Introduction

Reproduction is one of the most important biological functions for the survival and perpetuation of animal species. Understanding a species' specific reproductive and social patterns can facilitate both captive breeding and management of the species in its natural environment (Zimbler-DeLorenzo & Stone 2011), which can contribute to its propagation and conservation. Breeding species in captivity should be considered a last resort solution, used when wild populations cannot survive by other means and until more suitable conditions exist for conservation in their natural habitat (Amstislavsky et al. 2008). There are several obstacles associated with captive reproduction, however, including being able to recreate an adequate environment that meets the needs of the species (Clubb & Mason 2007), managing diseases derived from the entry of wild individuals that serve as founders or other animals housed at the centre (Thorne & Williams 1988) and the low genetic variability of captive populations due to inbreeding (Frankham 2003, 2005).

Homozygosity in individuals born in captivity results in the fixation of deleterious alleles, increasing susceptibility to diseases and causing reproductive problems leading to loss of fitness, high infant mortality rates and a low life expectancy (Amos & Balmford 2001, Schmalz-Peixoto & von Schmalz-Peixoto 2003). In addition, captive animals may be frequently exposed to stressful situations, such as confinement in limited spaces, restricted movement, reduced retreat space, lack of environmental stimulation, handling by caregivers and exposure to new stimuli interpreted as potentially aversive, such as loud noises (Morgan & Tromborg 2007). These factors can trigger a chronic physiological stress response in animals, negatively affecting their long-term health and well-being (Nájera 2022). The physiological response to chronic stress can affect the immune system (Webster Marketon & Glaser 2008), reproductive health (Dobson & Smith 2000), growth, development (Weary et al. 2008), and adaptive capacity of animals (Koolhaas et al. 2011).

In carnivores, the physiological stress response is regulated by a complex hormonal system responsible for the release of cortisol, which regulates reproduction through the release of sex hormones, such as testosterone, progesterone and oestrogens, in a bidirectional manner. These hormones are responsible for the maturation of the reproductive organs, the production and release of gametes, and sexual behaviour and mate selection (Morohashi et al. 2012, Jennings & Lecea 2020). Progesterone and oestrogens are female sex hormones that regulate the menstrual cycle and prepare the uterus for pregnancy (Jabbour et al. 2006, Wu et al. 2018). The hypothalamic-pituitaryadrenal (HPA) axis is a regulatory system involved in the stress response. Under stress, the hypothalamus releases corticotropin-releasing hormone (CRH), which stimulates the release of adrenocorticotrophic hormone (ACTH) from the pituitary gland. ACTH, in turn, acts on the adrenal glands to release cortisol, the primary stress hormone (Reeder & Kramer 2005, Barja 2015). The hypothalamic-pituitary-gonadal (HPG) axis is a complex feedback system that regulates the production and release of sex hormones. Here, the hypothalamus produces and releases gonadotropinreleasing hormone (GnRH), which stimulates the pituitary gland to produce and release two hormones, follicle-stimulating hormone (FSH) and luteinising hormone (LH). FSH and LH travel to the testicles in males or to the ovaries in females, stimulating the production and release of sex hormones (Dwyer & Quinton 2019). Since there is a bidirectional relationship between the HPA and HPG axis, chronic stress or distress can affect the production of sex hormones by changing their concentration. Hormone concentration measurements in animals are carried out using various techniques, using biological samples obtained either invasively (saliva, blood) or non-invasively (hair, faeces). Invasive methods require handling and immobilisation of the animal, which can alter basal glucocorticoid concentrations (Sheriff et al. 2011), while non-invasive methods avoid handling, thereby minimising measurement bias (Barja et al. 2012). A standard method used to measure glucocorticoid levels relies on quantifying metabolites in fresh faecal samples since cortisol is rapidly metabolised and eliminated from the body through faeces, with peak concentrations in mammals occurring between 12 and 48 hours after the stressor stimulus (Barja et al. 2012, Navarro-Castilla et al. 2021, Martín et al. 2023).

Ex situ conservation programmes for European mink (Mustela lutreola) have focused on captive breeding and reintroduction into their natural habitat. The main objective of these programmes is to maintain 85% of the original heterozygosity in the captive population for 50 years (Kneidinger et al. 2018); however, maintaining this genetic variability has its complications. Breeding pairs, specifically selected for their genetic load, have a 25% breeding success at each attempt, with breeding success depending more on males than females due to individual variation in behavioural features such as passiveness or aggressiveness (Kiik et al. 2013). The same authors verified that males born in captivity had less reproductive success than those born wild (35% and 89% success, respectively). Subsequently, Kneidinger et al. (2018) suggested that the captive environment may be having adverse effects on mink reproductive behaviour, while Blanchard et al. (2001) hypothesised that ‘social stress’ was a primary factor causing abnormal behaviour in captive-bred European mink. In nature, the mink is a solitary species (Youngman 1990); however, in captivity, individuals are usually placed in enclosures close to other mink, where they constantly receive signals (mainly olfactory) from neighbouring facilities that may cause chronic stress, leading to distorted reproductive behaviour (Kiik et al. 2013). In the study of Nagl et al. (2015), all captive females had normal heat cycles; thus, a priori, the focus is mainly on males. Nevertheless, for approximately three months before the start of the reproductive season, captive females are usually subjected to heat detection techniques. As this entails an increase in the frequency of invasive management, these techniques may constitute a potential stress-inducing factor.

To assess whether reproductive management for oestrus detection represented a stressful routine for female European mink, we evaluated whether routine vaginal cytology in females triggers an increase in the levels of faecal cortisol metabolites (FCM) that may then be altering metabolites levels of two sex hormones (progesterone and oestradiol), thereby physiologically hindering reproduction. Our hypotheses were: 1) frequent performance of vaginal cytology increases FCM levels in female European mink, as previously confirmed in several wild and domestic felid species (Genaro et al. 2007) and other mustelid species, such as the black-footed ferret (Mustela nigripes), subjected to veterinary tests requiring injections and/or animal handling (Young et al. 2001); 2) FCM levels are positively related to faecal progesterone metabolite (FPM) and faecal oestradiol metabolite (FEM) levels, as is the case with other mammalian species (Chatdarong et al. 2006, Barja et al. 2008, Liening et al. 2010); 3) older females will have lower FCM levels than younger females as they are more habituated to this type of handling (Rankin et al. 2009); and 4) variations in FCM levels will depend on individual variability based on previous experience and animal personality, among other factors (Fernández-Lázaro et al. 2019, 2023).

Material and Methods

Experimental methodology

The study took place between late March and mid-May. Typically, FIEB restricts research from late March to early September, making it impossible to carry out procedures during spring and summer. However, females usually come into oestrus around April (Amstislavsky & Ternovskaya 2000), though this period may start earlier (in February) or later (up to May), depending on latitude. Gestation takes 40-42 days, with births usually occurring between June and July (Palazón 2010). Given that the European mink is a highly stress-sensitive animal (Rozhnov & Petrín 2006) with low reproductive success, the risk of compromising reproduction was reduced by minimising caregiver visits and limiting animal manipulation to veterinary issues focused on health and reproduction only. The period between March and May is crucial for detecting oestrus in females and favouring mating by establishing genetically selected couples. Since the European mink is a solitary and territorial animal (Palazón 2010), it is housed in individual enclosures in conservation centres to reduce social stress (Blanchard et al. 2001). In our study, heat (oestrus) detection was performed by identifying cornified cells in a vaginal smear collected with a swab and examined under a microscope. Females were considered to have entered oestrus when vaginal smears showed 90% cornification (Amstislavsky et al. 2004). Vaginal cytology was performed over several days, sometimes consecutively and at other times intermittently, until female heat was detected, at which point caregivers brought the couples together in a larger, odour-free facility, which acted as neutral ground to reduce aggression caused by territorial behaviour.

Fig. 1.

Variation in hormone levels over time. x̄[ ]_w (ng/g) variation for A) faecal cortisol metabolites (FCM; ρ = –0.791, P = 0.0001); B) faecal progesterone metabolites (FPM; ρ = –0.608, P = 0.003); C) faecal oestradiol metabolites (FEM; ρ = –0.792, P = 0.0001).

Fig. 2.

Relationships between cortisol, progesterone and oestradiol metabolite levels. Correlation of x̄[ ]_w (ng/g and pg/g) between faecal cortisol metabolites (FCM), faecal progesterone metabolites (FPM) and faecal oestradiol metabolites (FEM): A) FCM-FPM (ρ = 0.566, P = 0.006), B) FCM-FEM (ρ = 0.899, P = 0.0001), C) FPM-FEM (ρ = 0.592, P = 0.003).

Fig. 3.

Correlation between the medians for metabolite levels of faecal cortisol metabolites (FCM), faecal progesterone metabolites (FPM) and faecal oestradiol metabolites (FEM): A) FCM-FPM (ρ = 0.467, P = 0.0001), B) FCM-FEM (ρ = 0.325, P = 0.001), C) FPM-FEM (ρ = 0.485, P = 0.0001). Mass-mass ratios are presented in pg/g for FEM, and in ng/g for FPM and FCM. Note the positive trend line for the three graphs with respect to R².

Fig. 4.

Median ± CI for individual level A) faecal cortisol metabolites (FCM), B) faecal progesterone metabolites (FPM) and C) faecal oestradiol metabolites (FEM).

If copulation was unsuccessful, the female was subjected to new tests until a new period of heat was detected, the European mink having a polyoestrous cycle, i.e. females may enter oestrus three times if not fertilised (Mead 1989).

Veterinarians and caregivers collected 117 fresh faecal samples from females during the screening days, the samples being collected the day cytology was performed and 24 h and 48 h after cytology to investigate whether repeated testing modified FCM levels and whether this affected the metabolite levels of two important sex hormones in the reproductive period, i.e. progesterone (involved in gestation) and oestradiol (involved in the ovulatory cycle). Faecal samples were classified as fresh if they showed no signs of dehydration and had a strong odour (Barja et al. 2012, Ortiz-Jiménez et al. 2022). All samples were stored in individually sealed bags and labelled with the date and name of the female. Immediately after collection, the samples were frozen at -20 °C until hormonal analysis.

Data analysis

Non-parametric statistical methods were used throughout, as the variables did not satisfy the assumptions of normality and equality of variance. To assess variation in physiological stress over time, we quantified FHM levels for cortisol (FCM, ng/g), progesterone (FPM, ng/g) and oestradiol (FEM, ρg/g) according to the duration and frequency of female handling. Since the duration and frequency of handling varied between females, more faecal samples were collected from some individuals than others. Based on this variability, two correction factors were used:

where [ ]_w was the weighted hormonal concentration of each sample, [ ]_i was the hormonal concentration of each sample from female i, and t represented time in days over which faecal samples were collected from each female. This allowed us to obtain the hormonal concentration of k samples for each female.

where x̄ [ ]_w was the mean weighted hormone concentration, ∑[ ]_wi was the sum of the weighted concentrations from i = 1 to k samples for each female, and n was the number of females (n = 8). In this way, we obtained the weighted mean of FHM of each female over the handling time.

Table 1.

Levels of oestradiol metabolites according to vaginal cytology frequency. Median and interquartile range (IQR = Q3-Q1) of faecal cortisol metabolites (FCM; pg/g) in relation to vaginal cytology number (ρ = –0.234, P = 0.013). The number of faecal samples analysed for each group is indicated.

We performed two Spearman's correlation tests to determine the influence of time on variation x̄ [ ]_w of FHM and the relationship between x̄ [ ]_w for pairs of hormones. In addition, we performed a separate Spearman's correlation to determine i) the influence of vaginal cytology frequency (n = 4 – 12) and age (measured in months) on FHM levels and ii) any correlation between the three hormones (FCM, FPM and FEM). Finally, we performed a Kruskal-Wallis test to determine the influence of individual variability on FHM levels, the results being shown as median ± confidence interval (CI).

Results

Variation in x̄ [ ]_w for the three hormones as a function of time was significant in all cases (FCM: ρ = –0.791, P = 0.0001, n = 22; FPM: ρ = –0.608, P = 0.003, n = 22; FEM: ρ = –0.792, P = 0.0001, n = 22; Fig. 1). Likewise, correlations between the x̄ [ ]_w for the three hormones were statistically significant: FCM-FPM: ρ = 0.566, P = 0.006, n = 22; FCM-FEM: ρ = 0.899, P = 0.0001, n = 22; FPM-FEM: ρ = 0.592, P = 0.003, n = 22; Fig. 2).

Table 2.

Median and interquartile range (IQR = Q3-Q1) of faecal cortisol metabolite (FCM) and faecal progesterone metabolite (FPM) levels (ng/g) in relation to age (months) of individuals (FCM-age: ρ = 0.378, P = 0.0001; FPM-age: ρ = 0.212, P = 0.023). The number of samples analysed for each group (No. of faecal samples) is indicated.

The correlation between FCM and FPM levels was both positive and significant (ρ = 0.467, P = 0.0001, n = 105), as was that between FCM and FEM (ρ = 0.325, P = 0.001, n = 106) and between FPM and FEM (ρ = 0.485, P = 0.0001, n = 112; Fig. 3). Frequency of vaginal cytology was not significantly correlated with FCM and FPM, but was with FEM (ρ = –0.234, P = 0.013, n = 114; Table 1). FCM levels were positively correlated with age (ρ = 0.378, P = 0.0001, n = 107), as were FPM levels (ρ = 0.212, P = 0.023, n = 115; Table 2). Individual variability had a strong and significant influence on FCM levels (H = 22.609, df = 7, P = 0.002, n = 108), FPM (H = 15.820, df = 7, P = 0.027, n = 108) and FEM levels (H = 20.136, df = 7, P = 0.005, n = 110; Fig. 4).

Discussion

Diagnostic tests for oestrus (vaginal cytology) in captive female European mink involve routine handling of individuals, with the potential to trigger changes in the levels of cortisol metabolites that vary over time. During this study, we observed an increase in FCM levels during such tests, reaching a peak on day four, i.e. 72 h after the first cytology was performed. Our results differ from those recorded in other wild mammal species, where faecal cortisol peaks were recorded between 24 and 48 h after induced stress, depending on the species (Wasser et al. 2000, Barja et al. 2012). In our study, however, we need to consider the potential cumulative effects of frequent cytology, with some females not having a break of more than 24 h between cytology tests. This break could explain the delay in reaching peak cortisol levels, since previous response tests to stressors have shown a cumulative effect over time surpassing a typical threshold for test subjects. Unfortunately, our study could not avoid this effect as the conservation centre established the cytology schedule based on the short period that oestrus lasts in females of this species, i.e. from one to ten days (Amstislavsky & Ternovskaya 2000). Such routine cytology tests could have become a stressor, triggering an increase in FCM; however, we observed habituation to veterinary handling over the period following the point at which maximum cortisol concentration was reached, i.e. four days after the start of testing. This cortisol peak may have been triggered by an acute response to novel handling, which only occurs during the three-month breeding season, with the youngest females experiencing such handling for the first time. The additional two peaks observed likely represent acute episodes experienced by females who received 24-72 h rest between cytology tests. This will have occurred when the cytology routine was interrupted by a female being moved to a neutral enclosure to allow copulation due to oestrus, the vaginal cytology routine restarting after the break. Note that female European mink can enter heat up to three times if they fail to copulate (Amstislavsky & Ternovskaya 2000).

At the conservation centre where this research took place, efforts were made to minimise unnecessary contact with the animals. Furthermore, this was especially true during the breeding season, as studies have shown that European mink are highly sensitive to stress at that time (Rozhnov & Petrín 2006). Despite avoiding routine interaction throughout the year, the handling of females necessarily increased during the three months when cytology examinations were conducted. This also meant a decrease in familiar handling practices and introducing a type of handling the females had not previously encountered or only briefly experienced each year. Though it appears that the females were able to gradually adapt to this more intensive handling over the three-month testing period, it may be beneficial to acclimatise the females throughout the year to this type of handling, especially in the case of younger females facing it for the first time or those with limited prior experience with this potential stressor. Despite the FIEB centre's preference for limiting regular contact with the animals to maintain their wild nature, it may prove beneficial to implement positive reinforcement training (Heidenreich 2007) for those females not intended for reintroduction and instead kept at the centre for breeding purposes. Numerous studies have demonstrated that mammals, including certain carnivores (Broder et al. 2008), are capable of voluntarily cooperating in medical and breeding procedures in zoo and research settings following such positive reinforcement training (Laule 2003, McKinley et al. 2003, Laule & Whittaker 2007, Coleman et al. 2008).

In this study, maximum FPM peaks occurred during the habituation days prior to handling (Rankin et al. 2009), coinciding with the lowest FCM levels. These FPM peaks coincided with supposed entry into oestrus, days during which there was no copulation, and days when gestation began. In these cases, FPM concentrations likely increased to inhibit the production of FSH, which causes follicular maturation and the production of LH, responsible for ovulation (Filicori et al. 2002, Murray & Orr 2020). Furthermore, the highest FEM peaks coincided with those of FCM but were opposite to those of FPM. If progesterone levels were lower, FSH inhibition would not occur, and the concentration of oestrogens and oestradiol would increase (Wiele et al. 1970). Note that our results showed a positive correlation between FCM and FPM when using weighted means and medians. This is consistent with other mammalian studies, especially those studying cortisol and progesterone interaction (Chatdarong et al. 2006) and those comparing all three hormones (Barja et al. 2008, Liening et al. 2010). For example, a study on the European marten (Martes martes) showed positive correlations between FCM levels and levels of FPM, testosterone, and oestradiol (Barja et al. 2011).

The number of cytology examinations was positively correlated with FEM levels but not with FPM or FCM levels. While this correlation does not confirm causality, it would appear that the total number of vaginal cytology examinations did not induce a continuous increase in FCM levels over the study period and, consequently, did not have an adverse effect, despite being a potential stressor during the reproductive season for an animal considered highly sensitive to stress (Rozhnov & Petrín 2006). In addition, older females showed higher FCM levels than young females, a result that contrasts with other studies, which tend to show younger females having higher FCM levels (Azevedo et al. 2019) or no significant differences between age groups (Stevenson et al. 2018). Some studies, however, have suggested that animals exhibit increased levels of glucocorticoids as they age, with ageing reducing the functionality of the endocrine system and affecting glucocorticoid secretion (Thompson et al. 2020). Studies with other mammals have also shown no differences in faecal progesterone (Burgess et al. 2012) and fat progesterone as a function of age in non-pregnant females (Atkinson et al. 2020). These conflicting conclusions point to the need for more accurate empirical data addressing deficiencies in how age influences the physiological stress response.

While variations in FPM and FEM levels are most likely attributable to differences in individual female menstrual cycles, variations in individual FCM levels can be attributed to a wide range of factors, such as gender, age, life experience and quality of life, which modulate an individual's behaviour over their lifetime (Fernández-Lázaro et al. 2019, 2023, Kirschbaum et al. 1992). These experience-based behaviours are essential for each individual and are responsible for shaping the individual's personality, i.e. sets of correlated behaviours expressed in different situations (Carere & Eens 2005). It has been suggested that animals with proactive personalities are likely to be more successful in a stable environment than those with reactive personalities, while the more cautious style of reactive animals may be more successful in a changing environment; thus, there is no one-size-fits-all personality for all situations (Cockrem 2005). Individual variability could be caused by habituation after several repeated tests or through variations in neuronal plasticity (Noer et al. 2015, 2016). For example, Fernández-Lázaro et al. (2019) have suggested that more active and aggressive individuals may have higher levels of FCM. Consequently, we consider it important to consider the significance of mink personality when designing management criteria, especially for breeding females, as different personalities may be more or less vulnerable to stressors and diseases (Carere et al. 2010). Regarding the management of animals intended for reintroduction, personality may also be crucial in selecting those best suited for survival. For example, individuals with proactive or cautious traits may be chosen after evaluating the area into which they will be reintroduced. Indeed, recent studies indicate that individual variation within a species can buffer against strong fluctuations in the species' natural habitat (Koolhaas & Van Reenen 2016).

In conclusion, repeated vaginal cytology does not appear to constitute a major stressor inducing increased FCM levels that could compromise reproductive success or alter behaviour during copulation. Furthermore, female age and individual variability were the main factors influencing cortisol and sex hormone metabolite levels. The results obtained in this study provide a first approximation for the elaboration of European mink management strategies during the reproductive season, especially for those individuals most susceptible to stress agents such as veterinary handling, in this case, adult females. Females older than one year are often prioritised for mating as some are already likely to have experience with males and will be sexually mature, while it is likely to be a new experience for younger females. With a view to future research, it would be interesting to convince conservation centres to standardise vaginal cytology methodology by testing stable groups of females receiving the same number of vaginal cytology tests and then increasing the sample size. We also consider it important to pay attention to the personality of each female since studies suggest that the most fearful individuals will be those exhibiting increased FCM levels, which could affect reproduction if remaining high over the long term. As such, we recommend tailoring management to each female to minimise the stressor's effect as much as possible. In the case of females destined solely for breeding (with no possibility of reintroduction), it could also be helpful to establish positive reinforcement training programs to facilitate habituation to this type of handling.