In Spain, several native dairy goat breeds are used, among which the Murciano-Granadina (MG) breed stands out due to its higher census (estimated at about 500,000 animals [1
]) and international diffusion (found in Mediterranean and Latin American countries and Western Africa). Goats of this breed have good milk production and quality records (average of the official milk recording in 2020: 547 kg in 210 days of lactation; 5.47% fat and 3.52% protein [2
]) and are well adapted to semi-intensive and intensive production systems and to being machine-milked once a day [3
]. From a reproductive standpoint, the MG breed does not present a deep seasonal anoestrus. Thus, in the spring months in the temperate latitudes of the southern Iberian Peninsula (36–38° N), the males maintain a high libido and the MG does are completely responsive to the male effect, without artificial manipulation of the photoperiod [4
]. Females present high fertility (usually around 80–90% in females naturally mated during breeding season) and prolificacy (herd averages around 1.8 kids/parturition) [5
]. However, a weak aspect of the MG breed is that kids have a low growth rate, generally between 100 and 140 g/day in artificial rearing [6
]. This could be explained by the limited mature size of the breed (males: 60–70 kg; females: 45–55 kg) [5
], as growth rate in goats is dependent on mature size of a particular breed [9
]. In Spain, at present, MG kids not destined for replacement are usually slaughtered for meat at a very young age, such as milk-fed kids at 35–45 days of age and a live weight of 7–9 kg. Kid meat is a product highly appreciated by consumers in the Spanish goat milk production regions, and the peaks of demand are mainly during the Christmas and Easter holidays. Despite this, farmers tend to consider MG kids as a byproduct and have little interest in raising them for meat, given the low profitability they obtain. This situation is undesirable, as it could lead to the emergence of practices that are incompatible with ethical goat production. By analogy to what happens in other species and goat breeds, the fastest way to improve the growth characteristics of kids and, consequently, to increase the profitability of their breeding for meat, would be the use of a terminal cross with males belonging to a breed with meat aptitude and high growth rate [10
]. This mating would apply only to MG goats that are not going to be the dams of replacement animals, and all crossbred kids should be slaughtered for meat production when they reach the appropriate weight for sale. However, it is surprising that such a cross has never been studied in the Murciano-Granadina breed.
The Boer goat is a meat aptitude breed of South African origin that has spread all over the world to be exploited as a pure breed (body mature weight: 70–80 kg female, 100–120 kg male [11
]), or for terminal crossbreeding with native goat breeds to improve growth and/or carcass quality of kids [10
]. According to the literature, the animals of this breed are fast growers (>200 g/d), hardy and adaptable, disease-resistant, fertile and produce meat of high quality [11
]. In crossbreeding, it has been shown to improve the growth of local breeds in some countries, such as: USA (Spanish breed [15
]), Mexico (Creole breed [17
]), Poland (White Improved breed [18
]), Brazil (local goats [19
]) or China (Chengdu Ma breed [20
On the basis of the cited precedents, we proposed to study the terminal crossbreeding of meat-suitable Boer breed males with MG females in experiments performed on two experimental farms. The aim of this paper is to study the reproductive traits of the dams and the growth of the crossbred kids in artificial rearing, also considering other factors that, according to the literature, could affect the variables mentioned above (reproductive traits: parity number of goats and season [21
]; growth traits: litter size and sex of kids [6
]). Our hypothesis is that this terminal crossbreeding improves the growth traits of the MG kids, without a negative effect on the reproductive traits of MG goats.
2. Materials and Methods
2.1. Experimental Design
Four experiments were carried out using a total of 354 MG goats (106 nulliparous and 248 lactating goats), half of which were mated with MG bucks (n = 12) and the other half with Boer bucks (n = 12; Table 1
). Experiments 1 and 2 were carried out at the experimental farm of the Universitat Politècnica de València (39°28′ N; 0°22′ E) and matings took place in February–March (non-breeding season). Experiments 3 and 4 were carried out at the experimental farm of the Universidad Miguel Hernández in Orihuela (38°05′ N; 0°56′ E), and matings took place in September–October (breeding season; experiment 3) and February–March (experiment 4). Both experimental farms are located in the Valencian Community (Spain) and the annual health checks performed by veterinary services showed that they were free from brucellosis, tuberculosis, Mycoplasma agalactiae
and caprine arthritis-encephalitis virus.
In each experiment the mating groups were organised in the same way: (a) nulliparous goats were separated into two homogeneous groups in terms of weight and age, and each group was randomly assigned to mating with Boer bucks (group 1) or MG bucks (group 2); (b) lactating goats were also separated into two homogeneous groups in terms of parity, live weight and milk production, and each group was randomly assigned to mating with Boer bucks (group 3) or MG bucks (group 4). Each group was located in different pens and no hormonal treatment was applied to either the bucks or the goats. Table 2
summarises the number of bucks and goats used in each of the four groups mentioned, for each experiment.
For each experiment, all the bucks were introduced to the mating groups on the same day, and they were in contact with the females for a total period of 40 days. At 70–75 days after introducing the bucks to the goats, pregnancy was determined by transabdominal ultrasound (5.0 MHz). Abortions, stillbirths, kiddings and litter size were recorded. All the kids were weighed at birth and raised artificially.
A total of 314 kids from single, double and triple parturitions were used for the growth study (145 BO × MG and 169 MG × MG; Table 1
). In experiment 1, kids were housed in pens with 2 or 3 animals grouped according to genotype (crossbred or purebred), sex and birth weight and were fed in artificial rearing in buckets; in 17 pens (8 with MG kids and 9 with BO × MG kids) the consumption of powdered milk was recorded daily to calculate the conversion rate. In the rest of the experiments, kids were reared with automatic suckling machines in separate groups according to genotype. All the kids that participated in the growth study were weighed twice a week until 9 kg live weight (slaughter weight).
2.2. General Management
Does of experimental groups were kept in indoor pens (size = 1.5 m2
/goat; feeder = 0.5 m/goat; 3-bowl water troughs per pen) and received a mixed ration twice daily (at 0900 and 1800 h) throughout the experimental period. The ration was formulated according to [23
] for goats in lactation and gestation using lucerne hay, straw and commercial feed. One month before kidding, does were dewormed and vaccinated (Clostridium perfringens
Types C and D plus Corynebacterium pseudotuberculosis
and tetanus toxoid). Following birth, kids received AD3
E·vitamin (1 mL/kid; Vit AD3
E oral, Industrial Veterinaria S.A., Esplugues de Llobregat-Barcelona, Spain) and were ear-tagged, separated from does and bottle-fed with colostrum for the first 48 h of life. Experimental kids were ad libitum artificially reared with warm (36–38 °C) commercial milk replacer (experiment 1: Nantamilk Supreme, Nanta S.A., Madrid, Spain; experiments 2, 3 and 4: Agno Chevro 63, Celtilait, Ploudaniel, France) reconstituted at 17% (w
). In the first days, kids were trained to suckle from an artificial nipple. Kids were incorporated into the experiments when it was confirmed that they could suckle directly from the nipples. In experiment 1, the reconstituted milk replacer was offered twice daily in containers fitted with teats, and kids were reared in straw-bedded pens (size: at least 0.3 m2
/kid). In the other experiments, milk was supplied with automatic suckling machines (JR AG-21, Industrias JR, Valdelafuente, Spain) and the kids were housed in artificial rearing slatted-floored rooms during the first 20 days of age, and thereafter were reared in straw-bedded pens (0.3 m2
2.3. Measured Variables
Three components of reproductive performance were considered: pregnancy rate (as %: number of pregnant does × 100/number of does exposed to bucks); kidding rate (as %: number of does kidding × 100/number of does exposed to bucks) and prolificacy (number of kids born/number of does kidding). Abortions, goat mortality, stillbirths and kid mortality during rearing were also recorded.
Kids were weighed using a digital dynamometer (KERN® HDB 10K10, readability ±10 g, Masnou, Spain) at birth and every 3–4 days (twice a week) thereafter until reaching slaughter weight (9 kg live weight). For each kid, average daily gain (ADG) and age to reach 9 kg live weight were calculated.
In experiment 1, daily milk replacer intake of each pen was determined by weighing milk supplied and milk refusals. A 50 mL sample of milk refusals was taken daily from each pen to determine dry matter, according to [24
]. This way, the daily dry matter intake of milk replacer was calculated for each pen. Feed conversion ratio for milk replacer was calculated for each pen as kg of dry matter (DM) milk replacer ingested/kg live weight gain. In the other experiments, the total weight of powdered milk delivered with the automatic suckling machines was recorded for each of the experimental groups and feed conversion ratio for milk replacer was calculated globally for each experimental group (crossed kids and MG purebred kids). However, in experiments 2 and 4, it was not possible to calculate the milk powder conversion rate due to accidents that caused losses of reconstituted milk during breeding (disconnection of the nipples).
2.4. Statistical Analysis
Binary variables (pregnancy rate, kidding rate, abortion rates, goat mortality and kid mortality) and ordinal variables (prolificacy: 1, 2 ≥ 3 kids/parturition) were analysed applying logistic regression models with the Logistic procedure of the SAS package [25
], considering the following factors: buck breed (Boer vs. MG), parity (nulliparous vs. lactating goats), experiment (1 to 4) and interactions (buck breed × parity and buck breed × experiment). Regression coefficients from the logistic regression were exponentiated to obtain the odds ratios and corresponding 95% confidence interval associated with each factor. For dichotomous factors (buck breed and parity), an odds ratio significantly higher (or lower) than 1 indicated an increased (or reduced) risk of pregnancy (or other binary variables) if the factor was present. For class factors (experiment), one class of each factor was considered as the reference, and an odds ratio significantly higher (or lower) than 1 for any other class of this factor was indicative of an increased (or reduced) risk of pregnancy (or the other binary variables) when compared to the reference class. For the prolificacy variable, an ordinal logistic regression model was used (proportional odds model).
Quantitative variables recorded in the kids (birth weight, ADG and age up to 9 kg live weight) were statistically analysed using the GLM procedure from the SAS package, considering the following factors: genotype (MG purebred vs. BO × MG) and sex (male vs. female) of the kids, litter size (single, twin and triplets), parity of the goats (1 and ≥2), experiment (1, 2, 3 and 4), the double interactions genotype × sex, genotype × litter size, genotype × parity and genotype × experiment and the triple interaction genotype × sex × litter size. With the effects that were significant (with more than two levels), pairwise comparison of means were made applying Student t-test. Finally, the consumption and the conversion ratio of powdered milk of 17 pens in experiment 1 (4 and 4 pens for MG male and female kids, respectively; 5 and 4 pens for BO × MG male and female kids, respectively) were analysed with the GLM procedure from the SAS package, considering the fixed effects of genotype (MG purebred vs. BO × MG), sex (males and females) and the genotype × sex interaction.
In this work, it has been shown that the use of Boer bucks instead of MG bucks for the mating of Murciano-Granadina goats causes a decrease in the fertility (pregnancy rate: −8.1 percentage points; kidding rate: −13.7 percentage points) and prolificacy (−0.2 kids/parturition) of the does. In addition, the greatest differences occurred when mating was carried out during seasonal anoestrus (experiments 1, 2 and 4). In breeding season, other authors [26
] did not find that the use of Boer bucks affected the fertility or prolificacy of a local breed (Spanish). However, we do not have any similar available work carried out in non-breeding season. On the other hand, the average kidding rate for MG goats (mated with MG bucks) in anoestrus season obtained in this work (78.62%) was similar to that obtained by [27
] (75%) and greater than that obtained by [28
However, we must point out that in experiment 4 the worst reproductive results were obtained, both with Boer bucks and MG bucks, but the cause is not clear. For example, in experiments 1 and 2, both also in the non-breeding season, the buck:doe ratio was similar to experiment 4, but fertility was higher.
In the temperate latitudes of the southern Iberian Peninsula, MG bucks are characterised by a high libido, even in the spring months [29
], and in non-breeding season the MG does are completely responsive to the male effect [4
]. In this regard, we can highlight that in our experiments it was observed that Boer males, mostly of German origin, showed a lower libido and interest in MG females than MG males did, especially in anoestrus season. This suggests that at the latitude at which the experiments were carried out (38–39° N), Boer males have lower sexual activity in non-breeding season and would have a lower capacity for stimulating the ovulatory process in MG goats. This hypothesis could explain, at least in part, the lower fertility of MG does mated to Boer bucks in the three experiments conducted in non-breeding season. A second hypothesis would be that Boer males suffer a greater deterioration of semen quality in non-breeding season. In MG bucks, some authors [30
] have found that the semen quality of MG males shows a seasonal variation, but even in non-breeding season it is of sufficient quality to guarantee adequate fertility on the farm. In South Africa, the libido and semen quality of Boer bucks also varies seasonally [12
]. However, at our latitude (Mediterranean area), there is no information available on the seasonal variation of semen quality of Boer bucks.
The lower prolificacy obtained with Boer males, compared to MG males, was due to a decrease in births with three and more offspring (14.7 vs. 22.4%), and an increase in births with one offspring (30.3 vs. 20.9%), the parturitions with twin births being similar (55.0 vs. 56.7%). We should point out that this result is not entirely negative, as MG breeders do not usually want their goats to have a high proportion of births with triplets, and especially quadruplets. In these cases, it is more common for goats to exhibit metabolic disease (pregnancy toxaemia) and increased mortality in pre-partum [33
]. In addition, the MG kids are born with an excessively low birth weight, which generally leads to poorer adaptation to artificial rearing, slower growth rate [35
] and more disorders and mortality during rearing [36
In any case, the fact that Boer males showed poorer reproductive results in anoestrous season, compared to MG males, should not be a serious drawback, as they could probably be improved by applying stimulation treatments prior to mating, such as artificial manipulation of the photoperiod [37
] or melatonin implants [28
]. Nevertheless, this hypothesis should be confirmed in new studies.
In the four experiments carried out, the advantages of crossbred kids over pure MGs became evident. First, both the birth weight and the ADG of crossbred kids were higher than those of pure MG kids, and this was observed in both males and females and in the three types of parturition (single, double and triple) studied. The increase in birth weight and/or in ADG in the crossbred kids has also been observed in other works in which terminal crossbreeding of Boer bucks was performed in goats of local breeds of sizes similar to the MG breed (females of 30–50 kg adult weight: Spanish in USA [15
], local breed from Mexico [17
], White Improved in Poland [18
], local breed from Brazil [19
], Feral in Australia [39
]). In contrast, the crossbreeding of Boer bucks with goats of larger adult size, such as Alpine or Saanen, scarcely improved the ADG of the crossbred kids [40
]. On the other hand, the average birth weight obtained in this work in the MG purebred kids (2.36 and 2.14 kg for males and females) is similar to that cited by other authors in this same breed (ranging from 2.03 to 2.8 kg [6
]). Likewise, the average ADG value (113 g/day) of the pure MG kids in artificial rearing falls within the range 96–125 g/d described in other works [7
], but is lower than that cited by [6
] (132 g/d), [42
] (159 g/d) and [43
] (153 g/d). Another interesting aspect of our study is that the ADG (both crossbred and purebred kids) in experiment 1 was higher than in other experiments. This result could have been, at least in part, because in experiment 1 the rearing system used (pens with two or three kids, artificial rearing in buckets, different commercial milk replacer) was also different from the other experiments.
Secondly, we can point out that the operators in charge of artificial rearing expressed their preference for crossbred kids, as they usually had to spend less time training them to drink milk from the teats in their first days of life. In addition, the mortality rate in artificial rearing for crossed kids was lower than that of the MG purebred kids. This last result could be explained, at least partially, by the higher birth weight of the crossbred kids and the lower frequency of triple and quadruple births, as in general, mortality is higher in kids with low birth weight [44
Finally, in crossbred kids, the milk replacer conversion rate improved by 16% (average experiments 1 and 3), which can be explained by the fact that these kids also showed a higher growth rate. This relationship between growth rate and feeding efficiency was also observed by [15
] in post-weaning growth of kids. Other authors [16
] also found that the Boer × Spanish cross kids exhibited superior feed efficiency (13%) during the pre-weaning period compared to Spanish kids. Based on the average data obtained in this work, we can estimate that a BO × MG crossed kid consumes 7.45 kg of powdered milk from birth to 9 kg ((9–birth weight) × conversion rate = (9–2.79) × 1.20), while a purebred MG kid consumes 9.72 kg ((9–2.25) × 1.43). Therefore, crossbred kids consume 2.20 kg less milk powder, which means a saving of around 6.16 euros per kid (2.80 euros/kg milk replacer).