Vitamin E and selenium concentrations of wild donkeys and their diets in the extreme desert ecosystem of Death Valley, California, versus captive donkeys

Introduction: Vitamin E and selenium are vital micronutrients that play key roles in metabolic functions and antioxidant defense, directly influencing equid health. There are no published reference intervals for concentrations of these micronutrients in healthy donkeys. Methods: We compared serum vitamin E and whole blood selenium concentrations in wild, recently captured donkeys with concentrations in donkeys who had been managed in captivity for at least half a year at the time the blood was drawn. In addition, we assessed hematology and serum biochemistry parameters of a subset of the donkeys. We also indirectly analyzed the diet of wild donkey herds using next-generation sequencing of fecal samples. Our last goal was to establish reference intervals for vitamin E and selenium concentrations in wild and rescued donkeys. To do so, we conducted a cross-sectional study of 113 donkeys from Saline Valley and Butte Valley in Death Valley National Park (DVNP), and captive donkeys from Davis, CA, and San Angelo, TX.

Results: Wild donkeys had significantly higher vitamin E concentrations (p < 0.001, difference = 180 μg/dL, CI = 132-228 μg/dL, Tukey multiple comparisons of means test). Captive donkeys, however, had significantly higher selenium concentrations (p < 0.001, difference = 4.97 μg/dL, 95% CI = 3.51-6.43 μg/dL, Tukey multiple comparisons of means test), particularly compared to the wild ones from the Butte Valley population. Reference intervals were established for these micronutrients in wild donkeys. The vitamin E reference range for wild female donkeys was 211-754 μg/dL and for wild male donkeys was 164-780 μg/dL; the selenium reference range for wild female donkeys was 8-21 μg/dL and for wild male donkeys was 12-26 μg/dL. The fecal analysis found a mixed feeding behavior in the wild donkeys, with both grazing and browsing elements, which may explain the relatively high concentrations of these nutrients.

Discussion: This study characterizes the dynamics of vitamin E and selenium concentrations in wild donkeys from Death Valley compared to captive populations, establishing species-specific reference intervals to enhance their management.

Introduction

Wild donkeys in extreme desert ecosystems face harsh climatic conditions and limited food availability. Yet proper nutrition is crucial for their survival and health. Donkeys evolved as both browsers and grazers and are capable of digesting highly fibrous plants consumed over 14–18 hours a day and across distances of 20–30 km (Burden and Thiemann, 2015). In the U.S., one common population control strategy of free-roaming donkey populations is removal from their natural habitat with subsequent placement into captivity. When wild donkeys are captured and brought into captivity for population management purposes, they are fed a hay-based diet (Cappai et al., 2020; Valle et al., 2018). We hypothesized that apparently healthy wild donkeys would exhibit significantly higher circulating vitamin E and selenium concentrations compared to their captive counterparts, indicating potential health disparities between the two populations.

Given the likelihood that wild donkeys consume a more varied roughage diet compared to the restricted forage diet of captive donkeys, the purpose of this study was to assess the micronutrients vitamin E as alpha-tocopherol and selenium in apparently healthy wild donkeys to better inform nutritional management strategies for captive equids. Therefore, our objectives were twofold: (1) to compare serum vitamin E and whole blood selenium concentrations between wild and captive donkeys, and (2) to establish preliminary reference intervals for these micronutrients in wild donkeys to support veterinary health assessments and management strategies.

Studies on free-roaming Przewalski horses in Ukraine (Dierenfeld et al., 1997) and feral Giara horses in Italy (Cappai et al., 2020) have demonstrated significant differences in vitamin E concentrations between free-ranging and captive animals. Deficiencies in these micronutrients in captive equids are common and can result in irreversible neuromuscular disease (Finno and McKenzie, 2025). Vitamin E and selenium are essential micronutrients that play crucial roles in metabolic pathways and protect against oxidative stress that impacts health and immune function. Antioxidant actions of vitamin E and selenium are complimentary; vitamin E is a lipid soluble essential vitamin with antioxidant actions in lipid membranes of the body; lipid peroxidation occurs when a free radical reacts with a molecular oxygen to form a peroxyl radical, which then removes a hydrogen from a polyunsaturated fatty acid, thus generating another free radical and leading to a domino effect. Vitamin E donates a hydrogen to the oxidized fatty acid, preventing propagation of the free radical damage (Traber and Head, 2021). Selenium is essential for the function of a number of important selenoproteins and enzymes, including glutathione peroxidase which act to scavenge free radicals in the cytoplasm (Battin and Brumaghim, 2009). Oxidized vitamin E is reduced and stabilized through actions of vitamin C (endogenously produced in horses), glutathione, and selenium-dependent glutathione peroxidase (Traber and Head, 2021). Both vitamin E and selenium are required in the equine diet and deficiency in either or both nutrients leads to neuromuscular damage and reduced fertility in affected equids (Finno and McKenzie, 2025).

Vitamin E deficiency is frequently reported in equids that do not have access to fresh green pasture, as the vitamin degrades over time in stored feed (Finno and McKenzie, 2025). Selenium is a naturally occurring element, but feed concentrations vary widely based on soil concentration and bioavailability, which both vary across the United States and throughout the world (https://mrdata.usgs.gov/geochem/doc/averages/se/usa.html). Excessive and highly bioavailable selenium in some geographic areas can lead to selenosis in horses, a toxic disease caused by too much dietary selenium, whereas low selenium or selenium bioavailability in other geographic areas leads to selenium deficiency in unsupplemented animals (Bischoff, 2022). Because of their critical role in equine health and performance and the frequency of nutritional imbalances of selenium and vitamin E, these nutrients are frequently monitored in horses. Both micronutrients can be easily measured at veterinary diagnostic laboratories. Despite this, there is a dearth of information available on vitamin E and selenium concentrations of free-roaming donkeys. Likewise, there are very few studies describing the concentrations of these micronutrients in apparently healthy captive donkeys (Bazzano et al., 2019; Quaresma et al., 2021).

Based on the known variation in vitamin E content in different types of forage, vitamin E degradation in stored feeds, as well as differences in selenium uptake in plants, diet was considered in this study. Diets of captive donkeys were recorded by their owners, while wild donkey diets were assessed utilizing environmental fecal sample analysis.

Interpreting the concentrations of micronutrients detected in donkeys is challenging since donkey-specific reference intervals have not yet been described. Most veterinarians rely on those established for horses, which may not be accurate for donkeys (Puls, 1994; Stowe and Pagan, 1998; Bazzano et al., 2019).

We compared vitamin E and selenium concentrations, along with health metrics, between wild and captive donkey populations to gain information relevant to the management of these populations. To aid in the overall health assessment of the animals in this study, we elected to perform hematology and serum biochemistry testing on a subset of the enrolled donkeys.

Methods

Study population and sample collection

Our study comprised a total of 113 donkeys distributed across four sites as detailed in Table 1. There were 63 wild (18 from Saline Valley and 45 from Butte Valley, both in Death Valley National Park (DVNP)) and 50 captive (12 jennets from Davis, CA and 38 jacks from San Angelo, TX) donkeys included. Age classification into foals and adults was determined visually by a veterinarian for wild donkeys and by rescue center staff for captive individuals. The wild group from Saline Valley was comprised of 13 males and 5 females, with most being adults except for 1 male and 1 female foal. The wild group from Butte Valley was comprised of 25 males and 20 females, again primarily consisting of adults with only 4 female foals and 1 male foal. The captive donkeys in San Angelo, TX were all adult males, while the captive donkeys in Davis, CA were all adult females.

Table 1

Table 1. Demographics of donkey populations and representation of dietary samples.

Venipuncture to obtain approximately 10 ml of blood was performed from the left or right jugular (IACUC # 2007-0146). Blood was collected on the 25th of October and 15th of November 2020 from the Butte Valley herd, on the 9th of December 2020 from Saline Valley herd, on the 5th of January 2021 from the captive Davis herd and on the 7th and 10th of March 2021 from the captive San Angelo herd. The captive donkeys, in San Angelo (38) and Davis (12), were originally free-roaming from the Southwestern US and were captured at different times. All 50 captive donkeys had been in captivity for at least half a year at the time the blood was collected. Upon collection, a portion of blood was first placed into red top tubes (10ml BD Vacutainer serum tubes, manufacturer Ontario, Canada) for vitamin E and serum biochemistry analyses. The remainder of blood was then placed into purple top tubes (4ml BD Vacutainer K2-EDTA) for selenium and hematologic analyses. To assist with the hematology interpretation two unstained, air-dried blood smears were prepared from EDTA whole blood obtained from 10 captive donkeys from Davis, CA at the time of venipuncture. Vitamin E and selenium assessments were performed on every donkey and the results from the adult wild donkeys were utilized to determine the selenium and vitamin E reference ranges. Hematology and serum biochemistry assessments were performed on 22 wild donkeys from Butte Valley (two foals; jack and jennet and 20 adults: 18 jacks and two jennets), 10 captive jennets from Davis, CA and 30 captive jacks from San Angelo, TX.

All wild donkeys sampled in DVNP were captured as part of ongoing population management activities conducted by authorities and not specifically for the purposes of this research. They were baited into the short-term pens (approximately 10m² x 10m² in size) with Bermuda grass hay. This pens were designed to hold wild donkeys before transfer to larger facilities. Males and females were placed in separate pens and foals stayed with jennets. Blood was collected from the wild donkeys anywhere from 1 to 30 days after capture, while they were housed in these short-term pens and fed only Bermuda grass hay. The captive jacks in San Angelo, TX were sampled at the time of their castration. The captive jennets in Davis, CA had their blood collected during routine health examinations. Information on the captive donkey diets was determined through interviews with owners.

Fecal samples from wild donkeys were collected in the areas where both wild herds were roaming. We analyzed 60 samples from the Butte Valley herd in November 2020 and 24 samples form Saline Valley herd in December 2020 (Table 1). Fresh fecal samples were collected from the ground into individual paper bags and dried in the sun during that same day. Freshness of the fecal sample was determined by the collector based on visual examination and interpretation of the feces collected from the ground. Approximately one week after collection, 10–20 mL of fecal material was placed in an empty vial and sent unfrozen to Jonah Ventures laboratory in Boulder, CO for dietary analysis using Next Generation Sequencing (NGS). Fecal samples were frozen upon receipt (4 weeks, -20°C) at the laboratory until they were processed.

Sample handling and laboratory analyses

Red top serum tubes were centrifuged within 30 minutes of collection. The serum was pipetted into a separate plain tube and protected from light exposure using aluminum foil. Both blood and serum samples were placed in an insulated box with ice packs for transport to the laboratory within 48 hours. The blood smear slides were placed in slide mailers and protected from contact with ice packs during transport to the laboratory. Blood samples were shipped to and analyzed at the Cornell University Animal Health Diagnostic Center (CU AHDC) for Vitamin E, selenium, biochemistry and hematology testing. Analyzers in the laboratory are evaluated daily for performance using commercially available quality control material and calibrated as needed. The laboratory participates in two external quality assurance programs.

Selenium concentrations in homogenous whole blood were determined using a PerkinElmer AAnalyst 600 atomic absorption graphite furnace (GFAA) with longitudinal Zeeman-effect background corrector. Palladium chloride solution was prepared as 1.667 g PdCl₂ added to 1L of 2% HNO₃ in deionized water. 12.5g Ni(NO₃)2*6 H₂O and 1mL Triton-X were added to 550 mL palladium chloride solution and q.s. with deionized water to 1 L to create palladium chloride-nickelous nitrate matrix modifier. Ammonium phosphate and magnesium nitrate matrix modifier was prepared as 0.2% NH₄H₂PO₄ and 0.05% Mg(NO₃) in 1% HNO₃ and 0.1% Triton-X in deionized water. 500 µl palladium chloride-nickelous nitrate modifier and 400 µl ammonium phosphate-magnesium nitrate matrix modifier were added to 100 µl aliquot whole blood. Standards were prepared in water at four concentrations ranging from 25–500 ng/mL (from a 10 µg/mL) selenium stock solution. 100 µl standard volumes were added to 500 µl palladium chloride-nickelous nitrate modifier and 400 µl ammonium phosphate-magnesium nitrate modifier to generate the calibration curve.

Vitamin E concentrations as α-tocopherol were determined using a high-pressure liquid chromatograph (HPLC) and fluorescence spectrometer with a Agilent Eclipse XDB-C18, 5-micron particle size, 15 cm x 4.6 mm internal diameter with the following specifications: column temperature of 40°C, flow rate of 1.4 mL/minute, excitation wavelength 291 nm (10nm slit width), and emission wavelength 330 nm (10 nm slit width). Retention time for α-tocopherol is approximately 10.5 minutes. Analysis was performed in subdued light. One mL serum was placed in a 15 mL centrifuge tube and weighed accurately. One mL ethanol was added to the sample and vortexed for 10 seconds. Two mL hexane was added and vortexed for 1 minute. Samples were placed in a rotating shaker for 5 minutes, then left undisturbed for layers to separate for 10 minutes. Hexane was transferred to a disposable glass culture tube with a transfer pipet. Extraction was repeated with one mL hexane, with samples centrifuged at 2000 rpm for 5 minutes after removal from rotating shaker. Hexane collection was repeated and the pooled hexane extracts evaporated to dryness at 30 to 40°C under nitrogen. 600 µL of solution containing 684 mL acetonitrile + 220 mL tetrahydrofuran + 70 mL methanol + 30 mL 1% ammonium acetate solution (W/V) was added to each tube and vortexed for approximately 1 minute. The resulting solution was transferred to an amber glass autosampler vial and capped for injection into the HPLC system. Aliquots of 30 µL of each working standard solution and each sample solution were injected. The peak area of the eluted peak for α-tocopherol for all standard solutions and sample solutions was recorded. Alpha-tocopherol standards were prepared in ethanol from a 97% purity α-tocopherol standard in a four-point calibration curve from 0.220 ug/mL to 13.193 ug/mL. The α-tocopherol content of each sample calculated by comparing sample peak area for α-tocopherol to the standard curve and adjusting by a factor of 1.667 (to correct for sample concentration from 1 mL to 600 µL).

The parameters analyzed in the complete blood count (CBC) were hematocrit value (HCT), hemoglobin concentration (HGB), red blood cell count (RBC), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), red cell distribution width (RDW), mean platelet volume (MPV), total white blood cell count (WBC), and platelet count, all using a Siemens Advia 2120i automated hematology analyzer (Siemens Medical Solutions, Malvern, PA, USA) Total protein was analyzed by refractometry. A differential WBC count was performed by manual count of 100 cells in a Wright Giemsa-stained smear.

The serum biochemistry panel included the following parameters (measured using a Cobas 501 automated serum analyzer by Roche Diagnostics): sodium (Na), potassium (K), chloride (Cl), calcium (Ca), magnesium (Mg), bicarbonate (HCO3), phosphate (PO4), bilirubin (total and direct), cholesterol, blood urea nitrogen (BUN), creatinine kinase, creatinine (Crea), total protein, albumin (ALB), globulin (GLOB), glucose (GLU), gamma glutamyl transferase (GGT), GLDH, SDH, aspartate aminotransferase (AST), and triglycerides (TRYG).

Fecal analysis for dietary assessment at the Jonah Ventures laboratory was performed using frozen fecal samples that were thawed for 1–2 hours before processing for NGS. Sample barcodes were recorded and assigned to a well within the 96 well plate or numbered extraction tube. Under a laminar flow hood, sterile cotton swabs (Fisher, cat# 22-363-173) were coated with fecal matter (swab was moistened with DNA-free water), and the swabs were placed in the corresponding extraction plate or tube. Sterile tweezers and pliers are used to handle cotton swabs and remove the wooden ends of the cotton swab before extraction. Plates or tubes were immediately processed or stored in -20C until the extraction process could be performed. Genomic DNA was extracted using the DNeasy PowerSoil HTP 96 Kit (Cat # 12955-4) according to the manufacturer’s protocol. Genomic DNA was eluted into 100µl vials and frozen at -20C. PCR was done following (Taberlet et al., 2007). Both forward and reverse primers also contained a 5’ adaptor sequence to allow for subsequent indexing and Illumina sequencing. Each 25 µL PCR reaction was mixed according to the Promega PCR Master Mix specifications (Promega catalog # M5133, Madison, WI) which included 0.4 µM of each primer and 1 µl of gDNA. DNA was PCR amplified using the following conditions: initial denaturation at 94°C for 3 minutes, followed by 40 cycles of 30 seconds at 94°C, 30 seconds at 55°C, and 1 minute at 72°C, and a final elongation at 72°C for 10 minutes. To determine amplicon size and PCR efficiency, each reaction was visually inspected using a 2% agarose gel with 5µl of each sample as input. Amplicons were then cleaned by incubating amplicons with Exo1/SAP for 30 minutes at 37°C following by inactivation at 95°C for 5 minutes and stored at -20°C. A second round of PCR was performed to complete the sequencing library construct, appending with the final Illumina sequencing adapters and integrating a sample-specific,12-nucleotide index sequence. The indexing PCR included Promega Master mix, 0.5 µM of each primer and 2 µl of template DNA (cleaned amplicon from the first PCR reaction) and consisted of an initial denaturation of 95°C for 3 minutes followed by 8 cycles of 95°C for 30 sec, 55°C for 30 seconds and 72°C for 30 seconds. Final indexed amplicons from each sample were cleaned and normalized using SequalPrep Normalization Plates (Life Technologies, Carlsbad, CA). 25 µl of PCR amplicon is purified and normalize using the Life Technologies SequalPrep Normalization kit (cat#A10510-01) according to the manufacturer’s protocol. Samples were then pooled together by adding 5 µl of each normalized sample to the pool. Sample library pools were sent for sequencing on an Illumina MiSeq (San Diego, CA) in the CU Boulder BioFrontiers Sequencing Center using the v2 500-cycle kit (cat# MS-102-2003). Necessary quality control measures were performed at the sequencing center prior to sequencing. Raw sequence data were demultiplexed using pheniqs v2.1.0 (Galanti et al., 2021), enforcing strict matching of sample barcode indices (i.e, no errors). Cutadapt v3.4 (Martin, 2011) was then used to remove gene primers from the forward and reverse reads, discarding any read pairs where one or both primers were not found at the expected location (5’) with an error rate < 0.15. Read pairs were then merged using vsearch v2.15.2 (Rognes et al., 2016), discarding resulting sequences with a length of < 100 bp or with a maximum expected error rate > 0.5 bp (Edgar and Flyvbjerg, 2015). For each sample, reads were then clustered using the unoise3 denoising algorithm (Edgar, 2016) as implemented in vsearch, using an alpha value of 5 and discarding unique raw sequences observed less than 8 times. Counts of the resulting exact sequence variants (ESVs) were then compiled and putative chimeras were removed using the uchime3 algorithm, as implemented in vsearch. For each final ESV, a consensus taxonomy was assigned using a custom best-hits algorithm and a reference database consisting of publicly available sequences (GenBank (Benson et al., 2005) as well as Jonah Ventures voucher sequences records. Reference database searching used an exhaustive semi-global pairwise alignment with vsearch, and match quality was quantified using a custom, query-centric approach, where the % match ignores terminal gaps in the target sequence, but not the query sequence. The consensus taxonomy was then generated using either all 100% matching reference sequences or all reference sequences within 1% of the top match, accepting the reference taxonomy for any taxonomic level with > 90% agreement across the top hits.

We followed the novel bioinformatics guidelines outlined in previous research (Littleford-Colquhoun et al., 2022) for our diet examination. In accordance with these guidelines, we included all sequences, both high and low in abundance, to ensure a thorough and accurate representation of dietary profiles. This approach was undertaken with the purpose of preserving the true composition of dietary profiles.

Statistics and experimental design

All data were maintained in a spreadsheet (Excel, Microsoft, Redmon, Washington, USA) and data analysis was performed using statistical software [R version 4.1.3 (2022-03-10)] in RStudio (version 2025.09.1 + 401, Posit Software, PBC, Boston, MA, USA), with a cutoff of p < 0.05 for determining statistical significance. All datasets used in this study are provided in the Supplementary Material. We performed a cross-sectional study comparing vitamin E and selenium concentrations in the blood of donkeys by locations (Butte and Saline Valley, DVNP, San Angelo, TX, and Davis, CA) and by group (wild, captive) for both genders (male, female), and both age groups (adult, foal). Results from the wild donkey groups (Saline Valley and Butte Valley) and captive groups (Davis and San Angelo) were combined accordingly to increase statistical power and to provide assessment of nutrient concentrations in wild donkeys compared to captive donkeys.

All donkeys tested for vitamin E and selenium concentrations were compared to those established for horses (Puls, 1994). The CBC and biochemistry results performed on a subset of donkeys, were interpreted using donkey-specific CBC and biochemistry reference intervals established by the CU AHDC for healthy adult donkeys (Goodrich and Webb, 2024).We compared concentrations of vitamin E and selenium across the four sites and later by two groups, combining Butte and Saline Valley sites for wild donkeys and Davis and San Angelo sites for captive donkeys. Results by groups (wild, captive) are presented in Table 2 and Figures 1 and 2.

Table 2

Table 2. Summary of vitamin E and selenium concentrations in wild and captive donkeys, organized by sex and age group (adults and foals).

Figure 1

Figure 1. Boxplots showing vitamin E concentrations in wild and captive donkeys, by age group and sex. The mean value is indicated by a horizontal line within each box. M, male; F, female; n, number of animals. *p < 0.05 (statistically significant); **p < 0.01 (highly significant); *** p < 0.001 (very highly significant).

Figure 2

Figure 2. Boxplots showing selenium concentrations in wild and captive donkeys, by age group and sex. The mean value is indicated by a horizontal line within each box. M, male; F, female; n, number of animals. *p < 0.05 (statistically significant); **p < 0.01 (highly significant); *** p < 0.001 (very highly significant).

Descriptive statistics (mean, median, range and standard deviation) were calculated for vitamin E, selenium and all hematological and biochemical parameters to assess for potential differences among sites and groups. Histograms and boxplots were used to visualize the distribution of data. Analysis of variance (ANOVA) was performed for vitamin E and selenium concentrations, across different sites and groups to assess for statistically significant differences in nutrient concentrations. Post-hoc testing for cases where ANOVA indicated significant differences was done using Tukey’s HSD (Honestly Significant Difference) to determine which specific groups exhibited statistically distinct nutrient concentrations. Geographic location and captivity status were partially accounted for by combining donkeys from similar environments into broader groups (wild: Butte and Saline Valley DVNP; captive: Davis, CA and San Angelo, TX). While we did not apply a full multivariate model such as ANCOVA or a linear mixed model due to sample size limitations and collinearity between some variables (e.g., sex and site), we stratified for age and sex to avoid confounder effect for both nutrients and conducted subgroup analyses to enhance validity and transparency.

To calculate reference intervals for vitamin E and selenium concentrations for wild donkeys, we initially used visualization methods such as histograms, density plots, and QQ (quantile–quantile) plots to evaluate the distribution of the data. Reference intervals (RIs) were calculated for adult wild males and females combined and for adult wild males and females separately for both selenium and vitamin E. Because both vitamin E and selenium data followed approximately normal distributions (Shapiro–Wilk test, p > 0.05), RIs were calculated using parametric methods (mean ± 2 SD), which correspond to the 2.5th and 97.5th fractiles representing the central 95% of values. We excluded foals from the RI calculations due to the limited number of individuals (only 2 male and 5 female donkeys). Although excluded from the RI calculations, foals were retained in comparative analyses to provide preliminary information on potential age-related differences, given the rarity of obtaining wild foal samples; these results should be interpreted with caution. Two extreme outliers in vitamin E concentrations (one adult female, one adult male) were excluded prior to RI determination. To account for the limited sample size (n = 54 for vitamin E, n = 56 for selenium) and associated uncertainty, 90% confidence intervals (CIs) were calculated around each reference limit (Friedrichs et al., 2012). Bootstrapping was not applied, as the 90% CIs were derived analytically from the parametric model; however, we acknowledge that it could be a useful approach for future studies with similarly constrained data.

We summarized fecal dietary data by plant family. While finer taxonomic resolution (e.g., genus or species) would be informative, especially if there was prior knowledge of species-specific plant nutrient concentrations, family-level dietary data is informative regarding broad axes of dietary differentiation in large herbivores, which generally reflect adaptations to grazing (e.g., the consumption of graminoids in the families Poaceae, Cyperaceae, and Juncaceae) and browsing (all other plant families).

Results

We summarized hematological and biochemical parameters for three populations of donkeys and compared them to reference intervals established by CU AHDC as shown in Table 3. Chloride, creatinine, hematocrit, sodium, total protein and white blood cells were normally distributed for Butte Valley donkeys. Albumin had both high and low outlier values. RBC, GGT, globulin, AST and hemoglobin were slightly skewed to the right with the last two both having one high outlier. Platelet count and potassium were very right skewed with platelet count having two extreme ouliers for adult wild donkeys. There was one extreme outlier identified as an adult wild female with a high triglycerides concentrations. Blood urea nitrogen was slightly skewed to the right with two outliers. No parasites or icterus were seen in the samples.

Table 3

Table 3. Hematological and biochemical profiles of three populations of donkeys by site – Butte Valley (wild) from Death Valley National Park (DVNP), Davis (captive) from California (CA), San Angelo (captive) from Texas (TX) – compared to reference intervals (RI) established by the Cornell University Animal Health Diagnostic Center (CU AHDC; Goodrich and Webb, 2024) for healthy adult donkeys.

Selenium was approximately normally distributed for all four sites. Vitamin E was positively skewed for all four sites with Butte Valley having two and Saline Valley having one extreme outlier. The three extreme outliers were identified as a nursing jennet with vitamin E concentration of 1439 µg/dL and her female foal, with 1519 µg/dL from Butte Valley and a jack with a vitamin E concentration of 1420 µg/dL from Saline Valley. All three were excluded from ANOVA and Tukey multiple comparisons of means test when testing vitamin E concentrations by sites and by groups. We found that wild donkeys from Butte and Saline Valley had wider ranges of vitamin E and higher mean and median values compared to captive donkeys from Davis, CA and San Angelo, TX, as shown in Table 2; Figure 1. Butte Valley donkeys had selenium concentrations with a wider range and lower mean and median compared to other sites (Table 2; Figure 2). Wild foals, specifically females, had higher concentrations of vitamin E compared to the adults but both sexes had lower selenium concentrations compared to the adults (Table 2; Figure 1), though this interpretation is limited by the small sample size (female foal n=5, male foal n=2). We observed that captive donkeys had lower range of vitamin E concentrations compared to the reference values established for horses (180 to 200 µg/dL for foal and 200 to 1000 µg/dL for adult, Puls, 1994). The majority of the Butte Valley population had selenium concentrations much lower than those established for horses (mature horses 14.0 to 24.0 μg/dL, foals 9.8 to 14.0 µg/dL, Stowe and Pagan, 1998).

Results comparing wild and captive groups

A Tukey multiple comparisons of means test was conducted after performing ANOVA for both vitamin E and selenium concentrations for donkeys by age and sex, comparing wild and captive donkey groups. The analysis revealed the following key findings:

Vitamin E

ANOVA revealed a significant effect of donkey’s site on serum vitamin E concentrations (F(1, 106) = 54.66, p < 0.001) and a marginal effect of age (F(1, 106) = 4.10, p = 0.045), while sex had no significant effect (F(1, 106) = 0.00, p = 0.99). Tukey multiple comparisons of means test confirmed that wild donkeys had significantly higher vitamin E concentrations than captive donkeys (mean difference = 180 µg/dL, 95% CI = 132-228 µg/dL, p < 0.001). No significant differences were found between sexes (mean difference = 0.33 µg/dL, 95% CI = -52-51 µg/dL, p = 0.99). Adults had marginally higher vitamin E concentrations than foals (mean difference = 104 µg/dL, 95% CI = -2-210 µg/dL, p = 0.054).

Selenium

ANOVA showed significant effects of donkey’s site (F(1, 109) = 45.31, p < 0.001), sex (F(1, 109) = 16.36, p < 0.001), and age (F(1, 109) = 19.88, p < 0.001) on whole blood selenium concentrations. Tukey multiple comparisons of means test confirmed that captive donkeys had higher selenium concentrations than wild donkeys (mean difference = 4.97 µg/dL, 95% CI = 3.51-6.43 µg/dL, p < 0.001). Males had higher selenium concentrations than females (mean difference = 3.12 µg/dL, 95% CI = 1.57-4.66 µg/dL, p < 0.001), and adults had higher selenium concentrations than foals (mean difference = 6.49 µg/dL, 95% CI = 3.48-9.50 µg/dL, p < 0.001).

Reference range establishment for vitamin E and selenium for adult wild donkeys from Death Valley

Vitamin E reference values

There were two vitamin E concentration outliers with particularly high values; one adult female (1439 µg/dL) from Butte Valley and one adult male (1420 µg/dL) from Saline Valley. After excluding these two outliers, vitamin E concentrations for wild adult donkeys showed an approximately normal distribution. For all adult wild donkeys (n = 54) calculated RI for vitamin E blood concentrations was 182-769 µg/dL, only for females (n = 19) RI was 211-754 µg/dL and only for males (n = 35) RI was 163-780 µg/dL. Other statistical metrics and results are shown in Table 4.

Table 4

Table 4. Reference intervals (RI) for vitamin E and selenium concentrations in adult wild donkeys from Death Valley National Park.

Selenium reference values

Selenium concentrations for wild donkeys showed approximately normal distribution for combined sexes and for distributions of males and females separated. For all adult wild donkeys (n = 56) calculated RI for selenium blood concentrations was 9.47-25.30 µg/dL, only for females (n = 20) RI was 8.36-20.82 µg/dL and only for males (n = 36) RI was 11.89-25.98 µg/dL. Other statistical metrics and results are shown in Table 4.

Results for dietary analyses

Diet analysis showed that wild donkeys have mixed feeding behavior, grazing and browsing on a variety of plant families. Summarized percentages of plant families across the diet of the Butte Valley herd showed that grasses (Poaceae) comprised 23% of diet, followed by Asteraceae (14.5%) and Chenopodiaceae (10.5%). In the diet of the Saline Valley herd, Asteraceae comprised 14.3% of diet, followed by Fabaceae (13.3%) and Boraginaceae (11.9%) as shown on Figure 3.

Figure 3

Figure 3. Distribution of plant families in wild donkey diets from two locations in Death Valley National Park (DVNP): Saline Valley and Butte Valley.

Discussion

We compared micronutrient concentrations and health metrics of wild and captive donkeys to gain information relevant to management of these populations. Vitamin E primarily functions as an antioxidant at the cellular membrane level (Fagan et al., 2020) and selenium is critical for the activity of cytosolic antioxidant enzymes, such as glutathione peroxidase (Mannervik, 1985). Although these nutrients operate in different parts of the cell and are not interchangeable, mild deficiencies in one may be somewhat offset by sufficient levels of the other. However, it is crucial to understand that severe deficiencies in either vitamin E or selenium cannot be fully compensated for by an abundance of the other, underscoring the importance of maintaining adequate concentrations of both nutrients for optimal equid health (Van Metre and Callan, 2001).

We assessed overall health of a portion of the donkeys in this study using hematology and biochemistry testing, however, due to logistical constraints and field conditions in DVNP and San Angelo, TX, blood smears were not prepared for these animals, which may have limited our ability to detect some subtle cellular morphological changes and characteristics that are best assessed in fresh blood samples rather than in the smears that would have been prepared upon receipt of the samples at the laboratory (Stokol, 2020).

Similar to the studies in horses (Dierenfeld et al., 1997; Cappai et al., 2020), there were significant differences in vitamin E and selenium concentrations between the wild and captive groups. There were also significant differences in blood selenium concentrations between male and female donkeys. Vitamin E plays an important role in mitigating oxidative stress and supporting neuromuscular function so maintaining appropriate concentrations is important for the health of the equid (Formigli et al., 1997; Fagan et al., 2020; Finno and Valberg, 2012). Adult wild jacks had higher concentrations of vitamin E than adult jacks in captivity, indicating the need for screening of serum vitamin E concentrations and supplementation of captive donkeys.

Compared to established whole blood selenium reference intervals reported for horses (Stowe and Pagan, 1998), wild jacks and captive donkeys were comparable. Some wild jennets and foals exhibited significantly lower selenium concentrations compared to the captive donkeys. Another study (Bazzano et al., 2019) acknowledges lower selenium concentrations as a normal finding for donkeys as compared to horses, however they measured plasma rather than whole blood selenium and their population of donkeys was small and exclusively from a dairy farm in Italy, which might not be representative of free-roaming or captive donkey populations in the U.S.

Diet analysis showed that wild donkeys have mixed feeding behavior, grazing and browsing on a variety of plant families. Some wild Asteraceae plants are known to have high vitamin E concentrations (Panfili et al., 2020). As wild donkeys in our study were browsing Asteraceae plant families, we speculate that this might have influenced their vitamin E concentrations. Vitamin E is important in plant physiology and found at high concentrations in reproductive and photosynthetic parts of plants, however the content degrades significantly once the plant is harvested and stored as hay (Niu et al., 2022; Pitel et al., 2020). The difference in captive donkeys was expected because serum vitamin E concentrations are known to decline in equids that have transitioned from fresh forage to hay (Cappai et al., 2020; Valle et al., 2018; Finno and Valberg, 2012). Further research is necessary to understand if the plants, leaves, and seeds identified in the feces of the wild donkeys in our study are the sources of higher-than-expected vitamin E concentrations in the serum of these wild donkeys.

The uptake of selenium by plants is influenced not only by the selenium content in the soil, which varies geographically, but also by the plant species, the soil fertility conditions, including pH, moisture, soil phosphate concentration, and the chemical form of selenium in the soil (Bauer, 1997; Delesalle et al., 2017). The typical estimation for selenium concentration in normal soils is 0.2.ppm (Bauer, 1997). Based on reported soil values (U.S. Geological Survey, n.d), Davis and both Death Valley sites would have normal soil selenium concentrations while San Angelo’s soil is slightly selenium deficient (Inyo County’s soil selenium is 0.27 +/- 0.23 parts per million (ppm), Yolo County’s is 0.19 +/- 0.06 ppm, and Tom Green County’s is 0.17 +/-0.05ppm).

The vitamin E and selenium concentrations from donkeys are normally compared to established reference intervals for horses (Puls, 1994; Stowe and Pagan, 1998). Our study suggests that vitamin E concentrations tend to be significantly higher in wild donkeys (182 – 769 µg/dL) compared to captive donkeys (mean difference = 180.08 µg/dL) and are slightly lower from the range previously established for horses (200 – 1000 µg/dL). The reference intervals our study determined for selenium concentrations in adult wild donkeys of (9 - 25 µg/dL) are similar but broader to those published for horses (17 - 25 µg/dL) (Puls, 1994). Selenium is an element, therefore stable in stored feed, thus the change from fresh forage to hay would not be expected to change the selenium status of these animals. However, geographic and plant factors may influence selenium status. The trend towards higher blood selenium concentrations in captive donkeys was likely due to the addition of selenium-containing mineral blocks to the diet.

There were several limitations to our study. The donkeys in this study are from a limited number of locations and may not be representative of all wild and captive populations. Bias could also occur due to inclusion of only male captive donkeys from San Angelo, TX and only female captive donkeys from Davis, CA, while wild populations included both sexes. Variation in donkey ages and sex among the different groups introduces the possibility for confounding effect. Age is known to impact nutrient metabolism, and if not adequately controlled for, it could confound the association between vitamin E or selenium concentrations and health outcomes. Another limitation of our study was that the number of foals was small (n = 7). This limited statistical power and precluded their inclusion in the reference interval calculations. However, we retained them in comparative analyses because wild foal samples are exceptionally rare, and even preliminary data may help guide future studies and management practices.

In our analysis, we controlled for age and sex through stratification and accounted for geographic differences by grouping donkeys based on site and management status (wild vs. captive). Other potential confounders, particularly diet composition, seasonal variation in forage vitamin E content, and stress associated with capture and confinement, were not directly measured and could not be included in multivariate models.

Fecal samples were collected from the environment and therefore represented herd-level rather than individual-level data (one herd from Butte and another from Saline Valley). Our dietary analysis also did not include direct quantification of vitamin E or selenium content in the identified forage species. The influence of diet on blood micronutrient levels in our study can therefore only be inferred. Future studies should include direct sampling and laboratory analysis of forage for vitamin E and selenium to more accurately assess nutrient intake and its relationship with serum and whole blood concentrations.

We also could not perform full clinical examinations and could only roughly estimate their ages due to the difficulties working with wild animals. For future studies it would be important to collect data from wild individuals in the wilderness, and to do follow-up data collection over time on wild donkeys that were placed in captivity. It would also be beneficial to collect more data from different age groups and genders at the same locations. Additionally, it would be valuable to collect diet data from wild individual animals at the same time that blood is collected for nutrient analysis, so that diet impacts on individuals could be considered, and not just at the herd level.

Conclusions

In the U.S., one common population control strategy of free-roaming donkey populations is removal from their natural habitat with subsequent placement into captivity. Our study demonstrates novel information about serum vitamin E and whole blood selenium concentrations in apparently healthy donkeys, and the diets that seem to be associated with those concentrations in wild donkeys. We determined that, similar to studies comparing free-ranging wild and feral horses with confined cohorts, there were differences in vitamin E concentrations between free-ranging and confined donkeys (Dierenfeld et al., 1997; Cappai et al., 2020). This information can be used to enhance the care and welfare of donkeys in captivity.

Data availability statement

The raw data has been uploaded to NCBI, accession number PRJNA1200696: https://www.ncbi.nlm.nih.gov/sra/PRJNA1200696.

Ethics statement

The animal studies were approved by The Institutional Animal Care and Use Committee (IACUC) at Cornell University, under protocol number 2007-0146. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent was obtained from the owners for the participation of their animals in this study.

Author contributions

SJ: Data curation, Formal analysis, Funding aquisition, Investigation, Methodology, Project administration, Resources, Software, Visualization, Writing – original draft and Writing -review & editing. KB: Conceptualization, Investigation, Methodology, Resources, Validation, Writing - review and editing. EL: Conceptualization, Formal analysis, Investigation, Methodology, Writing – review & editing, Data curation, Software, Visualization. JF: Writing – review & editing, Supervision, Validation, Visualization, Writing – original draft. EG: Supervision, Validation, Writing – original draft, Writing – review & editing, Conceptualization, Formal analysis, Funding acquisition, Investigation, Project administration, Resources, Methodology.

Funding

The author(s) declared that financial support was received for this work and/or its publication. The funding source for this research was Rural Veterinary Experience Teaching and Service (RVETS), a 501(c) 3 Non-profit, https://rvets.org. RVETS funding was used for travel for sample collection and laboratory analysis for vitamin E and Selenium. We express sincere gratitude to Dr. Eric Davis, Dr. John Madigan and Peaceful Valley Donkey Rescue (PVDR) for their generous financial contributions.

Acknowledgments

We appreciate Jonah Ventures laboratory for their cooperations and help with dietary analysis. We greatly appreciate Dr. Miguel Saucedo and Dr. Gerard Arino Estrada, who played key roles in the fieldwork, including professionally restraining wild donkeys and assisting with blood sample collection. We appreciate Dr. Lais Rosa Rodrigues Costa for her generous help and contribution to this research. We are grateful to the PVDR crew for their help in Death Valley and San Angelo, as well as to the Rural Veterinary Experience and Teaching Service crew in San Angelo for their valuable assistance. We also acknowledge the support of the Master of Preventive Veterinary Medicine (MPVM) Program at the University of California, Davis.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

Battin E. E. and Brumaghim J. L. (2009). Antioxidant activity of sulfur and selenium: A review of reactive oxygen species scavenging, glutathione peroxidase, and metal-binding antioxidant mechanisms. Cell Biochem. Biophysics 55, 1–23. doi: 10.1007/s12013-009-9054-7

PubMed Abstract | Crossref Full Text | Google Scholar

Bauer F. (1997). Selenium and soils in the Western United States. Electronic Green J. 1. doi: 10.5070/g31710269

Crossref Full Text | Google Scholar

Bazzano M., McLean A., Tesei B., Gallina E., and Laus F. (2019). Selenium and vitamin E concentrations in a healthy donkey population in central Italy. J. Equine Veterinary Sci. 78, 112–116. doi: 10.1016/j.jevs.2019.04.003

PubMed Abstract | Crossref Full Text | Google Scholar

Benson D. A., Karsch-Mizrachi I., Lipman D. J., Ostell J., and Wheeler D. L. (2005). GenBank. Nucleic Acids Res. 33. doi: 10.1093/nar/gki063

PubMed Abstract | Crossref Full Text | Google Scholar

Bischoff K. (2022). “Selenium toxicosis in animals,” in Merck Veterinary Manual. (Rahway, NJ, USA: Merck & Co., Inc.) Available online at: https://www.merckvetmanual.com/toxicology/selenium-toxicosis/selenium-toxicosis-in-animals (Accessed September 2, 2025).

Google Scholar

Burden F. and Thiemann A. (2015). Donkeys are different. J. Equine Veterinary Sci. 35, 376–382. doi: 10.1016/j.jevs.2015.03.005

Crossref Full Text | Google Scholar

Cappai M. G., Pudda F., Wolf P., Accioni F., Boatto G., and Pinna W. (2020). Variation of hematochemical profile and vitamin E status in feral giara horses from free grazing in the wild to hay feeding during captivity. J. Equine Veterinary Sci. 94. doi: 10.1016/j.jevs.2020.103220

PubMed Abstract | Crossref Full Text | Google Scholar

Delesalle C., Bruijn M., Wilmink S., Vandendriessche H., Mol G., Boshuizen B., et al. (2017). White muscle disease in foals: Focus on selenium soil content. A case series. BMC Veterinary Res. 13. doi: 10.1186/s12917-017-1040-5

PubMed Abstract | Crossref Full Text | Google Scholar

Dierenfeld E. S., Hoppe P. P., Woodford M. H., Krilov N. P., Klimov V. V., Yasinetskaya N. I., et al. (1997). Plasma α-tocopherol, β-carotene, and lipid levels in semi-free-ranging Przewalski horses (Equus przewalskii). J. Zoo Wildlife Med. 28, 144–147.

PubMed Abstract | Google Scholar

Edgar R. C. (2016). UNOISE2: Improved error-correction for Illumina 16S and ITS amplicon sequencing. bioRxiv. doi: 10.1101/081257

Crossref Full Text | Google Scholar

Edgar R. C. and Flyvbjerg H. (2015). Error filtering, pair assembly, and error correction for next-generation sequencing reads. Bioinformatics 31, 3476–3482. doi: 10.1093/bioinformatics/btv401

PubMed Abstract | Crossref Full Text | Google Scholar

Fagan M. M., Harris P., Adams A., Pazdro R., Krotky A., Call J., et al. (2020). Form of vitamin E supplementation affects oxidative and inflammatory response in exercising horses. J. Equine Veterinary Sci. 91. doi: 10.1016/j.jevs.2020.103103

PubMed Abstract | Crossref Full Text | Google Scholar

Finno C. J. and McKenzie E. C. (2025). Vitamin E and selenium-related manifestations of muscle disease. Veterinary Clinics North America - Equine Pract. 41, 77–93. doi: 10.1016/j.cveq.2024.11.001

PubMed Abstract | Crossref Full Text | Google Scholar

Finno C. J. and Valberg S. J. (2012). A comparative review of vitamin E and associated equine disorders. J. Veterinary Internal Med. 26, 1251–1266. doi: 10.1111/j.1939-1676.2012.00994.x

PubMed Abstract | Crossref Full Text | Google Scholar

Formigli L., Ibba Manneschi L., Tani A., Gandini E., Adembri C., Pratesi C., et al. (1997). Vitamin E prevents neutrophil accumulation and attenuates tissue damage in ischemic-reperfused human skeletal muscle. Histol. histopathology 12, 663–669.

PubMed Abstract | Google Scholar

Friedrichs K. R., Harr K. E., Freeman K. P., Szladovits B., Walton R. M., Barnhart K. F., et al. (2012). ASVCP reference interval guidelines: Determination of de novo reference intervals in veterinary species and other related topics. Veterinary Clin. Pathol. 41, 441–453. doi: 10.1111/vcp.12006

PubMed Abstract | Crossref Full Text | Google Scholar

Galanti L., Shasha D., and Gunsalus K. C. (2021). Pheniqs 2.0: accurate, high-performance Bayesian decoding and confidence estimation for combinatorial barcode indexing. BMC Bioinf. 22. doi: 10.1186/s12859-021-04267-5

PubMed Abstract | Crossref Full Text | Google Scholar

Goodrich E. L. and Webb J. L. (2024). Complete blood count and biochemistry reference intervals for healthy adult donkeys in the United States. Animals 14, 2018. doi: 10.3390/ani14142018

PubMed Abstract | Crossref Full Text | Google Scholar

Littleford-Colquhoun B. L., Freeman P. T., Sackett V. I., Tulloss C. V., McGarvey L. M., Geremia C., et al. (2022). The precautionary principle and dietary DNA metabarcoding: Commonly used abundance thresholds change ecological interpretation. Mol. Ecol. 31, 1615–1626. doi: 10.1111/mec.16352

PubMed Abstract | Crossref Full Text | Google Scholar

Mannervik B. (1985). “[60] glutathione peroxidase,” in Methods in enzymology: glutamate, glutamine, glutathione, and related compounds, vol. 113. (San Diego, CA: Academic Press), 490–495.

Google Scholar

Martin M. (2011). Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet. J. 17. doi: 10.14806/ej.17.1.200

Crossref Full Text | Google Scholar

Niu Y., Zhang Q., Wang J., Li Y., Wang X., and Bao Y. (2022). Vitamin E synthesis and response in plants. Front. Plant Sci. 13. doi: 10.3389/fpls.2022.994058

PubMed Abstract | Crossref Full Text | Google Scholar

Panfili G., Niro S., Bufano A., D’Agostino A., Fratianni A., Paura B., et al. (2020). Bioactive compounds in wild Asteraceae edible plants consumed in the Mediterranean diet. Plant Foods Hum. Nutr. 75, 540–546. doi: 10.1007/s11130-020-00842-y

PubMed Abstract | Crossref Full Text | Google Scholar

Pitel M. O., McKenzie E. C., Johns J. L., and Stuart R. L. (2020). Influence of specific management practices on blood selenium, vitamin E, and beta-carotene concentrations in horses and risk of nutritional deficiency. J. Veterinary Internal Med. 34, 2132–2141. doi: 10.1111/jvim.15862

PubMed Abstract | Crossref Full Text | Google Scholar

Puls R. (1994). Vitamin levels in animal health: diagnostic data and bibliographies. 2nd ed (Clearbrook, British: Sherpa International).

Google Scholar

Quaresma M., Marín C., Bacellar D., Nóvoa M., Navas F. J., and McLean A. (2021). Selenium and vitamin e concentrations in miranda jennies and foals (Equus asinus) in Northeast Portugal. Animals 11. doi: 10.3390/ani11061772

PubMed Abstract | Crossref Full Text | Google Scholar

Rognes T., Flouri T., Nichols B., Quince C., and Mahé F. (2016). VSEARCH: A versatile open source tool for metagenomics. PeerJ. 4, e2584. doi: 10.7717/peerj.2584

PubMed Abstract | Crossref Full Text | Google Scholar

Stokol T. (2020). Hematology red flags: the value of blood smear examination in horses. Veterinary Clinics North America - Equine Pract. 36, 15–33. doi: 10.1016/j.cveq.2019.11.001

PubMed Abstract | Crossref Full Text | Google Scholar

Stowe H. D. and Pagan J. D. (Eds.) (1998). Advances in Equine Nutrition: Selenium supplementation for horse feed (Columbia, Canada: Nottingham Univ Pr), 97–103.

Google Scholar

Taberlet P., Coissac E., Pompanon F., Gielly L., Miquel C., Valentini A., et al. (2007). Power and limitations of the chloroplast trnL (UAA) intron for plant DNA barcoding. Nucleic Acids Res. 35. doi: 10.1093/nar/gkl938

PubMed Abstract | Crossref Full Text | Google Scholar

Traber M. G. and Head B. (2021). Vitamin E: How much is enough, too much and why! Free Radical Biol. Med. 177, 212–225. doi: 10.1016/j.freeradbiomed.2021.10.028

PubMed Abstract | Crossref Full Text | Google Scholar

U.S. Geological Survey Selenium in counties of the conterminous states. Available online at: https://mrdata.usgs.gov/geochem/doc/averages/se/usa.html (Accessed January 2021).

Google Scholar

Valle E., Pozzo L., Giribaldi M., Bergero D., Gennero M. S., Dezzutto D., et al. (2018). Effect of farming system on donkey milk composition. J. Sci. Food Agric. 98, 2801–2808. doi: 10.1002/jsfa.8777

PubMed Abstract | Crossref Full Text | Google Scholar

Van Metre D. C. and Callan R. J. (2001). Selenium and vitamin E. Vet. Clin. North Am. Food Anim. Pract. 17, 373–402. doi: 10.1016/s0749-0720(15)30034-7

PubMed Abstract | Crossref Full Text | Google Scholar