The population of the Andalusian donkey is estimated at 793 individuals (DAD-IS-FAO, 2014), therefore it is considered as endangered (BOE, 2009). Equine embryo transfer (ET) permits to obtain several foals from a mare each year (Stout, 2006), becoming a valuable strategy for preserving endangered equid species (Summers et al., 1987). However, studies concerning ET in endangered donkeys are scarce and discouraging results have been obtained (Camillo et al., 2010; Panzani et al., 2012). Recipient´s pregnancy rate depends on the quality of recipients and their synchronization with donors (Carnevale et al., 2000; Squires et al., 2003). This implies needing a significant number of recipients for each donor, which is difficult to conduct in endangered species. Interspecific ET and the use of non-cycling recipients could be a successful long-term strategy to increase the number of recipients in ET programs.

The usefulness of cycling mules as recipients of horse (Camillo et al., 2003; Allen, 2005) and donkey (Silva et al., 2013; González et al., 2015) embryos has been previously demonstrated. Besides, intraspecific (Kaercher et al., 2013; Botelho et al., 2015) and interspecific (donkey-in-mare) (Allen et al., 1993; Peña-Alfaro et al., 2014) pregnancies have been established by ET in non-cycling mares. However, non-cycling recipient mules treated with oestradiol benzoate and progesterone before transfer of horse embryos failed to become pregnant (Camillo et al., 2003). To our knowledge, no studies have been conducted to evaluate if non-cycling recipient mules are able to carry a pregnancy to term after transfer of donkey embryos.

The study was started in the Centro Rural Malpica (Palma del Río, Cordoba, Spain) in May 2015. All animal procedures were performed in accordance with the Spanish laws for animal welfare and experimentation (BOE, 2013).

Two fertile Andalusian jennies (7-years-old) served as embryo donors. Estrus was induced with one intramuscular injection of 7.5 mg luprostiol (Prosolvin ^®, Virbac, Barcelona, Spain) during diestrous. Ovarian activity was evaluated daily during oestrus by transrectal ultrasound (Aloka SSD 500, ALOKA Co. Ltd., Tokyo, Japan). When a follicle = 35 mm in diameter was detected, ovulation was induced using 1500 IU of human chorionic gonadotropin (hCG; Veterin Corion ^®, Divasa-Farmavic S.A., Barcelona, Spain) intramuscularly. Next day, donor jennies were mated with a fertile jack every other day until ovulation.

Two healthy mules (4-years-old and in good body condition) in anoestrus were used as embryo recipients. When the donor jenny was mated the first time, the mule recipients received two intramuscular injections of 5 mg oestradiol benzoate (Sincrodiol ^®, Ourofino Saûde Animal, Cravinhos, Brazil) at an interval of 24 h. After the donor jenny ovulated, recipients were given 1500 mg of long-acting progesterone (P4-LA/IM, Brazil) intramuscularly every 7 days until day 120 of pregnancy (Day 0 = day of ET).

Embryo collections were done on day 7 after ovulation (Camillo et al., 2010). Embryos were evaluated under stereosmicroscope, washed 10 times in Syngro ^® holding (Bioniche Animal Health, Washington, USA) and loaded into 0.25 ml straws until transferred transcervically to the mule recipients.

Pregnancy diagnosis was performed on days 5 and 25 after ET by ultrasound. Pregnancy was monitored weekly until day 120 after ET and thereafter monthly until parturition. Gender determination was also performed at day 120 (Bucca, 2005).

Only one embryo (morula stage grade 1 with a diameter of 175 µm) was recovered and transferred to one of the non-cycling recipient mules at day 5 of progesterone treatment. The ET resulted in an ongoing pregnancy. Ultrasonography images of the conceptus at different gestational days are shown in Figure 1. Foetal viability was confirmed by ultrasound each month of pregnancy.

Parturition occurred spontaneously on day 375 of gestation and a live donkey foal was delivered without the need for assistance. The surrogate mule mother expelled the foetal membranes after birth, and presented strong protective behaviour and large amount of milk in the well-developed udders. Gross examination of the foetal membranes did not reveal any lesions of clinical importance; however, the donkey foal had a congenital osteodystrophy with facial hyperostosis. Similar findings were described by Bryant et al. (2012) in a Thoroughbred filly which was born with congenital aneurysmal bone cyst. Although the pathogenesis of these lesions remains unclear, possible causes are malformation of intraosseous vasculature secondary to trauma, bleeding disorders, fibrous dysplasia, hematomas or underlying neoplasia.

This is the first report of a foal born following transfer of a fresh donkey embryo into a non-cycling mule. Some authors have previously used cycling mule as donkey embryos recipients and obtained pregnancies and/or live donkey foals (Silva et al., 2013; González et al., 2015). Likewise, successful non-surgical transfer of horse embryos into cycling mules has been reported (Camillo et al., 2003), but failed to obtain pregnancies when non-cycling mules were used. These unsatisfactory results have been explained by the lower quality of the endometrium in permanently non-cyclic recipients (Camillo et al., 2003) and the lower fertility rates that have this recipients (Carnevale et al., 2000). The hormonal treatment used in this study showed to be effective for preparing anovulatory mares in deep anoestrus and transitional phase as embryo recipients (Rocha Filho et al., 2004; Botelho et al., 2015). Based on our results, the administration of this treatment to non-cyclic mules would be efficient for establishing and maintaining pregnancy after ET.

The eutocic birth of a live donkey foal suggests a normal development of the placenta (González et al., 2015). Unfortunately, the foal died few hours after the birth due to right-side heart failure, possibly related to asphyxia. In the previous case reported in literature (Bryant et al., 2012) the filly was euthanized, therefore the relationship between the facial lesions and the spontaneous death of the donkey foal cannot be ensured. We could not confirm the excellent maternal ability of mules reported previously (Camillo et al., 2003; González et al., 2015), but a strong protective behaviour and well-developed udders were observed.

We conclude that non-cycling mules treated with oestradiol benzoate and long-acting progesterone can establish and maintain pregnancy after non-surgical transfer of a 7-day donkey embryo. The pregnant mule can successfully carry the donkey conceptus to term and give birth to a mature donkey foal. If this result would be confirmed in further studies using more animals, the use of non-cycling mules in interspecific ET programs could be a good opportunity for the conservation of endangered donkey breeds.

We are indebted to the Centro Rural Malpica (Palma del Río, Córdoba) for providing the animals.