Reproduction in the Swamp Buffalo (Bubalus bubalis)

The domestic water buffalo (Bubalus bubalis) belongs phylogenetically to the tribe Bovini (Bovidae, Bovinae), the group includes domestic and wild cattle, bison, yak, African buffalo and saola [1,2]. Generally, they are classified into two main categories with distinctive morphology and karyotypes: the Murrah or river buffalo with 2n=50 and the Swamp buffalo with 2n=48 chromosomes, respectively [3-5].

The domestic water buffalo is typically found in tropical and subtropical forests, wet grasslands, marshes and swamps [6]. Combined available archaeological and molecular information suggests that these animals were domesticated independently from distinct wild populations [7,8]. Swamp buffalo was domesticated in the North-eastern parts of India, Bangladesh, China and South-East Asian countries approximately 5000 years ago and used mainly as draught animal, while the river buffalo was domesticated in the South- and Southwest Asia, Egypt, Balkan, Italy, South America and the Caribbean and was used as a dairy animal [9,10].

Despite the increase in the total world buffalo population (from 120 to 168 million) in the last two decades, the number of swamp buffaloes has declined due to mechanization of agriculture, the increase in meat demand and urbanization, as well as poor reproductive performance [2,11,12].

Reproduction in the swamp buffalo appears to be similar to the river buffalo with delayed puberty, poor estrus expression, prolonged calving intervals and seasonal breeding limiting the reproductive performance. Some differences, however, do exist between swamp and river buffaloes; swamp buffalo females are known to have a smaller number of primordial follicles in their ovaries [13], a silent estrus [14,15], a longer gestation period of 330 days [16] and a higher age at first calving [17,18] whereas swamp buffalo males evidence a higher age at puberty [19,20] and extremely low sperm concentration [19,21]. Many of the reproductive processes in swamp buffalo appear to be poorly studied and sparsely described. In this chapter we present the reproductive activity and assisted reproductive technology in swamp buffalo with the emphasis on the specific features of this species as a traditional breeder, which is capable of adapting to a harsh environment with minimal managerial input.

1. Reproductive Activity

1.1. Puberty

The reproductive activity of swamp buffalo, as in other mammals, begins with puberty, a complex physiological phenomenon which marks the maturation of the hypothalamus-pituitary –ovarian neuroendocrine axis. The integrated secretion of hypothalamic neuro-hormone GnRH (gonadotrophin releasing hormone), adenohypophysis gonadotropic hormones FSH (follicle stimulating hormone) and LH (luteinizing hormone), ovarian hormone progesterone, and estrogens stimulate the changes in primary and secondary sexual characters, the maturation of follicles and the release of the first oocyte in the female; the first mounting and erection in male, indicating the advent of puberty.

Swamp buffalo heifers begin puberty when they attain 55-60% of their adult body weight i.e., 200-300 kg [14]. It has also been found that the onset of puberty is highly variable, with an average between 3-4 years. In the Northeast of Thailand, the average age of first calving is 3.9 to 4.8 years, while it is 3.5-4.7 years for swamp buffalo in Vietnam [21-24] (Table 1). The puberty attainment and first estrus may vary depending on the level of nutrition, working condition and the period of the year. In the areas with better feeding quality and lower working regime, the age of puberty can be reduced to up to 18 to 24 months [23,25].

1.2. Utero-ovarian Anatomy

The ovaries of swamp buffaloes have an elongated egg-shape and a flat form, and are smaller in size and weight than that of river buffalo and cattle (Fig. 1; Table 1). The average dimensions of ovaries collected during summer (n = 205) were 2.07-2.17 cm long, 1.44-1.46 cm wide and 1.12-1.15 cm thick, for the left and right ovary respectively. Their weight varied from 2.03 to 2.12 g. For the ovaries collected during winter (n = 120), these data were 2.16–2.29 cm long; 1.52-1.57 cm wide and 1.01-1.04 cm thick and 2.58-2.77 g in weight, for the left and right ovary respectively [Uoc et al., unpublished data]. During the low breeding season, the ovaries were small (0.7 x 0.5 x 0.3 cm) with flat, solid stroma without growing follicles and corpora lutea.

Figure 1. Reproductive activity in the swamp buffalo: a) Group of swamp buffaloes b) Swamp buffalo during estrus; c) Mucus discharge from the vulva during estrus in a swamp buffalo d) Swamp buffalo's cervix with three cervical folds; e) Histological structure of cattle ovary; f) Histological structure of swamp buffalo ovary; g) Swamp buffalo ovaries after superovulation treatment with eCG; h) Swamp buffalo's ovaries after superovulation treatment with eCG plus estradiol supplementation; i) Uterus and ovaries of young swamp buffalo after treatment of eCG plus estradiol supplementation.

The number of visible surface follicles and corpus luteum (CL) varied, depending on the period of estrus activity or anestrus. The ovary of swamp buffalo during active estrous cycle contains few surface follicles compared to the ovaries of river buffalo or cattle. The CL of the swamp buffalo on the first days after ovulation is pinkish grey in colour with red veining; the regressing CL become dull grey in colour and is deeply embedded in the ovarian stroma. The average number of surface follicles per ovary was 5.48 ± 0.72 (for follicles size < 2 mm), and 1.85 ± 0.23 (for follicles size > 2 mm) during summer; and 1.92 ± 0.19 (for follicles size < 2 mm), and 5.85 ± 0.51 23 (for follicles size > 2 mm) during winter [Uoc et al., unpublished data].

The oviducts and bicornuate uterus of the swamp buffalo have similar shape compared to those of cattle, but differ in elasticity and structure. In addition, swamp buffaloes have a narrower cervix (with 3 cervical folds) (Fig. 1d) and a more tortuous lumen than in river buffalo and cattle (Fig. 1i). The passage through the cervix canal, therefore, is difficult in swamp buffalo heifers, even with an insemination catheter and it is impossible using a balloon catheter. The uterine endometrium and uterine wall of the swamp buffalo is thicker than that in cattle, but more flaccid and fragile.

1.3. Foliculogenesis

The histological pattern of follicular population and growth was studied in juvenile (4–9 months old) and adult swamp buffaloes. The distribution of follicular population in non-attretic and atretic follicles (more than five pyknotic bodies), which were grouped into six classes according to diameter, was found to be similar and resembled that of cattle, however, the population of antral follicles was poor (Table 1).

In swamp buffalo, the population of primordial follicles at age 2 years; 7-8 years and 12-14 years was decreased from 47.189 ± 39.23 to 5.996 ± 2.52 and 3.673 ± 1.97 respectively. In addition, follicle numbers were significantly lower than in cattle, from growing (4.233 vs 18.0 vs 17.0) to tertiary (62.7 vs 9.0 vs 6.67) stages [13,].

In comparison with cattle, the total number of healthy and atretic antral follicles in swamp buffalo ovaries was significantly lower, with an increased percentage of atresia (40±3% vs. 18±5%). The overall population of antral follicles in buffaloes was only 20% compared to that of cattle (47.5 ± 23.8 vs. 233.0 ± 95.8) [26,27]. Ocampo et al., [28] demonstrated that in swamp buffaloes, the average proportion of antral follicles which appeared to be attretic histologically (Fig. 1e, Fig. 1f), was 82% (71.4% during follicular phase and 94.4% during luteal phase), a higher percentage when compared to that reported in river buffaloes (67%) and in cattle (50%) [29,30].

The fundamental features of follicle recruitment, selection, dominance and atresia during the estrous cycle of the swamp buffalo present a typical wave-like pattern which is similar to river buffalo and cattle [31]. Each wave is characterized by wave emergence, growth, dominance and atresia or ovulation of a dominant follicle. The majority (77%) of cycles are characterized by 2 follicular waves and 22% of cycles are characterized by one follicular wave patterns (Fig. 2).

Figure 2. Follicular development in swamp buffalo (De Rensis et al., [135]).

Within cycles characterized by one wave of follicle development, emergence was recorded on Day 2.1 ± 0.5 (Fig. 2). The dominant follicle attained the largest size (15.5 ± 2.4 mm) on Day 13.1 ± 1.3. Within cycles with two follicular waves, first and second wave emerged on days 1.1±0.3 and 11.0±0.9, respectively, and the largest diameter (11.6 ± 1.2 mm) was recorded on day 19.3 ± 1.3 [32,33].

The presence of one follicular wave pattern in swamp buffalo (22% of estrous cycles) indicated that one out of three estrous cycles present a dominant follicle that persist for more than 10-14 days. The number of follicular waves during an estrous cycle in river buffalo (Murrah) was one in 3%, two in 63% and three in 33% of animals [31]. This follicular wave pattern does not change during the year but the average size of the dominant follicle during the low breeding season (14.5 ± 2.1 mm) is slightly reduced compared to that during the breeding season (16.4 ± 2.7 mm) [34].

The number of follicles recruited into a follicular wave was reported to be lower in swamp buffaloes than in river buffalo and cattle. The number of non-atretic follicles (> 1.7 mm) was between 1 and 5 (average of 3) for swamp buffaloes [26] and between 17 and 32 (average of 22) for cattle [20]. A recent study by Gimenes et al., [35] indicated that the number of follicles at wave emergence in river buffalo heifers (13.2 ± 1.9) was similar to that of Holstein heifers (Bos taurus) (17.20 ± 4.8), but lower than that of Nelore heifers (Bos indicus: 29.4 ± 4.2) however, the diameter of the dominant and ovulatory follicles did not differ.

1.4. Estrous Cycles

Previous studies reported a high variation in the length of the estrous cycle in swamp buffaloes e.g., 10.2 + 0.38 day [8], 21.32 days [36] and 19.9 + 4.4 days [15], slightly shorter compared to river buffalo and cattle (Table 1). The estrous cycle varied from 24-40 days in Vietnamese swamp buffalo [24]). Estrus duration in swamp buffalo is about 40 hours [14] (Table 1).

The luteal phase of swamp buffaloes lasts 15-16 days, and is characterized by the presence of a CL and elevated progesterone plasma levels, while the follicular phase lasts 4-6 days and is characterized by the absence of a CL, very low plasma progesterone levels and high plasma estrogen levels. Estrus activity in swamp buffaloes is characterized by a marked seasonal influence. The duration of estrus is shorter, estrus expression is less marked during the low breeding season from May to July (summer or dry season) and this may be mistaken for anestrus. Mucus secreted from the cervix during estrus is less copious (Fig. 1c); the vulval changes and expression of estrus mucus are less pronounced than in cattle. The animals without vaginal discharge may also be in estrus (Fig. 1a, Fig. 1b). In most swamp buffaloes, the discharge is observed during palpation of the genital organs per rectum before AI. The discharge can be found on the tail, as well as on the gluteal region. In buffaloes, the discharged mucus should be observed carefully in early hours of the morning or in late hours of the evening when the animal is in recumbent position.

1.5. Neuroendocrine Patterns

Neuroendocrine patterns which reveal the reproductive activity have been extensively studied in cattle and other domesticated species [62,63].

Despite numerous studies previously carried out in river buffalo, data about reproductive endocrinology in swamp buffalo is still limited. Based on results of different studies [64-67], a comparative study about plasma concentrations of progesterone, estradiol, prostaglandin and LH during the estrous cycle between cattle, river and swamp buffaloes was published previously [39].

The basic pattern of changes in the hormonal profile of the estrous cycle of swamp buffaloes is similar to that of river buffaloes and cattle. However, in this species, the circulating concentrations of progesterone, estradiol, and LH are always lower (Table 2).

The mean progesterone value during proestrus and estrus was 0.1-0.49 ng/ml and it increased rapidly from day of ovulation (Day 0) to Day 7-10, reaching a peak of 2.7 ng/ml at Day 15.3, then it declined rapidly to basal levels [67,74].

Plasma estradiol concentrations presented minor fluctuations for the first 15-16 days of the cycle and thereafter they increased with a peak of 9-13 pg/ml on the day of estrus or one day before LH pre-ovulatory surge, followed by a decline to 7-9 pg/ml within two days [67,74]. Maternal plasma estradiol concentrations were high before parturition in the swamp buffalo. Mean circulating concentrations of LH are low (0.8-1.2 ng/ml) and variable during the luteal phase and increase during the follicular phase with a preovulatory peak reaching 35 ng/ml at the onset of estrus, followed by a sharp decrease within a day. Peak LH concentrations were estimated to occur about 14.8 hours after the peak in estradiol concentration. The duration of the LH surge has been estimated to be 7 to 12 hours. Ovulation took place approximately 30 hours after the LH peak [67,74].

There is no information on the concentration of FSH in swamp buffalo. In river buffalo, a mean FSH value of about 120 ng/ml was reported with great variation, from 16.6 to 120 ng/ml during the follicular phase and 59.7 to 600 ng/ml during estrus [68,73,101].

The circulating prostaglandins and inhibin were investigated in river buffalo [15,70] and cattle [70,95,96,102], but no data are available for swamp buffalo. The plasma inhibin levels in relation to steroids and gonadotrophins during estrous cycle in river buffalo revealed that both inhibin and estradiol have a feedback regulatory effect on FSH secretion [90].

1.6. Fertilization, Embryo Development, Pregnancy and Parturition

The interval from the end of estrus to ovulation in swamp buffalo has been estimated to be 13.9 hours - a little bit longer compared to river buffaloes [67]. The optimal time for insemination has not been specifically studied but it is considered to be similar to river buffalo, i.e., 8-12 hours after the onset of estrus. Spermatozoa require at least 4-5 hours in the female tract to complete the process of capacitation before successful fertilization can take place [39]. There were only a few studies on the conception rate of swamp buffalo. Cockrill [4] reported an overall rate of 60% for buffalo in China. Chantalakhana [103] listed the common conception rates of 50 to 65% for swamp buffalo in Thailand, based on calving rates. The authors are of the view that under natural conditions, more than 80% of female animals come into estrus and get pregnant during the high breeding season (February to March). The average conception rates after natural mating for cyclic animals and with regular calving can range from 80-90%.

The development of buffalo embryos at early stage prior to implantation takes place in the oviduct and uterus which is similar to cattle, but a difference in its chronology does exist (Table 1). Few studies have been undertaken for in vivo development of swamp buffalo embryos. A few studies on Thai swamp buffaloes collecting embryos non-surgically [104,105] on Days 5.5, 6.0, 6.5, 7.0 and 7.5 after fertilization found embryos at the 16-cells stage, compact morula, blastocyst, hatched blastocyst and hatched expanding blastocyst, respectively. The percentage of normal embryos obtained with a single embryo collected after either natural or induced estrus (71% and 83%, respectively) was higher than that after superovulation (38%).

The gestation length of swamp buffalo is around 330 days, longer than that reported in river buffalo (315 days - [16]) or in Zebu and Red Sindhi cattle (285-292 days- [54]). The gestation length of Carabao and Murrah buffaloes breed in Philippines is 320 to 325 days and 314-317 days, respectively (Table 1). The gestation length in swamp buffaloes in Assam, India was 324.40 ± 0.21 days and was affected by the season of calving and gender of calf; buffaloes carrying male fetuses and calving during the breeding season had a significantly longer gestation period [106]. The gestation length for Vietnamese swamp buffaloes varied from 320-345 days [21]. Based on abattoir studies, the fetal growth dimensions during the different stages of gestation in swamp buffaloes have been described recently [107] and could serve as useful data for pregnancy diagnosis or calculation of fetal age. The maximum growth of fetus was recorded during 301-330 days of gestation [107].

The Carabao buffalo in Philippines had an average age at first calving of 43 months, slightly shorter than river buffalo (45-52 months) and considerably longer than the Red Sindhi cattle (35 months). The age at first calving for the Murrah x Carabao crossbreed females was recorded at 39 months, which is shorter than for both purebreds [108].

The calving interval of around 500 days has been reported for swamp buffalo. Swamp buffaloes with a calving interval of less than 500 days were considered as animals with a short calving interval [18]. Approximately 20% of the buffalo population in Malaysia had a calving interval greater than 500 days [17]. The process of parturition and postpartum events have not been described in detail for the swamp buffalo, yet are presumed to be similar to those in river buffalo.

1.7. Seasonal and Postpartum Anoestrous

Swamp buffaloes are polyoestrus but with a marked seasonal pattern of breeding activity, showing a distinct anestrus, and consequently, varied in display of estrus and calving rate. In Southeast Asia, the breeding frequency of swamp buffaloes is highest during the period from December to April (winter-spring) and lowest in the period from May to July (summer). The calving period usually last from December to January, the animals return to estrus and mating during spring. For the majority of remaining non pregnant buffaloes, they become anestrus due to the hot summer months (temperature extremes: 40°–46°C). Unobserved silent estrus and short duration of estrus are common in swamp buffaloes during this period [15,40,45,58]. The absence of large follicles and CL on the ovaries is sign of true anestrus. Palpation and direct observation of ovaries collected from slaughtered anestrous swamp buffaloes showed flat, solid ovaries which were significantly reduced in size.

The onset of ovarian activity during the postpartum period has been studied by ovarian palpation per rectum, expression of estrus, milk or plasma hormone profiles. There is an extremely high variability in the duration of postpartum period in the swamp buffalo. The first estrus can appear between 51.6 ± 15.3 to 144 ± 22 days after calving, depending on the region and breeding conditions. In the North East region of Thailand, this period can be up to 18.3 months [51].

Estrus behaviour is less evident in the postpartum swamp buffaloes. About 25-30% of estrus periods, which were associated with the first postpartum ovulation, were not detected, even when aided by a vasectomised bull [109]. It was reported that the maximum diameters of the first, second, third and fourth ovulatory follicles were 13.5 ± 2.09, 14.2 ± 1.58, 14.8 ± 2.38 and 14.0 mm, respectively. This partially explains why relatively few postpartum buffaloes displayed estrus signs at the time of the first ovulation (40 days postpartum), while all animals (100%) show estrus behaviour around the second, third, and fourth ovulation [47,110].

No follicular activity was present in swamp buffaloes during the first 30 days after calving. The results of plasma progesterone determination indicated that the first postpartum ovulation occurred at 96 ± 22 days in swamp buffaloes [57] and beyond 60-90 days in river buffaloes [93] (Table 1).

In swamp buffaloes, the major limiting factor in reproduction efficiency is the long period of postpartum anestrus. Under optimal conditions, the regression of the CL of pregnancy is very rapid, with a complete regression by 7 ± 2 days from calving and by Day 10; the postpartum CL is palpable as a small hard protuberance (< 3mm) over the ovarian surface.

The suckling process, level of nutrition, body condition score (BCS), milk yield and season can deeply influence the resumption of ovarian activity and estrus during the postpartum period. In suckled swamp buffaloes, the weaning of calf at 17-32 days postpartum induces early ovulation i.e., 42 ± 8 days, compared to 55 ± 10 days in suckled buffaloes [57]. In acyclic suckled buffaloes, the temporary calf removal during 72 h, at 91-93 days postpartum, induced ovarian cyclicity about 14 days earlier [111].

Body condition is also an important factor that can influence the postpartum resumption of ovarian activity. In fact, in animals with a mean BCS of 3.3 (0 = extremely thin, 5 = fat), ovarian activity is resumed within 90 days, while animals with a BCS of 2.8 continued to be acyclic [111]. In the region where buffaloes were fed with a constant balanced diet, were allowed to wallow and were provided with shade, under summer conditions, females returned to estrus in a short postpartum interval, with ovulation and showed a high calving rate (> 90%).

1.8. Pregnancy Diagnosis

Early pregnancy diagnosis is important for swamp buffalo breeding management due to their low conception rate and silent estrus expression. Attempts to develop laboratory methods for pregnancy diagnosis in swamp buffalo have been reported since the last three decades [112-114]. Studies in swamp buffaloes in Thailand and Malaysia [113] showed that progesterone levels can increase to 0.7–2.0 ng/ml during early pregnancy and the measure of serum progesterone can be used to detect early pregnancy between 19 and 30 days after insemination. The detection of estrone sulphate during gestation in swamp buffalo showed that the levels in the pregnant group were gradually increased and reached up to 1232 ± 249.46 pg/ml at Day 205 after AI. The estrone sulphate levels of pregnant swamp buffaloes and those that experienced early embryo loss at Day 20 after AI were 149.24 ± 97.03 and 55.30 ± 12.12 pg/ml, respectively [114].

The application of real-time ultrasonography was developed late for studying reproductive functions in buffalo at Days 30–35 of gestation [115]. The use of ultrasound guided transvaginal probe to detect pregnancy on Day 30 post breeding with 100% confirmation was reported [115].

The application of these techniques for pregnancy diagnosis is still very modest compared to river buffalo and cattle. Due to the low cost, fast and simple application, transrectal palpation of the genital tract is still the most popular method for pregnancy diagnosis in swamp buffalo at the village farms. The limitation of this technique is that gestation in buffalo is longer than in cattle and palpation is recommended to be carried out 4 months after insemination. Recently, the technique for pregnancy detection based on the measurement of PAG concentrations was studied in swamp buffaloes [116] (Fig. 3). PAG belong to a large family of glycoproteins that are synthesized in the superficial layer of the ruminant placenta and PAG-based pregnancy diagnosis has become popular and is accepted by many cattle farmers [117] and recently for river buffaloes [75,100,118]. In river buffaloes, according to Karen et al. [119], the sensitivity of PAG-RIA test was very low (11.1%) on Days 19–24; but increased to 80% on Days 25–30 and reached 100% on Days 31–35 of gestation. The results of the study designed to determine PAG concentrations in maternal and fetal plasma, allantoic and amniotic fluids in swamp buffalo species by RIA showed that the diagnosis was considered negative, doubtful, or positive according to the PAG concentrations of < 0.6 ng/mL, 0.6 to 0.8 ng/mL, and > 0.8 ng/mL, respectively [88].

Figure 3. Correlation of pregnancy-associated glycoprotein (PAG) concentrations measured by the three radioimmunoassay (RIA) systems in swamp buffalo samples: (a) Fetal plasma; (b) maternal plasma; (c) allantoic fluid and (d) amniotic fluid. RIA 1 (AS#497, raised against bovine PAG), RIA 2 (AS#706, raised against caprine PAG) and RIA 3 (AS#859, raised against buffalo PAG (Hanh et al., [98]).

1.9. Male Reproductive Activity

Male swamp buffaloes attain puberty at around 20–24 months of age [20,120]. Sexually mature swamp buffaloes can be trained for semen collection at about 3–4 years of age [20,]. The average testosterone concentrations varied from 0.1-0.33 for the younger to 0.55 ng/ml for the older bulls, whereas the mean LH concentrations for individual bull varied from 0.33 to 1.17 ng/mL [122].

The studies on the characteristics of the swamp buffalo semen have showed that the average ejaculate volume was 2.9 ml; the average sperm concentration was 1.06×109 cells/mL and the average motility was about 70% [123].

1.10.        Reproductive Disorders

Reproductive disorders are sparsely mentioned for the swamp buffaloes. Postpartum and seasonal anestrus appear to be the major reproductive disorder [109,124]. A study demonstrated that follicles of the postpartum swamp buffalo cow developed in a wave-like pattern that appeared to be irregular at the start. Poor estrus signs and a short subsequent cycle were observed to be related to the first ovulation after calving [8]. The completion of uterine involution occurs in 28-33 days and no difference in this interval between free- suckling and twice/day- suckling buffaloes has been reported [109,125]. Sparse reports mention reproductive problems such as ovarian cysts [126] or dystocia [127].

2. Assisted Reproductive Technology in Swamp Buffalo

The development of assisted reproductive techniques has been considered as an important approach for breeding and genetic improvement in swamp buffalo. However, due to the weak estrus expression, low efficiency in superovulation and embryo production, natural mating is still the main breeding mode in swamp buffalo [24,128].

2.1 Estrus Synchronization and AI

The early investigations on estrus synchronization in swamp buffalo were based on approaches developed in cattle using prostaglandins or progestagens as controls of the luteal phase. Estrus with mucous discharge was evident within 48-72 h from treatment and lasted for 4-5 days in swamp buffaloes which received a double prostaglandin treatment at 11-12 days interval [129-131]. The pregnancy rates, which varied from 18% to 35%, were obtained by using progestagens (CIDR®, PRID® or CRESTAR®) to synchronize estrus and ovulation in swamp buffaloes [132]. The level of nutrition and the season of initiation of synchronization are known to affect the outcome [133].

The use of newer protocols employing GnRH such as Ovsynch (GnRH-PGF2α-GnRH-AI), Select-Synch (GnRH-PGF2α-AI) or eCG and hCG plus estradiol, have yielded improved estrus and pregnancy rates varying from 64.71; 77.14 to 83.87% [134-137].

2.2. Superovulation and Embryo Transfer

The successful embryo transfers in swamp buffaloes were reported in Thailand in 1989 [138] and in the Phillippines in 1991 [139]. The calves were born from fresh collected embryos transferred into synchronized recipients. The freezing of swamp buffalo embryos was studied at different stages as morula, compact morula, early blastocyst and blastocyst [140] and the attempt to transfer cryopreserved embryos was carried out during the following years. However, the obtained pregnancy and calving rates were very low: 28.6% and 14.3%, respectively for fresh embryos. For frozen embryos, only one swamp buffalo (5.9%) was pregnant but the pregnancy was not carried to full term [94,141].

In swamp buffaloes, attempts to induce superovulation using treatment protocols currently applied in cattle resulted in a very low ovarian response. The average number of recovered embryos per donor is often less than 2.0, with the number of transferable embryos per donor averaging less than 1.0. [6,142-146]. Pre-treatment with the recombinant bovine somatotropin (rBST) can improve the total number of embryos recovered per collection (4.5 ± 1.6 vs. 2.3 ± 1.0), as well as the number of transferable embryos per collection (3.0 ± 1.0 vs. 0.8 ± 0.3) [108]. Significant improvement in estrus expression and ovarian response with an average 8.5 ovulation/ donor was reported for buffaloes that received estradiol supplementation during gonadotropin treatment (Fig. 4) and hCG to induce ovulation [147,148] (Fig 1g; Fig. 1h).

Figure 4. Superovulation protocols in swamp buffaloes reported by Techakumphu et al., [107] and Uoc et al., [110].

The efficiency of nonsurgical embryo collection for swamp buffalo was also previously evaluated. The development of embryos is more rapid and nonsurgical embryo collection is recommended at Day 5 or 6 after insemination [49,]. Following preliminary previous study [104], the recovery rate in buffaloes with a single embryo collected after natural estrus was higher than in buffaloes with induced estrus or superovulation (78% vs. 46% vs. 54.5%, respectively). The recovery rate of buffalo embryos was lower compared to cattle, even with high a superovulatory response [147,148,150]. In our experience, the loss of the embryo may also be due to the specific structure of the uterine wall in swamp buffaloes which is softer and easily absorbs the flushing medium.

2.3. In vitro Embryo Production and Ovum Pick Up

The in vitro production (IVP) of embryos based on the combination of in vitro maturation and fertilization (IVM-IVF) and ovum pick up (OPU) has been studied, aiming to overcome the reproductive inefficiencies in swamp buffalo breeding programs or cross-breed swamp-river buffaloes programs [80,151-154]. The early study on in vitro production in swamp buffalo was reported by Pavasuthipaisit et al., [155] with more than 60% of in vitro matured and fertilized oocytes attaining cleavage and developing to the 8-cell embryo stage. The maturation rates were 41% and 52%, and the fertilization rates were 21% and 21% for oocytes of adult swamp buffaloes and gonadotrophin-treated prepuberal swamp buffaloes, respectively [37]. The successful application of IVF-OPU to produce F1 (river x swamp) buffaloes or Murrah and Nili-Ravi buffaloes was reported recently [156]. The F1 embryos were produced by fertilization of swamp Cumulus oocyte complexes (COCs) recovered from abattoir ovaries co-incubated with river sperm cells. IVM-IVF resulted in 56.7% cleaved zygotes, 28.8% blastocysts, 41.4% pregnancies and 34.5% calvings after embryo transfer to 29 swamp recipients following natural estrus.

The available published data to date have therefore clearly indicated that the results of IVP in swamp buffaloes are modest compared to river buffaloes or cattle. Whatever method of oocyte recovery was used, buffalo ovaries could yield only a small number of quality oocytes. In a study using ovaries recovered from abattoir, the total yield of oocytes per ovary was 3.34- and 3.62 for the summer period and 4.48- and 5.26 for the winter period, for the left and right ovaries respectively. The number of oocytes among grade A and B quality were 47.7 and 7.7 for the summer period; and 29.0 and 46.6 for the winter period, respectively [ Uoc et al., unpublished data].

The application of gonadotropin hormone plus a progesterone ear-implant can increase the number of collected oocytes in prepubertal swamp buffalo calves. It was reported that 8.4-9.0 oocytes per animal were collected from buffaloes treated with FSH for 3 days and receiving GnRH on the third day, 24 hours before OPU at 2-week intervals [145,157]. It was also reported that 3.7 ± 2.7 and 5.9 ± 3.5 oocytes could be collected from, cycling and lactating postpartum buffaloes respectively, with animals being treated with a total of 400 mg, follicle stimulating hormone (FSH), administered twice daily over 3 days in decreasing doses, together with 100 μg of GnRH, 24 h after the last FSH injection before OPU at 2-week intervals [158]. According to Yindee et al., [159], the OPU treatment twice per week followed by once-weekly collection for 10 consecutive sessions without hormone stimulation is optimal for the collection of the oocytes from non-lactating multiparous swamp buffalo. The number of small and large follicles in the once-weekly OPU group (5.2 ± 0.7 and 0.9 ± 0.2, left and right ovaries respectively) was higher than in the twice-weekly OPU group (3.9 ± 0.5 and 0.5 ± 0.1, left and right ovaries respectively).

An important improvement in maturation and fertilization rates was achieved during the last few years. The maturation rate of 72.1% and fertilization rate of 54.8% were reported for COCs in vitro fertilized with frozen semen and cultured in 50 μL droplets of synthetic oviductal fluid containing 1% (v/v) fetal calf serum [160]. The IVF system with high rate of cleavage (77%) and development to blastocyst stage (33-51%) was set up by Parpnai’s group using co-culture in mSOF with bovine oviductal epithelial cells [161].

Season can affect considerably the quality of oocytes and sperm. The use of cryobanking for sperm and oocytes collected at favourable periods is recommended. Data from Koonjaenak’s study showed that the percentage of live sperm was significantly higher (54.6%) when collected during the winter months than during the rainy (43.5%) or summer seasons (46.7%) [37]. Recently, Liang et al., [162] reported the potential of buffalo oocytes vitrified at MII stage using the micro-drop method in 20% dimethylsulfoxide (DMSO) plus 20% ethylene glycol (EG) and 0.5M sucrose. Oocytes could be developed to the blastocyst stage following parthenogenetic activation and intracytoplasmic sperm injection. Although the development of blastocysts after intracytoplasmic sperm injection (ICSI) combined with chemical activation was similar to IVF oocytes, the production of a swamp buffalo calf from ICSI embryos has not been achieved to date. Male pronuclear formation failures, lacking expression of paternally-expressed gene has hampered the success of ICSI in swamp buffalo [160]. It has been suggested that sperm treated with dithiothreitol and detergent (DTT) with reacted acrosomes before ICSI, together with the activation of the ICSI oocytes could result in the successful male pronuclear formation [112].

2.4. Nuclear Transfer and Somatic Cloning

After the birth of Dolly the cloned ewe, several studies have been carried out on somatic cloning (SCNT) in swamp buffalo. The first successful production of blastocysts from swamp buffalo fetal fibroblast was reported by Parnpai and co-workers in 1999 [163,164]. The production of a swamp embryo by interspecies SCNT (between swamp buffaloes and cattle) was reported in the following years [165-168].

The birth of buffalo calves from somatic cloning was announced by a Chinese research group from Guangxi University in 2005 and by a Philippines research group at Carabao Centre in 2007. However, all these calves died soon after birth. The studies on somatic cloning in river buffalo have been carried out by Indian research group at NDRI in 2009. Repeat experiments have been undertaken for river buffalo and five live cloned calves were born in India in 2013 [169].

Attempts to improve somatic cloning efficiency in swamp buffalo using the combination of parthenogenetic activation and starved fetal fibroblasts was reported by Saikhun et al., [170]. The rates of cleavage and the percentage of blastocyst development of the embryos derived from starved cells were considerably higher compared to those of the serum fed cells (35% vs. 21%, respectively). The transfer of cloned blastocysts into recipients resulting in pregnancy, healthy calving and normal development was announced by a research team at Chulalongkorn University, in Thailand, in 2011 [171].

2.5. Cryopreservation

Sperm, oocytes, embryos and other cell types collected from swamp buffaloes have been deep-frozen in liquid nitrogen. In the early work reported by Dass [172] and Avenel [81], the French mini straw technique (I.M.V France) was used for semen freezing in 0.2M Tris-fructose-glycerol-egg yolk or in lactose based extender, resulting in 40-60% motility after thawing and an increase from 30% to more than 69% buffaloes conceiving to a single insemination with frozen semen [170,172]. In a study using different cooling systems, Sukhato et al., [173] demonstrated that the cooling procedures adopted before cryopreservation affect the post thaw survival of spermatozoa and the subsequent pregnancy rates. The pregnancy rates were improved from 28% for cows inseminated with frozen sperm using standard protocol (cooling at 10°C/min upto -40°C before plunging in liquid nitrogen) to 40-43% for cows inseminated with sperm cooled at 20 and 30°C/min up to -80°C or -120°C before plunging the semen in liquid nitrogen. In the recent work of Syahruddin and Tappa [136] the pregnancy rate of 50% and 71% -with successful calvings - were reported for buffaloes receiving treatment for estrus synchronization and AI with frozen sperm.

The successful cryopreservation of swamp buffalo embryos resulting in 23.1% (6/26) calving rate was reported [94,173-176]. The feasibility of cryopreservation of somatic cells and embryos for inter-subspecies cloning in buffaloes was reported by Yang et al., [177], who used river buffalo ear cryopreserved fibroblast and swamp buffalo oocyte cytoplasm for cloned embryo production. The cloned embryos that developed to blastocysts were selected for cryopreservation and embryos transfer. The cryo-survival rate was 30.3% and the transfer of survived blastocysts into recipient buffalos resulted in 33% pregnancy and one healthy calf delivered on day 320 of gestation. The ability of oocytes at germinal vesicle (GV)-stage and M2 stage to survive, mature and fertilize after a freeze-warm cycle using ethylene glycol (EG) was investigated [178].

2.6. Molecular Biotechnology

Buffaloes’ gene map and chromosomes were studied in both river and swamp buffaloes [6] and many molecular techniques have been applied in swamp buffalo for sexing control, genetic selection program and for improving the efficiency of somatic cloning.

A recent study reported that flow cytometric sorting of X- and Y-chromosome bearing sperm was feasible for production of sex-preselected calves in swamp buffalo. Results of calving with more than 85% females and 15% males calves was reported for buffaloes inseminated with preselected semen [179,180].

The work aimed at developing genetic markers specific for the calving interval (CI) trait of the domestic Thai buffalo (Bubalus bubalis) have been undertaken using the Amplified Fragment Length Polymorphism (AFLP) technique [18]. Using two groups of buffalo with extreme CI phenotypes, several AFLP markers were identified of which the AFE7M24-2 marker evidenced a high similarity with the prostaglandin F2-alpha receptor gene and the AFE16M25-2 and AFE4M28-3 markers had high similarity with the follicle stimulating hormone receptor gene [18].

The investigation on transcriptional activity was carried out to understand the developmental failure of cloned buffalo embryos. The results obtained by quantitative real time RT-PCR related to the expression profiles of five genes involved in DNA and histone modifications, DNMT1, DNMT3A, DNMT3B, HAT1 and HDAC1 indicated that the relative levels of DNMT3A and HDAC1 genes had dramatic differences between somatic cloning and in vitro fertilized embryos [181].