Unravelling the relative contribution of toxic metals to redox homeostasis in wild roe deer

Unravelling the relative contribution of toxic metals to redox homeostasis in wild roe deer | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 19 December 2025 V2 Latest version Share on Unravelling the relative contribution of toxic metals to redox homeostasis in wild roe deer Authors : Amandine Herrada 0009-0000-7279-213X [email protected] , Benjamin Rey , Jean-François Lemaître , Emmanuelle Gilot-Fromont , Marie-Laure Delignette-Muller , Gilles Bourgoin , François Debias , … Show All … , Paul Revelli , Rebecca Garcia , Sonia Saïd , Maryline Pellerin , Nadia Crini , Clémentine Fritsch , Renaud Scheifler , Jean-Michel Gaillard , and Pauline Vuarin Show Fewer Authors Info & Affiliations https://doi.org/10.22541/au.175697692.28236644/v2 283 views 137 downloads Contents Abstract Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Redox homeostasis, i.e. the balance between oxidative damage and antioxidant defences, is crucial for overall physiological function. Disrupting this balance may lead to oxidative stress. Non-essential metals (e.g. lead, cadmium) exhibit toxicity even at low concentrations and can increase the production of pro-oxidant molecules and impair antioxidant mechanisms. While the adverse effects of acute exposure are well documented, little is known about the impact of chronic, low-level exposure in the wild. Here, we investigated the relationship between toxic metal exposure and redox status in two roe deer populations (Capreolus capreolus), while accounting for the potential modulation by essential metals (e.g. selenium, zinc) and local weather conditions. We measured the concentrations of nine non-essential and thirteen essential metals, and assessed the oxidative status through seven biomarkers. Our findings revealed an elevation of oxidative damage (malondialdehyde and protein carbonyl levels, and reactive oxygen metabolites), and a depletion of an enzymatic antioxidant (SOD activity) but an increase in total antioxidant capacity, with increasing toxic metal content. While essential metals did not modulate these effects, individuals that had experienced higher temperatures or rainfall recently, exhibited stronger responses. Our study raises concerns regarding the broader consequences of chronic exposure to toxic metals for wildlife health and, ultimately, population dynamics. Keywords : non-essential metals, essential metals, weather conditions, redox status, oxidative stress, mammal physiology 1. Introduction Redox homeostasis, arising from the precise regulation of the balance between pro-oxidant molecules and antioxidant systems, is a cornerstone of physiological resilience across a broad range of taxa. Redox homeostasis is thought to influence the ability of organisms to cope with stressful conditions, supporting functions essential to reproduction and survival (Finkel and Holbrook 2000, Burton and Jauniaux 2011, Halliwell and Gutteridge 2015). Alteration of this balance, typically through over production of pro-oxidants such as reactive oxygen species (ROS), and/or depletion of intrinsic antioxidant systems, result in the detrimental accumulation of oxidative damage to macromolecules (proteins, lipids and nucleic acids) within cellular compartments, a physiological state referred to as oxidative stress (Burton and Jauniaux 2011, Halliwell and Gutteridge 2015). Oxidative stress may compromise cellular function and viability, as described in many vertebrate species, and on the long-run might contribute to an acceleration of the ageing process (e.g. López-Otín et al., 2023). Various environmental stressors have been shown to exacerbate ROS production, thereby increasing susceptibility to oxidative stress with possible cascading effects on individual health (e.g. Bilham et al., 2018; Rodríguez-Estival et al., 2016). Among them, organisms are increasingly exposed to various, and often toxic, chemicals steaming from human activities. In particular, non-essential metals (referred to as toxic metals hereafter), such as arsenic (As), cadmium (Cd), mercury (Hg) or lead (Pb), have no clear biological functions but are toxic to most living forms, even at low levels of exposure (Tchounwou et al. 2012, Govind 2014). Once released in the environment, these metals can bind to particulate matter and travel tens to hundreds of kilometres from their emission source, before falling out (Shahid et al. 2019, Azeh Engwa et al. 2019). These particles can pass the dermal barrier or be directly inhaled (Azeh Engwa et al. 2019). Additionally, they also deposit in water and soil, from which they are absorbed by organisms, eventually accumulating through the food chains (Shahid et al. 2019). Toxic metals become harmful when they accumulate in the body (i.e. when their absorption and storage rates exceed their detoxification or excretion rates (Govind 2014). They interact with cellular components, metabolic enzymes, as well as detoxification and repair mechanism pathways (Tchounwou et al. 2012). One key pathway through which these metals exert toxicity is by disrupting redox homeostasis, either directly by increasing ROS production, or indirectly by inhibiting antioxidant systems (Valko et al. 2005). For instance, several toxic metals may bind to crucial antioxidant enzymes such as superoxide dismutase (SOD), impairing their ability to neutralise the harmful pro-oxidant superoxide anions (Kumar et al. 2020). Beyond toxic metals, essential ones, such as copper (Cu), iron (Fe), selenium (Se) or zinc (Zn), known for their beneficial roles in various physiological processes, can influence the responses to toxic metal exposure. As cofactors of several antioxidant enzymes, these elements may mitigate or exacerbate metal-induced oxidative stress, underscoring their potential modulatory effects (e.g. Valko et al., 2005). Additionally, organisms in the wild are rarely exposed to isolated stressors. For example, variation in weather conditions, and the increasing occurrence of extreme temperatures, droughts or storms/severe rainfall, have been shown to influence redox homeostasis, as documented in both captive and free-ranging animal populations (see Chainy et al., 2016; Lushchak, 2011; Metcalfe & Alonso-Alvarez, 2010 for reviews). Such stressors may further exacerbate the physiological effects of chemical exposure. Yet, the interactive effects of multiple environmental stressors on oxidative stress remain poorly understood, especially in terrestrial mammals (but see Sokolova & Lannig, 2008; Xiao et al., 2024 for examples on aquatic ectotherms). Accounting for these potential interactions is important for assessing the ecological relevance of contaminant exposure in the wild. Most of our knowledge on the adverse health effects of toxic metals stems from studies on humans and laboratory animals, even if field studies have attracted increasing interest in the past decades (Shore and Rattner 2001). In the wild, aquatic and avian communities have received the most attention (Rattner 2009), with studies demonstrating that both acute and chronic exposure to toxic metals can lead to immunosuppression, endocrine disruption, and behavioural alterations (Iavicoli et al. 2009, Boyd 2010, Govind 2014, Saaristo et al. 2018), possibly underpinned by an overall increase in oxidative stress (Isaksson 2010). However, most existing literature on terrestrial mammals has concentrated on the effects of acute exposure to single metals (e.g. Rodríguez-Estival et al., 2011). Yet, it is more likely that the majority of organisms, including wildlife, are exposed to low levels of various metals through their life (e.g. Wirth & Mijal, 2010). Such exposure is also expected to affect individual performance on the long-term, as recently suggested for pesticides (Moreau et al. 2021). In this context, investigating whether low-level exposure to toxic metals affects wildlife physiology has become particularly relevant. In this study, we focused on the European roe deer ( Capreolus capreolus ), a widespread species known to be highly tolerant to human activities (Linnell and Andersen 1995, Bonnot et al. 2013). Roe deer are thus potentially exposed to various sources of chemicals including toxic metals through consumption of contaminated dust, water and/or plants, or through direct inhalation (Shahid et al. 2019). Moreover, as roe deer within a given population occupy small home ranges (Strandgaard 1972) which widely differ in terms of habitat structure and plant species composition (Pellerin et al. 2010), they are potentially subjected to varying toxic metal mixtures and exposure levels. Actually, roe deer are considered as reliable indicators of toxic metal contamination, and more broadly as sentinels for environmental health (García et al. 2011, Draghi et al. 2023). While a few studies have investigated toxic metal exposure in roe deer (Cappelli et al. 2020, Ludolphy et al. 2021, Draghi et al. 2023), none has examined the implications of such exposure for redox homeostasis nor did they consider its effects on roe deer inhabiting areas distant from major contamination sources, which could still be subject to chronic, low-level exposure. However, previous work highlighted spatial and temporal variations in toxic metal concentrations in roe deer, linking these variations to industries (Cappelli et al. 2020), urbanisation (Draghi et al. 2023) and agriculture (Ludolphy et al. 2021). In other large herbivores, some studies have linked toxic metal levels to redox status. For example, it was reported that increased Pb content was associated with reduced antioxidant levels in red deer ( Cervus elaphus ) living near a mining area in Spain. By contrast, wild boar ( Sus scrofa ) from the same area exhibited the opposite pattern (Rodríguez-Estival et al. 2011), highlighting species-specific responses to metals, and the need for additional field data to fully understand the consequences of toxic metal exposure for redox homeostasis. Here, we took advantage of the long-term monitoring of two French roe deer populations subject to different ecological contexts. The aim of the study was twofold: i) to decipher the relative contribution of toxic metal exposure on redox homeostasis in wild roe deer, and ii) to assess the modulatory effect of essential metals and weather conditions on the physiological responses to toxic metal exposure. Toxic metal concentrations were measured in roe deer hair because it is a non-invasive method and a reliable predictor of metal content in the environment (Rashed and Soltan 2005, Tête et al. 2014, Draghi et al. 2023). We focused on nine toxic metals whose harmful effects are widely recognised even at low level of exposure. We also measured three biomarkers of oxidative damage, three of antioxidant defences, and one marker of oxidative balance in plasma or in erythrocytes for the latter. We expected individuals with the highest toxic metal burdens to exhibit higher oxidative stress, i.e. an accumulation of oxidative damage on one hand, and a depletion of antioxidant defences on the other hand. Additionally, we hypothesised that these relationships would be modulated by the content in essential metals and weather conditions (i.e. cumulative rainfall and mean ambient temperature the week prior to capture). Specifically, we expected a stronger biomarker response to toxic metal exposure in individuals with minor concentrations of essential metals, as well as a contrasted response in individuals experiencing different weather conditions. 2. Materials and Methods 2.1. Study populations The studied roe deer populations are located in two enclosed forests managed by the Office Français de la Biodiversité (OFB): the Territoire d’Etude et d’Expérimentation of Trois-Fontaines (TF – 1360 ha), in north-eastern France (48°43’N, 4°55’E), and the Réserve Biologique Intégrale of Chizé (CH – 2614 ha), in western France (46°05’N, 0°25’W). The TF forest offers a high-quality habitat for roe deer, being characterised by a continental climate and fertile soils (Saïd et al. 2009), although increasing temperatures and drought frequency in spring-summer has decreased habitat quality since 2003 (Gaillard et al. 2013). On the other hand, the CH forest provides a poorer habitat quality for roe deer, and is defined by a temperate oceanic climate, poor soils and frequent summer droughts (Pettorelli et al. 2006). The two populations have been monitored for almost 50 years through a capture-mark-recapture program (since 1975 for TF and 1977 for CH; Gaillard et al., 1993). Captures take place every winter for 10 to 12 days between December and March. Roe deer are captured using drive nets and transported in wood boxes to a handling area. There, sex, age or age-class, hind foot length, and body mass (± 50 g) are determined for all individuals. Roe deer are fitted with ear tags upon their first capture and a sterilised microchip with a unique number is also inserted percutaneously at the back of the neck. For individuals of known age (i.e. roe deer first captured within their first year of life when age can be assigned with certainty based on the milk tooth sequence; Flerov, 1952), blood, hair and faeces samples are also collected. Hair are sampled from the white patch around the rump of the animal and stored in paper envelopes at room temperature. Blood (up to 1 mL/kg with a maximum of 20 mL) is collected from the jugular vein in 20 mL syringes with sterile hypodermic needles (gauge 19, 1.1 x 38 mm), and rapidly transferred to heparinised collection tubes of 4 mL. These tubes are then spun at 2000 g for 10 min and the plasma transferred into sterile cryogenic tubes. Blood cells are washed by adding ~2 mL of 0.9 % (v/v) NaCl solution, re-spun at 2000 g for 10 min and the buffy coat containing white blood cells is removed. All tubes (containing either plasma or washed erythrocytes) are frozen on site at - 80°C in a portable freezer (Telstar SF 8025) until laboratory analyses. At the end of the procedure, roe deer are released at their site of capture. 2.2. Essential and toxic metal concentrations The concentrations of 22 essential and toxic metals were assessed in roe deer hair samples. Roe deer undergo moult twice a year, in May-June and in October (Johnson and Hornby 1975, Sempéré et al. 1996). Consequently, metals measured in the hair represent what has been accumulated over the hair growth cycle. Since roe deer were captured during winter and the last moult occurred in autumn, the process of metal accumulation should be completed for all individuals. This prevents any bias regarding the timing of hair collection or potential seasonal effects (Combs et al. 1982). A total of 649 hair samples, representing 434 individuals captured between 2016 and 2019, were sent to the PEA²t platform of the Chrono-Environnement laboratory (UMR CNRS 6249, Université Marie et Louis Pasteur, France) to measure the concentration of a large set of essential and toxic metals. Eighteen elements were measured by inductively coupled plasma mass spectrometry (aluminium (Al), As, Cd, cobalt (Co), chromium (Cr), Cu, Fe, Hg, manganese (Mn), molybdenum (Mo), nickel (Ni), Pb, antimony (Sb), strontium (Sr), Se, titanium (Ti), thallium (Tl), Zn), and four elements were measured by inductively coupled plasma atomic emission spectrometry (calcium (Ca), potassium (K), magnesium (Mg), sodium (Na)) following the protocol given in Herrada et al . (Herrada et al. 2024a). In brief, hair samples were first washed in alternating baths of acetone and ultrapure water using an ultrasonic cleaner to remove any potential external contamination. After drying at 60°C, hair samples were digested in nitric acid and hydrogen peroxide, with reagent volumes adjusted according to sample weight. The resulting solutions were diluted with ultrapure water and filtered. Procedure blanks (acid, hydrogen peroxide and ultra-pure water) and certified reference materials were prepared and analysed using the same methods as the samples. The concentrations of essential and toxic metal concentrations in hair are expressed as µg/g dry mass. For values below quantification limits, half of the quantification limit value was assigned for the statistical analyses (USEPA 2000). As recovery rates were considered satisfactory between 75 % and 125 %, caution should be paid when considering the values for Hg and Na since they exhibited a high (173 %) and a low (46 %) average recovery, respectively (see also Herrada, Bariod, et al., 2024) for detailed metrology). The following elements were considered as non-essential toxic metal(loid)s: Al, As, Cd, Hg, Pb, Sb, Sr, Ti, and Tl, since the harmful effects of these elements is widely acknowledged (Tchounwou et al. 2012). The following elements were considered as essential metals: Ca, Co, Cr, Cu, Fe, K, Mg, Mn, Mo, Na, Ni, Se, and Zn. Although essential metals can become toxic at high concentrations, this is unlikely to be the case in our studied populations because essential metal concentrations are relatively low compared to those recorded in other populations of large herbivores (Herrada et al. 2024a). 2.3. Oxidative stress biomarkers Roe deer redox status was assessed by measuring three complementary biomarkers of oxidative damage in plasma, namely the total amount of reactive oxygen metabolites (d-ROMs test, Diacron International), the levels of thiobarbituric Acid Reactive Substances (TBARS), expressed as malondialdehyde (MDA) concentrations and reflecting lipid peroxidation (TBARS Assay kit, Cayman Chemical), and the protein carbonyl content as an estimate of oxidative damage to proteins (carbonyl content assay kit, Sigma-Aldrich). We also measured two biomarkers of antioxidant defences in plasma, namely the overall non-specific antioxidant capacity of plasma to in vitro oxidation (OXY-Adsorbent test, Diacron International) and the activity of an enzymatic antioxidant (superoxide dismutase (SOD) activity determination, Sigma-Aldrich); as well as two biomarkers of antioxidant defences in erythrocytes, namely a non-enzymatic antioxidant, reduced glutathione (GSH quantification, Sigma-Aldrich), and the reduced-to-oxidised glutathione ratio, GSH:GSSG (oxidised (GSSG) and reduced glutathione quantification, Sigma-Aldrich). All biomarkers were assayed spectrophometrically (SAFAS MP96 microplate reader, Monaco) following protocols previously described elsewhere (Herrada et al. 2024b), except for the oxidised and reduced glutathione quantification that will be briefly presented below. Raw values of biomarkers and their correlations are presented in Table S1 and Fig. S1, respectively. Reduced glutathione (GSH, γ-L-glutamyl-L-cysteinylglycine) is a tripeptide with high antioxidant properties that is converted to its oxidised form (GSSG) in the presence of ROS, making the GSH:GSSG ratio a widely used index of oxidative stress. GSSG is measured after masking GSH, and both GSSG and total glutathione (GSH + GSSG) concentrations are determined spectrophotometrically at 412 nm using DTNB (5,5’-dithiobis (2-nitrobenzoic acid)) and an enzymatic recycling system. GSH concentration is then calculated by subtracting the amount of GSSG from the total amount of glutathione. Levels of GSH are expressed in µmol/L and the GSH:GSSG ratio is in arbitrary units (AU). Measurements were performed in erythrocytes, hence limiting the analysis to individuals for whom sufficient erythrocyte volume was available, and reducing the final sample size. Average intra-plate CV was 5 % for the total glutathione assay and 8 % for the GSSG assay (kit reference number: 38185). 2.4. Plasma haemoglobin concentrations Plasma haemoglobin concentrations were measured to account for the impact of haemolysis on the values of oxidative stress biomarkers (Herrada et al. 2024b). The haemoglobin assay kit (Sigma-Aldrich) utilises the Triton/NaOH method, wherein haems undergo oxidation in alkaline conditions to produce a colorimetric product detectable via spectrophotometry at 400 nm. Levels are determined by referencing against a standard curve and expressed in mg/dL. Average intra-plate CV was less than 10 %. 2.5. Weather conditions Weather data, i.e. daily temperatures and rainfall, were obtained from the Météo-France database (“Météo-France” 2023), which provides records with a spatial resolution of 8x8 km throughout France since 1958. After targeting each study site (i.e. CH and TF forests), we averaged temperatures and summed rainfall over the week preceding the capture date of each individual, as these values reflect short-term, local weather conditions experienced by each individual prior to sampling. This time frame was chosen to capture recent weather conditions, while taking into account the relatively rapid dynamics of oxidative stress responses (Costantini 2019). Meanwhile, we are aware that weather conditions may also influence metal accumulation in hair. Although not the primary focus of our study, we explored the relationships between toxic metal concentrations in roe deer hair and average weather conditions (i.e. temperature and rainfall) during hair growth and thus metal accumulation. These results, along with a brief discussion, are provided in the supporting information (Fig. S2). 2.6. Statistical analyses All statistical analyses were performed using R (version 4.2.1; R core Team, 2021). Statistical analyses were performed on a maximum of 649 samples corresponding to 434 different individuals. A total of 362 plasma samples were collected in CH (198 females and 164 males) and 287 in TF (147 females and 140 males). Between 2016 and 2019, 87 individuals were captured twice, 46 individuals were captured three times, and 12 individuals were captured every year (i.e. four times). Age ranged from 1 to 14 years. We first summarised inter-individual variation among all 9 toxic metals, and among all 13 essential metals, respectively, using Principal Component Analyses (PCA) performed with the dudi.pca function from the ade4 R-package (Thioulouse et al. 2018). Element concentrations were first log-transformed to respect normality assumptions, and then standardised to minimise variance between variables. We considered all data from 2016 to 2019, including repeated measurements on the same individuals, as well as both populations, as PCA results remain similar without repeated measurements, and when considering populations separately, respectively (Figs. S2 & S3). The advantages of this multivariate approach are three-fold: (i) minimising the number of potential explanatory variables, (ii) discarding the problem of collinearity between explanatory variables by obtaining uncorrelated synthetic variables, and (iii) capturing the covariation between variables across individuals (Thioulouse et al. 2018). To preserve as much variance as possible, we chose to keep the axes with the largest eigenvalues. For the PCA on toxic metals, the first two principal components (PCT1 and PCT2) were selected as they captured a substantial proportion of the variance (46.8% and 17.4%, respectively), while PCT3 only explained a limited amount of that variance (i.e. 8.8%). The first two dimensions were therefore used as integrative indices of inter-individual variation in toxic metal exposure. By contrast, for the PCA on essential metals, the first principal component (PCE1) captured 31.1% of the variance, while the proportion of variance explained by successive axes dropped from the second axis onwards and did not differ much between those axes (i.e. PCE2 = 13.0%, PCE3 = 11.0%, PCE4 = 8.4%), making it less straightforward to justify the inclusion of additional components. To avoid overfitting and facilitate interpretability of the following analyses, we chose to keep only the first principal component to be used as an integrative index of inter-individual variation in essential metal content. Then, for each oxidative stress biomarker, we fitted two separate models to analyse the relationships between metal concentrations and redox homeostasis: one testing the effect of PCT1 and another testing the effect of PCT2. Using linear mixed-effects models performed with the lmer function from the lme4 R-package (Bates et al. 2015), each model included as fixed effects either PCT1 or PCT2, PCE1, cumulative rainfall, mean ambient temperature, as well as their interactions (PCT x PCE1, PCT x cumulative rainfall, and PCT x mean ambient temperature). Indeed, we expected the influence of toxic metals on redox homeostasis to be stronger in individuals with minor concentrations of essential metals, and to differ in individuals that had experienced contrasted local weather conditions prior to capture, such as heavy rainfall and/or harsh winter temperatures. Since factors other than essential element content and recent weather conditions may influence biomarkers values, we also included the additive effects of body mass, age, capture year (2016 to 2019) and plasma haemoglobin levels (except for GSH and GSH:GSSG ratio, which were measured in erythrocytes) as quantitative explanatory variables, and sex and population as qualitative explanatory variables. We effectively expected an increase in oxidative damage and a decrease in antioxidant defences with increasing age, notably considering that alteration of redox homeostasis in response to chemical pollution may be age-dependent (Martin et al. 2024), and with decreasing body mass (Montes et al. 2011). We also expected higher damage and lower defences in males compared to females, and in CH compared to TF due to its lower habitat quality (see Metcalfe & Alonso-Alvarez, 2010 for review). Individual identity (ID) was set as a random effect to consider the non-independence between repeated measurements on the same individuals. Note that entering ID as a random effect resulted in singularity errors for models with MDA, protein carbonyls, SOD, GSH and GSH:GSSG ratio as dependent variables. In that case, a linear model was computed with the lm function from the stats package, and a single measurement per individual was randomly selected for the analysis. All continuous variables were scaled before analysis. MDA, protein carbonyls, GSH and GSH:GSSG ratio were also log-transformed before scaling to respect normality assumptions. For fawns (< 1 year old), body mass was recalculated as their predicted mass on the 27 th of January (i.e. the median date of captures), to account for their growth over winter and thus to ensure comparability between individuals captured at different dates (Douhard et al. 2017). We implemented model selection from the full model based on the Akaike Information Criteria (AICc), using the dredge function from the R-package MuMin (Bartoń 2013). We retained all models within 2 ΔAICc (i.e. difference in AICc between a given model and the model with the lowest AICc ; Burnham and Anderson, 2002), excluding those with uninformative parameters (Arnold 2010). We then selected the model with the fewest parameters in accordance with the principle of parsimony. When models had the same number of parameters, we chose the one with the lowest ΔAICc value. The retained models within 2 ΔAICc, as well as the results of the selected models, are presented in Table S2 and S3, respectively. Normality and homoscedasticity of models’ residuals were checked by visual inspection with the check_model function from the performance R-package (Lüdecke et al. 2021). For each variable, we present its coefficient (Estimate), as well as its 95 % confidence intervals (95% CI). We also provide the marginal and conditional R² for all models. 3. Results 3.1. Integrative toxic metal exposure indices The first two principal components explained 64.26 % of the total inter-individual variation in toxic metal burdens (Fig. 1). The first axis (PCT1) captured 46.8 % of the variation in toxic metal concentrations. PCT1 received loadings particularly from Al, As, Pb, Ti, Tl, and to a lesser extent from Sb and Sr. The second axis (PCT2) captured 17.4 % of the inter-individual variation in toxic metal concentrations and received loadings mostly from Cd and Hg. Overall, PCA results revealed that individuals with a high PCT1 and/or PCT2 coordinate were characterised by higher toxic metal burden. PCT1 and PCT2 were then interpreted as integrative indices of toxic metal exposure. Note that, individuals from the two populations are not homogeneously distributed along PCA axes, particularly the second one, suggesting higher concentrations of Cd and Hg at TF than at CH. Figure 1 . First factorial plan from the Principal Component Analysis (PCA) performed on 9 toxic metals in roe deer (n = 649). (a) Loadings on the first two axes (PCT1 and PCT2), with colours representing the contribution of the different metals to the first axis, (b) individuals, with green triangles representing individuals from Trois-Fontaines (TF) and lilac dots representing individuals from Chizé (CH), and (c) proportion of inertia contributed by each PC. 3.2. Integrative essential metal content index The first principal component (PCE1) explained 31.1 % of the total inter-individual variation in essential metal concentrations (Fig. 2). PCE1 received loadings mainly from Ca, Co, Cr, Fe, Mn and Ni. Overall, the PCA results revealed that individuals with a high PCE1 coordinate were characterised by higher essential metal concentrations. PCE1 was therefore interpreted as an integrative index of essential metal content. Note that individuals from TF also exhibited higher essential metal content than those of CH (t-test, t = -4.264, df = 647, p < 0.001). Figure 2 . Principal Component Analysis performed on 13 essential metals in roe deer (n = 649). (a) Loadings on the first axis (PCE1), (b) differences in PCE1 score between populations, with green colour representing individuals from Trois-Fontaines (TF) and lilac colour representing individuals from Chizé (CH), and (c) proportion of inertia contributed by each PC. 3.3. Relationships between oxidative stress biomarkers, toxic and essential metals, and weather conditions 3.3.1. Relative contribution of toxic metal exposure on oxidative stress biomarkers PCT1 coordinates were associated with two antioxidant defence biomarkers (SOD activity and OXY levels), and with the oxidative status marker (i.e. GSH:GSSG ratio). By contrast, PCT2 coordinates were associated with all biomarkers but non-enzymatic antioxidant levels (GSH) and the GSH:GSSG ratio (Tables S3,S4; Figs. 3,4). More specifically, higher toxic metal burdens resulted in higher oxidative damage on lipids (MDA) and proteins (Carbonyls), as well as higher reactive oxygen metabolite levels (dROMs) (Table S4; Fig. 4a-c). Associations between toxic metals and antioxidant defences were contrasted as higher toxic metal concentrations were associated with higher total antioxidant capacity (OXY), but lower SOD activity (Tables S3,S4; Fig. 3a,b; Fig. 4d,e). Note that for most markers, effect sizes suggested relatively small associations (Fig. 5). Since significant interactions between toxic metals and weather variables were also detected, the associations between toxic metals and biomarkers of oxidative stress may vary according to the weather context and should be interpreted accordingly. 3.3.2. Essential metals and their interactions with toxic metals PCE1 values were positively associated with plasma total antioxidant capacity (OXY), negatively associated with the GSH:GSSG ratio, and showed either positive or negative associations with enzymatic antioxidant activity (SOD), depending on which toxic metal index is considered (Tables S3,S4). However, the effects on SOD and OXY levels were particularly small (Fig. 5). As for interacting terms, the interaction between toxic and essential metal indices was never retained in the final models (Table S2). 3.3.3. Weather conditions and their interactions with toxic metals Regarding the effects of weather variables alone, rainfall prior to capture was never statistically significant. On the contrary, higher ambient temperatures resulted in higher oxidative damage on lipids (MDA) and a higher GSH:GSSG ratio, but lower reactive oxygen metabolite levels (dROMs) and enzymatic antioxidant activity (SOD) (Tables S3,S4). Significant interactions between toxic metals and weather conditions were found for six biomarkers, indicating that individuals subjected to higher toxic metal burdens and higher ambient temperatures or higher cumulative rainfall showed a greater increase in oxidative damage (dROMs, MDA, Carbonyls), antioxidant defences (OXY) and the GSH:GSSG ratio, as well as lower enzymatic antioxidant activity (SOD) (Tables S3,S4; Fig. 3c; Fig. 4). This supports that the relationship between toxic metals and oxidative stress may be partly modulated by local weather conditions. Effect sizes revealed, however, that these associations were small to almost null (Fig. 5). Figure 3. Relationships between individual coordinates on the first axis of the PCA performed on toxic metals (i.e. PCT1, mainly supported by Al, As, Pb, Sb, Sr, Ti and Tl) and levels of antioxidant defence biomarkers (SOD, OXY) and oxidative status (GSH:GSSG). Points are raw data and lines with shaded areas represent the predictions from linear models with 95 % confidence intervals. Relationships between PCT1 and (a) OXY, (b) SOD, and (c) GSH:GSSG ratio, where individuals are coloured according to the mean ambient temperatures over the week preceding capture. For the purpose of graphical representation, individuals were divided into 2 groups based on the mean ambient temperatures: low ( median of mean temperatures). Figure 4. Relationships between individual coordinates on the second axis of the PCA performed on toxic metals (i.e. PCT2, mostly supported by Cd and Hg) and levels of biomarkers of oxidative damage (dROMs, MDA, Carbonyls) and antioxidant defences (OXY, SOD). Points are raw data and lines with shaded areas represent the predictions from linear models with 95 % confidence intervals. Relationship between PCT2 and (a) dROMs, (b) MDA, (c) Carbonyls, (d) OXY, and (e) SOD. Individuals are coloured according to the mean ambient temperatures or the cumulative rainfall over the week preceding capture. For the purpose of graphical representation, individuals were divided into 2 groups based on the weather factor: low ( median value). Figure 5. Standardised regression coefficients (β) and 95 % confidence intervals for the effects of toxic metal exposure indices (i.e. (a) PCT1 and (b) PCT2), essential metal content index (PCE1) and their interaction (PCT:PCE1), cumulative rainfall (Rain) and their interaction (PC:Rain), mean ambient temperature (Temp) and their interaction (PCT:Temp), and capture date on biomarkers of oxidative damage (dROMs, MDA, Carbonyls), antioxidant defences (OXY, SOD, GSH) and oxidative status (GSH:GSSG). Each oxidative stress biomarker is depicted by a specific colour and symbol. 4. Discussion We hypothesised that individuals with substantial toxic metal burdens would be more prone to oxidative stress than those with lower burdens, hence displaying higher oxidative damage and lower antioxidant defences. While essential metal content did not appear to modulate any of the associations between toxic metals and oxidative stress biomarkers, weather conditions prior to capture slightly modulated them. 4.1. The relative contribution of toxic metals to oxidative stress Our results strongly support that exposure to toxic metals promotes oxidative stress in wild roe deer from both studied populations, as evidenced by the positive relationships between toxic metal concentrations in hair and the accumulation of oxidised lipids and proteins, and to a lesser extent, other derivatives of reactive oxygen metabolites in plasma. The shift in oxidative balance was also evident from the negative correlation between toxic metal concentrations and the SOD activity, a key antioxidant enzyme acting as a major endogenous ROS scavenger. Only the overall non-specific antioxidant capacity of plasma deviated from our prediction, showing an unexpected positive association with toxic metal concentrations. These associations were particularly pronounced with PCT2, an integrative index of exposure to Cd and Hg, which emerged as a predictor for all but two biomarkers. When focusing on PCT1, reflecting the exposure to Al, As, Pb, Ti, Tl, Sb and Sr, an association was detected with only three biomarkers. Overall, our findings are particularly compelling as it was recently reported that the concentrations of toxic metals in hair of roe deer from the two populations studied here are relatively low compared to other populations of roe deer and other large ungulates (Herrada et al. 2024a). This therefore suggests that even low-level exposure to toxic metals has the potential to affect redox homeostasis in a wild mammal. Field studies on the impact of toxic metals on mammal physiology remain limited. A few studies nevertheless focused on long-term contamination, showing results consistent with ours. For example, Pb exposure has been associated with increased lipid peroxidation in the liver of red deer ( Cervus elaphus ; Rodríguez-Estival et al., 2011) and decreased SOD activity in the spleen of wild boars ( Sus scrofa ; Rodríguez-Estival et al., 2013). Total antioxidant capacity (TAC) also increased with increasing Cd and Pb renal concentrations in brown bears ( Ursus arctos ; Lazarus et al., 2024). However, other trends were also evidenced, such as an increase in glutathione content and a decrease in reactive oxygen species with increasing Cd and Pb concentrations in brown bears (Lazarus et al. 2024). As well as a decrease in GSH levels in the spleen of red deer and wild boars with increasing Pb concentrations (Rodríguez-Estival et al. 2013). Note, however, that the above-cited studies measured metals in organs rather than hair. While hair concentrations are a reliable predictor of metal content in the environment, and have been used to estimate organ concentrations, correlations between hair and other matrices remain often limited (e.g. D’Havé et al., 2006; Tête et al., 2014; Vuarin et al., 2025). Additionally, toxic metal concentrations in roe deer hair reflect accumulation over a few months (e.g. Sempéré et al., 1996), when hair grew and their roots were still connected to the bloodstream (McLean et al. 2009). By contrast, oxidative stress biomarkers measured in blood are transient and reflect circulating levels within a much narrower period. Although this temporal mismatch may affect the interpretation of the observed associations, the fact that we found significant relationships between toxic metals and oxidative stress biomarkers, and that those relationships are mostly of similar direction for markers that represent comparable processes (e.g. oxidative damage), supports that the observed oxidative profiles likely reflect the cumulative effects of chronic exposure. The fact that almost all plasma and cellular oxidative stress biomarkers were related to hair toxic metal content highlights the broad impact of these elements and indicates that various components of redox homeostasis are affected. Considering that each biomarker may reflect distinct biochemical processes (Christensen et al. 2015), this also underscores the complexity of the interactions between oxidative stress and metals, and the varying sensitivities of each biomarker to different metals. For example, the bimetallic CuZn-SODs that can be found in the cytosol (SOD1 isozyme) and in the extracellular space (SOD3 isozyme), are directly affected by redox-inactive metals such as As, Cd, Hg, and Pb. Indeed, these metals may interact with the SODs’ cofactors, preventing their normal functioning at least in mammals, which normally consists in the dismutation of superoxide anions into less harmful hydrogen peroxides (H 2 O 2 ) (Nzengue et al. 2011, Kumar et al. 2020). This mechanism may therefore play a role in the reduced SOD activity observed in roe deer with relatively elevated toxic metal content. In such case, other endogenous, but also exogenous, antioxidants may help compensate for specific antioxidant decline, resulting in differences between antioxidants’ response to metals (e.g. Losdat et al., 2011). Such mechanisms may explain the observed increase in the overall non-specific antioxidant capacity (OXY) observed in roe deer. 4.2. The role of essential metals in toxic metal-induced oxidative stress Several essential metals are directly or indirectly involved in maintaining redox homeostasis. They can even partly modulate the effects of toxic metals on the oxidative balance. For instance, Se is a component of glutathione peroxidase (GPx) and has thus an indirect protective role against oxidative damage, especially from hydrogen peroxides (Tapiero et al. 2003, Soetan et al. 2010). As already mentioned, Cu and Zn are crucial for redox homeostasis since they act as cofactors in SOD1 and SOD3 enzymes, as well as Mn which acts as a cofactor in the mitochondrial SOD2 enzyme (Kumar et al. 2020). Therefore, individuals with substantial concentrations of such essential metals would benefit from their protective effect and be expected to exhibit less pronounced relationships between oxidative stress biomarkers and toxic metals. However, our analyses did not reveal any mitigation of toxic effects by essential metals. Our results might therefore indicate that the physiological disturbance induced by toxic metals did not exceed a threshold where compensation mechanisms involving essential elements would be mobilised. At first, we highlighted relationships between most biomarkers and toxic metals, and it was particularly evident for Cd and Hg. However, while observing these associations, we need to consider that they may not necessarily translate into long-term physiological consequences. It is indeed plausible that the antioxidant repair mechanisms may efficiently counteract oxidative damage. This raises the question of the biological relevance of the observed relations between biomarkers and toxic metals and would require identifying threshold levels beyond which ageing hallmarks would be affected, such as mitochondrial dysfunction or inflammation (López-Otín et al. 2023), with expected health consequences. Unfortunately, such thresholds are poorly documented and difficult to determine (Lushchak 2014). Considering effect sizes when interpreting our results may ultimately be the safest and most accessible way to characterise the degree of impairment of redox homeostasis in our populations. 4.3. The influence of weather conditions on metal toxicity Oxidative stress may occur and persist for hours, days or weeks following exposure to a stressor, depending on its nature and intensity. Among potential stressors, weather conditions have received some attention, including extreme temperatures and rainfall (Metcalfe and Alonso-Alvarez 2010, Lushchak 2011, Chainy et al. 2016). In our study, warmer local temperatures the week preceding capture were associated with higher oxidative damage to lipids and lower antioxidant capacity, and milder local temperatures correlated with greater oxidative damage to lipids, a higher GSH:GSSG ratio, and lower reactive oxygen metabolites and antioxidant activity, though most of these effects depended on interactions with toxic metal concentrations. No significant effect of cumulative rainfall was detected. Importantly, we observed differences between populations, with individuals from TF exhibiting higher oxidative damage, lower antioxidant defences and a lower oxidative status than individuals from CH. While this may reflect contrasted environmental conditions between the two study sites, it could also result from differences in metal exposure, as we showed higher toxic metal content in TF on average, especially for Cd and Hg. When examining the influence of weather conditions on toxic metal-induced oxidative stress, our results suggest that it is indeed, albeit modestly, modulated by temperature and rainfall. Specifically, we found that the effects of toxic metals on oxidative stress biomarkers were overall amplified under higher local temperatures and higher cumulative rainfall. Although effect sizes were small, these interactions seem consistent with the limited literature on the subject. For example, studies have shown that increased temperatures tend to exacerbate the toxic effects of metals on organisms, notably by intensifying oxidative stress (see Sokolova & Lannig, 2008; Xiao et al., 2024 for reviews on aquatic organisms). Hence, these findings underscore that short-term weather conditions may subtly influence oxidative stress responses to toxic metals, an effect that should not be overlooked in future field studies. 4.4. Limits of the study Since each oxidative stress biomarker captures a different facet of the redox status, interpretation of results may remain incomplete if biomarkers are analysed separately. For instance, the TBARS assay, which is based on absorbance spectrophotometry and supposedly estimates MDA as a proxy for lipid peroxidation, has well-documented drawbacks. Indeed, since TBA may react non-specifically with substances other than MDA in the sample, it can artifactually generate lipid peroxidation and result in an overestimation of MDA levels (Halliwell & Gutteridge, 2015; Murphy et al., 2022; but see Aguilar Diaz De Leon & Borges, 2020 for an updated use of this method). Other absorbance spectrophotometry-based biomarker assays used in this study might also be criticised for their low reproducibility and non-specificity (e.g. OXY, dROMs; Murphy et al., 2022). Nevertheless, the use of seven oxidative stress biomarkers that reflect different but complementary processes (i.e. lipid and protein oxidation, enzymatic and non-enzymatic defences, overall oxidative status through the GSH:GSSG ratio), should, at least partly, mitigate the above-mentioned issues, and provide an integrative view of the redox status when combined and interpreted together. Most importantly, the results show that the biomarkers follow the expected patterns, and co-vary on a given side of the balance (e.g. all markers of oxidative damage increase with PCT2 values), hence reinforcing the conclusions drawn from the findings of this study. Since redox reactions can be cell- and tissue-specific (e.g. Costantini, 2019), we are aware that the measurements of biomarkers in blood (either plasma or erythrocytes) may not always reflect oxidative dynamics elsewhere in the body. Ideally, oxidative stress biomarkers should be measured in other tissues as well (e.g. liver, muscle, kidney, etc.), and ultimately, so should toxic metals (see section 4.1.). However, this study uses data from a long-term capture-recapture program, where only non-destructive techniques can be applied to collect biological samples. While hair and blood may remain imperfect proxies for internal tissue exposure and oxidative status assessments, they are classically used in ecophysiology studies, making the comparison between studies easy. We also sought to control known biases associated with these matrices, for instance by accounting for the effect of haemolysis on the measurement of plasma biomarkers (Herrada et al. 2024b). Finally, although several detected effects were statistically significant, their magnitude was often modest, raising the question of their biological relevance. Future studies could evaluate whether such physiological alterations have measurable impacts on individual fitness and demographic trajectories (e.g. Ouyang et al., 2016). It is also important to keep in mind that the reported associations between toxic metal exposure and oxidative stress biomarkers do not imply causal relationships. We cannot rule out that the observed oxidative profiles could be shaped, at least partially, by unmeasured environmental factors (e.g. availability and quality of resources) or individual state (e.g. parasite burden). 5. Conclusion Key results of the present study revealed a clear contribution of toxic metals on redox homeostasis in wild roe deer, despite relatively low-level exposure. There was, however, no clear support for a modulation of these effects by essential metals, and only a moderate one by local weather conditions. Our study therefore adds valuable data to the scattered knowledge on the relationship between exposure to toxic metals and oxidative stress in natural populations. It also improves our understanding of the potential, although overlooked, impact of low-level metal exposure on wildlife physiology. Finally, as oxidative stress has been reported to promote ageing, to impact health and ultimately survival (Finkel and Holbrook 2000, Halliwell and Gutteridge 2015), the next step would be to analyse whether toxic metal-induced disruption of redox homeostasis actually contributes to these processes. Only a handful of studies have suggested an implication of toxic chemical exposure for population dynamics in the wild, focusing on marine birds (e.g. Goutte et al., 2014, 2018). Our study thus calls for greater attention to the physiological costs of chronic toxic metal exposure in terrestrial animals. Acknowledgements We are grateful to all the technicians and volunteers who participated in the captures and data collection in Chizé and Trois-Fontaines forests. We also would like to thank Louise Cheynel, Lou Decosne, Alexia Gache and Yoan Regnaud for their participation in the assessment of oxidative stress biomarkers. Competing interests The authors declare no conflicts of interest. Funding This work was carried out with the financial support of the University of Lyon's IDEXLYON Project as part of the “Programme Investissement d'Avenir” (ANR-16-IDEX-0005). Ethical notes Game captures were conducted in accordance with European and French laws. The experiment was designed to minimise animal stress and handling time, and to ensure animal welfare, as defined in the guidelines for the ethical use of animals in research. Captures are carried out in accordance with the French Environmental Code (Art. R421-15 to 421-31 and R422-92 to 422-94-1) and approved by the Prefecture of Paris (agreement n°2009-14, n°2013-118, n°2019-02-19-003). The Trois-Fontaines population is part of the long-term Studies in Ecology and Evolution (SEE-Life) program of the CNRS. Data accessibility The datasets and codes supporting the findings of this study are available via a zenodo repository: https://doi.org/10.5281/zenodo.17867014 . References Aguilar Diaz De Leon, J. and Borges, C. R. (2020). Evaluation of oxidative stress in biological samples using the thiobarbituric acid reactive substances assay. J Vis Exp . Arnold, T. W. (2010). Uninformative parameters and model selection using Akaike’s Information Criterion. The Journal of Wildlife Management 74 , 1175–1178. Azeh Engwa, G., Udoka Ferdinand, P., Nweke Nwalo, F. and N. Unachukwu, M. (2019). Mechanism and health effects of heavy metal toxicity in humans. In Poisoning in the Modern World - New Tricks for an Old Dog? (ed. Karcioglu, O.) and Arslan, B.), p. IntechOpen. Bartoń, K. (2013). MuMIn: Multi-model inference. R package version 1.10.0. Bates, D., Mächler, M., Bolker, B. and Walker, S. (2015). Fitting linear mixed-effects models using lme4 . Journal of Statistical Software 67 ,. Bilham, K., Newman, C., Buesching, C. D., Noonan, M. J., Boyd, A., Smith, A. L. and Macdonald, D. W. (2018). Effects of weather conditions on oxidative stress, oxidative damage, and antioxidant capacity in a wild-living mammal, the European badger ( Meles meles ). Physiological and Biochemical Zoology 91 , 987–1004. Bonnot, N., Morellet, N., Verheyden, H., Cargnelutti, B., Lourtet, B., Klein, F. and Hewison, A. J. M. (2013). Habitat use under predation risk: Hunting, roads and human dwellings influence the spatial behaviour of roe deer. European Journal of Wildlife Research 59 , 185–193. Boyd, R. S. (2010). Heavy metal pollutants and chemical ecology: Exploring new frontiers. Journal of Chemical Ecology 36 , 46–58. Burnham, K. P. and Anderson, D. R. (2002). Model selection and multimodel inference: A practical information-theoretic approach . 2nd ed. New York: Springer. Burton, G. J. and Jauniaux, E. (2011). Oxidative stress. Best Practice & Research Clinical Obstetrics & Gynaecology 25 , 287–299. Cappelli, J., Frasca, I., García, A., Landete-Castillejos, T., Luccarini, S., Gallego, L., Morimando, F., Varuzza, P. and Zaccaroni, M. (2020). Roe deer as a bioindicator: Preliminary data on the impact of the geothermal power plants on the mineral profile in internal and bone tissues in Tuscany (Italy). Environmental Science and Pollution Research 27 , 36121–36131. Chainy, G. B. N., Paital, B. and Dandapat, J. (2016). An overview of seasonal changes in oxidative stress and antioxidant defence parameters in some invertebrate and vertebrate species. Scientifica 2016 , 6126570. Christensen, L. L., Selman, C., Blount, J. D., Pilkington, J. G., Watt, K. A., Pemberton, J. M., Reid, J. M. and Nussey, D. H. (2015). Plasma markers of oxidative stress are uncorrelated in a wild mammal. Ecology and Evolution 5 , 5096–5108. Combs, D., Goodrich, R. and Meiske, J. (1982). Mineral concentrations in hair as indicators of mineral status: A review. Journal of animal science 54 , 391–8. Costantini, D. (2019). Understanding diversity in oxidative status and oxidative stress: The opportunities and challenges ahead. Journal of Experimental Biology 222 , jeb194688. D’Havé, H., Scheirs, J., Mubiana, V. K., Verhagen, R., Blust, R. and De Coen, W. (2006). Non-destructive pollution exposure assessment in the European hedgehog ( Erinaceus europaeus ): II. Hair and spines as indicators of endogenous metal and As concentrations. Environmental Pollution 142 , 438–448. Douhard, F., Gaillard, J.-M., Pellerin, M., Jacob, L. and Lemaître, J.-F. (2017). The cost of growing large: Costs of post-weaning growth on body mass senescence in a wild mammal. Oikos 126 , 1329–1338. Draghi, S., Agradi, S., Riva, F., Tarhan, D., Bilgiç, B., Dokuzeylül, B., Ercan, A. M., Or, M. E., Brecchia, G., Vigo, D., et al. (2023). Roe deer ( Capreolus capreolus ) hair as a bioindicator for the environmental presence of toxic and trace elements. Toxics 11 , 49. Finkel, T. and Holbrook, N. J. (2000). Oxidants, oxidative stress and the biology of ageing. Nature 408 , 239–247. Flerov, K. K. (1952). Fauna of USSR, Mammals Vol. 1, No. 2. Musk Deer and Deer . USSR Academia Science. Gaillard, J.-M., Delorme, D., Jean-Marie, B., Van Laere, G., Boisaubert, B. and Pradel, R. (1993). Roe Deer Survival Patterns: A Comparative Analysis of Contrasting Populations. Journal of Animal Ecology 62 , 778–791. Gaillard, J.-M., Mark Hewison, A. J., Klein, F., Plard, F., Douhard, M., Davison, R. and Bonenfant, C. (2013). How does climate change influence demographic processes of widespread species? Lessons from the comparative analysis of contrasted populations of roe deer. Ecology Letters 16 , 48–57. García, M. H. de M., Moreno, D. H., Rodríguez, F. S., Beceiro, A. L., Álvarez, L. E. F. and López, M. P. (2011). Sex- and age-dependent accumulation of heavy metals (Cd, Pb and Zn) in liver, kidney and muscle of roe deer ( Capreolus capreolus ) from NW Spain. Journal of Environmental Science and Health, Part A 46 , 109–116. Goutte, A., Barbraud, C., Meillère, A., Carravieri, A., Bustamante, P., Labadie, P., Budzinski, H., Delord, K., Cherel, Y., Weimerskirch, H., et al. (2014). Demographic consequences of heavy metals and persistent organic pollutants in a vulnerable long-lived bird, the wandering albatross. Proceedings of the Royal Society B: Biological Sciences 281 , 20133313. Goutte, A., Meillère, A., Barbraud, C., Budzinski, H., Labadie, P., Peluhet, L., Weimerskirch, H., Delord, K. and Chastel, O. (2018). Demographic, endocrine and behavioral responses to mirex in the South polar skua. Science of The Total Environment 631–632 , 317–325. Govind, P. (2014). Heavy metals causing toxicity in animals and fishes. Research Journal of Animal, Veterinary and Fishery Sciences 2 , 17–23. Halliwell, B. and Gutteridge, J. M. C. (2015). Free radicals in biology and medicine . Oxford University Press. Herrada, A., Bariod, L., Saïd, S., Rey, B., Bidault, H., Bollet, Y., Chabot, S., Débias, F., Duhayer, J., Pardonnet, S., et al. (2024a). Minor and trace element concentrations in roe deer hair: A non-invasive method to define reference values in wildlife. Ecological Indicators 159 , 111720. Herrada, A., Vuarin, P., Débias, F., Gache, A., Veber, P., Pellerin, M., Cheynel, L., Lemaître, J.-F., Gilot-Fromont, E. and Rey, B. (2024b). Haemolysis overestimates plasma oxidative stress biomarkers in free-ranging roe deer. Comparative Biochemistry and Physiology Part A: Molecular & Integrative Physiology 298 , 111750. Iavicoli, I., Fontana, L. and Bergamaschi, A. (2009). The effects of metals as endocrine disruptors. Journal of Toxicology and Environmental Health, Part B 12 , 206–223. Isaksson, C. (2010). Pollution and its impact on wild animals: A meta-analysis on oxidative stress. EcoHealth 7 , 342–350. Johnson, E. and Hornby, J. (1975). Seasonal changes of pelage in the roe deer Capreolus capreolus and its role in thermoregulation. Journal of Natural History 9 , 619–628. Kumar, A., Khushboo, Pandey, R. and Sharma, B. (2020). Modulation of superoxide dismutase activity by mercury, lead, and arsenic. Biological Trace Element Research 196 , 654–661. Lazarus, M., Sergiel, A., Orct, T., Ferenčaković, M., Tariba Lovaković, B., Žunec, S., Rašić, D., Oster, E., Reljić, S., Macur, J., et al. (2024). Apex carnivores coping with metal(loid) pollution and oxidative stress: Biological and environmental drivers of variation in kidney of European brown bear. Environmental Pollution 363 , 125285. Linnell, J. and Andersen, R. (1995). Site tenacity in roe deer: Short term effects of logging. Wildlife Society Bulletin 23 , 31–36. López-Otín, C., Blasco, M. A., Partridge, L., Serrano, M. and Kroemer, G. (2023). Hallmarks of aging: An expanding universe. Cell 186 , 243–278. Losdat, S., Helfenstein, F., Gaude, B. and Richner, H. (2011). Reproductive effort transiently reduces antioxidant capacity in a wild bird. Behavioral Ecology 22 , 1218–1226. Lüdecke, D., Ben-Shachar, M., Patil, I., Waggoner, P. and Makowski, D. (2021). performance: An R package for assessment, comparison and testing of statistical models. Journal of Open Source Software 6 , 3139. Ludolphy, C., Kierdorf, U. and Kierdorf, H. (2021). Lead concentrations in antlers of European roe deer ( Capreolus capreolus ) from an agricultural area in Northern Germany over a 119-year period - a historical biomonitoring study. Environmental Science and Pollution Research 28 , 56069–56078. Lushchak, V. I. (2011). Environmentally induced oxidative stress in aquatic animals. Aquatic Toxicology 101 , 13–30. Lushchak, V. I. (2014). Classification of oxidative stress based on its intensity. Experimental and Clinical Sciences Journal 13 , 922–937. Martin, C., Capilla-Lasheras, P., Monaghan, P. and Burraco, P. (2024). The impact of chemical pollution across major life transitions: a meta-analysis on oxidative stress in amphibians. Proc. R. Soc. B. 291 ,. McLean, C. M., Koller, C. E., Rodger, J. C. and MacFarlane, G. R. (2009). Mammalian hair as an accumulative bioindicator of metal bioavailability in Australian terrestrial environments. The Science of the Total Environment 407 , 3588–3596. Metcalfe, N. B. and Alonso-Alvarez, C. (2010). Oxidative stress as a life-history constraint: the role of reactive oxygen species in shaping phenotypes from conception to death. Functional Ecology 24 , 984–996. Météo-France (2023). Montes, I., Newman, C., Mian, R. and Macdonald, D. W. (2011). Radical health: Ecological corollaries of body condition, transport stress and season on plasma antioxidant capacity in the European badger. Journal of Zoology 284 , 114–123. Moreau, J., Monceau, K., Crépin, M., Tochon, F. D., Mondet, C., Fraikin, M., Teixeira, M. and Bretagnolle, V. (2021). Feeding partridges with organic or conventional grain triggers cascading effects in life-history traits. Environmental Pollution 278 , 116851. Murphy, M. P., Bayir, H., Belousov, V., Chang, C. J., Davies, K. J. A., Davies, M. J., Dick, T. P., Finkel, T., Forman, H. J., Janssen-Heininger, Y., et al. (2022). Guidelines for measuring reactive oxygen species and oxidative damage in cells and in vivo. Nature Metabolism 4 , 651–662. Nzengue, Y., Candéias, S. M., Sauvaigo, S., Douki, T., Favier, A., Rachidi, W. and Guiraud, P. (2011). The toxicity redox mechanisms of cadmium alone or together with copper and zinc homeostasis alteration: Its redox biomarkers. Journal of Trace Elements in Medicine and Biology 25 , 171–180. Ouyang, J. Q., Lendvai, Á. Z., Moore, I. T., Bonier, F. and Haussmann, M. F. (2016). Do hormones, telomere lengths, and oxidative stress form an integrated phenotype? A case study in free-living tree swallows. Integrative and Comparative Biology 56 , 138–145. Pellerin, M., Calenge, C., Saïd, S., Gaillard, J.-M., Fritz, H., Duncan, P. and Van Laere, G. (2010). Habitat use by female western roe deer ( Capreolus capreolus ): Influence of resource availability on habitat selection in two contrasting years. Canadian Journal of Zoology 88 , 1052–1062. Pettorelli, N., Gaillard, J.-M., Mysterud, A., Duncan, P., Chr. Stenseth, N., Delorme, D., Van Laere, G., Toïgo, C. and Klein, F. (2006). Using a proxy of plant productivity (NDVI) to find key periods for animal performance: The case of roe deer. Oikos 112 , 565–572. R core Team (2021). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Rashed, M. N. and Soltan, M. E. (2005). Animal hair as biological indicator for heavy metal pollution in urban and rural areas. Environmental monitoring and assessment 110 , 41–53. Rattner, B. A. (2009). History of wildlife toxicology. Ecotoxicology 18 , 773–783. Rodríguez-Estival, J., Martinez-Haro, M., Monsalve-González, L. and Mateo, R. (2011). Interactions between endogenous and dietary antioxidants against Pb-induced oxidative stress in wild ungulates from a Pb polluted mining area. Science of The Total Environment 409 , 2725–2733. Rodríguez-Estival, J., de la Lastra, J. M. P., Ortiz-Santaliestra, M. E., Vidal, D. and Mateo, R. (2013). Expression of immunoregulatory genes and its relationship to lead exposure and lead-mediated oxidative stress in wild ungulates from an abandoned mining area. Environmental Toxicology and Chemistry 32 , 876–883. Rodríguez-Estival, J., García-de Blas, E. and Smits, J. E. G. (2016). Oxidative stress biomarkers indicate sublethal health effects in a sentinel small mammal species, the deer mouse ( Peromyscus maniculatus ), on reclaimed oil sands areas. Ecological Indicators 62 , 66–75. Saaristo, M., Brodin, T., Balshine, S., Bertram, M. G., Brooks, B. W., Ehlman, S. M., McCallum, E. S., Sih, A., Sundin, J., Wong, B. B. M., et al. (2018). Direct and indirect effects of chemical contaminants on the behaviour, ecology and evolution of wildlife. Proceedings of the Royal Society B: Biological Sciences 285 , 20181297. Saïd, S., Gaillard, J.-M., Widmer, O., Débias, F., Bourgoin, G., Delorme, D. and Roux, C. (2009). What shapes intra-specific variation in home range size? A case study of female roe deer. Oikos 118 , 1299–1306. Sempéré, A. J., Sokolov, V. E. and Danilkin, A. A. (1996). Capreolus capreolus . Mammalian Species 1–9. Shahid, M., Natasha, Dumat, C., Niazi, N. K., Xiong, T. T., Farooq, A. B. U. and Khalid, S. (2019). Ecotoxicology of heavy metal(loid)-enriched particulate matter: Foliar accumulation by plants and health impacts. In Reviews of Environmental Contamination and Toxicology (ed. de Voogt, P.), pp. 65–113. Springer International Publishing. Shore, R. F. and Rattner, B. A. (2001). Ecotoxicology of wild mammals . Wiley. Soetan, K. O., Olaiya, C. O. and Oyewole, O. E. (2010). The importance of mineral elements for humans, domestic animals and plants: A review. African Journal of Food Science 4 , 200–222. Sokolova, I. and Lannig, G. (2008). Interactive effects of metal pollution and temperature on metabolism in aquatic ectotherms: Implications of global climate change. Climate Research 37 , 181–201. Strandgaard, H. (1972). The roe deer ( Capreolus capreolus ) population at Kalø and the factors regulating its size. Danish Review of Game Biology 7 , 205. Tapiero, H., Townsend, D. M. and Tew, K. D. (2003). The antioxidant role of selenium and seleno-compounds. Biomedicine & pharmacotherapy 57 , 134. Tchounwou, P. B., Yedjou, C. G., Patlolla, A. K. and Sutton, D. J. (2012). Heavy metal toxicity and the environment. In Molecular, Clinical and Environmental Toxicology (ed. Luch, A.), pp. 133–164. Basel: Springer Basel. Tête, N., Afonso, E., Crini, N., Drouhot, S., Prudent, A.-S. and Scheifler, R. (2014). Hair as a noninvasive tool for risk assessment: Do the concentrations of cadmium and lead in the hair of wood mice ( Apodemus sylvaticus ) reflect internal concentrations? Ecotoxicology and Environmental Safety 108 , 233–241. Thioulouse, J., Dray, S., Dufour, A.-B., Siberchicot, A., Jombart, T. and Pavoine, S. (2018). Multivariate analysis of ecological data with ade4 . Springer New York, NY. USEPA (2000). Guidance for data quality assessment. Valko, M., Morris, H. and Cronin, M. (2005). Metals, toxicity and oxidative stress. Current Medicinal Chemistry 12 , 1161–1208. Vuarin, P., Rey, B., Herrada, A., Débias, F., Crini, N., Scheifler, R., Fritsch, C., Lemaître, J.-F. and Saïd, S. (2025). Exposure to toxic trace metals is negatively associated with testis mass in wild male roe deer. Environmental Pollution 126825. Wirth, J. J. and Mijal, R. S. (2010). Adverse effects of low level heavy metal exposure on male reproductive function. Systems Biology in Reproductive Medicine 56 , 147–167. Xiao, W., Zhang, Y., Chen, X., Sha, A., Xiong, Z., Luo, Y., Peng, L., Zou, L., Zhao, C. and Li, Q. (2024). The easily overlooked effect of global warming: Diffusion of heavy metals. Toxics 12 , 400. Information & Authors Information Version history V1 Version 1 04 September 2025 V2 Version 2 19 December 2025 Copyright This work is licensed under a Non Exclusive No Reuse License. Keywords essential metals mammal physiology non-essential metals oxidative stress redox status weather conditions Authors Affiliations Amandine Herrada 0009-0000-7279-213X [email protected] Université Claude Bernard Lyon 1 View all articles by this author Benjamin Rey Université Claude Bernard Lyon 1 View all articles by this author Jean-François Lemaître Université Claude Bernard Lyon 1 View all articles by this author Emmanuelle Gilot-Fromont Université Claude Bernard Lyon 1 View all articles by this author Marie-Laure Delignette-Muller University Claude Bernard Lyon 1 View all articles by this author Gilles Bourgoin Université Claude Bernard Lyon 1 View all articles by this author François Debias Université Claude Bernard Lyon 1 View all articles by this author Paul Revelli Université Claude Bernard Lyon 1 View all articles by this author Rebecca Garcia Université Claude Bernard Lyon 1 View all articles by this author Sonia Saïd Université Claude Bernard Lyon 1 View all articles by this author Maryline Pellerin Office francais de la biodiversite View all articles by this author Nadia Crini Universite Marie et Louis Pasteur View all articles by this author Clémentine Fritsch Universite Marie et Louis Pasteur View all articles by this author Renaud Scheifler Universite Marie et Louis Pasteur View all articles by this author Jean-Michel Gaillard Université Claude Bernard Lyon 1 View all articles by this author Pauline Vuarin Université Claude Bernard Lyon 1 View all articles by this author Metrics & Citations Metrics Article Usage 283 views 137 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Amandine Herrada, Benjamin Rey, Jean-François Lemaître, et al. Unravelling the relative contribution of toxic metals to redox homeostasis in wild roe deer. Authorea . 19 December 2025. DOI: https://doi.org/10.22541/au.175697692.28236644/v2 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.175697692.28236644/v2","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9feadb6babe352ad',t:'MTc3OTI3NTAzMA=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();