The Fat-tailed Dunnart
Sminthopsis crassicaudata



Establishment and Maintenance of a Laboratory Colony


J.H. Bennett1, Meredith J. Smith2, R.M. Hope1 and C.M. Chesson1.

1Genetics Department, University of Adelaide, Adelaide.
2Zoology Department, University of Adelaide, G.P.O. Box 498, Adelaide, S. Aust. 5001




Line drawing courtesy Sue Stranger.
Click on thumbnail to enlarge

The Fat-tailed Dunnart Sminthopsis crassicaudata (Gould) is an insectivorous marsupial widely distributed in southern Australia. Adults weigh about 14 g and have a head-and-body length of about 90 mm and tail length about 60 mm. Martin (1965) bred this species in the laboratory and recognised its potential value as an experimental animal.    Females are polyoestrous, with an oestrous cycle length of 31. I standard error0.7 days(n = 25) and a gestation period of 13-16 days. A maximum of 10 young can be suckled, although more than 10 may be born. The young are suckled for about 70 days and the mother has an oestrous 1-2 days after suckling ceases. Hence a second litter may be born 82-90 days after the first. Young females may have their first oestrous when 115 days old (Godfrey and Crowcroft 1971; Ewer 1975).


The first recorded breeding of S. crassicaudata in captivity was in 1962 by G. Gregory, a medical practitioner at Leigh Creek, South Australia (Martin 1965). In 1964 Martin set up a colony of S. crassicaudata in the Zoology Department, R.A. Fisher Laboratories, University of Adelaide, with seven animals captured near Warramboo in the upper Eyre Peninsula district of South Australia (Fig. I). Some additional wild-caught animals were subsequently introduced into this colony and in 1967 it was amalgamated with another colony established at the South Australian Museum in 1965. The colony in the Zoology Department formed the focal point of successive studies by Martin (1965), Ewer (1968), Hope and Godfrey (1968), Smith and Godfrey (1970), and Godfrey and Crowcroft (1971). This colony was eventually disbanded in 1970.


In 1965, nine animals from the Zoology Department and three wild-caught were used to start a colony at the Genetics Department, R . A. Fisher Laboratories, University of Adelaide. This colony, which has been maintained continuously since then, has been supplemented from time to time with animals taken from the wild and also with animals bred in captivity by G. Gregory, by P.R. Birks of Adelaide, or in the Zoology Department colony. Between 1968 and 1974, 54 animals caught in the southeast of South Australia; Eyre Peninsula; the lower Flinders Ranges; the Loxton Renmark area of the River Murray; or in central Australia were introduced into the colony at the Genetics Department. Of these, 12 females and 9 males reproduced. The last animal from the wild was introduced in October 1974, and the last from any source was on the 10 January 1975, when four captive-bred animals were received from P.R. Birks.


Some factors affecting reproduction in female S. crassicaudata in captivity have been identified (Smith, et al.) 1978). Continuous long-day illumination (16 hours light; 8 hours dark) for more than 6 months leads to most females entering anoestrous. They remain anoestrous during a following 3-week period of short- day illumination (8L:l6D) but come into oestrous again 20-30 days after the onset of l6L:8D.   Significantly more litters are born in pairings that have experienced short days than in those continuously exposed to long days (Smith eta!. 1978). Analysis of the colony’s breeding record from 1968 to 1976 showed that in 23 of the 44 fertile first pairings the litter was conceived within 10 weeks of pairing. Moreover, of 77 females that eventually produced a litter, 33 did not do so as a result of their first pairing and some did not give birth until their third or even fourth pairing (Smith et al. 1978).   Thus a female that does not produce a litter after being paired with one male for two or three oestrous cycles is more likely to reproduce if paired with a different male than if she stays with the first male. Females older than 30 months produced less than 4% of total litters (Smith et al. 1978).  In this paper we describe the management of the colony and consider the success of reproduction in 1977-78.


Physical Environment

Until early 1978 the whole colony was maintained in one windowless room 3.7 mx 2.2 mand2.5 mhighlit by eight 40 W fluorescent tubes hung vertically, four at each end of the room, and six 60 W incandescent globes on the ceiling. A continuously glowing neon pilot light (0.05 W) prevented absolute darkness when the main lights were off. The lights were always switched off at 1300. The room was ventilated by ducted air, the temperature of which was maintained at 25C by a thermostat operating in the return air vent. Some of the air was exhausted by a continuously run- fling exhaust fan. The system was not influenced by, and did not affect, any other room.


Fig 2
click to enlarge

The cages were positioned on five shelves of Dexion framework fitted on the two long walls of the room. The cages, 53 cm x 36 cm x 22 cm high, had galvanized iron sides, galvanized mesh top with hinged lid, and plate-glass front. They were bottomless and each stood in a tray containing loam 2-3 cm deep. Each cage contained a nest box in the form of a plastic container 20 cm x 10 cm x 5 cm, with the lid perforated and a 4 cm diameter entrance hole in one end. The nest box was loosely filled with shredded paper (Fig.2).    Exercise wheels were provided in most cages.


The cages were washed in germicide (Surgidene, Gibson Chemicals Ltd. Adelaide) and the loam was renewed about every 4 weeks.   Mite infestations were controlled with flea powder (Bayer, Sydney).


The weekly diet which supported the growth and reproduction of at least four generations of S. crassicaudata consisted of two meals of jellied raw egg; three meals of raw beef heart, fat-free and minced; one of meal- worms (larvae of Tenebrio sp.), and one of mealworms if available or processed lambs brains (The Nestle Company, Sydney).  The jellied egg and the minced beef heart were presented in paper patty pans (National, Thebarton) which were discarded after use.   Each day fresh food (10 g beef per animal), more than the animals needed, was put in the cages at about midday so that it was fresh when the animals became active after the lights were switched off.   Food left over from the previous day was removed. Water was always available in every cage.


Vitamin and mineral supplements were added to the minced beef heart once a week, as follows: ABDEC (Parke Davis, Caringbah, N.S.W.) diluted 60 times to form a stock solution given as 10mL/1.4 kg beef, that is 10mL/140 animals; potassium iodide stock solution in distilled water (0.015% by weight), given as 5mL/1.4 kg beef, that is 5.mL/140 animals, and a mixture of 10 parts by weight of calcium orthophosphate powder (Ca.,(P03)4) and 1part of vitamin E powder [d-α tocopheryl acid succinate (1000 i.u./l6 g): White-E, Medical Research (Marketing) Pty Ltd, North Ryde, N.S.W.].     The liquid supplements were mixed with the beef mince: the powder supplement was sprinkled over the beef in the patty pans.


In June 1978 the minced beef heart was replaced by tinned pet food (Whiskas, Uncle Ben’s of Australia, Wodonga. Victoria).   Later the jellied raw egg was replaced by jellied powdered egg.   In January 1981 the calcium orthophosphate was replaced by calcium carbonate powder.



The young of a litter were separated from their mother when 70 days old and were caged in groups of three or four of the same sex.    Females were first paired when 6 months old and males when 8 months old.

Before being paired the male was placed in a clean cage, and after about I week the female was introduced into the male’s cage.  Sometimes two females were caged with one male.  The pouches of paired females were examined once per week (Fig. 3) and as soon as a litter was found, the male was removed. Lactating females were examined weekly but the pouch was not opened unless the absence of a bulge suggested that the litter had been lost.

As soon as the litter had been separated from the mother, the mother was returned to the male so that mating might occur at the post-weaning oestrous.


Fig 3
click to enlarge

For the past 2 years, pairs that have not produced a litter within 3 months of pairing have been separated and each animal has been paired with another. Usually males and females were removed from pairs when they reached 30 months old.

The lighting regime was 16 hours of light and 8 hours of dark each day (l6L:8D) for 6 months, then a period of 3 weeks of 8 hours of light and 16 hours of dark(8L:16D) followed by a return to l6L:8D.

Special techniques

Collection of urine and determination of oestrous.

Animals taken from their nest boxes during the light period usually urinated immediately they were caught.


At oestrous, cornified epithelial cells were abundant in the urine, and could be seen in a drop of urine examined with a microscope.  Cells were rare or absent throughout the rest of the oestrous cycle (Godfrey 1969).


A I : I mixture of oxygen and nitrous oxide containing ethrane, or enflurane (Abbott Laboratories, Sydney), 4%, administered from an “Enfluratec” anaesthetic machine at 1 L/min was suitable for anaesthesia of short duration.

In a room heated to 37C and with an exhaust fan running, one animal was placed in each of two anaesthetic vessels. After 10 minutes, one animal was taken out of the vessel and its muzzle was put into a facemask” supplied with gas from the anaesthetic machine. A third animal was placed in the vacated anaesthetic vessel while the first animal was removed from the face-mask and a sample of its blood was taken. As soon as the blood had been obtained, the animal was placed in a small recovery vessel, open to the air, and in which drinking water was available. Recovery took about 5 minutes. Meanwhile the second animal was being bled.

Blood samples.

Blood was taken from anaesthetised animals by puncturing the orbital sinus with a 2 I-gauge needle and drawing the resulting drops of blood into a heparinized syringe (Hope and Godfrey 1968).   About 0.25-0.3 mL of blood were collected. Blood could be obtained from 15 animals in 1.5-2 hours, i.e. the time involved is about 6-8 minutes per animal.   Nearly 400 blood samples have been taken since ethrane has been used, and 10 animals have died during anaesthesia and bleeding, to give a de4th rate of 2.6%.



Fertility of females

During 1977 and 1978, 66 females produced 117 litters, and in 66 pairings that existed for 10 weeks or longer, no litter was born. The proportion of females giving birth was particularly high among young adults. Twenty-five females paired when less than 7 months old bore a litter in their first pairing, 20 of them within 12 weeks of pairing. Fourteen other females paired when less than 7 months old remained nulliparous in first pairings lasting more than 9 weeks; 13 of these nonfertile pairings lasted more than 12 weeks.


The parous females comprised 3 first-generation laboratory-bred, 27 second-generation, 31 third generation and 5 fourth-generation laboratory-bred females.


The maximum number of litters borne by a female was 5 but neither of the two females that produced 5 litters reared them all.   Only one female reared all of 4 litters: four females reared all of 3 litters (Table 1).

Fig 4
click to enlarge

Mortality during lactation

Of the 117 litters produced by females that gave birth during 1977-78 in 87 litters at least one young survived to weaning, but 30 litters were lost completely before weaning, that is in about only two-thirds of the litters born were any young reared to weaning.   The loss of young within litters has not been analysed because of the difficulty in obtaining the correct number for the initial size of the litter. As the animals were examined weekly, some litters were 6 days old when first counted and some young could have been lost by then.


Eleven females lost their first litter.   12 their second.   4 their third.   2 their fourth litters and I female lost her fifth litter.


Of the 30 litters lost completely, 11 died after their mother died.   Necropsy of these females revealed that the death usually resulted from pneumonia or pleurisy.


Litter size

In the 87 litters in which at least one young survived, the modal litter size at weaning was 6 (Fig. 4) and the mean litter size was 5.2, standard deviation 2.0.   Although several litters of 10 were found in the pouch, 9 was the maximum number reared to weaning.   If the number of litters lost completely are included in the calculation, the mean litter size is 3.9, standard deviation 2.9.


The mean number of young reared, summed over all litters (120) for each parous female, was 7.0, standard deviation 5.7.   However, this mean under-estimates the reproductive success of the females because 19 of the females were still of reproductive age at the time of analysis and several of them did rear more young.   The maximum number of young reared by one female was 23, but very few reared more than 15 (Fig. 5).


Fig 5 & 6
click to enlarge

Rate of reproduction

Eighty-two days is the minimum interval between successive parturitions when the first litter is reared to weaning. The observed interval ranged from 82 days to 545 days, and 3 I times out of 43 (72%) the second litter was born less than 120 days after the previous one (Fig. 6). Hence 72% of females who produced a subsequent litter after a successful lactation were fertilised at either their first or second post-weaning oestrous.


females that experienced 16L: 8D from birth and whose mothers received l6L 8D during pregnancy became mature by the expected age of 6 months. Twenty females that gave birth in 1978, when less than 9 months old, were themselves born in 1978 when the main colony was never exposed to short days. Thus females can mature without experiencing short days.


Parous females that produced second and third litters in near minimum time did so while experiencing long days continuously.   Of the 12 females who produced a litter more than 119 days after the birth of the preceding litter that was reared, 4 had not received 8L:l6D, one had received it early in lactation and 7 had experienced an 8L:l6D period that ended 31-75 days before the second parturition.

the 6 months after the fresh beef heart in the diet was replaced by tinned pet food, reproduction did not appear to decline but the reproductive rates of animals on the two diets cannot be compared because the new diet has not been used long enough.


The rate of increase of the colony’s population was limited by an unfortunately high death rate not only among pouch young, but also among juveniles and adults (Fig. 7).    The cause of death was usually an infection of the respiratory tract.    Death was rapid and we were seldom aware of the animal’s debility.

Fig 7
click to enlarge

Diet has been implicated in the failure of laboratory colonies of Sminthopsis to thrive.   Nutritional deficiency is believed to have caused the extinction of a colony of S. macroura (synonym S. larapinta) (Godfrey 1969).   In a colony of S. crassicaudata started by the Radiobiology Unit of the Medical Research Council in Berkshire, England. many animals died with enlarged thyroid glands.    The symptoms suggested an iodine deficiency, but although the animals were given much more iodine, reproduction ceased and the colony died out (Breckon and Hulse 1972).

When prenatal and early postnatal mortality in the colony at the Zoology Department became very high at the beginning of the 1967 breeding season, one massive dose of both Vitamin E and iodine given to all individuals was sufficient to return the breeding to normal 6 weeks later (Godfrey and Crowcroft 1971).

Management is important in maintaining a high rate of reproduction.   Pairing of young females as soon as they reached 6 months of age and reconstitution of the pair as soon as the litter was 70 days old were important in management.   Small cages (floor area 0. 19m2) were large enough to allow the animals to breed and to rear young.   Larger cages (floor area 0.56 m2) as are used to house S. macroura at the U.S.A. National Zoological Park (Collins 1973) are not necessary, nor are the 4m2 pens used to provoke resistant S. crassicaudata into breeding (Godrey and Crowcroft 1971).


S. crassicaudata bred in the laboratory resembled their wild conspecifics in that oestrous cycles stopped during short days, either autumn in the wild, or 8L:16D in the laboratory, and resumed during long days (Morton 1978: Smith et al. 1978).   Laboratory-bred S. crassicaudata differed from the wild populations in that the females became sexually mature at 6 months and without having experienced short days, whereas the wild animals do not reproduce until after the winter following their birth (Morton 1978).   Also, laboratory females can produce and rear as many as five litters successively whereas wild females are not known to produce more than two litters per season nor to reproduce in a second season (Morton 1978; Smith et al. 1978).


Colonies of few insectivorous marsupials have been established in the laboratory but of these, S. crassicaudata compares favourably in rate of reproduction and in viability of the colony.   Colonies of Dasyuroides byrnei and of Dasycercus cristicauda have been established at La Trobe University and two generations of Dasycercus and three of Dasyuroides had been raised by 1971 (Woolley 1971). S. macroura (syn. S. Iarapina) breed readily in captivity and although Godfrey’s (1969) colony died out after two generations, a colony of the same species was thriving at the U.S.A. National Zoological Park, where a fourth generation litter of captive bred animals had been born (Collins 1973).   Quantitative data for rates of reproduction and of survival in these colonies of dasyurids are unavailable at present.   In colonies of the omnivorous didelphid Marmosa robinsoni, litters have been produced in the ratio of one for every 3-4.6 pairings when the animals were paired only when females were in oestrous (Barnes and Barthold 1969; Godfrey 1975).    In the Brookfield Zoo colony the mean litter size was 7.8, standard deviation 3.1, in the parental generation, 9.1, standard deviation 2.7, in the second laboratory-bred generation (Godfrey 1975); the number weaned from each litter was not given.   Females bred for only 1 year and a low proportion of laboratory-bred females contributed to the next generation, with only 8 of 52 first generation and 13 of 39 second generation females becoming parous (Godrey 1975).    S. crassicaudata’s mean reproductive rate of more than 7 young weaned per parous female, and with more than half of the paired females producing, can be favourably compared with the data for Marmosa robinsoni.


Of the several persons who have at times cared for the colony we especially acknowledge the assistance of Mrs D. Golding.   We thank Dr P. Baverstock, IMVS, Adelaide, for his advice on Ethrane anaesthesia.   We are grateful to M. Archer, F. Aslin, P. Birks, L. Corbett, T. Dennis, B. Eves, G. Godfrey, 0. Gregory, 1. Hann, R. Lang,D. Pratt and J. Ramsey for (providing us with animals and to the South Australian National Parks and Wildlife Service for their permission to take wild S. crassicaudata.



Breckon, G. and Hulse, E.V. (1972). Difficulties in the management of Sminthopsis crassicaudata due to iodine deficiency and thyroid disease. Lab. Anim. 6: 109-18.

Barnes, R.D. and Barthold, S.W. (1969) Reproduction and breeding behaviour in an experimental colony of Marmosa mitls Bangs (Didelphidae). J. Reprod. Fertil., Suppl. 6:477-82.

Collins, L.R. (1973) Monotremes and marsupials: a reference for zoological institutions. (Smithsonian Institution Press: Washington.)

Ewer. R.F. (1968) A preliminary survey of behaviour in captivity of the dasyurid marsupial, Sminthopsis crassicaudata (Gould). Z. Tierpsychol. 25: 319-65.

Godfrey, G. K. (1969) Reproduction in a laboratory colony of the marsupial mouse Sminthopsis larapinta (Marsupialia: Dasyuridae). Aust. J. Zool. 17: 637-54.

Godrey, G.K. (1975) A study of oestrous and fecundity in a laboratory colony of Mouse Opossums Marmosa robinsoni. J. Zool., Lond. 175: 541-55.

Godfrey, O.K. and Crowcroft, P. (1971) Breeding the fat-tailed marsupial mouse in captivity. In: Lucas, J. (ed.), Inter. Zoo. Yearb. 11: 34-8.

Hope. R.M. and Godfrey, G.K. (1968) Transferrin polymorphism in the Australian marsupial mouse Sminthopsis crassicaudara (Gould). Aust. J. Biol. Sci. 21: 587-91.

Martin, P.G. (1965) The potentialities of the fat-tailed marsupial mouse, Sminthopsis crassicaudara (Gould) as a laboratory animal. Aust. J. Zool. 13.

Morton, S.R. (1978) An ecological study of Sminthopsis crassicaudara (Marsupialia: Dasyuridae). lii. Reproduction and life history.
Aust. Wild. Res. 5: 22 1-48.

Smith. M.J. and Godfrey. G.K. (1970) Ovulation induced by gonadotrophins in the marsupial Sininthopsis crassicaudata (Gould). J. Re prod. Fertil. 22: 41-7.

Smith. M.J., Bennett,J.H. andChessonC.M. (1978) Photoperiod and some other factors affecting reproduction in female Stninrhopsis crssicaudata (Gould) (Marsupialia: Dasyuridae) in captivity. Aust. J. Zool. 26: 449-63.

Woolley, P. (1971) Maintenance and breeding of laboratory colonies. Dasvuroides bvrnei and Dasvcercus cristicauda. In: Lucas. J. (ed.). Inter. Zoo Yearb. 11:351-4.


This is reproduced from an article from ‘The Management of Australian Mammals in Captivity’, by D. D. Evans, 1982 with many thanks and acknowledgments.


Bennett's Wallaby
Juvenile NT Brushtail Possum
Swamp Wallaby
Golden Brushtail Possum
Red Kangaroos
Yellow-footed Rock-wallabies
Baby Squirrel Glider
Sugar Glider

Copyright © The Marsupial Society of Australia Inc. 2003 - 2006 All rights reserved. Privacy Statement

Email Webmaster