Reproductive characteristics of the female grasscutter (Thryonomys swinderianus) and formulation of colony breeding strategies

P G Addo, B Awumbila*, E Awotwi* and N-A Ankrah

Noguchi Memorial Institute for Medical Research, Box LG581, University of Ghana, Legon, Ghana
* Department of Animal Science, Box LG 25, University of Ghana, Legon, Ghana
paddo@noguchi.mimcom.net   ;   adokaley@yahoo.com

Abstract

The reproductive characteristics of 34 female grasscutters (Thryonomys Swinderianus) were investigated to identify common features that could be used to formulate breeding strategies for grasscutters in captivity. The female's ability to manifest an oestrous cycle, mate, conceive, deliver and wean its young in captivity, as well as elicit signs of successful mating, conception and imminent parturition were investigated.

Thirty-one (91.2%) females cycled erratically and mated within 2 weeks of pairing with the male, irrespective of the oestrous phase. Twenty-seven (87.1%) conceived, 2 aborted and 25 (92.6%) gave birth after 148-157 days to 2-6 pups per litter, totalling 98 neonates, which included 9 stillborn. The remaining 89 pups were successfully weaned at 6 weeks of age. All mated females presented post-mating vulval congestion, while those that became pregnant bled intermittently on the 32^nd-38^th day after mating, Eleven (44%) of the pregnant females adopted the 'penguin posture' 2-3 days before delivery. The variations among the grasscutters with respect to all the parameters investigated (except the oestrous cycle length) were statistically not significant; suggesting that a common breeding programme could be put in place for management of grasscutters in captivity. Furthermore, since mating, pregnancy and imminent parturition were accompanied by visible cues, they could be used to enhance the practical management of the breeding colonies.

In conclusion, the overall findings of this study showed that though the grasscutters of the study are still undergoing domestication, they exhibit fairly uniform characteristics and therefore the formulation of a common breeding programme is practicable.

Keywords: Breeding strategies, captivity, grasscutter (Thryonomys swinderianus), reproductive characteristics

Introduction

The grasscutter (Thryonomys swinderianus) is a wild hystricomorph rodent found currently only in Africa (Adoun 1993). It is among several kinds of animals that are used locally and regionally for meat (Anonymous 1993). The grasscutter is widely distributed and therefore exploited in most areas south of the Sahara, particularly West Africa (Mensah and Baptist 1986). It is the preferred and perhaps the most expensive meat in West Africa (National Research Council 1991; Asibey and Addo 2000). The high demand for grasscutter meat and the economic benefit that accrues from its sale has resulted in aggressive hunting with complete disregard for conservation of the species and the environment. Domestication of the grasscutter is being encouraged in West Africa to help address these problems. Progress has however been slow due to paucity of information on its biology (Ewer 1969; Yeboah and Adamu 1995), its nervous temperament and the difficulty in getting it acclimatized in captivity (Hemmer 1993). Recent publications have identified husbandry, health and reproductive biology as the areas that need to be investigated for the successful domestication of the grasscutter (National Research Council 1991; Yeboah and Adamu 1995).

Of the three areas identified for research, reproductive characterisation appears to be the most crucial since the propagation of the species depends on it. Above all, experts on animal domestication have identified the maintenance of reproductive competence in captivity as the cardinal marker of successful domestication (Zeuner 1963; Fox 1987; Dukelow 1978; Adams 1989). The reproductive characteristics noted to be crucial to facilitating and maximizing productivity are: testes size, semen and sperm quality (Betteridge 1986), oestrous cycle, ovulation, mating, pregnancy, parturition, embryonic mortality, foetal mortality, perinatal mortality (Betteridge 1986; Short 1989), gestation length, litter size, birth weight, age at puberty, birth interval, mating system, breeding season and reproductive lifespan (Short 1989). The aim of this study was therefore to characterize female reproductive parameters of the grasscutter, since they constitute a greater part of the aforementioned parameters and identify common features that could be employed in the formulation of breeding strategies to enhance the breeding of grasscutters in captivity.

This paper therefore reports on reproductive characteristics of the female grasscutter specifically, oestrous cycle, mating, pregnancy, parturition, gestation length, litter size, birth and weaning weights, as well as the visual cues that accompany mating, pregnancy and imminent parturition. It also presents recommendations on the use of these reproductive characteristics for the formulation of colony breeding programmes.
 

Materials and methods

Experimental animals

Thirty-four female and 18 male grasscutters were chosen from a colony of wild and laboratory-bred grasscutters maintained in the conventional animal facility of the Noguchi Memorial Institute for Medical Research (NMIMR). The grasscutters of wild origin were of an indeterminate age but had been captured as adults (having 4 cheek teeth), weighed between 1.47-3.62 kg and been maintained in captivity for three months prior to their use in the study. The laboratory-bred stock was 6 -7½ months old and weighed between 1.5-2.4kg.

Animal management

The grasscutters of wild origin were initially quarantined for 1 month and maintained for an additional two months to get them acclimatized to the laboratory animal housing conditions in accordance with the guidelines of the Institutional Animal Care and Use Committee of NMIMR. The housing conditions were: caging in laboratory cages measuring 50cm (H) by 40cm (W), by 40cm (L) for adults and 55cm (H) by 65cm (W), by 70cm (L) for weaners, placed in a room with an ambient temperature and relative humidity ranging between 25-30ºC, and 60-85% respectively. Also provided were a 12-12 hour light-dark cycle by fluorescent lighting and 24 hour natural ventilation. The grasscutters were provided feed and water ad libitum. The feed consisted of rodent pellets, guinea grass (Panicum maximum) and sugar cane (Saccharum spp.) and pre-boiled water was provided in sipper bottles. The laboratory-bred grasscutters were maintained in the same manner as the wild stock. Both the wild and laboratory stock were given fresh provision of feed and water daily and their animal room was swept and disinfected daily with 1% Virkon S (A broadspectrum bactericidal, fungicidal and virucidal disinfectant produced by Antec International Ltd, England. Composition: potassium peroxymonosulfate, sulfamic acid, sodium alkyl benzene sulfonate). Cage trays, feeding troughs and water bottles were washed and autoclaved twice weekly, while cages were scrubbed in hot water monthly. The animals were experimentally manipulated in accordance with the guidelines of the Institutional Animal Care and Use Committee of NMIMR.

Detection of oestrous cycle

The females were individually caged and monitored for three months for manifestation of the periodic vaginal membrane perforation phenomenon, which is the sign of oestrous cycling and sexual maturity in hystricomorph rodents (Weir 1974). The females were placed in the dorsal decubitus position and the perineum examined daily for intermittent rupture and restoration of the vaginal closure membrane. The oestrous cycle length was defined as the interval between the first day of spontaneous vaginal membrane perforation (evidenced by a patent vagina or secretion-sealed opening), through to the period of vaginal closure by an epithelial membrane, to the day before the next perforation (Weir 1974). The females that manifested an oestrous cycle became eligible for inclusion in the study.

Hand-mating and detection of mating

The females chosen for the study were non-pregnant and non-lactating, while the males presented a darkened perineum, the sign of sexual maturity in male grasscutters (Adjanohoun 1989). Grasscutters breed all year round (Asibey 1974); therefore no consideration was given to the time of mating. Hand-mating was conducted by transferring the female to the male's cage. Prior to the transfer of the female into the male's cage, its vaginal status (open, secretion sealed, closed) and date of the transfer were noted, after which it was left with the male until it had been mated. During the female's stay with the male it was examined daily for post-mating perineal changes as described by Addo (2002). On observing any of the mating signs, the female was immediately separated from the male, transferred to its own cage and the day of the appearance of the mating-sign(s) noted.

Detection of pregnancy, imminent parturition and determination of gestation length

Pregnancy diagnosis was conducted daily and weekly for two months by the methods described by Addo (2002) namely daily monitoring for pregnancy-associated intermittent vaginal bleeding and weekly abdominal palpation of developing foetuses in utero. Animals diagnosed pregnant by any of the two methods were observed unobtrusively for signs of imminent parturition as described by Addo (2002) until parturition. Gestation length was computed to be from the day of the appearance of the mating sign (same as day of separation from the male) until the day of parturition.

Determination of litter size, birth and weaning weights

Each mother and her litter were maintained in the same cage from birth to weaning, and information gathered on each litter was recorded next to that of the respective mother. The neonates were temporarily identified with a non-toxic dye (picric acid or gentian violet), counted, sexed and weighed within 24 hours of birth. The mother and her litter were observed daily and each neonate was weighed weekly until weaning at six weeks of age. Thereafter, they were maintained in groups of three until the 5^th month when they were categorized as sub-adults and the males were transferred to individual cages to avoid fighting.

Reproductive performance

The overall reproductive performance (i.e. ability to manifest an oestrous cycle, be sexually receptive to the male, conceive and nurse young to weaning age) of the grasscutters as a unit in captivity was assessed at the conclusion of the study. The objective was to determine the percentage homogeny in the presentation of selected reproductive characteristics which occurred and progressed without veterinary intervention; and as such could be used in the formulation of a uniform breeding programme for grasscutter colonies in captivity. The following reproductive characteristics were assessed:

Percentage of animals that manifested an oestrous cycle during 90 days of observation in captivity

Percentage of animals that accepted to mate within 2 weeks of exposure to a male in captivity

Percentage of animals that became pregnant within 2 weeks of exposure to a male in captivity

Percentage of animals that delivered 148-158 days after mating in captivity

Percentage of females that weaned at least 60% of their babies after 6 weeks of nursing in captivity

Percentage of babies successfully weaned in captivity

Percentage of weaners that survived to the 5^th month in captivity

Percentage of females that elicited post-mating visual cues in captivity

Percentage of females that elicited pregnancy-associated visual cues in captivity

Percentage of females that elicited parturition-associated visual cues in captivity

Data analyses

Statistical analysis was conducted with the Statistical Package for the Social Sciences (SPSS), Standard version, Release 12.0.1 (SPSS Inc. 1989-2003). The case summaries procedure was used to determine the means, standard deviations, minimum and maximum values of the oestrous cycle lengths, gestation lengths, litter sizes and birth and weaning body weights. After conducting the above-mentioned determinations, the oestrous cycle lengths were screened for normality. The one-way analysis of variance (ANOVA) procedure was used to compare the oestrous cycle lengths among the females. The relationship between gestation length and litter size, gestation length and birth weight, litter size and birth weight and birth weight and weaning weight were analysed by the Pearson's correlation coefficient and linear regression, after determining that they were normally distributed and a linear relationship had been confirmed by scatter plot. The overall reproductive performance of the grasscutters as a colony was analysed by the case summaries procedures.
 

Results

Oestrous cycle, mating and detection of mating

Thirty-one (91. 0%) of the 34 females manifested an oestrous cycle. The oestrous cycle occurred erratically in all the 31 females. The mean length of the oestrous cycle was 27.8 ± 22.1 days and differed significantly (P< 0.05) among the females. All the 31 (100%) females accepted mating after 1-14 days of pairing with the males, but irrespective of the oestrous phase (active oestrous phase: open or secretion-sealed vagina; inactive oestrous phase: membrane-closed vagina). Each of the 31 mated females also presented post-mating vulval congestion. The reproductive details are presented in table 1.

Detection of pregnancy, imminent parturition and determination of gestation length

Twenty-seven (87.1%) of the 31 successfully mated females were definitively diagnosed pregnant by abdominal palpation 28 days after separation from the male, in conjunction with detection of intermittent vaginal bleeding on the 32^nd-38^th day after mating. Two females aborted in the 3^rd month of pregnancy as a result of improper handling prior to a cage washing activity. The remaining 25 (92.6%) grasscutters delivered on the 148^th -157^th day after mating. All 25 grasscutters reduced their feed intake and were less responsive to their surroundings in the last 7-11 days before parturition, while 11 (44%) also adopted the 'penguin posture' (Addo 2002) 2-3 days before delivery.

Litter size, sex, birth and weaning weights

The 25 (92.6%) grasscutters gave birth to 2-6 precocious (fully-furred, open eyes, open ears and walking about in the cage) pups per litter. In all, 98 neonates that included 9 stillborn were delivered. All the remaining 89 pups, made up of 46 females and 43 males were successfully weaned at 6 weeks of age. The neonates weighed on average120.5 ± 22.1g and 450.9 ± 42.9g at birth and weaning respectively.

Gestation length was negatively but weakly correlated with litter size (r = - 0.37). The relationship was not statistically significant (P> 0.05) and gestation length could not be used to predict litter size (B: Constant = 22.7, Gestation length = -123, P> 0.05; Regression = 2.69; Residual = 22.0, P> 0.05; R=0.33, R² =0.11, P> 0.05).

Gestation length was positively but weakly correlated with birth weight (r=0.25). The relationship was not statistically significant (P> 0.05) and gestation length could not be used to predict birth weight (B: Constant = -140, Gestation length =1.70, P> 0.05; Regression= 513 Residual = 11,159.9, P> 0.05; R=0.21, R² =0.04, P> 0.05).

There was a statistically significant negative correlation between litter size and birth weight (r = - 0.71, P< 0.01) and litter size could be used to predict birth weight (B: Constant = 180 Litter size = -15. 5, P< 0.01; Regression = 5,882.8, Residual = 5,790.4, P< 0.01; R=0.71, R² =0.50, P< 0.01).

There was also a statistically significant negative correlation between litter size and weaning weight (r = - 0.47, P< 0.05). However, litter size could not be used to model the value of weaning weight since the residual sum of squares was greater than the regression sum of squares (B: Constant = 528, Litter size = -19.9, P< 0.01; Regression = 9,761.9, Residual = 34,428, P< 0.05; R=0.47, R² =0.22, P< 0.05).

There was a strong and statistically significant positive correlation between birth weight and weaning weight (r = 0.85, P< 0.01). Birth weight was found to be a good predictor of weaning weight (B: Constant = 252.9, birth weight = 1.64, P< 0.01; Regression = 31,534.5, Residual = 12,656.1, P< 0.01; R=0.85, R² =0.71, P< 0.01).

The overall reproductive performances are summarized in Table 1.

The detailed data on the reproductive parameters investigated are presented in Table 2.

Discussion

The grasscutters manifested the periodic vaginal membrane perforation phenomenon known to occur in sexually mature hystricomorph rodents (with the exception of the coypu - Myocastor coypus) (Weir 1974). The manifestation of the phenomenon by 91.2% of the females and their subsequent conception (87.1%) suggests that the phenomenon is a dependable parameter for determining the sexual maturity of grasscutters in captivity. Additionally, manifestation of the phenomenon could also be used as a criterion in the selection of breeders; since the phenomenon is unequivocally absent in sexually immature grasscutters (Addo 1997). The erratic manifestation of the phenomenon in all the females confirms the inference that the grasscutter is an induced ovulator, an assertion made from behavioural (Stier et al 1991; Adjanohoun 1993) and ovarian histological studies (Addo et al 2001) of the grasscutter.

Given that all the females were sexually receptive within a few days (1-14 days) of pairing, and their receptivity was irrespective of the oestrous phase suggests that (i) grasscutters (apart from breeding all year round) could be mated at whatever time is convenient to the farmer without reference to the oestrous cycle; and (ii) the permanent mating system commonly practised by grasscutter farmers (Asibey 1974; Yewadan 1993; Adjanohoun 1993) to achieve sexual receptivity and conception could be discontinued, since the practice often results in cannibalism of the neonates by the male (Asibey 1974). As an alternative, the couples could be paired for 2-4 weeks and thereafter separated.

It is advantageous in practical animal production to know when females have been successfully mated, when pregnant, and when delivery is due. The information enables farmers provide appropriate care during pregnancy and prepare adequately towards the day of parturition, thus cutting down on perinatal losses. Confirmation of conception is also important in situations where animals have been mated but fail to conceive, while in the case of a negative pregnancy diagnosis, a quick rescheduling of mating could be done to ensure continued productivity (Arrington 1972). Fortunately, all the grasscutters studied elicited visible cues when mated, pregnant and about to deliver, while a few also adopted the 'penguin posture' when parturition was imminent, suggesting that these signs could be detected to aid in the management of grasscutter colonies. However, since the penguin posture was manifested by less than 50% of the pregnant stock and as such cannot be relied upon in all cases, we suggest that farmers observe their animals closely for signs of imminent parturition (dullness and disinterest in their surroundings), immediately on entering the 5^th month, in order to prepare adequately for parturition.

The findings of this study suggest that females given for mating should be examined once daily for the post-mating sign (vulval congestion) and be separated from the male when detected. However, daily examination may be inconvenient for both the animals and the farmers; therefore since all the receptive females mated within 2 weeks of pairing and bled intermittently from the 32^nd-38^th day, farmers could adopt a system in which the animals are paired for 4 weeks without being examined for post-mating signs. The females are separated after the 4^th week and immediately thereafter examined daily for intermittent vaginal bleeding for 2 weeks. This system would cut down on the number of times the animals are handled, since the 14 day examination period would provide information on both mating and conception. Females that do not bleed intermittently within the 2-week examination period could be serviced again thereafter. In view of the panicky nature of the grasscutter we do not recommend abdominal palpation by farmers without the proper technical know-how, since it could result in abortion.

The gestational length (152.8 ± 2.77 days; 148-157 days) obtained in this study confirmed those established by Mensah and Baptist (1986) who reported a gestational length of 152 ± 2 day; Adjanohoun (1989), reported a gestational length of 156 ± 3 day, Stier et al (1991) obtained a gestational length of 152 days and Addo (2002) reported a gestational length of 152 ± 2.98 days (148-158 days; 5.3 months-5.6 months). Though the long gestation length observed in grasscutters is usual of hystricomorph rodents, it is a disadvantage to production, because with the best of strategies only two deliveries could be obtained per female per year, and worse still, only 4 -12 neonates could be obtained per female within the period. To ensure that two deliveries are obtained per female per year, the nursing period should not be longer than 6 weeks. The benefits of lactation notwithstanding, weaning at 9 or 12 weeks as practised by some farmers (Stier et al 1991; Yeboah and Adamu 1995) is too long and disadvantageous to productivity, since by this practise only one delivery could be obtained per female per year. Moreover, the grasscutter neonates are born precocious like most hystricomorph rodents and as such could be weaned at 3-4 weeks, as is already being practised by some grasscutter farmers (Asibey 1974; Skinner and Smithers 1990; Alexander 1992). However, in view of the high post-weaning survival rate 100% (n = 89) obtained in this study, we recommend that grasscutters are weaned at 6 weeks to enable them obtain the full benefits of lactation (adequate nutrition and protection against infection by way of maternal antibodies (Cowie 1990)) and thereby enhance their post-weaning survival in captivity, especially since the species is still undergoing domestication.

The litter size of grasscutters was inversely related to birth weight, while birth weight was directly related to weaning weight. The inverse relationship between litter size and birth weight shows that large litters are bound to be very small; and taking into consideration the direct relationship between birth and weaning weights, it would be beneficial to foster babies from large litters to nursing mothers with small litters, so that by weaning age they (those from large litters but with small body weights) would have gained enough weight to thrive on their own after weaning. This is important since birth weight has been found to be a good predictor of weaning weight, and therefore if those with low weights are not assisted during the pre-weaning period they could become the 'disadvantaged' in the colony. In view of the need to adopt, we suggest that several females should be mated at the same time in order to synchronize their deliveries/lactation.

In summary:

1. sexually mature female grasscutters manifest the periodic vaginal membrane perforation phenomenon observed in hystricomorph rodents; consequently the phenomenon could be used as a criterion for selecting breeders for use in captivity;
2. grasscutter females are reflex ovulators, breed all year round, are sexually receptive irrespective of the status of the vaginal membrane (oestrous cycle) and would conceive irrespective of the status of the vaginal membrane (oestrous cycle); consequently they could be mated at the convenience of the farmer and without having to know the oestrous cycle;
3. the females elicit visual cues when mated (vulval congestion), when pregnant (intermittent vaginal bleeding) and when parturition is imminent (unrest, disinterest in surroundings, 'penguin posture') and as such these could be used to aid in the management of the colony in captivity. However, since daily examination for post-mating signs as well as daily examination for intermittent bleeding would be inconvenient for both the animals and the farmers, we suggest that the animals should only be examined for intermittent vaginal bleeding, and females that do not elicit the sign should be re-given for servicing;
4. grasscutters have a long gestation length which is disadvantageous to high productivity, therefore nursing of the young should be limited to a maximum of 6 weeks, which would permit the re- servicing of the females and consequently enable the farmer obtain 2 deliveries per year and
5. litter size and birth weight reflect inversely and directly respectively on weaning weight, therefore deliveries should be synchronized to facilitate adoption of babies from large litters to enhance their survival after weaning.

Conclusion

The overall findings of this study show that though the grasscutters of the study are still undergoing domestication, they exhibit fairly uniform characteristics, and in view of the very high reproductive performance (above 90%) obtained in the study, the formulation of a common grasscutter breeding programme for use in captivity is considered practicable. We therefore suggest that the recommendations are incorporated into grasscutter breeding programmes to enhance grasscutter production in particular and the up and coming grasscutter industry in general.
 

Acknowledgements

The authors thank the Japan International Cooperation Agency (JICA) and the Noguchi Memorial Institute for Medical Research for funding the study and also thankfully acknowledge the technical assistance of Messrs Emmanuel Atta Tioh, and David Appiah for the care of the grasscutters.
 

References

Adams C E 1989 The laboratory rabbit. In: Trevor B Poole (editor), UFAW Handbook on the care and management of laboratory animals. 6th Edition. Longman Scientific & Technical, Essex, England. 415-435

Addo P G 1997 Domesticating the wild grasscutter (Thryonomys swinderianus-Temminck, 1827) under laboratory conditions. Ph. D Thesis, University of Ghana. xviii+245 pp.

Addo P 2002 Detection of mating, pregnancy and imminent parturition in the grasscutter (Thryonomys swinderianus); Livestock Research for Rural Development (14) 4: http://www.cipav.org.co/lrrd/lrrd14/4/addo144.htm

Addo P, Dodoo A, Adjei S, Awumbila B and Awotwi E 2001 Determination of the ovulatory mechanism of the grasscutter (Thryonomys swinderianus). Animal Reproduction Science 71/4 -2: 125-137.

Adjanohoun E 1989 Contribution au developpment de l'élevage de l'aulacode (Thryonomys swinderianus-Temminck, 1827) et a l'étude de sa reproduction. Agence de Coopération Culturelle et Technique du Bénin. Para Graphic 31240 L'Union Toulouse, France. 211pp

Adjanohoun E 1993 Quelques aspects du cycle sexuel de l'aulacode (Thryonomys swinderianus Temminck, 1827) et leurs conséquences pratiques sur la conduite des élevages. In: Schrage R and Yewadan L T (éditeurs), 1^ère Conférence Internationale sur l'aulacodiculture: Acquis et perspectives 111- 118.

Adoun C 1993 Place de l'aulacode (Thryonomys swinderianus) dans le regne animal et sa repartition geographique In: Schrage R and Yewadan L T (editors), 1^ère Conférence Internationale sur l'aulacodiculture: Acquis et perspectives 35-40.

Alexander A J 1992 Thryonomys swinderianus (grasscutter/cane rat/vondo) and peoples of Southern Africa. Can this be a mutualistic relationship? Acres Editions, Rabat 171-184

Anonymous 1993 The grasscutter: an African delicacy. In: Livestock for the landless. Spore CTA No. 46 August 1993 p.4.

Arrington L R 1972 Management of laboratory animals. In: The breeding, care and management of experimental animals.141-159. Interstate Printers and Publishers, Danville, Illinois.

Asibey E O A 1974 Reproduction in the grasscutter (Thryonomys swinderianus Temminck) in Ghana. Symposium of the Zoological Society of London No. 34, 251-263.

Asibey E O A and Addo P G 2000 The grasscutter, a promising animal for meat production. In: Turnham D, (editor) African perspectives: Practices and policies supporting sustainable development. (Scandinavian Seminar College, Denmark, in association with Weaver Press, Harare, Zimbabwe).p120

Betteridge K J 1986 Increasing Productivity in Farm Animals. In: Austin C R and Short R V (editors), Reproduction in mammals: Manipulating reproduction 2^nd edition. University Press. Cambridge, 1-47

Cowie A T 1990 Lactation. In: Austin C R and Short R V (editors), Reproduction in mammals: Hormonal control of reproduction. Volume 3. 2^nd Edition University Press. Cambridge 195-231.

Dukelow W R 1978 Ovulation detection and control relative to optimal time of mating in non-human primates. Symposium of the Zoological Society of London No. 43, 195 - 206.

Ewer R F 1969 Form and function in the grasscutter Thryonomys swinderianus. Temminck (Rodentia, Thryonomyidae). Ghana Journal of Science 9: 131-149.

Fox R R 1987 Taxonomy and genetics. In: Weisbroth S H, Flatt R E and Kraus A L (editors), The biology of the laboratory rabbit., Academic Press Incorporated., Orlando, Florida 1-22.

Hemmer H 1993 Domestication: Concept and consequences. In: Schrage R, and Yewadan L T (editors), 1^ère Conférence Internationale sur l'aulacodiculture: Acquis et perspectives 189-202.

Mensah G A, Baptist R 1986 Aspects pratiques en elevage d'aulacode (Thryonomys swinderianus). 1. Modes d'accouplement et durée de la gestation. Revue d'Elevage et de Médecine. Vétérinaire des Pays Tropicaux 39: 239-242.

National Research Council 1991 Grasscutter In: Microlivestock: Little-known small animals with a promising economic future. 233-240. National Academy Press, Washington, D.C. XVII + 449pp

Short R V 1989 Specific differences in reproductive mechanisms. In: Austin C R and Short R V (editors), Reproduction in mammals: Reproductive fitness volume 4, 2^nd Edition. University Press, Cambridge 24-61

Skinner J D and Smithers R H N 1990 The mammals of the Southern African subregion. Pretoria University Press, Pretoria 79-97

SPSS Inc 1989-2003 SPSS 12.0.1 for Windows, Release 12.0.1, 11 November 2003. Developed with input from Lead Technologies Inc. 1991-2000 and Apache Software Foundation 2000. Copyright SPSS Inc. Headquarters, 233 S Wacker Drive, Chicago, Illinois 60600

Stier C H, Mensah G A and Gall C F 1991 Elevage d'aulacodes (Thryonomys swinderianus) pour la production de viande. World Animal Review 69, 44-49 http://www.fao.org/docrep/U5700T/u5700T0j.htm

Weir B J 1974 Reproductive characteristics of hystricomorph rodents. In: Rowlands I W and Weir B J (editors), The Biology of hystricomorph rodents. Symposium of the Zoological Society of London, 34: 265-301.

Yeboah S and Adamu E K 1995 The cane rat. Biologist 42 (2) 86-87

Yewadan L T 1993 Système d'élevage en aulacodiculture In: Schrage R and Yewadan L T (editors), 1^ère Conférence Internationale sur l'aulacodiculture: Acquis et perspectives 131-142.

Zeuner F E 1963 A history of domestic animals. 560pp. London, Hutchinson.