First successful intra-uterine artificial insemination in a cownose ray (Rhinoptera javanica) using chilled semen

Abstract

Reproduction of endangered wildlife, including elasmobranchs, is a fundamental mission of modern aquariums. Captive breeding programs necessitate artificial insemination to better utilize the limited stock in each facility. A cownose ray (Rhinoptera javanica) was successfully inseminated producing a live offspring. The semen was collected under chemical restraint, extended with a custom-made solution and kept chilled at 6°C for 24 hr. The quality of the semen was assessed at collection and just prior to insemination, noting color, volume, subjective sperm motility, sperm viability and concentration. Intrauterine insemination was carried out under chemical restraint. The monitoring of pregnancy was performed by serial ultrasound scans of the fetus assessing gilling and heartbeat. A healthy female pup was born after a gestation of 429 days (61 weeks and 2 days). Genetic microsatellite analysis confirmed that the pup was not a product of parthenogenesis and identified the donor male as the sire. This is the first report of a successful artificial insemination in an aplacental viviparous ray, suggesting these techniques can be applied to other species in this clade.

INTRODUCTION

There are approximately 500 species of skates and rays, collectively known as batoid elasmobranchs and characterized by their dorsoventrally flattened bodies. All batoid species are internal fertilizers. Skates lay eggs and are thus oviparous; rays give birth to live pups in what is known as aplacental viviparity or ovovivparity [4]. The cownose ray, Rhinoptera javanica, is a marine pelagic and coastal aplacental viviparous batoid elasmobranch naturally distributed in the Indo-West Pacific Ocean [25]. It is classified as endangered by the International Union for the Conservation of Nature [23]. This species is commonly housed in public aquariums where it presents a number of challenges, including rubbing injuries, erratic reproductive success and poor pup survival [1, 17].

As the health of the oceans diminishes and as the threats to [aquatic] biodiversity including elasmobranchs multiply [11,12,13, 39], captive breeding programs for elasmobranchs are increasingly pertinent [3, 26, 27, 31, 34, 46]. Assisted reproductive techniques (ART) including semen collection and preservation and artificial insemination (AI) are being developed by a number of aquariums to overcome limitations such as the unavailability of suitable mates and poor individual reproductive performance. ART provide additional tools for population management of captive or wild elasmobranchs [2, 6,7,8,9, 15, 16, 46].

The objective of this study was to explore the feasibility of inseminating cownose rays using chilled semen to produce viable offspring. We report the first live birth by successful AI of an aplacental viviparous ray, using extended semen preserved chilled at 6°C for 24 hr.

MATERIALS AND METHODS

Ethical considerations

This study was approved by the Animal Welfare and Ethics Committee of Ocean Park Hong Kong monitoring the use of novel techniques and research in our animals, under ‘Ethics Approval Number VET-RES_2017-M0-E1.

General husbandry

The animals were housed in a 100,000-L seawater tank. Parameters of the water quality were tested daily. Parameters included temperature (23.5–25.5°C), pH (7.8–8.0), salinity (30–33 ppt), total ammonia/NH3/NH4+-N (0.00–0.06 ppm), dissolved oxygen saturation (97.0–99.5%), dissolved oxygen (6.5–6.9 ppm). All parameters remained within this range for a year prior and after the event. There is no natural sunlight and the light cycle is a 16-hr light cycle all year.

Chemical restraint

The collection of semen, the inseminations and some of the ultrasound examinations to monitor pregnancy were performed under chemical restraint to facilitate handling and to minimize stress and risks of trauma [31, 43, 44]. The sedation protocol consisted of a bath of phenoxyethanol at an initial concentration of 0.25 to 0.3 mL/L [21, 22, 41] achieving a level of sedation allowing all manipulations. Depth and quality of sedation were monitored by observing righting reflex, response to stimuli, respiration [gilling frequency and depth] as well as by ultrasonographic visualization of the heart rate and contractility.

Female recipients

Six months prior to insemination, 3 adult female cownose rays living at Ocean Park Hong Kong for 21, 15 and 12 years respectively, with body weights between 15 and 20 kg and disc width ranging from 90 to 120 cm were sedated as described above. After a general examination, ultrasonography (Easote mylab delta, curvilinear 1–8 MHz) was performed to ascertain their non-gravid status and verify the mature status of the ovaries and follicular activity. Thereafter, the 3 females were housed together in the absence of males in a fiberglass tank 5 m × 10 m × 2 m (100,000 L). Ultrasonography were repeated 4 and 6 months after isolation and prior to AI to verify that they were not gravid and to check follicular diameter in the ovaries. The females had follicular diameters during that period ranging from 0.5 cm to 1.5 cm and no pre-ovulatory follicular development was seen.

Male donors

Two adult male cownose rays living at Ocean Park Hong Kong for 15 and 12 years, weighing 18 and 19 kg respectively, with disc widths of approximately 100 cm, were guided to a smaller pool off the main exhibit tank. In turn they were manually lifted into a 300 L container in a phenoxyethanol solution as described above. They were placed in dorsal recumbency (supine) and the cloaca was digitally spread to reveal the two separate openings leading to the seminal ampullae (Fig. 1). Semen was collected directly from each seminal ampulla using a French 8 feeding tube (Safeed/feeding tube, Terumo, Beijing, China) as shown in Fig. 1. Sperm assessment was conducted at the time of collection, noting color, volume and subjective sperm motility. Sperm viability was assessed using nigrosine eosin stain. Two μL of stain were mixed on a microscope slide with 3 μL of semen for 90” and then spread and air dried. Using oil immersion and 100× magnification a minimum of 200 sperm cells are counted, the live spermatozoa appear clear while the dead ones are pink. Semen concentration is calculated using a microscope at 40× magnification and counting sperm cells a hemocytometer after 200× dilution, (Table 2). The raw semen was diluted shortly after collection using an equal volume (1:1) of an in-house extender solution modified based on Daly & Jones 2017 [6]. The formula of the extender is detailed in Table 1. The extended semen was kept chilled at 6°C overnight. The semen was assessed again just before insemination (Table 2).

Fig. 1.

Semen collection using a French 8 feeding tube (Safeed/feeding tube, Terumo, China) inserted in one of the paired genital openings cranial and lateral to the urinary papilla.

Table 2. Results of the quality of the semen of the two donor males on the day of collection and the next day on the day of insemination. AQ130 was used in Dam AQ533 who gave birth to a live pup and AQ204 who failed to conceive. AQ 638 semen was used in AQ160 who failed to conceive.

Animal ID	Seminal ampulla	Color	Volume before extender (mL)	Total Motility (%)	Total raw concentration (10^6 sperm/mL)	Viability (%)
AQ130 (sire)	Left	Pearl white	4	80	475	Collection day 96%
						Insemination day 93%
	Right	Pearl white	3	80	155	Collection day 93%
						Insemination day 90%
AQ638	Left	Pearl white	3	80	115	Collection day 82%
						Insemination day 82%
	Right	Pearl white	3	80	80	Collection day 86%
						Insemination day 81%

Table 1. Formula of the semen extender solution currently used for elasmobranchs at Ocean Park Hong Kong.

Elasmobranch semen extender solution ingredients, g/100 mL
NaCl	1.403
KCl	0.052
NaHCO3	0.019
MgCl2	0.047
CaCl2	0.111
Na2HPO4	0.007
Na2SO4	0.007
Urea	2.162
Glucose	0.991
TMAO	0.451
pH	7.800

Artificial insemination

Artificial inseminations were performed using the chemical restraint described above, with the rays in dorsal recumbency (supine). The cloaca was lifted above the water line while the head and gills remained submerged. The cloaca was parted digitally and a 10G, 20 cm long sterile stainless steel crop needle (Fig. 2) was passed through the opening to the left uterus and inserted as far as it would go. A crop needle is atraumatic and it was found superior to a soft or semi-rigid tube in finding the opening to the uterus as well as feeling the proximal end of the uterus, located approximately halfway in the coelomic cavity. The chilled semen was loaded in a 5 mL syringe and passed through the crop needle, followed by another 2 mL of extender and 2 mL of air. The cloaca was observed for semen reflux which was considered minimal in all females. The females were then returned to their tank where they remained isolated from males.

Fig. 2.

Insemination of female AQ355. A 10G, 20 cm long sterile stainless steel crop needle was inserted as far as it will go in the uterus.

RESULTS

Semen quality

Semen characteristics are presented in Table 2. It is worth noting that the characteristics of the semen collected differed between seminal ampullae. Both testicles and their associated structures are not expected to behave exactly the same at all times, as in other paired organs. N.B., mammal ejaculates are a mix of both testicles and differences between testicles are not pertinent to semen evaluation.

Pregnancy monitoring

Twenty weeks after insemination, the 3 females were coaxed into a small floating bag and were examined by ultrasonography for pregnancy without sedation. In conscious stingrays, ultrasound scans were carried out through the dorsal wall. In chemically restrained stingrays, ultrasound scans were performed both from the dorsal and ventral sides. One female (AQ533) was found to have a viable fetus, the other two female rays did not show any uterine content. Identical findings were made at week 41 after which only AQ533 was scanned. The other two females remained in the tank to provide company for the gravid female, enhancing the welfare of this gregarious species. On week 46, 48, 50 and 54 ultrasound scans were performed without chemical restraint and showed a growing fetus with visible gilling or heart beat or both. An ultrasound scan without chemical restraint at weeks 60 failed to provide satisfactory visualization of the heartbeat or gilling to confirm the viability of the fetus. As we were already 4 to 10 weeks beyond the published gestation period for the genus [29, 30, 35, 37, 40], the female AQ533 was chemically restrained on week 61 to allow a thorough ultrasound examination. During that procedure a live fetus with normal gilling and heart beat could be seen (Fig. 3). Three days later, on the 23rd April 2023, after a gestation of 429 days, cownose ray AQ533 gave birth to a live pup in fine health, named AQ1316.

Fig. 3.

Sixty-one weeks after AI. a. Ultrasound of a viable fetus. b. Line drawing of the Ultrasound in Fig. 3a.

Genotypes were analyzed for AQ1316 (pup), AQ533 (dam), and AQ130 (donor male) and AQ638 (control male). Microsatellite analysis of the parents and offspring by two different laboratories confirmed that the offspring (AQ1316) resulted from the fertilization of female AQ533 by male AQ130 (Tables 3 and 4). AQ1316 and AQ533 showed different genotypes at several loci determining that the offspring is not a parthenote (Tables 3 and 4).

Table 3. Results from two laboratories using microsatellite markers to confirm paternity by AQ130 as well as exclude the possibility of parthenogenesis. Laboratory 1.

	AQ638		AQ130		AQ533		AQ1316
Rbon37	156	156	148	148	148	148	148	148
Rbon56	269	271	267	271	187	227	187	267
Rbon80	266	266	208	217	232	241	208	241

Eleven microsatellite markers (McDowell and Fisher 2013) were used to genotype all four samples. One locus (Rbon30) failed to amplify. Seven loci (Rbon1, Rbon38, Rbon41, Rbon52, Rbon75, Rbon78, and Rbon79) were monomorphic and thus uninformative. The remaining three loci were variable and confirm paternity by AQ130 as well as exclude the possibility of parthenogenesis.

Table 4. Results from two laboratories using microsatellite markers to confirm paternity by AQ130 as well as exclude the possibility of parthenogenesis. Results from laboratory 2.

	AQ638		AQ130		AQ533		AQ1316
Rb01	83	103	103	103	83	103	103	103
Rb02	166	175	166	175	175	175	166	175
Rb03	249	283	249	249	249	283	249	283
Rb04	110	110	110	110	102	110	102	110
Rb04-2	226	228	230	232	244	246	230	244
Rb05	192	242	192	192	192	215	192	215
Rb06	103	155	177	200	142	150	150	200
Rb09	84	98	98	98	98	98	98	98
Rb10	99	101	99	101	99	101	99	101

Ten microsatellite markers were used, eight could be amplified and scored. Marker Rb04 amplified two loci. Results confirm paternity by AQ130 and exclude the possibility of parthenogenesis.

DISCUSSION

Semen collection and processing of fresh and chilled semen is increasingly better understood in elasmobranchs [15, 24, 28, 34, 45] and is relatively simple [9, 16]. Techniques for freezing semen are also being developed with reported post-thaw motility of up to 42% [15], 35%, [18], 45% [Martelli unpublished]. While these results are comparable to human frozen-thawed semen [33], further trials and refinement of extenders and freezing/thawing protocols should produce improved reliable cryopreservation protocols.

Parthenogenesis is a breeding strategy that is common in elasmobranchs [2, 5, 10, 14, 19, 38]. Sharks follow an XY chromosomal sex determination system in which all the offspring of parthenogenesis are females [20, 42]. Given that the pup referred to in this paper is a female, it was necessary to verify that this was not a parthenogenic event but a true sexual reproduction resulting from AI. Blood samples of the mother, the pup and the two males used for the insemination were sent to two separate labs for genetic testing. Both laboratories reached the same conclusion that the pup could not be a parthenote, and was the result of sexual reproduction. Paternity was attributed by both laboratories to the same male (AQ130) which we knew to be the donor used on this female (AQ533) (Table 3).

This report describes the successful intra-uterine insemination of a cownose ray using semen chilled at 6°C collected 24 hr prior to insemination. This is the first report of a successful fetal development and live birth in an aplacental viviparous ray resulting from artificial insemination. The gestation period was 61 weeks and 2 days, considerably longer than the 47–53 weeks reported in the literature [40]. We are unable to determine whether the birth occurring so close after that last ultrasound examination under chemical restraint was the result of the end of a normal gestation, or was precipitated by the stress of the restraint, or was triggered by a typhoon that coincidentally hit our facility on the date of birth. An April birth is consistent with what is reported in wild populations [29, 30, 35, 40]. It is possible that gestation period is variable. It may also be that, although not documented in stingrays, cownose rays too are capable of storage of sperm in the oviducal gland, as is the case for other elasmobranch species [36]. Repeated ultrasound examinations confirming the presence of a live fetus were instrumental in informing decisions along the unexpectedly long gestation period.

Challenges remain in the ART of elasmobranchs, not least the question of the best timing for AI. In contrast with studies by Nozu [32] we have so far been unable to match hormones with follicular ultrasonographic changes and naturally occurring mating behaviors. Mating activity in this species, as suggested by mating injuries or witnessed directly, occurs at seemingly random times of year in our aquarium. At the time of insemination none of the three female rays had large follicles. The timing of insemination was chosen to match the timing of the best semen quality of the males in our collection, in this case semen was collected on the 17th February 2022 followed by AI on the 18th February 2022.

A viable offspring was successfully produced for the first time by artificial insemination in an aplacental viviparous ray, validating efforts by numerous groups perfecting ART techniques in elasmobranchs. This demonstrates that straight forward techniques for semen collection and chilled semen processing as well as intrauterine artificial insemination can be used in this clade, opening option for genetic management of collections and captive propagation of the species.

CONFLICTS OF INTEREST

The authors have nothing to disclose.