• CASA ;
  • detached heads;
  • DNA integrity;
  • elephant;
  • H33342;
  • morphology;
  • SCSA


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Ejaculates from nine Asian and two African elephants were analysed to gain a further understanding of mechanisms underlying variable semen quality after transrectal massage. Semen analysis was performed after collection (0 h; subjective motility parameters only) and after 24 h of chilled storage at 10 °C (24 h; all ejaculate and sperm characteristics). Ejaculates with ≤50% total motility (TM) at 24 h, which represented >90% of collection attempts, contained a sperm population with a high degree of DNA damage (64.2 ± 19.2% fragmented DNA) and an elevated incidence of detached heads (43.3 ± 22.5%). In contrast, good quality ejaculates designated as those with >50% TM at 24 h displayed higher (< 0.05) values of sperm kinetic parameters, DNA integrity and normal morphology. Fertility potential was high for good quality ejaculates from two males (one Asian and one African bull) based on in vitro characteristics after chilled storage for up to 48 h post-collection. Urine contamination of semen, as assessed quantitatively by creatinine concentration, was confirmed as a significant factor in reduced elephant ejaculate quality. However, the identification of considerable DNA damage and morphological degeneration in the majority of ejaculates after only 24 h of chilled storage indicates that sperm ageing could be a primary contributor to inconsistent semen quality in the elephant.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

In conjunction with concerted efforts towards natural breeding, artificial insemination (AI) is assisting institutions to increase the genetic representation of target males and females under the guidance of the relevant Species Survival Plans, and a total of 42 African and Asian elephant calves have been born worldwide by AI using fresh, chilled semen (D. L. Schmitt, personal communication) since the first successful attempts in the late 1990s (Hildebrandt et al., 2000). Nevertheless, a considerable proportion of adult males in the managed ex situ North American population aged more than 20 years have only one or no living offspring (Asian: 11/16, 69%; African: 12/17, 71%; Keele, 2010; Olson, 2011).

Continued success of AI using chilled semen, and the future integration of cryopreserved and/or sex-sorted spermatozoa into elephant breeding programs relies on the ability to collect high quality semen. Ejaculated elephant spermatozoa have been collected for characterization using a variety of methods (post-coital: Landowski & Gill, 1964; electroejaculation [EEJ]: Jones, 1973; Howard et al., 1984; Mar et al., 1992; non-sedated transrectal massage: Heath et al., 1983; Price et al., 1986; Schmitt & Hildebrandt, 1998; transrectal massage under sedation: Portas et al., 2007). Of all the aforementioned semen collection methods, only non-sedated transrectal massage can be conducted on a regular basis. For studies utilizing large numbers of males, which better capture the status of the general captive population, only a low proportion of ejaculates collected using the transrectal massage method are of suitable quality (e.g. ≥60% initial total motility) for use in chilled storage/AI or sperm preservation and sex-sorting research (Saragusty et al., 2009 [10/30, 33%]; Kiso et al., 2011 [15/125, 12%]; O'Brien et al., 2011 [19/118, 16%]). Poor semen quality has been attributed to urine contamination during the transrectal massage collection technique and to socially mediated reproductive suppression experienced by subordinate bulls (Hildebrandt et al., 2000). The objective of this study was to gain a further understanding of potential factors responsible for the inconsistent quality of elephant semen collected by the transrectal massage technique. This was undertaken using an extensive panel of ejaculate and sperm quality parameters across multiple elephants following chilled semen storage.

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Semen collection procedures described within were reviewed and approved by each institution's Animal Care and Use Committee.


Ejaculates were collected from nine adult captive-born Asian elephants at seven institutions (aged 9–37 years), and two African elephants at two institutions (aged 9  and 27 years) during 2007–2012 (Table 1). One of the African bulls and seven of the Asian bulls were of proven fertility. All males were in visual, auditory and olfactory contact with cyclic female elephants, and were managed under a protected contact or semi-protective contact training regimen.

Table 1. Description of elephants and samples collected in Study 1
SpeciesMale numberFacilityProven breederAge at semen collection (year)Ejaculate quality type and number analysed at 0 and 24 h post-collectionaTotal number of LOW ejaculates analysed at 0 h onlybTotal number of collection attemptsc
  1. a

    LOW: 0 h TM ≤ 35%, 24 h TM ≤ 5%; INTERMEDIATE: 0 h TM ≥ 50 and ≤80%, 24 h TM < 5 and ≤50%; HIGH: 0 h TM ≥ 70%, 24 h TM > 50%.

  2. b

    A minimum of one and a maximum of four LOW quality ejaculates were analysed per bull; if subsequent collections only yielded samples with initial TM ≤ 35% (LOW), such samples were not shipped for further analysis.

  3. c

    Of all collection attempts 89.8% (141/157) yielded semen and 82.3% (116/141) were poor quality (0 h TM ≤ 35%).

  4. d

    Collections occurred after Male 7 was transferred to Facility 5.

74 and 5Yes1910011214
Total 16 6 7 29 63 100
Total 4 6 6 16 33 57

Reagents and media

All chemicals were of analytical grade and cell culture tested where possible by the manufacturer. Unless otherwise stated, all media components were purchased from Sigma-Aldrich (St Louis, MO, USA) and were prepared with tissue culture grade water (Sigma-Aldrich; or Millipore, Billerica, MA, USA). Diluents containing egg yolk were prepared by ultracentrifugation (10 000 g) for 1 h at 10 °C. The supernatant was filtered (0.22 μm; Millipore) and frozen at −80 °C for a maximum of 24 months.

Semen collection and processing

Semen was collected by transrectal massage of the pelvic urethra and ampullae (Price et al., 1986; Schmitt & Hildebrandt, 1998). Transrectal ultrasound was used to examine the accessory glands prior to semen collection, in particular the ampullae, to ensure an adequate semen reservoir was present and to avoid overstimulation and urination (Hildebrandt et al., 2000). In addition, semen collection was typically attempted after less than 10 days of sexual abstinence to minimize the collection of aged spermatozoa. The penis was washed with warm (~30 °C) tap water to remove debris and residue urine, then dried with sterile cotton gauze. Semen was collected into sterile WHIRL-PAK bags (NASCO, Fort Atkinson, WI, USA). Bags were changed often resulting in up to 10 separate fractions obtained during each semen collection attempt.

Semen was diluted (between 1 : 2 and 1 : 3, semen : diluent, v/v) with a N-tris(hydroxymethyl)methyl-2-aminoethane sulfonic acid (TES)-TRIS yolk buffer (TYB) diluent, modified from Graham et al. (1972) (315 ± 5 mOsmol/kg and pH 7.2–7.4) containing hen's egg yolk (20% v/v), gentamicin sulfate (50.0 μg/mL), tylosin tartrate (8 μg/mL), lincomycin hydrochloride (0.1 mg/mL) and spectinomycin sulfate tetrahydrate (0.2 mg/mL, Linco-Spectin; Pharmacia & Upjohn, Kalamazoo, MI, USA).

Semen was transported overnight to the reproductive laboratory using an Equitainer (Hamilton Research, South Hamilton, MA, USA; cooled at ≈ −0.3 °C/min to ~5 °C; temperature on arrival was 7–13 °C) placed in a box lined with styrofoam packing material for additional insulation against temperature fluctuations.

Evaluation of semen and sperm characteristics

Semen characteristics were determined after collection (0 h; semen volume, colour) or after overnight shipment (at 24 h post-collection; sperm concentration, pH, osmolality, creatinine concentration). Sperm characteristics determined after collection (0 h) comprised subjective assessment of motility parameters only (total motility, progressive motility, kinetic rating, agglutination rating). All other sperm characteristics as described in the following sections (computer-assisted sperm motility parameters, agglutination rating, membrane and acrosome integrity and morphology) were assessed at 24 h post-collection. Initial findings from a concurrent study in our laboratory on the development of sperm sorting technology for the elephant indicated that the resolution of X and Y sperm nuclei populations during flow cytometric analysis differed between high and low quality ejaculates. It was hypothesized that differences in flow cytometric parameters among ejaculate quality groups may reveal attributes of DNA quality in addition to those provided by the chromatin structure assay conducted herein. Thus, evaluations of in vitro sperm characteristics in Study 1 also included two flow cytometric parameters as described in a subsequent section.

Semen characteristics

After overnight shipment, diluted semen was examined microscopically (×400 and ×1000) and those samples displaying an excess amount of debris or round cells (n = 2) underwent standard cytological analysis to exclude the presence of bacteria or leukocytes (ejaculates were negative for both cell types). Sperm concentration (haemocytometer, Reichert, Buffalo, NY, USA) and volume were determined from diluted semen, and a raw semen aliquot was assessed for pH (20 μL; pH indicator strips; EM Science, Gibbstown, NJ, USA) and osmolality (30 μL; Advanced Instruments Inc., Norwood, MA, USA).

Semen creatinine

Creatinine (Cr) concentration of an aliquot of semen (100 μL) was measured using a modification of the Jaffe reaction (colorimetric method: modified from Taussky, 1954). In brief, creatinine and picric acid react in an alkaline environment to produce creatinine picrate, a red-orange tautomer, which was measured by spectrophotometry 30 min after commencement of the reaction. Inter-assay coefficient of variation for the creatinine assay was 4.96%. For five ejaculates, there was an insufficient volume of raw semen for analysis so Cr concentration of diluted samples was determined for all ejaculates. A Cr threshold was determined and diluted samples with a Cr value below this threshold were considered to be representative of urine-free semen. The formula for the threshold was:

  • display math

where: A = raw semen volume (mL); B = extended semen volume (mL); 14 = the Cr concentration (μg/mL) of ejaculated elephant semen with negligible urine contamination [which is comparable to that reported in the horse (5.0 μg Cr/mL, Turner et al., 1995) and the bull (14.0–23.7 μg Cr/mL, Cevik et al., 2007)]; C = the Cr concentration (μg/mL) of the extender used in this study (Cr is derived from the extender's egg yolk component, and ranged from 7.0 to 13.0 μg Cr/mL). For example, for an extender with a creatinine concentration of 13.0 μg Cr/mL, the Cr threshold for semen diluted at 1 : 2 or 1 : 3 (semen:extender) was 13.3 μg/mL using the above formula.

Sperm motility and agglutination characteristics

Sperm motility [total motility (TM), progressive motility (PM)], kinetic rating (KR: 0–5; 0 = no movement; 5 = rapid, forward progressive movement) and agglutination rating (0–5; 0 = no agglutination; 5 = 100% of motile spermatozoa displaying head-head agglutination) of diluted samples were assessed subjectively by multiple evaluators after collection by analysing four to five fields of view of diluted spermatozoa placed on heated slides (37 °C) using bright-field optics (×400 magnification). The agglutination rating was also estimated at all evaluation time points after semen collection. After overnight shipment, motility was objectively assessed using computer-assisted sperm analysis (CASA; HTM-IVOS Version 12.2; Hamilton-Thorne, Inc., Beverly, MA, USA) as described previously for the elephant (O'Brien et al., 2012). Five randomly selected microscopic fields were scanned (30 and 60 frames/sec) to calculate: average pathway velocity (VAP, μm/sec), straight-line velocity (VSL, μm/sec), curvilinear velocity (VCL, μm/sec), amplitude of lateral head displacement (ALH, μm), beat cross frequency (BCF, Hz), straightness of sperm movement (STR,%, calculated as VSL/VAP) and linearity (LIN,%, VSL/VCL). For progressive cells, the VAP was >25 μm/sec and STR was >75% (PM). Spermatozoa with a VAP >5 μm/sec were considered motile (TM). An additional group based on velocity of movement was also determined: rapid velocity (VAP >25 μm/sec,%).

Sperm plasma membrane and acrosome integrity

Plasma membrane integrity was assessed after chilled storage using a light microscopy live–dead exclusion stain (eosin-nigrosin; IMV International Corp., Maple Grove, MN, USA). Ten microlitres of sample was mixed with 10 μL of the stain for 5 sec. A smear was made and allowed to air dry for evaluation within 60 min (100 spermatozoa per sample, ×1000). Spermatozoa were classified as viable (no stain uptake) or non-viable (partial or complete stain uptake).

Plasma membrane and acrosome integrity of diluted chilled-stored samples were also assessed by fluorescence microscopy (×400) using the fluorochromes propidium iodide (PI) and fluorescein isothiocynate-conjugated Arachis hypogaea (peanut) agglutinin (FITC-PNA), respectively (Robeck et al., 2011). The assay has been validated for use with Asian elephant spermatozoa (O'Brien et al., 2012). Briefly, an aliquot of sperm sample (~0.2 × 106 spermatozoa) was stained with 86 μg/mL of PI and 69 μg/mL of FITC–PNA and evaluated under a low bright-field setting (400×) to permit visualization of non-fluorescent spermatozoa (previously immobilized by the addition of glutaraldehyde at a 0.06% final concentration).

A total of 100 cells were classified per sample using the following staining patterns: no stain (intact plasma membrane and an intact acrosome), green staining in the acrosome region including the equatorial segment (intact plasma membrane and a damaged or reacted acrosome), red staining (non-intact plasma membrane and an intact acrosome), red and green staining (non-intact plasma membrane with a damaged or reacted acrosome). Data on spermatozoa with an intact plasma membrane and acrosome are presented.

Sperm morphology

Sperm morphology was determined after chilled storage by placing 10 μL of diluted semen in 100 μL of 2% (v/v) glutaraldehyde fixative (in PBS, pH 7.0–7.4, 290 ± 10 mOsmol/kg). Spermatozoa (100 per sample) were examined by phase-contrast microscopy (×1000) for structural abnormalities. The proportion of spermatozoa with heads detached from the tail (‘detached heads’) was determined at the time of sperm concentration evaluation whereby semen diluted in 6% saline (1 : 100, semen: saline solution, v/v) was mixed by vortexing for 2–3 sec and examined on a haemocytometer (×200, ≥400 spermatozoa/sample). This method was deemed acceptable for measuring the percentage of detached heads after it was demonstrated that even prolonged vortexing (1–2 min) did not cause an increase in the incidence of this morphological parameter (J. K. O'Brien, unpublished observations).

Sperm DNA integrity

After 24 h of chilled storage, an aliquot of diluted semen (≥55 μL) was snap frozen in liquid nitrogen and retrospectively analysed for the susceptibility of spermatozoa to DNA denaturation using the sperm chromatin structure assay (SCSA, reviewed by Evenson et al., 2002). Use of the SCSA has been described previously in an Asian elephant (O'Brien et al., 2012). Thawed spermatozoa were diluted in a TRIS buffer (1 mm disodium EDTA, 0.01 m Tris-HCl, 0.15 m NaCl, pH 7.4) and exposed to an acid-detergent solution (0.08 n HCl, 0.1% Triton X-100, 0.15 m NaCl, pH 1.2) for 30 sec after which the solution was quenched with an acridine orange solution [0.1 m citric acid monohydrate, 0.2 m Na2HPO4, 0.15 m NaCl, 1 mm disodium EDTA, 4 μg/mL acridine orange stock solution (Polysciences Inc, Warrington, PA, USA), pH 6] and immediately analysed (5 000 events per sample; FACScan; Becton Dickinson, Mountain View, CA, USA; 100–200 cell/sec after a 30 sec equilibration in the cytometer). Flow cytometer settings were calibrated using a control sample, comprising spermatozoa from a known fertile Asian elephant (%COMPαt: 7.6–8.7%). SCSA values were calculated using WinList software (Verity Software House, Topsham, ME, USA). Analysis endpoints included: (i) the percentage of cells displaying denatured DNA (Cells Outside the Main Population, %COMPαt), determined by selecting those spermatozoa located to the right of the control main population, and represents a percentage of the total number of spermatozoa with denatured, or fragmented, DNA (with the denaturation occurring preferentially at sites of pre-existing DNA strand breaks), (ii) the mean of αt (Meanαt), and (iii) the standard deviation of αt (SDαt). The inter-assay coefficient of variation (= 5) using the elephant control sample was 16.5% for COMPαt, 4.5% for Meanαt and 8.0% for SDαt.

Flow cytometric parameters of X and Y chromosome-bearing spermatozoa after H33342 staining

For Study 1 only, a sperm nuclei sample was prepared from each ejaculate for flow cytometric analysis. Following chilled storage for 24 h, diluted semen was filtered (35 μm), and centrifuged (850 g, 10 min) twice to remove egg yolk extender. The pellet was resuspended after each centrifugation in Androhep Enduraguard (AE; Minitube of America, Verona, WI, USA), then sonicated (Branson Ultasonics, Danbury, CT, USA) until >95% of midpieces and tails had been removed from heads. Following sonication, samples were centrifuged (150 g, 10 min) and the pellet resuspended in AE. This process was repeated three times to permit removal of >95% of midpieces and tails from the final sperm nuclei sample and maximum resolution of X and Y chromosome-bearing spermatozoa. Samples were then incubated with varying concentrations of Hoechst 33342 (Sigma) (range: 17.8–89.0 μm, 100 × 106 sperm nuclei/mL) at 33.5 °C for 1 h. A high-speed cell sorter (MoFloSX; Dako, Fort Collins, CO, USA) modified for sperm sorting (reviewed by Sharpe & Evans, 2009) operating at 40 psi was used to analyse sperm nuclei at 3 000 events per second. H33342 was excited by UV light from a diode pumped solid state pulse laser (Vanguard 350 HMD-355; Spectra Physics, Mountain View, CA, USA) running at 175 mW. Prior to all analyses, the MoFloSX was calibrated and aligned with domestic bull nuclei. Assessments included the degree of: (i) X–Y resolution (depth of split between the X and Y populations on the histogram output) and, (ii) correct orientation (the percentage of nuclei with their flat surface oriented towards the laser beam).

Study 1: Semen characteristics, sperm characteristics and X–Y sperm flow cytometry parameters across ejaculates of varying quality

A total of 157 semen collection sessions were conducted across eleven bulls (nine Asian and two African; Table 1) and ejaculates were used to determine in vitro characteristics of semen and spermatozoa immediately after collection (= 141; subjective motility assessments only) and after storage at ~10 °C for 24 h (= 45; assessment of all in vitro characteristics as described in subsequent sections) to permit transport to the reproductive laboratory. Flow cytometry parameters of X and Y chromosome-bearing sperm nuclei were also examined after 24 h of chilled storage.

Study 2: Effect of prolonged chilled storage on in vitro sperm characteristics of high quality ejaculates from one Asian and one African elephant

A preliminary study was conducted to examine the effect of a 48 h chilled storage period on in vitro sperm quality of high quality ejaculates. A total of six high quality ejaculates from Study 1 (three ejaculates from one Asian and three ejaculates from one African elephant; initial total motility: ≥70%) were used. Because of the low number of high quality ejaculates available for this research (Table 1), each ejaculate was divided into three equal aliquots to create the minimum of nine treatment replicates necessary for statistical analyses. While this method partially violates the assumption of independence of samples for ANOVA, it allowed for sufficient statistical power to perform these preliminary comparisons. In vitro characteristics were assessed within 20 min of semen collection (0 h) and at 24, 30 and 48 h after chilled storage.

Statistical analysis

Data for ejaculate and sperm quality parameters were analysed using ANOVA (SigmaStat, Version 3.5; SSPS Inc., San Rafael, CA, USA). Data for in vitro sperm characteristics were analysed across different time points using one-way repeated measures ANOVA (Study 2). Prior to ANOVA, data were normalized, when necessary, by logarithmic or arc-sin transformations (total spermatozoa per ejaculate, initial agglutination rating). All pairwise multiple-comparison procedures between means were conducted using the Tukey test. For the comparison of ejaculate parameters (sperm concentration and initial TM) among ejaculate quality groups in the African elephant data set, the Dunn's test was used after performing Kruskal–Wallis ANOVA on Ranks because of unequal variances. Pearson correlation coefficients were used to examine the relationships among sperm quality parameters. < 0.05 was considered to be significant unless otherwise noted. Data are presented as the untransformed mean ± SD.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Study 1: Semen characteristics, sperm characteristics and X–Y sperm flow cytometry parameters across ejaculates of varying quality

Twenty-nine Asian elephant ejaculates and sixteen African elephant ejaculates underwent complete analysis (Table 1). Total semen volume of fractions shipped to the reproductive laboratory was 18.6 ± 19.2 mL (range: 1.0–85.0 mL), raw semen concentration was 102.6 ± 92.0 × 107 spermatozoa/mL (range: 4.9–300.0 × 107 spermatozoa/mL) and the total number of spermatozoa was 14.2 ± 19.2 × 109 (range: 0.2–86.3 × 109 spermatozoa). Three ejaculate quality groups were determined for both species based on total motility (TM) after collection (0 h) and after 24 h of chilled storage (24 h): (i) LOW: 0 h TM ≤50%, 24 h TM ≤5%, (ii) INTERMEDIATE: 0 h TM ≥50% and ≤80%, 24 h TM > 5 and ≤50%, (iii) HIGH: 0 h TM ≥ 70%, 24 h TM > 50% (Table 2).

Table 2. Ejaculate and motility characteristics of Asian (= 29) and African elephant (= 16) samples of varying quality after collection and chilled storage for 24 h at 10 °C
ParameterEjaculate quality group (n)a
  1. a

    Ejaculate quality classification was based on total motility after collection and following 24 h of chilled storage at 10 °C.

  2. b

    pH, osmolality and creatinine were determined from a raw semen aliquot after 24 h of chilled storage.

  3. c

    Kinetic rating: 0 = no movement, 5 = rapid, steady forward progression.

  4. d

    Motile sperm agglutination rating: 0 = none, 5 = 100% of motile cells agglutinated.

  5. e

    Motility parameters were determined by computer-assisted sperm analysis (CASA). CASA parameters: Rapid velocity, >25 µm/s; VAP, average pathway velocity; VSL, straight-line velocity; VCL, curvilinear velocity; ALH, amplitude of lateral head displacement; BCF, beat cross frequency; STR, straightness of sperm movement [STR (%) = VSL/VAP].

  6. f

    Combined data from both species for parameters where there was no significant difference (p < 0.01) between species within ejaculate quality group.

  7. g,h,i Values without a common superscript within the same row within each species, or within combined dataf are different (< 0.01).

Ejaculate and initial sperm characteristicb
Total fraction volume (mL)20.8 ± 22.516.5 ± 14.533.7 ± 23.512.8 ± 6.35.0 ± 5.110.9 ± 8.018.4 ± 20.210.8 ± 12.023.2 ± 21.0
Sperm conc. (107/mL)99.2 ± 86.2g66.9 ± 65.1g6.3 ± 5.2h97.5 ± 67.1150.8 ± 115.6173.4 ± 90.1110.8 ± 82.0108.9 ± 99.683.5 ± 104.5
Total spermatozoa (×109)19.4 ± 25.2g14.8 ± 20.4g2.2 ± 1.6h5.2 ± 2.79.7 ± 14.617.1 ± 16.118.5 ± 23.012.2 ± 17.19.1 ± 13.0
pH7.1 ± 0.67.8 ± 0.37.6 ± 0.27.6 ± 0.37.4 ± 0.37.3 ± 0.27.2 ± 0.67.5 ± 0.57.5 ± 0.2
Osmolality (mOsmol/kg)296.5 ± 42.7298.0 ± 19.1281.6 ± 8.3377.0 ± 127.3269.0 ± 5.7272.5 ± 14.3306.6 ± 58.4284.2 ± 19.6278.3 ± 11.1
Creatinine (µg/mL)35.2 ± 21.8g29.8 ± 12.4g,h5.0 ± 3.3h31.5 ± 14.9g15.2 ± 3.3g,h11.5 ± 1.6h34.7 ± 20.6g21.0 ± 10.7g,h8.0 ± 4.2h
Total motility (%)15.5 ± 14.5g57.1 ± 10.5h80.0 ± 7.1i3.0 ± 2.3g79.2 ± 2.0h85.0 ± 5.5h13.6 ± 13.8g82.3 ± 6.7h
Progressive motility (%)10.2 ± 10.7g48.5 ± 14.1h74.2 ± 5.7i0.6 ± 0.3g58.0 ± 5.5h67.2 ± 4.8h9.0 ± 10.6g53.2 ± 11.3h71.0 ± 6.3i
Kinetic ratingc1.8 ± 1.2g3.7 ± 0.8h4.3 ± 0.4h1.3 ± 0.3g4.2 ± 0.8h4.2 ± 0.4h1.7 ± 1.1g3.9 ± 0.8h4.2 ± 0.4h
Agglutination ratingd0.2 ± 0.3g1.3 ± 0.6h0.5 ± 0.4g,h0.2 ± 0.3g1.2 ± 0.4h1.0 ± 0.6g,h0.2 ± 0.3g1.2 ± 0.5h0.7 ± 0.6h
Sperm characteristic post-chilled storage for 24 he
Total motility (%)0.5 ± 1.3g14.2 ± 16.9h66.0 ± 7.8i2.6 ± 4.3g33.0 ± 14.1h66.3 ± 11.1i1.0 ± 2.4g23.6 ± 17.8h66.2 ± 9.1i
Progressive motility (%)0.3 ± 1.0g9.5 ± 7.6h28.7 ± 13.4i0.2 ± 0.5g14.7 ± 10.3h34.7 ± 10.1i0.3 ± 0.9g12.1 ± 9.1h31.5 ± 11.9i
Kinetic ratingc0.3 ± 1.0g2.5 ± 1.8h4.1 ± 0.7i0.7 ± 1.0g2.8 ± 0.6h4.7 ± 0.3i0.4 ± 1.0g2.7 ± 1.3h4.4 ± 0.6i
Agglutination ratingd0.0 ± 0.0g1.2 ± 0.9h0.6 ± 0.4h0.1 ± 0.1g0.7 ± 0.5g,h1.1 ± 0.5h0.1 ± 0.1g0.9 ± 0.8h0.8 ± 0.5h
Rapid velocity (%)0.2 ± 0.4g11.0 ± 8.5h58.8 ± 9.4i1.4 ± 2.2g28.3 ± 13.3h62.0 ± 12.2i0.5 ± 1.2g17.5 ± 16.3h60.7 ± 10.7i
VAP(µm/s)33.8 ± 5.5g45.3 ± 4.0g103.3 ± 7.4h32.3 ± 8.8g63.7 ± 20.1g112.2 ± 13.2h33.2 ± 5.9g59.1 ± 19.1g108.6 ± 11.7h
VSL µm/s)24.6 ± 3.3g29.9 ± 2.4g62.3 ± 13.0h20.7 ± 7.6g48.2 ± 16.2g84.0 ± 13.0h23.0 ± 4.9g43.6 ± 16.1g75.3 ± 16.6h
VCL (µm/s)52.4 ± 23.2g108.6 ± 16.3g213.7 ± 30.7h85.9 ± 26.4g132.1 ± 34.7g221.3 ± 35.5h65.8 ± 24.8g126.2 ± 31.8h218.3 ± 32.1i
ALH (µm)4.9 ± 0.65.5 ± 3.78.8 ± 2.23.2 ± 2.2g6.8 ± 2.5g10.0 ± 2.2h4.2 ± 1.5g6.4 ± 2.6g9.5 ± 2.2h
BCF (Hz)33.7 ± 5.230.8 ± 6.837.4 ± 3.934.2 ± 2.332.7 ± 5.329.6 ± 2.133.9 ± 3.932.2 ± 5.232.7 ± 4.8
STR (%)69.0 ± 2.065.0 ± 9.962.0 ± 11.765.0 ± 4.272.8 ± 9.573.0 ± 5.167.4 ± 3.470.9 ± 9.668.6 ± 9.6
LIN (%)42.3 ± 11.529.0 ± 5.731.8 ± 7.630.0 ± 1.436.7 ± 11.239.3 ± 7.837.4 ± 10.434.8 ± 10.336.3 ± 8.3

Across all collection attempts, 89.8% (141/157) yielded semen, and 82.3% (116/141), 8.5% (12/141) and 9.2% (13/141) were of LOW, INTERMEDIATE and HIGH quality, respectively. Age of males varied considerably (Table 1) and did not differ across ejaculate quality groups (> 0.05; LOW: 20.4 ± 11.7 years; INTERMEDIATE: 20.7 ± 10.5 years; HIGH: 20.6 ± 8.8 years; data combined across species), nor did sperm concentration and the total number of spermatozoa (Table 2). For Asian elephants, sperm concentration and total sperm number for HIGH quality ejaculates were lower (< 0.05) than those of LOW and INTERMEDIATE, but it should be noted that six of the seven HIGH ejaculates were derived from one male (Male 1). Ejaculates of LOW and HIGH quality displayed similar characteristics between species (> 0.05; Tables 2 and 3). For INTERMEDIATE ejaculates, initial total motility was higher (> 0.05) for the African than the Asian elephant (Table 2). However, African elephant ejaculates were derived from only two males and species comparisons presented in this study are considered to be of preliminary in nature.

Table 3. In vitro sperm characteristics (morphology, membrane and DNA integrity) and flow cytometric parameters of Asian and African elephant ejaculates of varying quality after chilled storage for 24 h at 10 °C
ParameterEjaculate quality group (n)a
  1. a

    Ejaculate quality classification was based on total motility after collection and following 24 h of chilled storage at 10 °C.

  2. b

    Light microscopic analysis of plasma membrane integrity after eosin-nigrosin staining.

  3. c

    Fluorescence microscopic analysis of plasma membrane and acrosome integrity after staining with propidium iodide & FITC-PNA.

  4. d

    X–Y sperm nuclei resolution determined by H33342 staining and flow cytometric analysis.

  5. e

    Spermatozoa with their flat surface oriented at 0° to the laser beam during flow cytometric analysis.

  6. f

    Combined data from both species for parameters where there was no significant difference (< 0.01) between species within ejaculate quality group.

  7. g,h,i Values without a common superscript within the same row within each species, or within combined dataf are different (< 0.01).

Detached heads (%)51.0 ± 26.8g31.0 ± 19.9g8.7 ± 7.5h44.0 ± 5.9g35.6 ± 13.0g13.0 ± 6.4h49.5 ± 24.0g33.3 ± 16.2g10.7 ± 7.1h
Morphologically normal (%)30.1 ± 17.9g51.6 ± 16.7h75.2 ± 9.8i42.3 ± 1.5g54.7 ± 6.7g68.7 ± 9.8h32.9 ± 16.4g51.5 ± 9.6g71.9 ± 9.9h
Pl. membrane intact (%)b27.5 ± 28.3g37.2 ± 20.1g79.1 ± 5.7h17.0 ± 15.3g49.3 ± 13.4h72.7 ± 10.8i25.6 ± 26.3g43.8 ± 17.1g76.2 ± 8.7h
Pl. membrane & acr. intact (%)c24.6 ± 24.8g29.5 ± 10.6g71.3 ± 13.9h16.3 ± 15.6g45.5 ± 18.4h69.9 ± 9.5i23.1 ± 23.1g40.2 ± 17.1g70.4 ± 10.2h
Denatured DNA (%COMPαt)68.7 ± 18.3g69.4 ± 21.6g7.2 ± 6.5h62.8 ± 10.3g40.6 ± 13.2g18.3 ± 5.8h67.9 ± 17.2g55.0 ± 22.5g12.3 ± 8.3h
Meanαt405.9 ± 121.9g407.0 ± 185.4g226.6 ± 28.7g425.0 ± 27.0g226.9 ± 14.5h230.8 ± 37.0h408.4 ± 113.3g316.9 ± 153.5g,h228.3 ± 30.9h
SDαt140.4 ± 71.5g134.9 ± 71.1g34.9 ± 14.0h156.3 ± 22.4g45.8 ± 5.7h53.8 ± 20.0h142.5 ± 66.7g90.3 ± 66.5g,h42.7 ± 18.6h
X–Y sperm resolution (% split)d28.3 ± 29.1g21.4 ± 17.8g60.7 ± 21.1h39.2 ± 20.7g40.7 ± 3.5g75.3 ± 3.3h29.6 ± 28.0g26.9 ± 17.3g68.2 ± 11.8h
Oriented spermatozoa (%)e57.4 ± 18.272.3 ± 7.173.5 ± 17.761.8 ± 11.066.5 ± 6.464.0 ± 3.557.9 ± 17.470.3 ± 6.968.5 ± 9.0

Of the biochemical parameters analysed (pH, osmolality and creatinine), only Cr varied among ejaculate quality groups, with higher (> 0.05) raw semen Cr concentrations observed for LOW than HIGH, and similar (> 0.05) values for INTERMEDIATE and HIGH (Table 2). Urine contamination of semen, as indicated by Cr values exceeding the Cr threshold of 14 μg/mL (raw semen) or 13.3 μg/mL (diluted semen; using an extender with a Cr concentration of 13.0 μg/mL), occurred in 100.0% (20/20) of LOW, 50.0% (6/12) of INTERMEDIATE and did not occur in HIGH ejaculates (0/13). Colour of all ejaculates was white to slightly off white/amber hue, with the exception of two ejaculates in the LOW group which displayed a yellow hue and the highest Cr values across the whole data set (60.0 and 89.0 μg/mL, respectively). Creatinine was positively correlated with osmolality (= 0.59, < 0.01), but not pH (= −0.11, = 0.62). Osmolality of non-contaminated and urine-contaminated ejaculates was similar (> 0.05; 300.5 ± 52.7 mOsm/kg and 278.7 ± 11.6 mOsm/kg, respectively, data pooled across species and ejaculate types).

As expected, all initial motility parameters (TM, PM, KR) were higher (< 0.05) for HIGH and INTERMEDIATE than LOW ejaculate types, and HIGH ejaculate types also possessed increased (< 0.05) values of 0 h PM than INTERMEDIATE and LOW. After 24 h of chilled storage, all motility parameters (TM, PM, KR) and the velocity parameter VCL for HIGH exceeded (< 0.05) those of INTERMEDIATE and LOW (Table 2). Overall, the degree of motile sperm agglutination after collection and chilled storage was low across all groups (0.5 ± 0.6, data pooled across species and ejaculate types), but at both time points the agglutination rating was higher (< 0.05) for HIGH and INTERMEDIATE than LOW (Table 2).

After 24 h of chilled storage, the proportion of spermatozoa with an intact plasma membrane using eosin-nigrosin staining was higher (< 0.05) for HIGH than INTERMEDIATE and LOW ejaculates (Table 3). Plasma membrane and acrosome integrity using the dual staining method (PI and FITC-PNA staining) of HIGH was similar (> 0.05) to INTERMEDIATE, but greater (< 0.05) than LOW where less than 25% of cells displayed an intact membrane and acrosome (Table 3). Chilled-stored HIGH ejaculates contained a greater proportion of morphologically normal spermatozoa and a reduced proportion of cells with detached heads than INTERMEDIATE and LOW (< 0.05; Table 3).

The HIGH ejaculate type displayed high DNA integrity at 24 h of storage, with an average of 12.3% of spermatozoa displaying fragmented DNA (%COMPαt, Table 3; Fig. 1). DNA integrity of LOW and INTERMEDIATE ejaculate types was substantially poorer (< 0.05) than HIGH, where an average of 67.9 and 55.0% of spermatozoa in the former groups exhibited fragmented DNA after 24 h of chilled storage (Table 3; Fig. 1). Meanαt and SDαt values were reduced (< 0.05) for HIGH than LOW (Table 3). Greater resolution of X and Y sperm populations was observed for HIGH than INTERMEDIATE and LOW ejaculate types (< 0.05, Table 3). There were no differences (> 0.05) in the percentage of correctly oriented spermatozoa among the three ejaculate groups (Table 3).


Figure 1. Sperm chromatin structure assay (SCSA) cytograms [left; green (double-stranded DNA] vs. red (single-stranded DNA) fluorescence] and corresponding αt frequency histograms (right) representative of a HIGH (top panel; %COMPαt: 10.0%, Meanαt: 289.2, SDαt: 69.0) and a LOW (bottom panel; %COMPαt: 43.0%, Meanαt: 367.0, SDαt: 86.0) quality Asian elephant ejaculate.

Download figure to PowerPoint

Across all ejaculates, there were significant correlations between ejaculate/sperm characteristics [Cr concentration, motility parameters, plasma membrane integrity and morphology (normal morphology and detached heads)] and SCSA variables (COMPαt, Meanαt, SDαt), and between those same ejaculate/sperm characteristics and X–Y resolution (Table 4). A striking positive correlation was observed between the incidence of detached heads and fragmented sperm DNA (< 0.001), and both these parameters displayed an inverse relationship (< 0.001) with the degree of resolution of X and Y sperm populations (Table 4). Correct orientation of the sperm nucleus during the assessment of X–Y resolution was negatively correlated with the proportion of detached heads and all SCSA variables (< 0.01–0.001) (Table 4).

Table 4. Pearson correlation coefficients (r) between ejaculate and sperm characteristics, SCSA and flow cytometric variables (n = 45; data combined across ejaculates from Asian and African elephants)
In vitro sperm characteristicSCSA variableaSperm flow cytometric variablea
%COMPαtMeanαtSDαtX–Y resolution (%)bCorrect orientation (%)c
  1. a

    Determined after 24 h of chilled storage.

  2. b

    X–Y sperm resolution determined by H33342 staining and flow cytometric analysis.

  3. c

    Correctly oriented spermatozoa were spermatozoa with their flat surface oriented at 0° to the laser beam during flow cytometric analysis.

  4. d

    < 0.05.

  5. e

    < 0.01.

  6. f

    p < 0.001.

Ejaculate and initial sperm characteristic
Total motility (%)−0.71f−0.64f−0.65f0.380.17
Progressive motility (%)−0.75f−0.61f−0.63f0.310.14
Kinetic rating−0.66f−0.39d−0.59e0.370.23
Agglutination rating−0.32−0.45d−0.370.100.31
Sperm characteristic post-chilled storage for 24 h
Total motility (%)−0.88f−0.69f−0.70f0.51e0.11
Progressive motility (%)−0.81f−0.63f−0.65f0.51e0.17
Agglutination rating−0.17−0.15−0.056−0.09−0.42
Plasma membrane integrity (%)−0.85f−0.75f−0.84f0.77f0.50
Plasma membrane & acrosome integrity (%)−0.85f−0.80f−0.85f0.77f0.47
Normal morphology (%)−0.88f−0.76f−0.83f0.78f0.49
Detached heads (%)0.82f0.80f0.86f−0.85f−0.64f
Degree of X–Y sperm resolution (% split)−0.75f−0.70e−0.80f
Correctly orientated spermatozoa (%)−0.46d−0.60e−0.74f
Fragmented DNA (COMPαt)−0.71f−0.41

Within ejaculate and in vitro sperm characteristics, Cr displayed a positive correlation with the proportion of spermatozoa displaying detached heads (= 0.49; < 0.01) and a negative correlation (= −0.53 to −0.74; < 0.01–0.001) with all 24 h motility parameters except agglutination rating. As expected, 24 h total motility, progressive motility, plasma membrane integrity and the proportion of plasma membrane intact cells with an intact acrosome displayed positive correlations (< 0.001) between each other, and a negative correlation with the incidence of detached heads (< 0.001).

An example of X–Y sperm resolution from a LOW and HIGH quality ejaculate from the same Asian elephant bull (Male 6) is shown in Fig. 2. The LOW quality sample was collected on 5th March 2009 (1% total motility at 0 h, 77% detached heads and 81% fragmented DNA at 24 h); the HIGH quality sample was collected 3 weeks later on 2nd April, the day after the bull had bred a cow for three consecutive days (85% total motility at 0 h, 5% detached heads and 19% fragmented DNA at 24 h).


Figure 2. Flow cytometric dotplot and histogram outputs showing fluorescence signals generated by X and Y chromosome-bearing sperm nuclei (Asian elephant, Male 6) of a LOW (a) and a HIGH (b) quality ejaculate (collected 3 weeks apart), displaying low and high resolution of X and Y populations, respectively. The HIGH quality ejaculate was collected from the male after he had bred a female for three consecutive days. Fluorescence signals from correctly oriented sperm nuclei shown in Region 1 (R1) of the dotplot output are displayed in the histogram output, and the degree of resolution is indicated by the depth of the split between X- and Y sperm populations.

Download figure to PowerPoint

Individual profiles of ejaculate quality

Between one and five ejaculates per bull were analysed in this study, with the exception of one Asian (Male 1) and one African bull (Male 10), where 10 and 15 ejaculates, respectively, were assessed (Table 1). Profiles of ejaculate quality (of the highest quality fractions for each collection attempt) over time for those two bulls are presented in Fig. 3. For the Asian elephant bull, LOW quality ejaculates were produced during his first year of semen collection attempts at 9 years of age. Semen quality underwent a notable improvement over the 4 year collection period of this study (2007–2010), with HIGH quality ejaculates first collected at 11 and 12 years of age (Fig. 3a). The majority of ejaculates collected from this bull during 2011–2012 for other research (19/23, 83%) have been of HIGH quality (data not shown). The male was in contact with cyclic females during the study (but was unable to successfully copulate owing to his height) and successfully achieved a conception via AI in September 2011.


Figure 3. Profiles of in vitro sperm parameters in an Asian (a, Male 1) and an African (b, Male 10) elephant to demonstrate the relationship among in vitro sperm characteristics. Note: additional ejaculates displaying poor initial motility (TM ≤ 35%) were collected from both males during Study 1 (= 21, Male 1; = 22, Male 2); these ejaculates did not undergo chilled storage or further analysis and are therefore not shown in the above graphs.

Download figure to PowerPoint

For the African elephant bull, LOW, INTERMEDIATE and HIGH quality ejaculates were documented from 2009 (at 27 years age) to 2012 (Fig. 3b). This male had been producing semen of varying quality for ~11 years prior to commencement of the current study. Ejaculates collected from this bull following the completion of the current study in February 2012 have been of LOW (16/18, 88.9%) or of INTERMEDIATE quality (2/18, 11.1%), despite having visual, auditory and olfactory contact with females in oestrus during that time (March–August, 2012).

Study 2: Effect of prolonged chilled storage on in vitro sperm characteristics of HIGH quality ejaculates from one Asian and one African elephant

Total volume of combined viable fractions for HIGH quality ejaculates from an Asian (Male 1) and an African (Male 10) elephant was 14.9 ± 7.2 mL (range: 3.0–23.0 mL), raw semen concentration was 99.0 ± 131.9 × 108 spermatozoa/mL (range: 154.0–300.0 × 10spermatozoa/mL) and the total number of spermatozoa per collection was 107.6 ± 171.1 × 108 (range: 3.1–450.0 × 108 spermatozoa; data pooled across both males).

Membrane integrity (Fig. 4h) and all motility (Fig. 4a–c) and velocity parameters (Fig. 4d–f) decreased (< 0.05) over the 48 h chilled storage period for both males with the exception of STR and LIN for Male 1 which remained unchanged (Fig. 4g). Total motility was adequately maintained during chilled storage for Male 1 (57.6 ± 4.8% of 0 h TM was maintained at 48 h) and Male 10 (41.0 ± 6.7% of 0 h TM was maintained at 48 h). Rapid velocity was similarly well-maintained between 24 and 48 h of chilled storage (Male 1: 67.0 ± 5.2%; Male 10: 57.8 ± 6.2%). In contrast, only 15.6 ± 6.7% and 15.1 ± 4.0% of 0 h PM was maintained at 48 h for Male 1 and Male 10, respectively. Velocity parameters (VAP, VSL, VCL) were well-maintained between 24 and 48 h of chilled storage for both males (Male 1: 73.3 ± 4.3%); Male 10: 57.1 ± 10.1%; data pooled across VAP, VSL and VCL).


Figure 4. In vitro characteristics of spermatozoa from an Asian (Male 1, ‘Ema’) and an African (Male 10, ‘Laf’) elephant across a 48 h chilled storage period; (a) Total motility, (b) progressive motility, (c) rapid velocity, (d) average path velocity, (e) straight-line velocity, (f) curvilinear velocity, (g) straightness and linearity and (h) plasma membrane and acrosome integrity. a,b,cValues without a common letter within each male are significantly different across time points (p < 0.05). Note: motility evaluations were subjective (0 h) and objective (computer-assisted sperm analysis, 24–48 h).

Download figure to PowerPoint

Motile sperm agglutination rating was low and did not change (> 0.05) during storage (Male 1: 0.8 ± 0.3, Male 10: 1.0 ± 0.1 data pooled across evaluation times). The proportion of spermatozoa with an intact plasma membrane and acrosome declined (> 0.05) over time for Male 1 and Male 10 (Fig. 4), whereas membrane integrity as assessed by eosin-nigrosin remained unchanged (Male 1: 72.7 ± 5.1%, Male 10: 73.8 ± 5.3%, data pooled across evaluation times). Normal morphology and the incidence of detached heads also remained unchanged for Male 1 (79.7 ± 6.0% and 3.6 ± 0.9%, respectively, data pooled across evaluation times) and Male 10 (74.2 ± 7.8% and 10.9 ± 5.0%, respectively, data pooled across evaluation times). Similarly, all SCSA variables did not change over the 24–48 h storage period for Male 1 (%COMPαt: 15.0 ± 8.6%, Meanαt: 222.8 ± 19.1, SDαt: 49.1 ± 13.7) or Male 10 (%COMPαt: 19.2 ± 5.5%, Meanαt: 224.3 ± 15.6, SDαt: 60.9 ± 32.3, data pooled across evaluation times).


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Analyses conducted in this study of semen characteristics, sperm motility, morphology and membrane integrity underscore the well reported large intra- and inter-individual variation in elephant ejaculate quality collected by transrectal massage (Saragusty et al., 2009; Kiso et al., 2011; O'Brien et al., 2011). Results herein revealed that the majority of elephant ejaculates displayed a high degree of sperm DNA damage by 24 h of chilled storage, and an accompanying elevation in the proportion of spermatozoa with detached heads. Conversely, ejaculates designated as high quality (≥50% TM at 24 h of chilled storage), which accounted for 8.3% (13/157) of collections, displayed a low incidence of sperm DNA fragmentation and detached heads, and appear to exhibit high fertility potential based on in vitro characteristics of ejaculates from one Asian and one African bull, including DNA integrity, during chilled storage over 48 h. Urine contamination as measured by semen creatinine concentration was confirmed as an important factor in reduced semen quality in the elephant, but considerable sperm DNA damage and gross morphological degeneration was also observed in non-contaminated ejaculates. Our collective findings on sperm quality, including DNA integrity and morphology, provide evidence for the presence of aged spermatozoa in elephant ejaculates collected by transrectal massage, even when collections are performed on a regular basis, and indicate that this is the primary challenge associated with the collection of high quality ejaculates for use in AI and other assisted reproductive technology.

Ejaculates collected in the current research were divided into HIGH, INTERMEDIATE and LOW quality categories based on motility parameters after collection and following 24 h of chilled storage. Striking differences among characteristics of the three ejaculate quality types examined herein were observed for the proportions of spermatozoa exhibiting DNA fragmentation and the detached head form after 24 h of chilled storage. For ejaculates of HIGH quality, a low incidence of fragmented DNA and detached heads was observed for both species (Asian: 7 and 9%; African: 18 and 13%, respectively). In contrast, LOW and INTERMEDIATE ejaculate groups for both species displayed a higher frequency of spermatozoa with fragmented DNA (68 and 55%, respectively) and detached heads (50 and 33%, respectively). Degree of sperm DNA fragmentation in LOW and INTERMEDIATE ejaculates exceeded the proportion of spermatozoa with detached heads (by 18 and 22% respectively) at 24 h post-collection, implying that even intact spermatozoa in those ejaculates contained fragmented DNA.

The susceptibility of sperm DNA to undergo denaturation during the SCSA has been demonstrated to be inversely correlated to fertility in several species (reviewed by Evenson & Wixon, 2006). Although the aetiology of DNA strand breaks remains unclear, oxidative stress and abnormal/abortive apoptosis are hypothesized to be primary causes (reviewed by Evenson, 1999; Aitken & Koppers, 2011). With sizable pregnancy data, it is possible to determine a threshold of the SCSA's DNA fragmentation index (DFI) above which a detrimental impact on fertility is observed (Evenson & Wixon, 2006). A DFI threshold value of 30% is related to a poor pregnancy outcome in humans (Evenson et al., 2002; Evenson & Wixon, 2006). The DFI threshold value for reduced fertility varies among domesticated species, with estimations published for horses (~28%, Kenney et al., 1995), cattle: (~10–20%, Rybar et al., 2004) and pigs (~6%, Didion et al., 2009). The degree of fertility loss as a consequence of a DFI being above the aforementioned values is also species-specific, and is likely influenced in part by the degree of selection for sperm quality/fertility that species has undergone (Love & Kenney, 1998; Didion et al., 2009). Recently, application of the SCSA was performed with three ejaculates from one Asian elephant (O'Brien et al., 2012). Those preliminary results demonstrated that high quality semen (86.3 ± 2.5% initial TM, 2.7 ± 0.6% detached heads) displayed a low incidence of spermatozoa with denatured DNA (1.2 ± 0.4%) during 24 h chilled storage before cryopreservation, and up to 6 h after thawing and room temperature storage. Semen analysis results of this study which similarly incorporated multiple in vitro characteristics, but included a large number of males and differing ejaculate quality types, highlight the potential utility of DNA integrity analysis in helping to discriminate elephant ejaculates with poor and high fertility potential. In view of the strong evidence in humans for an increased rate in spontaneous abortions for assisted fertility patients whose sperm DNA integrity exceeds the DFI threshold (Zini et al., 2008), determination of the DFI threshold relating to subfertility in the elephant, and threshold values for other elephant sperm quality parameters such as morphology, would be warranted to maximize the efficiency of AI. This will require large scale, cooperative studies whereby stringent quality controls are placed on in vitro sperm assessment and insemination methods.

Measurement of sperm DNA fragmentation has also been recently reported in the elephant for poor quality ejaculates using the Sperm Chromatin Dispersion (SCD) test (Imrat et al., 2012a). In that study, five ejaculates (20.0 ± 15.4% initial TM, 16.6 ± 18.8% detached heads) displayed a high proportion of spermatozoa with fragmented DNA immediately after collection (32.2 ± 9.8%; = 5 ejaculates, Imrat et al., 2012a). In a study by the same group on ejaculates collected a month later, presumably from the same four bulls, ejaculate quality based on criteria used herein was low (32.8 ± 8.4% initial TM; incidence of detached heads was not reported), but sperm DNA quality was high at both 0 and 24 h of chilled storage, with less than 15% of spermatozoa displaying fragmented DNA (0 h: 5.1 ± 0.9%, 24 h: 11.6 ± 1.9%; Imrat et al., 2012b). Intrinsic methodological differences of the SCSA and SCD tests are a possible source of the contrasting findings between those of the current study and Imrat et al. (2012b), and direct comparisons of both DNA integrity analysis methods are warranted in forthcoming trials.

Socially driven reproductive suppression has been hypothesized to contribute to poor semen quality in the elephant (Hildebrandt et al., 2000). Those authors examined semen quality and reproductive tract morphology in 65 Asian and African elephant bulls. Similar to the findings of the current study, semen quality from intact adult males was highly variable (0–90% motility; Hildebrandt et al., 2000). Notably, the majority of adult bulls (38/56, 68%) produced spermatozoa with little or no motility, and such bulls did not display any ultrasonographically detectable reproductive pathologies. All these bulls were classified as non-breeders owing to their social status (subordinate to other animals or keepers) and reproductive history (failure to successfully breed during natural breeding attempts, or no access to cyclic females) (Hildebrandt et al., 2000). In contrast, 14 dominant bulls in successful breeding situations produced consistently high quality semen (Hildebrandt et al., 2000). The presence of a dominant con-specific or the absence of cyclic females in the same facility can profoundly depress a stallion's reproductive function (McDonnell, 2000). Indeed, overall testosterone concentrations of elephants at three multi-male facilities were higher for the older, dominant male compared with their subordinate counterparts (Brown et al., 2007). In view of the considerable intra-individual fluctuations of serum testosterone that occur over time, regardless of musth or social status (e.g. Brown et al., 2007), simultaneous, longitudinal monitoring of semen quality, reproductive hormone concentrations and reproductive tract morphology (Hildebrandt et al., 2000), ideally in conjunction with in vivo fertility, is required to help characterize mechanisms underlying socially mediated effects on elephant reproduction including semen quality. Multiple causal factors are presumed to exist, as dominant bulls housed in visual and auditory contact with cyclic females in this study (= 6; Males 2, 4, 7, 8, 9, 10) continued to produce poor quality ejaculates (LOW and INTERMEDIATE) with a high incidence of fragmented DNA and detached heads by 24 h of chilled storage.

Of the semen characteristics examined after chilled transport to the reproductive laboratory, only Cr concentration differed significantly among the three ejaculate quality groups with HIGH ejaculates displaying lower values than LOW. There was a tendency for the osmolality of LOW ejaculates to be higher than that of HIGH, which is in line with the positive correlation observed between Cr and osmolality. However, combined results indicate that Cr concentration is a sensitive indicator of urine contamination, whereas, osmolality (as reported herein), visual assessments of colour, microscopic presence of urine salts or odour of elephant ejaculates (Schmitt & Hildebrandt, 1998; Hildebrandt et al., 2000), are suitable only for the detection of a high degree of urine contamination.

The observed positive correlation of elevated Cr concentration with detached heads and all three SCSA variables necessitates consideration of the impact of exposure to urine during and after the collection process on sperm quality. To our knowledge there exist no controlled studies on the effect of urine on elephant sperm quality, nor on the effect of urine on sperm head detachment for any species. A prior investigation in an Asian elephant showed that elevated Cr concentration was associated with reduced motility parameters and increased agglutination (O'Brien et al., 2012). Addition of 5% urine of average concentration to stallion semen resulted in an immediate decrease in progressive motility (from 64% to 16%, Griggers et al., 2001). Based on an average Cr concentration of elephant urine of 930 μg/mL (range: 323–2380 μg/mL; Miller, 2006), urine contamination of LOW and INTERMEDIATE ejaculates represented 3.7 and 2.3% of the total ejaculate volume, respectively, and this was further reduced after dilution (at least 1 : 2, v/v) of semen with extender after collection. While results indicate that urine is detrimental to elephant sperm motility and plasma membrane integrity, it seems unlikely that the aforementioned degree of urine contamination (<4%) observed for LOW and INTERMEDIATE quality samples in this study would induce the high values of detached heads and fragmented DNA observed at 24 h post-collection. In cases where the elephant bull is located at a distance from the site of AI or semen banking, inclusion of the relatively rapid (~40 min) creatinine assay into initial on-site evaluations of semen quality would be useful to exclude ejaculates with creatinine concentration exceeding the threshold described in this study (>14.0 μg/mL for raw semen), as such samples would likely have poor fertility potential after a typical interval of chilled transport (~12–24 h).

The hypothesized existence of an aged sperm population in elephant ejaculates based on the aforementioned DNA integrity and morphological data is supported by results of H33342 staining and ensuing X–Y resolution analysis. Under the staining conditions described herein, H33342 binds specifically and quantitatively to double-stranded (intact) DNA (reviewed by Garner, 2009), and has previously enabled the determination of the difference in DNA content between X and Y spermatozoa in the elephant during the development of sex-sorting technology (O'Brien et al., 2009). The finding of reduced resolution of X and Y sperm populations for LOW and INTERMEDIATE samples compared with HIGH, particularly for ejaculates from the same male (Fig. 2), is likely indicative of non-uniform binding of H33342 to double-stranded DNA because of high DNA fragmentation in the former two ejaculate types.

Normal ejaculates exhibit a low incidence of the detached head form (<15%; cattle: Barth & Oko, 1989; horse: Love, 2011), but elevated proportions of this abnormal form have been attributed to a number of causes including the accumulation of senescent spermatozoa owing to extragonadal duct system irregularities (Barth & Oko, 1989; Love et al., 1992; Pozor et al., 2011). In the stallion, partial or complete obstruction of the ampullae resulting in the accumulation of spermatozoa displaying poor motility and a high incidence of detached heads has been resolved with transrectal massage of the ampullae and treatment with oxytocin or a prostaglandin analogue to enhance smooth muscle contraction followed by intensive semen collection (e.g. 11 ejaculates collected over 7 days using a phantom mount and an AV; Love et al., 1992; Blanchard et al., 2012). While transrectal ultrasound imaging of the ampullae is recommended prior to and after semen collection in the elephant to determine if emptying of this organ occurred during the massage/semen collection process (Hildebrandt et al., 2000), it remains possible that incomplete emptying of the ampullae and the distal epididymis, combined with perturbations in the passive loss of unejaculated spermatozoa, may contribute to the formation of an aged sperm population as described herein. Administration of oxytocin prior to transrectal massage would be of interest to include in future semen collection trials in the elephant.

In cattle, transrectal massage has been used to deliver good quality semen (e.g. Palmer et al., 2005). Interestingly, one study found that the proportion of detached heads was higher for transrectal massage samples than that of samples collected directly from the epididymis, although the overall incidence of this defect was low (5.3 and 2.3%, respectively; Persson et al., 2006). A subsequent study in yearling bulls also showed that semen characteristics were similar for transrectal massage and AV collection methods (Persson et al., 2007). The considerable anatomical differences between cattle and elephant bulls warrant a comparison of semen characteristics after transrectal massage and AV collection in the elephant. Semen collection using a phantom mount/AV set-up with additional stimulatory agents (e.g. urine or faeces from an oestral female) may provide a more normal ejaculatory response compared with transrectal massage, and this basic approach is under assessment by other workers (J. Andrews, personal communication). The collection of a LOW quality ejaculate from an elephant in this study, followed by a HIGH quality ejaculate 4 weeks later after the male had mated with a female over the three preceding days (Fig. 2), supports evaluation of this approach.

Notable findings of this study included the significant correlation between the incidence of sperm DNA damage and detached heads, which to our knowledge has not been reported previously for any species, and the poor DNA integrity and morphology after 24 h of chilled storage of non-contaminated ejaculates displaying values of initial motility (≥50% total motility) considered suitable for use in AI. Collective results on in vitro sperm characteristics and flow cytometric DNA parameters indicate that the predominant ejaculate types collected using the transrectal massage method (LOW and INTERMEDIATE) contain an aged sperm population with a high degree of DNA damage and probable poor fertility potential. Further research is required to determine if irregularities in the ejaculatory response, and a prolonged interval of sperm storage in the male genital tract prior to ejaculation, possibly under the influence of socially and hormonally mediated effects, contribute to inconsistent semen quality in the elephant.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This research was supported by SeaWorld Parks and Entertainment Inc. (SEA). We thank the many curatorial, keeper and veterinary staff at the following institutions for their collaborative efforts which made this research possible: Albuquerque Biological Park, Cincinnati Zoo and Botanical Garden, Dickerson Park Zoo, Disney's Animal Kingdom, Fort Worth Zoo, Indianapolis Zoo, International Elephant Foundation, Oregon Zoo, Taronga Conservation Society Australia, Tulsa Zoological Park. The SeaWorld and Busch Gardens Reproductive Research Center staffs Michelle Buescher and Angela Ho, and Texas A&M's Sheila Teague, are thanked for technical assistance, as is Dr Terri Roth for useful comments on the manuscript. Brad Andrews (SEA) is also thanked for institutional support of this research. This project was funded by SeaWorld Corporation and is a SeaWorld Technical Contribution no. 2012–0X-C.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Aitken RJ & Koppers AJ. (2011) Apoptosis and DNA damage in human spermatozoa. Asian J Androl 2011, 3642.
  • Barth AD & Oko RJ. (eds) (1989) Photomicrographic features of bovine sperm cell abnormalities. In: Abnormal Morphology of Bovine Spermatozoa, pp. 89129. Iowa State University Press, IA.
  • Blanchard TL, Varner DD, Brinsko SP & Love CC. (2012) Azoospermia in stallions: determining the cause. Compend Contin Educ Vet 34, E1E8.
  • Brown JL, Somerville M, Riddle HS, Keele M, Duer CK & Freeman EW. (2007) Comparative endocrinology of testicular, adrenal and thyroid function in captive Asian and African elephant bulls. Gen Comp Endocrinol 151, 153162.
  • Cevik M, Tuncer PB, Tasdemir U & Ozgurtas T. (2007) Comparison of spermatological characteristics and biochemical seminal plasma parameters of normozoospermic and oligoasthenozoospermic bulls of two breeds. Turk J Vet Anim Sci 31, 381387.
  • Didion BA, Kasperson KM, Wixon RL & Evenson DP. (2009) Boar fertility and sperm chromatin structure status: a retrospective report. J Androl 30, 655660.
  • Evenson DP. (1999) Loss of livestock breeding efficiency due to uncompensable sperm nuclear defects. Reprod Fertil Dev 11, 115.
  • Evenson DP & Wixon R. (2006) Clinical aspects of sperm DNA fragmentation detection and male infertility. Theriogenology 65, 979991.
  • Evenson DP, Larson KL & Jost LK. (2002) Sperm chromatin structure assay: its clinical use for detecting sperm DNA fragmentation in male infertility and comparisons with other techniques. J Androl 23, 2543.
  • Garner DL. (2009) Hoechst 33342: the dye that enabled differentiation of living X-and Y-chromosome bearing mammalian sperm. Theriogenology 71, 1121.
  • Graham EF, Crabo BG & Brown KI. (1972) Effect of some zwitterion buffers on the freezing and storage of spermatozoa I. Bull J Dairy Sci 55, 372378.
  • Griggers S, Paccamonti DL, Thompson RA & Eilts BE. (2001) The effects of pH, osmolarity and urine contamination on equine spermatozoal motility. Theriogenology 56, 613622.
  • Heath E, Jeyendran RS & Graham EF. (1983) Ultrastructure of spermatozoa of the Asiatic Elephant (Elephas maximus). Zbl Vet C Anat Histol Embryol 1983, 245252.
  • Hildebrandt TB, Göritz F, Hermes R, Schmitt DL, Brown JL, Schwammer H et al. (1999) Artificial insemination of African (Loxodonta africana) and Asian (Elephas maximus) elephants. Proceedings of the American Association of Zoo Veterinarians, Columbus, Ohio. pp. 8386.
  • Hildebrandt TB, Hermes R, Pratt NC, Fritsch G, Blottner S, Schmitt DL et al. (2000) Ultrasonography of the urogenital tract in elephants (Loxodonta africana and Elephas maximus): an important tool for assessing male reproductive function. Zoo Biol 19, 333345.
  • Howard JG, Bush M, de Vos V & Wildt DE. (1984) Electroejaculation, semen characteristics and serum testosterone concentrations of free-ranging African elephants (Loxodonta Africana). J Reprod Fertil 72, 187195.
  • Imrat P, Hernandez M, Rittem S, Thongtip N, Mahasawangkul S, Gosálvez J et al. (2012a) The dynamics of sperm DNA stability in Asian elephant (Elephas maximus) spermatozoa before and after cryopreservation. Theriogenology 77, 9981007.
  • Imrat P, Mahasawangkul S, Gosálvez J, Suthanmapinanth P, Sombutputorn P, Jansittiwate S et al. (2012b) Effect of cooled storage on quality and DNA integrity of Asian elephant (Elephas maximus) spermatozoa. Reprod Fertil Dev 24, 11051116. doi: 10.1071/RD11309
  • Jones RC. (1973) Collection, motility and storage of spermatozoa from the African elephant Loxodonta africana. Nature 243, 3839.
  • Keele M. (2010) Asian Elephant (Elephas Maximus) North American Studbook. Association of Zoos and Aquariums/Oregon Zoo, Silver Spring, MD/Portland, OR.
  • Kenney RM, Evenson DP, Garcia MC & Love CC. (1995) Relationship between sperm chromatin structure, motility, and morphology of ejaculated sperm and seasonal pregnancy rate. Biol. Reprod. Monogr. Ser. 1. Equine Reprod VI. (eds DC Sharp, IW Bazer), pp. 647653. Society for the Study of Reproduction, Madison WI.
  • Kiso WK, Brown JL, Siewerdt F, Schmitt DL, Olson D, Crichton EG et al. (2011) Liquid semen storage in elephants (Elephas maximus and Loxodonta africana): species differences and storage optimization. J Androl 32, 420431.
  • Landowski VJ & Gill J. (1964) Einige Beobachtungen uber das Sperma des Indischen Elefanten (Elephas maximus L.) Zoologische Garten Lpz 29, 205212.
  • Love CC. (2011) Relationship between sperm motility, morphology and the fertility of stallions. Theriogenology 76, 547557.
  • Love CC & Kenney RM. (1998) The relationship of increased susceptibility of sperm DNA to denaturation and fertility in the stallion. Theriogenology 57, 955972.
  • Love CC, Riera FL, Oristaglio-Turner RM & Kenney RM. (1992) Sperm occluded (plugged) ampullae in the stallion. Proc Ann Meet Soc Theriogenology, 117127.
  • Mar UK, Thein M, Khaing AT, Tun W & Nyunt T. (1992) Electroejaculation and Semen Characteristics in Myanmar Timber Elephants. Forestry Science Research Paper, Union of Myanmar, Ministry of Forestry, Rangoon.
  • McDonnell SM. (2000) Reproductive behavior of stallions and mares: comparison of free-running and domestic in-hand breeding. Anim Reprod Sci 60–61, 211219.
  • Miller RE. (2006) Urinary System In: Biology, Medicine and Surgery of Elephants. (eds ME Fowler & SK Mikota), pp. 389392. Blackwell Publishing, Ames, IA.
  • O'Brien JK, Roth TL, Stoops MA, Ball RL, Steinman KJ, Buescher MY et al. (2011) Sperm sorting and preservation technologies for sex ratio modification in the elephant and rhinoceros: an update. In: Proc. 2011 Int Elephant and Rhinoceros Conservation and Research Symposium. (ed D Olson), pp. 3940. International Elephant Foundation, Rotterdam, The Netherlands.
  • O'Brien JK, Steinman KJ & Robeck TR. (2009) Application of sperm sorting and associated reproductive technology for wildlife management and conservation. Theriogenology 71, 98107.
  • O'Brien JK, Steinman KJ, Montano GA, Love CC, Saiers RL & Robeck TR. (2012) Characteristics of high quality Asian elephant (Elephas maximus) ejaculates and in vitro sperm quality after prolonged chilled storage and directional freezing. Reprod Fertil Dev. doi: 10.1071/RD12129.
  • Olson D. (2011) North American Region Studbook for the African Elephant (Loxodonta Africana). Association of Zoos and Aquariums, Silver Spring, MD.
  • Palmer CW, Brito LFC, Arteaga AA, Söderquist L, Persson Y & Barth AD. (2005) Comparison of electroejaculation and transrectal massage for semen collection in range and yearling feedlot beef bulls. Anim Reprod Sci 87, 2531.
  • Persson Y, McGowan M & Söderquist L. (2006) Comparison between the sperm morphology in semen samples obtained from yearling beef bulls by transrectal massage of the ampullae and cauda epididymal dissection. Reprod Dom Anim 41, 233237.
  • Persson Y, Strid GM, Håård M & Söderquist L. (2007) Comparison of semen samples collected from beef bulls by transrectal massage or artificial vagina. Vet Rec 161, 662663.
  • Portas TJ, Bryant BR, Göritz F, Hermes R, Keeley T, Evans G et al. (2007) Semen collection in an Asian elephant (Elephas maximus) under combined physical and chemical restraint. Aust Vet J 85, 425427.
  • Pozor MA, Macpherson ML, Troedsson MH, Klein C, Diaw M, Buergelt C et al. (2011) Midline cysts of colliculus seminalis causing ejaculatory problems in stallions. J Equine Vet Sci 31, 722731.
  • Price P, Bradford J & Schmitt D. (1986) Collection and semen analysis in Asian elephants. Proc Am Assoc Zool Parks Aquar, 310313.
  • Robeck TR, Gearhart SA, Steinman KJ, Katsumata E, Loureiro JD & O'Brien JK. (2011) In vitro sperm characterization and development of a sperm cryopreservation method using directional solidification in the killer whale (Orcinus orca). Theriogenology 76, 267279.
  • Rybar R, Faldikova L, Faldyna M, Machatkova M & Rubes J. (2004) Bull and boar sperm DNA integrity evaluated by sperm chromatin structure assay in the Czech Republic. Vet Med 49, 18.
  • Saragusty J, Hildebrandt TB, Behr B, Knieriem A, Kruse J & Hermes R. (2009) Successful cryopreservation of Asian elephant (Elephas maximus) spermatozoa. Anim Reprod Sci 115, 255266.
  • Schmitt DL & Hildebrandt TB. (1998) Manual collection and characterization of semen from Asian elephants (Elephas maximus). Anim Reprod. Sci 53, 309314.
  • Sharpe JC & Evans KM. (2009) Advances in flow cytometry for sperm sexing. Theriogenology 71, 410.
  • Taussky HH. (1954) A microcolorimetric determination of creatine in urine by the Jaffe reaction. J Biol Chem 208, 853861.
  • Turner RM, Love CC, McDonnell SM, Sweeney RW, Twitchell ED, Habecker PL et al. (1995) Use of imipramine hydrochloride for treatment of urospermia in a stallion with a dysfunctional bladder. J Am Vet Med Assoc 207, 16021606.
  • Zini A, Boman JM, Belzile E & Ciamp A. (2008) Sperm DNA damage is associated with an increased risk of pregnancy loss after IVF and ICSI: systematic review and meta-analysis. Human Reprod 23, 26632668.