Bernd Hoffmann, DVM, Prof. Dr. med. vet.; Schuler G, DVM; Hdoz, Dr. med. vet.
Klinik für Geburtshilfe, Gynäkologie und Andrologie der Groß-und Kleintiere mit Tierärztlicher Ambulanz
Introduction and quality criteria of native semen
The first recorded and successful artificial insemination (AI) was performed in the dog in 1785 by Spallanzani, the first successful use of frozen canine semen was reported by Seager in 1969. However, only in the last two decades AI in the dog became a rather widespread and generally accepted breeding technique. It may be applied for medical reasons, to prevent infections, to overcome travel and quarantine restrictions and to avoid long, expensive and stressful travels with the animals. In conjunction with cryopreservation AI is especially indicated to build up stocks from excellent studs and for conservation of semen from rare breeds. Only semen meeting the quality criteria outlined in table 1 should be used for conservation. The same quality criteria apply when studs are examined for breeding soundness.
The most common method for semen collection in dogs is by manual dissemination, needing no further elaborated equipment. It is generally no problem in young and possibly hypersexual and in trained dogs; it may be problematic in experienced males used for natural breeding. In these cases the presence of a teaser bitch at the optimal stage of oestrus may be essential and efficient stimulation may only be obtained when the stud is allowed to mount. To our experience the preparation of bitches not in heat with methyl, a component of vaginal secretions of female dogs in oestrus, has no satisfying effect on male dogs not responding to manual stimulation. Semen collection should be performed in a quiet and non-distracting environment. Typically, canine ejaculates are collected as three fractions: a small pre-sperm fraction, the sperm rich fraction and a more voluminous post-sperm fraction mainly provided by the prostate. A lack of the milky sperm rich fraction may be due to hypo-/azoospermia, obstruction of seminiferous ducts, retrograde ejaculation or an irregular course of the ejaculation reflex due to insufficient stimulation. Retrograde ejaculation may be confirmed by the detection of numerous spermatozoa in urine sediments after mild centrifugation. First results suggest that in dogs alkaline phosphatase activities in the sperm rich fraction can be used as a marker to monitor depletion of epididymal sperm stores during semen collection, as this enzyme is highly expressed in the epididymal tail (Gobello et al., 2000). In the dog semen quality in general is not improved in a short termed second ejaculate; however, 70% more spermatozoa may be obtained compared to a single ejaculate (England, 1999). Alternatively, canine semen can be collected applying electroejaculation under general anesthesia, and pregnancies have been obtained after the use of cryopreserved epididymal sperm, even after postmortem extraction (Marks et al., 1994, Hori et al., 2004).
Semen preservation and dose for insemination
Except for transfer of native semen only the sperm rich fraction should be used for semen preservation.
Native semen may be transferred due to medical or behavioural reasons without any further handling. The ejaculate may be split, however, transfer should be performed as soon as possible after semen collection and evaluation, although fertilization rates of 70% have been observed with unextended canine semen after two days of storage at 4°C (Tsutsui, 2002).
Table 1. quality criteria for dog semen (Hoffmann, 2003)
> 75 %
< 20 %
total number of spermatozoa
> 0.3-1.0 x 106 (1)
1. depending on breed; e.g., Dachshund > 0.3; Dobermann > 1.0
2. Sperm rich fraction
The use of chilled semen requests the availability of the donor on a short call and semen transport within a short term. A number of semen extenders has been described; generally accepted is a tris-buffered solution containing 20% egg yolk, 12.5% fructose and an admixture of an antibiotic. The cooling down to about 5°C is over a period of 2 hours. Depending on the concentration of spermatozoa, the ratio (volume) of ejaculate to extender may vary between 1:3 to 1:5. Mild centrifugation might become necessary when a selective collection of the sperm rich fraction was not possible. The total dose is stored in an appropriate sealed vial. The whole sperm rich fraction may be extended and used as one dose. If the ejaculate has to be split, one dose should at least contain 100-200x106 spermatozoa (Linde-Forsberg, 1991). The fertilization capacity of chilled semen decreases progressively, and it may be difficult to estimate the time after which the use of a certain semen portion may become unpromising, as after 5 days of storage motility may be still high, when fertilization capacity has considerably decreased (Tsutsui et al., 2002).
Whereas pregnancies may be obtained with fresh semen of inferior quality, albeit with smaller litter sizes, insemination of frozen semen of poor quality is unpromising. Thus, raw material of adequate quality and appropriate preservation procedures are indispensable for acceptable results. There are various protocols for cryopreservation of canine semen. Basically the same extender as used for the preparation of chilled semen may be used. The concentration of the cryoprotectant glycerin is at 6%. In our clinic semen meeting the criteria given above is diluted and packed in 0.5 ml straws. Following cooling to 5°C over 2 hours they are left for 10 min at-140°C (vapor of liquid nitrogen). Final storage is at-196°C. It is an absolute requirement to test the quality of each individual batch by thawing (37°C; 20 seconds) and reexamining a frozen aliquot, as the suitability for freezing may vary considerably between different ejaculates of an individual stud. Loss in forward motility should not exceed 20%, pathomorphology may be increased up to 30% (Hoffmann 2003). For insemination, a number of straws corresponding to approximately 1.5x108 progressively motile spermatozoa is thawed and pooled.
As with natural mating, precise determination of the optimal time of insemination is less critical using native or chilled semen compared to cryopreserved semen due to the longevity of fresh canine spermatozoa in the female genital tract which is between 5-7 days, yielding a broad biological leeway for pregnancy even under suboptimal insemination management (Jeffcoate, Lindsay, 1989). Thus, using native or freshly provided chilled semen, a single insemination around the time of ovulation may be sufficient, although significantly higher pregnancy rates have been found with two inseminations (Linde-Forsberg et al., 1993). In contrast, the limited life-span of frozen-thawed spermatozoa in the female genital tract, which is estimated to be shorter than 12-24 hours (Concannon et al., 1989), requires a precise definition of the time of insemination in the individual bitch, which is about 2-5 days after ovulation due to a comparatively long postovulatory process of oocyte maturation. Repeating the AI after 24-48 hours may result in significantly higher pregnancy rates and litter sizes (Tsumagari et al., 2003) probably due to the occurrence of an extended ovulation and maturation period. Indications for the appropriate time of insemination may be obtained by observation of behavioural changes, vaginal cytology, vaginoscopy and determination of the electrical impedance. Repetitive examinations are necessary and the variability inherent to these approaches and the underlying parameters is high. It is therefore recommended to directly determine the underlying hormonal changes by applying a reliable quantitative assay for progesterone. Progesterone levels closely reflect the pre and postovulatory luteinization of granulosa cells with about 1-2 ng/ml during the LH-peak and about 5 ng/ml at ovulation. Termination of oocytes maturation is more difficult to assess due to the variability in the increase of progesterone levels with formation of the corpora lutea. However, progesterone levels between 10-12 and even up to 20 ng/ml are generally considered as optimal for insemination with cryopreserved semen (Linde-Forsberg et al., 1989).
Technique of semen transfer
Due to the anatomical structure of the vagina intrauterine insemination in the dog is more difficult than in other species and requires quite some experience. This is largely due to a dorsal median fold of tissue that extends caudally from the vaginal portion of the cervix. At vaginoscopy the caudal portion of the fold and the constrictions of the lateral and ventral vaginal walls give a distinct appearance, referred to as pseudocervix. The true vaginal portion of the cervix is cranial to the pseudocervix. This makes intrauterine cannulation difficult, also because the cervical canal is nearly perpendicular to the longitudinal axis of the vagina and the uterine body.
Currently four methods are used for semen transfer in dogs: deep intravaginal insemination, transcervical insemination using an rigid catheter (Norwegian Catheter) with transabdominal fixation of the cervix, transcervical endoscope-aided insemination and intrauterine insemination using a laparoscopic approach. Freshly ejaculated spermatozoa and spermatozoa from chilled semen may readily penetrate the cervix and reach the site of fertilization in the oviduct. Hence deep intravaginal insemination with this type of semen is common practice. As a substitute of the vaginal plug by the erected bulbus glandis it has been generally recommended to elevate the rear end of the bitch at a 45° to 60° angle for 5-20 minutes after AI. However, pregnancy rates and litter size was not different comparing elevation times of 1 and 10 min (Pinto et al., 1998). In respect to cryopreserved semen it was widely accepted that semen deposition must be intrauterine for optimal results. However, data from literature are contradictory as in some studies no significant difference was found between precervical and transcervical insemination using frozen thawed semen (Rota et al., 1999). Yet up to now the statistical basis is very weak not allowing any definite conclusions (see below). Hence we still recommend intrauterine semen deposition using the transcervical approach since the laparoscopic approach is not accepted by many pet owners, which consider this method as unethical and too risky for the bitch. Yet it must be anticipated that the transcervical approach may be difficult in maiden dogs and in those cases with a high discharge of cervical mucus, blinding the optical system. In these cases the semen should be placed as close as possible to the orificium externum of the cervix.
Pregnancy rates and litter sizes
Success of AI in dogs depends on various parameters such as type and quality of semen used, sperm dosage, the method and site of semen deposition, insemination regime and breed.
Provided semen of adequate quality is used, success of AI using native or chilled semen is equal to that after natural mating (Pinto et al., 1999).
Several studies are available reporting pregnancy rates and litter sizes after AI with frozen-thawed semen in bitches using intravaginal and/or intrauterine insemination. However, results vary substantially obviously due to the high variability of the conditions for insemination. However, in a relatively large comprehensive study using semen of defined quality, pregnancy rates/liter size after a single insemination per cycle were 84.3%/5.0 ± 3.0 puppies after intrauterine semen deposition, but only 34.8% / 2.5 ± 1.3 puppies after intravaginal insemination. These data show that--provided a good method of cryopreservation and intrauterine semen deposition are applied--results with AI using frozen-thawed semen may be comparable to those obtained with natural mating. In those studies yielding no difference between intravaginal vs. intrauterine insemination, the benefit of intrauterine insemination may be overridden by higher doses of spermatozoa or by higher numbers of AI's per cycle. Accordingly, in the above mentioned study whelping rate/litter size after intravaginal deposition of frozen-thawed semen increased to 63.9%/5.1 ± 3.1 when three inseminations per cycle were applied (Linde-Forsberg et al., 1999).
By now AI in the dog must be considered a well established breeding technique with the use of frozen semen gaining constantly more importance. Techniques for semen preservation, the quality assessment of semen and the dosages used are fairly standardized. Appropriate semen quality and AI management provided, results with native, chilled and even frozen-thawed semen may be comparable to natural mating. It is predicted that the use of AI will continue to rise in dogs with more AI centers becoming available, increasing acceptance by breeders associations and the abolition of import restrictions.
1. Concannon PW, Battista M (1989): In: Current Veterinary Therapy X: Small Animal Praczice (Ed.: RW Kerk), W.B. Saunders, Philadelphia, p 1247
2. England GC (1999): Semen quality in dogs and the influence of a short-interval second ejaculation. Theriogenology, 52 981-986.
3. Gobello C, Castex G, Corrada Y (2002): Serum and seminal markers in the diagnosis of disorders of the genital tract of the dog: a mini-review. Theriogenology, 57, 1285-1291.
4. Hoffmann B (2003): Andrologie; Physiologie, Pathologie und Biotechnologie der männlichen Fortpflanzung. Lehmanns Media.
5. Hori T, Ichikawa M, Kawakami E, Tsutsui T (2004): Artificial insemination of frozen epididymal sperm in beagle dogs. J. Vet. Med. Sci., 66, 37-41.
6. Jeffcoate IA, Lindsay FE (1989): Ovulation detection and timing of insemination based on hormone concentrations, vaginal cytology and the endoscopic appearance of the vagina in domestic bitches. J. Reprod. Fertil. Suppl., 39, 277-287.
7. Linde-Forsberg C, Forsberg M (1989): Fertility in dogs in relation to semen quality and the time and site of insemination with fresh and frozen semen. J. Reprod. Fertil. Suppl., 39, 299-310.
8. Linde-Forsberg C, Forsberg M (1993): Results of 527 controlled artificial inseminations in dogs. J. Reprod. Fertil. Suppl., 47, 313-23.
9. Linde-Forsberg C, Strom Holst B, Govette G (1999): Comparison of fertility data from vaginal vs intrauterine insemination of frozen-thawed dog semen: a retrospective study. Theriogenology, 52, 11-23.
10. Linde-Forsberg C (1991): Achieving canine pregnancy by using frozen or chilled extended semen. Vet. Clin. North. Am. Small. Anim. Pract., 21, 467-485.
11. Marks SL, Dupuis J, Mickelsen WD, Memon MA, Platz CC Jr. (1994): Conception by use of postmortem epididymal semen extraction in a dog. J. Am. Vet. Med. Ass. 204, 1639-1640.
12. Pinto CR, Paccamonti DL, Eilts BE (1999): Fertility in bitches artificially inseminated with extended, chilled semen. Theriogenology, 52 609-616.
13. Rota A, Iguer-Ouada M, Verstegen J, Linde-Forsberg C (1999): Fertility after vaginal or uterine deposition of dog semen frozen in a tris extender with or without Equex STM paste. Theriogenology, 51, 1045-1058.
14. Seager SWJ (1969): Successful pregnancies utilizing frozen dog semen. AI Digest, 17, 6-7.
15. Tsumagari S, Ichikawa Y, Toriumi H, Ishihama K, Morita M, Kanamaki M, Takeishi M (2003): Optimal timing for canine artificial insemination with frozen semen and parentage testing by microsatellite markers in superfecundency. J. Vet. Med. Sci., 65, 1003-1005.
16. Tsutsui T, Tezuka T, Mikasa Y, Sugisawa H, Kirihara N, Hori T, Kawakami E (2002): Artificial insemination with canine semen stored at a low temperature. J. Vet. Med. Sci., 65, 307-312.