium-housed snakes. UVB fluorescent lighting was absent in all terraria housing Python regius and Boa imperator, and in 90.91% of terraria housing Pantherophis guttatus. As expected, none of the snakes maintained in rack systems were provided with UVB lighting. Heat sources used in terraria varied among species. In terraria housing P. regius, heating was provided primarily by heating mats (64.71%), followed by overhead heating panels (29.41%) and heating bulbs (5.88%). For B. imperator, heating mats were most commonly used (66.67%), with smaller proportions of terraria equipped with overhead heating panels (11.11%), heating bulbs (11.11%), or a combination of heating mats and heating bulbs (11.11%). In terraria housing P. guttatus, heating was predominantly supplied by heating mats (86.36%), while heating bulbs (4.54%) or a combination of heating mats and bulbs (4.54%) were used less frequently. Owners reported the frequency of enclosure manipulation in the animal’s tank (opening the tank, cleaning, etc. without removing the snake from the tank). 58.82% were taken care of twice a week, 17.65% once a week, 23.53% once a month (pythons); daily 11.11%, once a week 55.56%, twice a week and more often 33.33 (boas); once a week 31.82%, twice a week 40.90%, 3 times a week 13.64%, once a month 13.64% (corn snake). For rack system snakes, all species were serviced 2 times a week. Manipulations involving deliberate removal of snakes from tanks to habituate them to human contact (handling) were not performed in snakes maintained in rack systems. In contrast, regular handling was used in snakes maintained in terrariums by 66.18% of pythons, 66.67% of boas and 90.91% of corn snake.

Sample preparation

The snake shed skins were rinsed in distilled water for 2 min to remove unnecessary biological materials and the substrate on which the snake had passed through the shedding process. After drying, the shed skins were sprayed with 100% methanol and then left to evaporate the alcohol at room temperature for 12 h. The prepared shed skins were milled using a grinder, and in this form, the milled shed skins were packed into zipper bags and frozen (−20 °C) until analysis.

After a sufficient number of shed skins had been collected for determination, the material was thawed and set aside to reach room temperature. The studied material was weighed in a ratio of 0.1 g to 9 mL DMSO into screw-capped Falcon polystyrene centrifuge tubes (50 mL) and shaken for 2 h at room temperature on a rotator.

Determinations of corticosterone in snake shed skin were performed with ELISA, using the Cort (Corticosterone) ELISA Kit (Reed Biotech, Catalog No.: RE10113, Wuhan, PRC). The claimed detection range is 1.57–100 ng/ml, with a sensitivity of 0.51 ng/ml. Determination of corticosterone levels in snake skin was performed according to the assay procedure provided with the ELISA kit. Optical density (OD) was measured spectrophotometrically at 450 nm on a microplate reader (TECAN, Sunrise, Austria). The concentration of corticosterone in the tested snake skin samples was calculated by comparing the OD of the samples with a standard curve. Each sample was tested in duplicate, and the average of the OD readings was used for analysis. For several samples, the OD exceeded the upper limit of the standard curve, so a retest was conducted with dilution of the samples tested.

The content of corticosterone was not determined (they were < LOQ) in the shed skins of 8 boa from a rack system, 26 shed skins from a python (19 from a rack system, 7 from a terrarium). These samples were removed from analysis due to the lack of data on their corticosterone concentration. On the other hand, in the case of the corn snake, 28 shed skin samples (18 from the rack system, 10 from the terrarium) required retesting due to excessive concentrations of the hormone under study. For this purpose, the corn snake samples were diluted 5 times and retested.

Statistical analysis of data

Corticosterone concentrations for each species were assessed for normality using the Shapiro–Wilk test and showed significant deviations from a normal distribution. Therefore, non-parametric statistical methods were applied throughout the analysis. Statistical analyses focused on within-species comparisons to evaluate the effect of housing system (terrarium vs. rack) on corticosterone concentrations. Differences between housing systems within each species were analysed using the Mann–Whitney U test with two-tailed significance testing. In addition to reporting U statistics and corresponding p-values, effect sizes (r) were calculated to quantify the magnitude of differences independently of sample size. The effect size was derived from the standardized test statistic (Z) and interpreted according to Cohen’s guidelines (small: 0.10; medium: 0.30; large: ≥ 0.50). Because only two housing conditions were compared within each species, post hoc multiple comparison procedures were not required. Differences in corticosterone concentrations between species were not subjected to formal statistical testing and are presented descriptively, due to the lack of control over species-specific baseline levels, metabolic differences, and the potential influence of captivity on interspecific hormone profiles. Descriptive statistics included the mean, standard error (SE), median, and interquartile range (IQR; 25th–75th percentile). Associations between enclosure volume and corticosterone concentration were analysed separately for each species using Spearman’s rank correlation coefficient. The level of statistical significance was set at α = 0.05. Statistical analyses and figure preparation were performed using Statistica 13.3 (TIBCO Software Inc.).

Authors’ contributions

DZ: Conceptualization Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Validation; Resources, Visualization; Writing - original draft; and Writing - review & editing, Supervision, Project administration; Resources. AS: Methodology; Validation, Resources, Writing - review & editing MJ: Resources Writing - review & editing. All authors read and approved the final manuscript.

Funding

The study was funded by the National Science Centre in Poland as part of the MINIATURA 7 grant: 2023/07/X/NZ9/00093.

Data availability

Data are available from the corresponding author upon request and subject to institutional approval.

Declarations

Ethics approval and consent to participate

The study was conducted in accordance with the guidelines of the Institutional Review Board (Animal Welfare Commission) of the Faculty of Animal Science and Bioeconomy at the University of Life Sciences in Lublin, Poland. The research involved only privately owned snakes and was limited to the collection of naturally shed skins, without any experimental procedures or intervention involving animals; therefore, formal ethics approval was not required. In accordance with the Polish Act of 15 January 2015 on the protection of animals used for scientific or educational purposes, approval from an ethics committee is not required for research involving this type of material and procedure. All owners of the privately kept snakes voluntarily provided informed consent by collecting and submitting the sheds and completing the questionnaire accompanying the declaration.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.