Livestock Research for Rural Development 13 (3) 2001

Citation of this paper

Reproductive performances of Creole goats in Guadeloupe (French West Indies)
1. Station-based data

  G Alexandre, G Matheron*, P Chemineau** , J Fleury 
and A Xandé

  Unité de Recherches Zootechniques, INRA Antilles-Guyane, Domaine Duclos, Prise d'Eau  97170 Petit-Bourg, Guadeloupe, French West Indies
Telephone: 0590255933; Fax: 0590255936; 
* Président du Centre CIRAD (Centre International de Recherche Agronomique pour le Développement) de Montpellier, Avenue Agropolis 34398  Montpellier
** Directeur de l’UMR (Unité Mixte de Recherche) INRA-CNRS-Université de Tours Physiologie de la reproduction et des comportements 37380 Nouzilly 


Reproductive performances of the Creole goat, the meat breed of Guadeloupe in the Caribbean, are described from a database generated for 15 years (1973 to 1988) at an experimental farm of the Institut National de la Recherche Agronomique (INRA, Guadeloupe, French West Indies). The data contain information on 599 females, 2259 attempted matings, 2135 achieved matings and 1904 kiddings. Animals were reared under semi-intensive management conditions (on the basis of feeding system, health and culling policy). Main sources of variation were analysed: mating type (buck in permanence or buck effect), mating season (dry, intermediate and humid season) and rank of kidding. The Creole goat is a continuous breeder. On average 94.5 percent of exposed females were mated (defined as mating rate) and 90.5 percent of exposed females achieved a kidding (fertility rate). Their prolificacy reached 1.98 kids/kidding, with only 3.9 percent of kids born dead. These reproductive parameters varied essentially with rank of kidding but very poorly with the mating season. Productivity of these females was regularly high during their whole lifetime production. From the results obtained, it is concluded that  Creole goats have very high reproductive performances, provided that animal husbandry and nutrition are adequate.

Keywords : Creole goat, fertility rate, litter size, neonatal mortality, buck effect, humid tropics, Guadeloupe


In the Caribbean, as in many other tropical countries, goats are mainly raised under the suckling system for meat production (Devendra and Burns 1983; Alexandre et al 1991). In spite of the fact that the goat is of paramount importance to smallholders and to the local meat markets, only a few isolated attempts at improving its production through controlled management have been made. In addition, little attention has been paid to the indigenous tropical breeds (Taneja 1982; Devendra and Burns 1983; Le Gal and Planchenault 1993), that have been considered adapted to the tropics but of low productivity.

It is generally reported that increasing the reproductive rate and/or reproductive performances are important ways of improving meat production in the tropics (Terril 1983; Wilson and Light 1986). Therefore, an assessment of the general reproductive characteristics of native breeds is necessary prior to developing strategies aimed at improving meat supplies. The assessment of the production potential requires a large number of animals, raised over several years and under controlled management.  Only a few studies on the tropical goat’s reproductive performances have been reported under experimental station conditions (Kochapakdee et al 1994; Montaldo et al 1995).

In Guadeloupe, the local meat breed is called the “Creole goat” as it is called in other Caribbean regions. Recent studies in the field of genetic markers indicated that it was derived from the West African Dwarf Goat (Pepin 1994). Experiments have been carried out in Guadeloupe on the Creole goat evaluating reproductive physiology (Chemineau 1986) and preweaning performances (Alexandre et al 1999). The objectives of this present study were to determine the reproductive potential of the Creole goat and to examine the main factors responsible for variations observed.

Materials and methods

This study was based on data generated from 1973 to 1988 (15 years) in the Animal Production Unit (APU) at the INRA Research Centre in Guadeloupe. Guadeloupe is a humid tropical island of the Caribbean area (16.1° N; 61.6° W). The experimental farm of the INRA-APU is located in the dry zone. Annual rainfall averages 1280 mm, with a marked dry season from January to July (less than 70 mm per month). Maximum air temperature varies from 27°C (January) to 32°C (August) and the minimum  from 21°C to 25°C, respectively. The relative humidity is usually above 70% and the day length ranges from 11 h to 13 h.

Animals and their management

All the females used in the experimental flock were Creole goats, which have been described as small-sized animals by Chemineau et al (1984). The adult females had a mean (± SD) liveweight of 25 ± 7 kg and a mean (± SD) wither height of 51 ± 4 cm. Initially, animals were maintained under extensive management conditions from 1973 to 1979. This involved grazing dry pastures, no supplementation and limited preventive health management. Males were permanently kept with the herd. The stock was built up from 1979 with animals coming from different zones of Guadeloupe. The flock size increased from about 50 to 100 does. Since 1979 semi-intensive reproductive management has been practised, achieving a frequency of three kiddings in two years and three mating seasons per year (March, July and November). From 1979 to 1983 several experiments took place in the field of reproductive performance and sexual activity of the does and bucks (Chemineau 1986). Since 1984 the research program has dealt with feeding of the does during the suckling period; and the goat flock has expanded from 120 to 150 does.

From 1979 to 1988, the buck effect was systematically used for inducing ovulatory and oestrus activities. At each season, after more than three weeks of complete separation from females, vasectomised males, harnessed with raddle blocks, were introduced in the doe flock at a ratio of one male to ten females. The mating period lasted for 45 days from 1979 to 1983 and lasted for 30 days from 1984 to 1988. Every day, freshly marked females were exposed to a specific male chosen according to a planned mating. Hand mating was then conducted once per day, from 1979 to 1983 and then twice per day, from 1984 to 1988.

Dipping was done every 2 weeks for tick control, while anthelminthic drenchings were carried out every month for the suckling kids and every 1.5 to 2 months for the weaned animals and adults, from 1979 to 1988

Feeding management

Year round grazing was on Digitaria decumbens pastures, which were rotated at 35 days age of regrowth. These pastures were irrigated after 1984 and were fertilised with 300 kg N/ha/y. As regards supplementation, no supplementary feed was provided from 1973 to 1979. After 1979, reproductive and suckling does were supplemented with commercial pellets (10.3 MJ ME and 180 g CP per kg DM) which were supplied at various levels depending on the experiment.

Data base and statistical analysis

Regular animal performance monitoring was carried out and this allowed the development of a database containing information on 599 females, involving 2259 attempted matings, 2135 achieved matings and 1904 kiddings. Unknown reproduction management occurring during the first four years and management with buck running permanently with the flock represented 26% of the total of attempted matings. At each mating season, the number of females exposed to teaser males was noted. At every reproductive event, i.e. oestrus, mating, or service, dates and the number of the animal involved were recorded. If abortion occurred, this was recorded along with the size of the aborted litter. At birth, animals were ear tagged, then the number of the parents, the sex and number of stillborn and living kids were collected. From 1973 to 1979 and for the 20 % of the foundation flock, the females had unknown birth date, then age at first kidding was not estimated. So intervals between different reproductive events have been calculated as the interval related to the first kidding; this term has been defined as the reproductive age of the doe. Rank of kidding has been defined as the parity of the doe. Mating rate (MR) has been calculated, for each mating season, as the ratio of the number of matings to the number of exposed females. Fertility rate (FR) is the frequency of exposed females giving birth to kids (born dead or alive). Between two successive kiddings, the number of matings which were necessary to induce a kidding were added and named number of services per kidding. The litter size (LS) was defined as number of total born kids reported on total kidding does. General linear model (SAS 1988) procedures were used to adjust data to the following sources of variation: year, experiments nested to the year, mating season, mating type combined with year and rank of kidding.


Mating type effect

During three successive years the two mating types, buck effect and buck running permanently with the flock, were used for different experiments in two different groups. The main results are reported in Table 1. The mean (± SD) fertility rate varied from 86 ± 4 % to 90 ± 3 %, while the mean (± SD) kidding intervals ranged from 284± 28 days to 246 ± 10 days and the mean (± SD) litter sizes were 1.95 ± 0.13 and 1.80 ± 0.05 total kids born alive, respectively. The kid mortality at birth was 0.02 to 0.04 %.

Table 1:  Reproductive performances of the Creole goat at the INRA Guadeloupe according to the year and the mating type: M1, buck effect and M2, buck kept permanently with the flock (definitions in the text)

Mating year




Mating type







Number of does       93       27      151        71      157        30
Mating rate * (%)       94a       84b       84A      100B       92A      100B
Fertility rate ** (%)       82a       88a       85 A        95 B       90a        88b
Kidding interval (days)      319A     223B      262a      252a     270a      254b
Kids born alive      2.02A    1.75B     2.00a     1.82b     1.76a     1.82a
Kids born dead      0.03    0.04     0.03




* Mating rate = ratio of the number of matings to the number of exposed females
** Fertility rate = frequency of exposed females giving birth to kids (dead or alive) 
Values within same row and same year with different superscripts are significantly different: a and b for P<0.05 and for A and B for P<0.01
Mating season effect

Number of exposed females, matings and kiddings were generally at the same level for the three mating seasons (Table 2). Average mating and fertility rates were 95.0 and 90.5 %, respectively. Kidding interval was 267 days and number of total kids born per kidding varied from 1.92 to 2.00. During mating in the intermediate season of July, lower services per kidding (1.06) and higher fertility rates (92.7 %) where observed than in the two other seasons. No other mating season effect upon reproductive parameters was observed.

Table 2:  Reproductive performances of the Creole goat at the INRA Guadeloupe according to the mating season (adjusted data for mating type and year): mating and fertility rates (%), number of services per kidding, kiddings interval (days) and  prolificacy at birth (number of kids born alive or dead)
Parameter\Mating season*





Number of joinings





Number of matings





Number of kiddings





Mating rate ** (%)





Fertility rate ** (%)


92.7 b

89.6 a


Number of services ***

1.09 a

1.06 b

1.08 a


Kidding interval (days)





Number of kids born alive





Number of kids born dead





* Dry season (mid March), intermediate season (mid July) and rainy season (mid November)
** Mating rate = ratio of the number of matings to the number of exposed females; fertility rate = frequency of exposed females giving birth to kids (dead and/or alive)
*** Number of services per kidding = total number of matings, between two successive kiddings, necessary to induce a kidding
Values within same row with different superscripts are different P<0.01
Mating year effect

The buck effect was applied between 1979 and 1988 with a 2-year break in 1982 and 1983 where other experimental techniques were conducted. The year effect was studied for this mating type. Significant differences (P<0.01) occurred from one year to another for mating and fertility rates (Figure 1a).  

Figure 1a: Yearly variations of reproductive performances of Creole goats in Guadeloupe: mating rate (ratio of the number of matings to the number of exposed females, %) and fertility rate (frequency of exposed females giving birth to kids, dead and/or alive, %)   

Mating rates ranged from 84 % to 98 %, while fertility rates varied from 82 % to 96 %. Except for the years 1979 to 1981, when specific experiments on reproduction occurred, kiddings interval decreased from 280 to 245 days and litter size was regularly distributed around 1.9 kids born alive (Figure 1b).  

Figure 1b: Yearly variations of reproductive performances of Creole goats in Guadeloupe: kidding interval (days) and  litter size at birth (number of kids born alive / kidding)
Reproductive history of the doe

The flock comprised more than 50 % of the total females having achieved 3 kiddings and having an 18 month-reproductive age (Table 3). Almost one third of the flock (29 % females) had their rank of kidding equal to or higher than 6 kiddings with a reproductive age reaching 43 or 52 months, respectively.

Table 3:  Creole goats distribution according to rank of kidding and age of reproduction of the does
Rank of kidding







³ 7 *

Number of kiddings








Frequency (%)








Reproductive age **








* Rank of kidding equal or higher than 7  
** Reproductive age is calculated in months as the interval between one given kidding to the next kidding (definitions in the text)

The mating and the kidding rates (Figure 2) ranged from 91 % and 76 % to 93 % and 82 %, from the first to the fifth rank of kidding. These figures then declined regularly to 85 % and 68 % for the last rank of kidding, respectively.  The litter size at birth increased from 1.65 to 2.35 total kids born, from the first to the seventh kiddings, respectively (Figure 2). The values for the number of kids born alive were 1.52 and 2.20, respectively. The difference between the total number of kids born and the number of kids born alive (0.05) remained stable during most of the reproductive life of the doe, except for the first kidding (0.13) and for the last rank of kidding (0.15).

Figure 2: Kidding rate (%) and litter size (number of kids born alive or total kids born / kidding) of Creole goats in Guadeloupe according to the rank of kidding

Variations in the mating and fertility rates according to the kidding interval are represented in Figure 3. The best results were observed for intervals ranging from 225 to 280 days. The mating and fertility rates reached an optimum of 97 % and 92 %, respectively. For longer post-partum intervals, the fertility rate decreased (P<0.01), with more than 10 points of difference for intervals equal or higher than 500 days. 

Figure 3:  Mating and kidding rate (%) of Creole goats according to the kidding interval

The relationships between performances of two successive kiddings of the same doe are indicated in Table 4. The litter size (number of kids born alive) of a doe during it’s kidding number "n + 1", increased with its litter size during its kidding number "n". No significant differences were observed for the number of still born kids or for the fertility rate.

Table 4:  Reproductive performance of Creole goats in Guadeloupe during their kidding number "n+1" according to their litter size during their kidding number "n": mating and fertility rates (%) and  prolificacy at birth (number of alive and dead born kids)
Litter size at kidding number n (alive born kids)




³ 3 *

Number of kiddings 43 375 845 208  
Mating rate ** (%) 89 96 93 98 P<0.01
Fertility rate ** (%) 82 90 90 91 NS
Litter size at kidding number n +1  (kids born alive) 1.82 1.86 2.03 2.20 P<0.01
Litter size at kidding number n +1 (kids born dead) 0.05 0.06 0.05 009 NS
*  Litter size equal or higher than 3 kids/kidding  
** Mating rate = ratio of the number of matings to the number of exposed females; fertility rate = frequency of exposed females giving birth to kids (dead and/or alive)


We observed classical effects of year and rank of kidding on almost all the variables studied; on the other hand there was a significant year*season interaction. From one year to another, different management conditions occurred and that is the reason why analyses have been carried out on separate parts of the database. Bad management conditions occurring from 1973 to 1979 (as described in the materials and methods section: grazing dry pastures, no supplementation and limited preventive health management) induced low and irregular performances such as 80 % and 198 % for fertility and prolificacy rate, respectively. In the second phase from 1979 to 1983, all reproductive parameters were improved (87 and 203 %, respectively), owing to adequate concentrate supply levels. In fact, good nutritional levels of the does appeared to be responsible for the high reproductive performances observed, which agrees with what was stated by Lindsay et al  (1993) and by Walkden-Brown and Restall (1996). In the third phase from 1984 to 1988, irrigated grazing conditions and good reproductive management promoted equally good results (93 and 191 %, respectively) with short kidding intervals and very few seasonal variations.

In many parts of the Tropics, natural mating in goats generally occurs with bucks that run freely with the does all year round or during the once-a-year breeding season.  The sudden introduction of bucks induces fertile cyclic reproductive activity in seasonally anovulatory does (Chemineau 1987). This technique, termed the male effect, has considerable potential for the induction of early breeding, the synchronisation of breeding activity, and possible enhancement of the ovulation rate. The buck effect was very efficient for the Creole goat under our tropical conditions for the following reasons: (i) Creole goats are known to have low seasonality of sexual activity (Chemineau 1983), similar to breeds which originated from tropical and subtropical latitudes (Delgadillo et al 1997); in fact, at tropical latitudes responsiveness to photoperiod is lost or masked and most indigenous breeds of goats are able to exhibit oestrus cycles year-round; and (ii) in these environments nutritional and social factors are the major regulators of reproductive function. In our experimental conditions, does are mated after a flushing period is carried out thus ensuring they have good body condition, as recommended by Lindsay et al  (1993). Moreover, the mating period lasts for a sufficient time (30 to 45 days of contact) with a good proportion of teaser males (10 to 15 %) being present.

Information on reproductive performances of does based on the systematic use of the buck effect is limited. Nevertheless, the performance of the Creole goats seems to be exceptional. Kidding intervals were short and regular allowing for intensive reproduction management which resulted in high kidding frequencies (3 kiddings within 2 years) followed without failures along the whole reproductive career of the does. The fertility rate was  higher than 90 %, comparable with reports for the best tropical breeds such as the West African Dwarf  in Africa  (Osuagwuh 1992), the Kambing Katjang in Asia (Devendra and Burns 1983), the Thai goat reared under intensive feeding (Kochapakdee et al 1994), the Criollo of Venezuela (Gonzalez-Stagnaro 1983) and the Nubian in Northern Mexico (Mellado et al 1991). The frequency of multiple births was remarkably high (greater than 80 % for all sources of variations). Devendra and Burns  (1983) in their exhaustive review on goat reproduction  evaluated the Creole goat as a highly prolific breed. Chemineau (1983, 1986) has studied its reproductive abilities on the physiological point of view. Our work, based upon long-term animal monitoring, confirms the high reproductive potential of this local and hardy breed. The prolificacy rate (up to 1.98 kids/kidding) is higher than that of other native or tropical breeds reared in Latin America (Gonzalez-Stagnaro 1983) or for Nubian goats in Mexico (Mellado et al 1991) and for local Mexican goats even those up-graded by exotic breeds (Montaldo et al 1995). Our value is in fact similar to the litter size of 1.93 to 2.25 for the famous Boer goat reported by Casey et al in 1988.

The mortality rate observed at birth was very low as compared to the frequently high level recorded for goats in adverse environments (Morand-Fehr et al 1984; Sherman 1987). Only 3 % of the kiddings observed had zero live kids while for 95 % there were no stillborn kids. In tropical conditions mortality varies according to the viability of the new-born kids  (birthweight and adaptability to post-natal period, as stated by Morand-Fehr et al 1984 and by Hussain et al 1995), to nutritional level of the does or to their milk yield (Alexandre 1991; Osuagwuh 1992; Hussain et al 1995) and to disease constraints that are numerous. So this trait can be considered as very useful for this hardy breed, and indicates their very good reproductive abilities. On average Creole goats gave birth to a total of 14.2 live kids from their first mating to their last kidding, 53 months later, i.e. 3.22 live kids per year of reproductive life. It seems that does can be culled after 6 kiddings thus giving the opportunity to apply a 20 % culling rate.


Creole meat goats exhibit very good reproductive abilities - continuous breeding with high conception rates - provided nutrition and health preventive management are adequate, which is not very frequent in tropical conditions. The analyses of these data from a Guadeloupean research station, nonetheless, show the significant effect that management exerts on reproductive potential of a hardy native breed such as the Creole goat. Kidding interval has been successfully reduced while fertility and mating rates have been increased by imposing a regular and efficient "buck effect", and culling policies based on poor reproductive performances. Litter size at birth was remarkably high while neonatal mortality was low. Thus there exists considerable scope for improving goat meat production in the tropics, as we know that the most important requirement of a meat-producing animal is a consistently high rate of reproduction. On the other hand, this reproductive potential also makes a strong evidence for goats in the tropics as a milk producer as was argued in the extensive review of Knights and Garcia (1997).

The main non-genetic effects have been estimated and this allows a well-documented characterisation of this native breed. More statistical analysis is now being conducted to determine the genetic parameters of the productive traits in order to define a breed improvement program. Further studies are required in order to assess levels of reproductive performances and factors of variation under farm conditions.


The authors express thanks to H Borel, V Gartizer, H Mogne-Mali, G Paul-Urbain and to D Renard for their collaboration and technical assistance. They are also very grateful to Dr Gary Garcia of the University of the West Indies for the consistent work done in reviewing the manuscript.


Alexandre G 1991 Elevage présevrage des chèvres Créoles. Revue Elevage Medecine Veterinaire Pays tropicaux (special issue), 45 - 64

Alexandre G, Aumont G, Mainaud J C, Fleury J and Naves M 1999 Reproductive performances of Guadeloupean Creole goats during the suckling period.  Small Ruminants Research 34, 155-160

Alexandre G, Borel H, Matheron G and Remy C 1991 Elevages caprins en Guadeloupe . Revue Elevage Medecine Veterinaire Pays tropicaux (special issue), 27-40

Casey N H and Van Niekerk WA 1988 The Boer goat: 1 Origin, adaptability, performance testing, reproduction and milk production. Small Ruminant Research 1, 291-302

Chemineau P 1983 Effect on oestrus and ovulation of exposing Creole goat to the male at three times of the year. Journal of  Reproduction and Fertility 67, 65-72

Chemineau P 1986 Influence de la saison sur l'activité sexuelle du cabrit Créole mâle et femelle. Thèse de doctorat UST Languedoc, pp 63

Chemineau P 1987 Possibilities for using bucks to stimulate ovarian and oestrus cycles in anovulatory goats - a review. Livestock Production Science 17, 135- 147

Chemineau P, Cognie Y, Xande A, Peroux F, Alexandre G, Levy F, Shitalou E, Beche JM, Sergent D, Camus E, Barre N and Thimonier J 1984 Revue Elevage Medecine Veterinaire Pays tropicaux 37, 225-238

Delgadillo J A, Malpaux B and Chemineau P 1997 La reproduction des caprins dans les zones tropicales et subtropicales. INRA Productions  Animales 10, 33-41

Devendra C and Burns M 1983 Goat production in the Tropics. Commonwealth Agricultural Bureaux pp 183

Gonzalez-Stagnaro C 1983 Comportamiento reproductivo de las razas locales de rumiantes en el Tropico Americano In Reproduction des ruminants en zone tropicale.(Editors:  Chemineau P Gauthier D, Thimonier J) 8-10 Juin 1983 Pointe-à-Pitre Guadeloupe INRA Publications n° 20 , 1-80

Hussain S S, Horst P and Islam A B M M 1995 Effect of different factors on pre-weaning survivability of Black Bengal kids. Small Ruminant Research 18, 1-5

Knights M and Garcia G W 1997 The status and characteristics of the goat (Capra hircus) and its potential role as a significant milk producer in the tropics: a review. Small Ruminant Research 26, 203-215

Kochapakdee S, Pralomkarn W, Saithanoo A, Lawpetchara A and Norton B W 1994 Grazing management studies with Thai goats. II Reproductive performances of different genotypes of does grazing improved pasture with or without concentrate supplementation. Asian-australasian Journal of Animal Science 7, 563-570

Le Gal O and Planchenault D 1993 Utilisation des races caprines exotiques dans les régions chaudes Contraintes et intérêts. CIRAD-EMVT pp 261

Lindsay D R, Martin G B and Williams I H 1993 Nutrition and reproduction In: Reproduction in domesticated animals (Editor: King G J).  World Animal Science Series, 459-491 Elsevier Science Publishers Amsterdam

Mellado M,  Foote R H and Gomez A 1991 Reproductive efficiency of Nubian goats throughout the year in Northern Mexico. Small Ruminant Research 6,151-157

Montaldo H, Juarez A, Berruecos J M and Sanchez F 1995 Performance of local goats and their backcrosses with several breeds in Mexico. Small Ruminant Research 16, 97-105

Morand-Fehr P, Vilette Y and Chemineau P 1984 Effets des facteurs non pathologiques sur la mortalité des chevreaux. In Proceedings les maladies de la chèvre Niort France 9-11/octobre Les colloques de l'INRA 28, 31-46

Osuagwuh A I A 1992 Effects of strategic feed supplementation during pregnancy on birthweight and perinatal survival of West African Dwarf kids. Journal of Agricultural Science Cambridge 119, 123-126

Pépin L 1994 Recherche de polymorphisme génétique chez les caprins Applications à l'étude de la diversité des populations au contrôle de filiation et à la résistance génétique à la cowdriose Thesis Paris XI pp

SAS 1988 SAS user's guide Statistics. SAS Inst Inc Cary NC

Sherman D M 1987 Causes of kid morbidity and mortality: an overview In: Proc IVth International Conference on Goats Brasilia Brazil Vol I, 335-354

Taneja G C 1982 Breeding goat for meat production.  Proceedings of the IIIrd International Conference on Goat Production and Disease Tucson Arizona 10-15 January 1982, 27-30

Terril C E 1983 Ruminant meat production in the tropics. In: Reproduction des ruminants en zone tropicale (Editors: Chemineau P, Gauthier D and Thimonier J). INRA Publications 20, 391-398

Walkden-Brown S W and Restall B J 1996 Environmental and social factors affecting reproduction. In VIth International Conference on Goats. 6-11 May 1996 Beijing China Vol 2, 762-775

Wilson R T and Light D 1986 Livestock production in Central Mali: economic characters and production indices for traditionally managed goats and sheep. Journal of Animal Science 62, 567-575

  Received 2 May 2001

Go to top