REPRODUCTION IN FEMALE WILD GUINEA-PIGS

J. P. ROOD and BARBARA J. WEIR Department ofZoology, MichiganState University, East Lansing, Michigan, U.S.A. and Wellcome Institute of Comparative Physiology, Zoological Society of London, Regent's Park, London, jV. W. 1 (Received 10th December 1969) Summary. The breeding characteristics of three species of wild guinea- pig (F. Caviidae) are reported. Cavia aperea, Galea musteloides and Micro- cavia australis were studied in Argentina in the field and in outdoor pens, and laboratory colonies of the two former species were also established in England. Pens of domestic guinea-pigs (Cavia porcellus) and of C. aperea \m=x\C. porcellus hybrids were maintained in Argentina for comparisons with C. aperea. C. aperea and G. musteloides gave birth in every month but there was a breeding peak in spring (September to December). Microcavia had a more restricted breeding season ; in the field study area, births occurred only between August and April. Gestation length in C. aperea was variable but the mode was at 61 days, while the modes of Galea and Microcavia were much shorter at 53 and 54 days, respectively. All three species exhibited a post-partum oestrus and Galea may experience a lactation anoestrus. Oestrous cycle lengths in C. aperea and Galea varied consider- ably but the mean length in Cavia was 20\m=.\6\m=+-\0\m=.\8days and in Galea it was 22\m=.\3\m=+-\1 \m=.\4days; in the latter species, the presence of a male in the same cage was necessary for the induction of oestrus. Average litter size was 2\m=.\2for C. aperea, 2\m=.\6for Galea and 2\m=.\8for Microcavia. In the Argentine colonies, the age at first conception in C. aperea and Galea varied with the time of year of birth, but this variation was not maintained under the more equable laboratory conditions.

INTRODUCTION Although the domestic guinea-pig, Cavia porcellus, has been kept in captivity for over 400 years, little information has been published on wild guinea-pigs. Those studies which have been reported have mainly been concerned with the taxonomy and distribution of cavies (Cabrera, 1954) and with the genetics of the genus Cavia (Detlefson, 1914; Castle & Wright, 1916; Ubisch & Mello, 1940). These authors give some data on reproduction in the species which they studied and Pearson (1951) gives a little information on Galea musteloides in Peru. Nothing else appears to have been published on the reproductive biology of any wild members of this family (Caviidae) of hystricomorph rodents. The * Enquiries concerning reprints should be addressed to Dr . J. Weir. 393 Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access 394 J. P. Rood and Barbara J. Weir literature on reproduction of the domestic guinea-pig is, of course, extensive (see Asdell, 1964; Marshall, 1952-66). Various other hystricomorphs have been studied, mainly in captivity, and several unusual features of their reproductive physiology have been reported (Weir, 1967a). All except the coypu (Myocastor coypus) have a vaginal closure membrane which is perforate at oestrus and parturition. Apart from the sub-family Caviinae, all other hystricomorphs investigated have very long pregnancies ofover 100 days and this is not associated with a delay ofimplanta¬ tion such as is found in various other species (e.g. roe deer, stoat, bear, badger, armadillo). The oestrous cycle is also longer in other hystricomorphs than the 16^-day cycle of the domestic guinea-pig. For example, both the chinchilla (Chinchilla laniger) and the plains viscacha (Lagostomus maximus) have a mean cycle length of 41 days (Weir, 1966; Weir, in preparation) and the cycle length in the dasyproctids (Dasyprocta and Myoprocta) is about 35 days (Weir, 1967b; Kleiman, 1969). During the course of investigations into the comparative behaviour of wild guinea-pigs in Argentina (Rood, in preparation), information was obtained on the breeding characteristics of three species of different genera. Colonies of two of these species were also established in London and the breeding data compiled at both places are compared below. The species identifications for Galea musteloides (PI. 1, Fig. 2) and Microcavia australis (PI. 1, Fig. 1) are based on Cabrera (1954); but for Cavia aperea (PL 1, Fig. 3) the recommendation of Hiickinghaus (1962) has been followed.

MATERIALS AND METHODS Observations in Argentina (December 1965 to February 1968) The three wild species (Cavia aperea, Galea musteloides and Microcavia australis) were observed in the areas shown in Text-fig. 1 and limited trapping and observations were also made in several other localities. Colonies of each of these three species, and of the domestic guinea-pig (Cavia porcellus) and of C. apereaxC. porcellus hybrids were maintained in large pens at Sierra de la Ventana, Buenos Aires Province. Cavies were trapped live in the study areas by mesh, treadle-type traps (National Live Traps) placed in their runways. Each animal was individually marked by toe clipping and with Nyanzol dye. Total length and weight were recorded and compared with data on animals of the field and pen populations of known age in order to assess approximate ages. Pregnancy was determined by palpation. The pens were initially stocked with Galea and C. aperea trapped within a 40-km radius of Sierra de la Ventana; the Microcavia colony was established with animals brought from the study area in southern Buenos Aires Province. The original C. porcellus stock was donated and the hybrids were obtained by crossing female C. porcellus with a C. aperea male caught in the wild. The colonies were allowed to grow naturally until about twenty-five animals were present and were then controlled by removing newly weaned animals. Such removals were only necessary for the Galea population as, although all the species did

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access PLATE 1

Fig. 1. Microcavia australis ( 0-5). Fig. 2. Galea musteloides ( 0-5). Fig. 3. Cavia aperea ( 0-3).

(Facing p. 394)

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access Reproduction infernale wild guinea-pigs 395 breed in the pens, the reproductive success varied considerably and was highest in Galea and lowest in Microcavia. The wire mesh pens were each 5 5 2 m high ; the mesh was extended 75 cm underground to prevent escape but only the Microcavia were found to burrow. The principal shelter in each pen consisted of four waterproof nest boxes which were kept supplied with fresh sawdust. Additional shelter was provided by an inverted metal tub in the Microcavia pen and open wooden boxes in the other pens. All the pens were supplied with branches for climbing. Commercial rabbit

Text-fig. 1. Map of part ofArgentina to show the areas ofstudy and the sources of cavies investigated in captivity: (1) Cavia aperea study area; 'Estancia El Destino', 20 km S.E. Magdalena. (2) Microcavia australis study area; 'Estancia La Cristina', 14 km S.E. Cardenal Cagherò. (3) Galea musteloides study area ; 'Estancia La Sorpresa', 5 km E. Torn- quist, and source of G. Musteloides for London colony. A, Source of C. aperea for London colony. ·, All three species in the penned colonies, Sierra de la Ventana. pellets (Purina), whole oats and water were always available; a selection of grasses, carrots, lettuce, apples and potatoes was given once a day. Lawn grass was grown under wire mesh frames to supplement the supply of green food. Observations in London (August 1967 to August 1969) The C. aperea and G. musteloides colonies were derived from three C. aperea (one male, two females) caught in the wild and from ten Galea (four males, six

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access 396 J. P. Rood and Barbara J. Weir females) which were either from the pen colonies (see above) or had been trapped immediately before their transfer to London. Both species were initially kept in small groups in standard guinea-pig cages, but C. aperea is now housed in floor pens and Galea in mesh cages with solid floors. Details of the housing and husbandry are given by Weir (1970). The animals were checked daily for reproductive condition and all events, such as vaginal opening, copulatory plugs and litters, were recorded. Oestrous cycle length was defined as the interval from the 1st day ofvaginal opening ofone cycle up to, but not including, the 1st day of opening at the next cycle. The day of finding a copulatory plug was recorded as Day 1 of pregnancy; most gestation lengths were, however, determined from the interval between births of litters as there is an immediate post-partum oestrus in these species. Standard statistical tests were applied and mean values are followed by the standard error of the mean.

RESULTS

A. CAVIA SPP. 1. Annual reproductive cycle Under laboratory conditions, domestic guinea-pigs breed throughout the year, although slight seasonal variations in reproductive performance have been reported (Rowlands, 1949; Wright, 1960). Castle & Wright (1916) reported that their colony of Cavia cutleri, obtained from Peru and maintained in the United States, bred in all months, but thirty-eight litters were born from January to June as against fifteen born from June to January. It is clear from Text-fig. 2 that, in Argentina (B, C and D), Cavia species breed throughout the year, although there is an indication that reproductive activity is reduced in C. aperea (B) during the autumn and winter months (April to August). In the combined data for the 2-year period studied, 42% to 87% of C. aperea females over 3 months old were pregnant in each month from August to February whereas only 9% to 25% were pregnant from March to July. In the field study area, five of thirteen females in March 1967, none of twelve in August 1967 and four of four in January 1968 were visibly pregnant or lactating. The incidence of winter breeding differed in the 2 years studied. In the pen, several C. aperea females produced litters in the winter of 1966 but none did so in the unusually cold winter of 1967. In the field, one of six females trapped in August 1967 at Balcarce, Buenos Aires Province, was visibly pregnant whilst all females trapped in the study area in August 1967 were non- breeding. No seasonal pattern of breeding was apparent in the London colony of C. aperea (Text-fig. 2A). 2. Oestrus Observations of the vaginal closure membrane in twenty-one laboratory C. aperea (Text-fig. 3) gave a mean cycle length for sixty-seven cycles of 20-6+ 0-8 days. Females of all species conceived most frequently at the post-partum oestrus and one laboratory C. aperea had six consecutive pregnancies resulting from

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access Reproduction infernale wild guinea-pigs 397 post-partum matings. In Argentina, the maximum number of litters produced by a female in 1 year was five in the Fj hybrids, five in C. aperea and four in C. porcellus. When conception did not occur at the post-partum oestrus, there was no evidence of a lactation anoestrus in five of the laboratory animals.

Text-fig. 2. Monthly distribution of births in species of Cavia: A, C. aperea in London (August 1967 to August 1969) ; B, C. aperea in Argentina (February 1966 to February 1968) ; C, C. porcellus in Argentina (February 1966 to February 1968); D, C. aperea C. porcellus Fj hybrids in Argentina (January 1967 to January 1968).

Text-fig. 3. Distribution of oestrous cycle length in C. aperea in London. 3. Gestation The variation of gestation length in Cavia species is shown in Text-fig. 4. It can be seen that the range is very much greater in laboratory C. aperea (A) although the mode at 60 days is near the mean of 60-9+0-3 days for C. aperea

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access 398 J. P. Rood and Barbara J. Weir in Argentina (B). In this small sample, the difference between the mean gestation length afterpost-partum matings (62-7+ IT days) and after conceptions at other oestrous periods (65-5+1-4 days) was not significant (P>0-05). Although there is no overlap for gestation lengths in the Argentine colonies of C. aperea (B) and C. porcellus (C), recorded gestation lengths in large colonies

Text-fig. 4. Distribution of gestation length in species oí Cavia: A, C. aperea in London (mean = 63-9±0-9 days); B, C. aperea in Argentina (mean = 60-9 + 0-3 days); C, C. porcellus in Argentina (mean = 67-0 + 0-8 days); C, C. aperea

Litter size

Mean gestation length (days) ±S.E. 64-6+1-6 63-5+1-0 61-6+1-0 63-5±3-l* 59-5 ±003 No. of observations 8 13 8 4 2

* This group contains the longest recorded gestation of 74 days; if this value is excluded, the mean gestation length is 60-0 + 0-6 days. of domestic guinea-pigs range from 54 to 72 days (Blandau & Young, 1939; Rowlands, 1949). However, the mean gestation length in C. aperea is consider¬ ably shorter than that of C. porcellus and hybridization between the two species results in intermediate values (D and E). There is some indication that gestation length in C. aperea is inversely proportional to litter size (Table 1 ), as in domestic guinea-pigs (Rowlands, 1949; Goy, Hoar & Young, 1957).

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access Reproduction infernale wild guinea-pigs 399 4. Litter size and sex ratio In the present study, litter size was similar in all species (Text-fig. 5) and the overall mean was 2-3. Of 117 C. aperea born in the pens in Argentina and in the laboratory, forty- nine were males and sixty-eight were females. 5. Age at onset of reproduction Although a few exceptional female guinea-pigs kept under laboratory conditions have been reported to come into first oestrus at 20 days (Reid, 1958;

Text-fig. 5. Distribution of litter size in species of Cavia: A, C. aperea in London (mean = 2-5 young/litter); B, C. aperea in Argentina (mean = 2-1 young); C, C. porcellus in Argentina (mean = 2-1 young); D, C. aperea JxC.porcellus $ in Argentina (mean = 2-3 young) ; E, Hybrid C. aperea C. porcellus FjS in Argentina (mean = 2-5 young).

Table 2 age at first conception in relation to season of birth in Cavia species

Estimated age at conception (days) C. aperea C. porcellus Month born C. aperea C. porcellus Fts F2s March 143 June 102, 109 July 45 August 73 73 September 62,30 36, 33 October 89 31 29 November 48,29 43

Ibsen, 1950), the majority probably do not do so until they are over 30 days old. Young, Dempsey, Hagquist & Boling (1939) recorded a mean age at first heat of 67-8 days (range 33 to 134 days). In wild forms, Castle & Wright (1916) reported that female C. cutleri could breed at 2 months and Ubisch & Mello (1940) recorded two female C. rufescens which littered at 3 months and thus conceived at about 1 month of age. The estimated ages at first conception Cavia females born in the pens during different months are shown in Table 2. Of the ten females born from September to November, six (representing all forms studied) conceived at their first oestrus when 29 to 36 days old. The three animals, all C. porcellus, born between

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access 400 J. P. Rood and Barbara J. Weir March and June did not conceive until they were more than 3 months old suggest¬ ing that onset of reproduction may be delayed in autumn- and winter-born females. In the laboratory colony of C. aperea, mean age at first vaginal opening in twenty-seven animals was 58-6+3-0 days but only three females conceived at this first oestrus, the youngest being 27 days of age. The mean age at conception in ten animals was 84-4+9-9 days and first conception occurred up to the fifth oestrus after puberty. There was no evidence in the laboratory colony that the age at puberty or conception varied according to the time of year of birth.

B. GALEA MUSTELOIDES 1. Annual reproductive cycle The field observations suggested that wild populations of Galea may produce litters throughout the year and that most litters are born in spring and summer. All five adult females caught in December 1965, two often in June 1966 and three ofseven in April 1967 were pregnant or lactating. In the pens in Argentina (Text-fig. 6A) and in the laboratory in England (Text-fig. 6B), births occurred in

Text-fig. 6. Monthly distribution of births in Galea musteloides: A, In Argentina in the pens (February 1966 to February 1968); B, In London (August 1967 to August 1969).

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access Reproduction infernale wild guinea-pigs 401 every month. In the Argentine colony, more than 50% of females over 3 months of age were pregnant in any one month but at least 90 % were pregnant from September to February, indicating increased reproductive activity in spring and early summer. 2. Oestrus Galea also has a vaginal closure membrane which is open for 1 to 6 days at oestrus and parturition, although closure is very rapid after mating has taken place. The majority of animals mate at the post-partum oestrus and once maturity is reached, most females are perpetually pregnant. A female caught in the wild in Argentina produced twelve litters in 21 months and in London, a female had fifteen consecutive pregnancies.

—r-r 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 Length of cycle (days) Text-fig. 7. Distribution of oestrous cycle length in laboratory G. musteloides; black columns denote cycles in the presence of a vasectomized male, open columns denote other cycles.

Spontaneous oestrous cycles are, therefore, uncommon and only three cycles of 17, 17 and 18 days were recorded in Argentina. In England, the mean cycle length of fifty-six observations on twenty-six animals (Text-fig. 7) was 22-3+1-4 days and most of these cycles occurred before the first conception (see below). When females were isolated from males, either singly or in groups, spontaneous cycles did not occur even if the male was separated only by a mesh partition. For example, a pregnant female was removed from the male and littered on 23rd January 1968; no periods of vaginal opening were observed until the male was re-introduced on 14th May 1968 and a fertile oestrus occurred 2 days later. To follow consecutive cycles in individual females, it was necessary to use a vasectomized male. All four females run with this male then cycled at regular intervals. The mean length of twenty-one cycles in the presence of the vasectom¬ ized male was 19-0+1-2 days and this was not significantly different from the mean length for sporadic cycles. It is clear that the presence of the male is very

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access 402 J. P. Rood and Barbara J. Weir important in this species for the initiation of oestrus. Adult males 'chin-rump follow' a female with increasing frequency as she approaches oestrus, suggesting that this pattern may be instrumental in inducing a pro-oestrous condition (Rood, 1970). In five laboratory females which did not conceive at the post-partum oestrus, the mean interval to the next oestrus was 41-8+4-0 days. This was significantly different (P<0-001) from the mean oestrous cycle length and indicates that Galea may have a lactation anoestrus.

3. Gestation Since there was no difference in the lengths of gestation in the Argentine and London colonies, the data are combined in Text-fig. 8. There was no indication that the size of the litter affected the length of gestation in this species.

Text-fig. 8. Combined distribution of gestation length in G. musteloides in Argentina and London.

4. Litter size and sex ratio The litter size ranges from one to five and the distributions for the English and Argentine colonies are combined in Text-fig. 9. Of all Galea born, 182 were males and 179 were females.

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access Reproduction infernale wild guinea-pigs 403 5 Onset of reproduction In Argentina, nine females born from June to December conceived at a mean age of 58 days (range 28 to 93 days) while six females born from February to April did not conceive until a mean age of 167 days (range 113 to 255 days). This suggests that females born in the autumn do not breed until the following spring, while females born in the spring breed that summer. Under laboratory conditions, there was no evidence of any seasonal effect of the time of birth on age at first conception. In thirty-four animals, all of whom were housed with a male from birth, the mean age at conception was 60-6 days (range 17 to 152 days) ; twenty-one of these conceived at the first oestrus. The mean age at puberty (as determined by first vaginal opening) in fifty-two females was 48-5 days (range 15 to 170 days).

Text-fig. 9. Combined distribution of litter size of G. musteloides in Argentina and London.

Three females were housed at 21 days of age with an adult male in an indoor cage in Argentina and were checked twice daily for vaginal opening. One came into oestrus and mated at 32 days of age and the other two were then separated from the male by a mesh partition. No vaginal opening occurred within the next 20 days, but when the females were replaced with the male both came into oestrus within 2 days (at the ages of 53 and 54 days). Although these values fall within the normal range for the interval to puberty, the experiment indicated that the initiation of puberty may also be dependent upon the presence of a male (see above).

C. MICROCAVIA AUSTRALIS 1. Annual reproductive cycle In a field study area, litters were born from August to April with the majority of births between September and December (Rood, 1970). In the pens, litters were born from October to February and none of the females was pregnant

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access 404 J. P. Rood and Barbara J. Weir from March to July. Therefore, both the field and pen data indicate that Microcavia has a more restricted breeding season than Cavia or Galea. Three females resident in the pen in the winter of 1966 all produced their first spring litters within 3 days of each other (from 7th to 10th October) and the only survivor of the winter of 1967 also gave birth in this period (on 7th October), suggesting that reproductive activity may be triggered by a specific environmental stimulus.

2. Oestrus Microcavia also has a post-partum oestrus and the vaginal closure membrane is perforate for about 5 days after parturition. One female came into oestrus and mated 15 days after giving birth and three other animals, which were thought to have aborted upon capture, conceived within 15 days of introduction into the pen. 3. Gestation This was estimated from the interval between consecutive litters and the mean length for six pregnancies was 54-2+0-5 days (range 53 to 55 days). Gestations of 51 and 52 days were also recorded but the young appeared to be premature and died within a few hours of birth.

4. Litter size and sex ratio The distribution oflitter size is shown in Text-fig. 10. In the pens, twenty-one males and nineteen females were born.

Text-fig. 10. Distribution of litter size in Microcavia australis.

5. Onset of reproduction Information on the age at which breeding first commences is limited since mating of young females at normal oestrus was not observed in the pen colony and only one female born in the pen subsequently produced a litter. This

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access Reproduction infernale wild guinea-pigs 405 parturition took place when the female was 138 days old and, therefore, con¬ ception occurred when she was about 85 days old. In the field study area, two young females born in the spring were observed mating when they were 40 to 50 days old (Rood, 1970).

DISCUSSION The breeding record for the penned Cavia species indicated that, under identical conditions, the most successful breeders were the hybrids. The four hybrid Ft females were continuously pregnant after they reached maturity whereas some of the C. porcellus and C. aperea females were not. C. porcellus appeared better than C. aperea but the reproductive performance of the latter was lowered by four females which lived in the pens for 6 months without breeding. The C. aperea also appeared to be more sensitive to climatic conditions because in the very cold winter of 1967, breeding ceased in the pens as well as in the field study area. The capture of a pregnant female in the middle of the milder winter of 1966 shows that winter breeding does occur when conditions are favourable. Wild Cavia may have been affected by food shortage but this explanation was not applicable to the penned animals which were fed regularly. The Argentine populations of Galea are similar to the Cavia species in the characteristics of the annual reproductive cycle, although Pearson's (1951) data suggest that G. musteloides in Peru may be a seasonal breeder as breeding condition was found in males and females trapped in December, but not in animals trapped in July and September. There was no peak of breeding activity in the laboratory Galea colony, probably because, under the constant environ¬ mental conditions, females became sexually mature at approximately the same age, regardless of the time of year of birth. In the Argentine colony, winter breeding was restricted to females which conceived at the post-partum oestrus. The southern population of Microcavia australis studied has a restricted breeding season and it would be interesting to know if the Cavia species which live on the Altiplano also breed at a limited season of the year. The indications are that the breeding season of cavies may be determined by climatic conditions, most probably by temperature. The occurrence of a post-partum oestrus and ovulation is a common feature of the Caviinae. It has been found in the three species described in this paper and inC. cutleri (Castle & Wright, 1916) andC*. rufescens (Ubisch & Mello, 1940). In favourable conditions, therefore, females of these species are constantly pregnant after reaching maturity. The rapidity of re-impregnation must offset the rela¬ tively long gestation period and, together with the fact that the young are precocious at birth, must contribute greatly to the success of these species in competition with the more prolific myomorph rodents which have shorter pregnancies. This post-partum characteristic is used extensively to ensure maximum output of young in most large-scale breeding colonies of domestic guinea-pigs. The possible association, in Galea, of a male sexual pattern (the chin-rump follow) with the initiation of oestrus, and possibly also of puberty, causes speculation as to whether a similar phenomenon will be discovered in Microcavia

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access 406 J. P. Rood and Barbara J. Weir australis since this species also exhibits this distinctive behaviour pattern (Rood, 1970). The introduction of a male has also been shown to affect the time ofonset of oestrus in the mouse (Whitten, 1956) but in this species, unlike Galea, the females undergo spontaneous oestrous cycles in the absence of a male and the excreta is effective in modifying the cycle. It will be interesting to determine if a post-partum ovulation occurs in the absence of a male in Galea. This species was the only one studied by us which may have a lactation anoestrus, a phenomenon also found in another hystricomorph rodent, the chinchilla (Weir, 1967a). The oestrous cycle lengths recorded for C. aperea and Galea are very variable and it may be argued that these are abnormal because the animals were disturbed under laboratory conditions. However, the similarity of all the data compared for the English and Argentine colonies suggests that the laboratory studies are valid. It has also been shown that most cycles in the chinchilla are 30 to 50 days (Weir, 1966) and in the plains viscacha, only 75% of cycles are between 25 and 55 days (Weir, in preparation). Therefore, this variation is probably normal for other wild and newly domesticated species of hystricomorph rodent. Table 3 comparison of litter size in species of the sub-family caviinae

Litter size Species Mean Range Authors Microcavia australis 2-8 1 to 5 Present study Galea musteloides 2-6 Ito 5 Present study Cavia aperea (Argentina) 2-1 1 to 5 Present study Cavia aperea (London) 2-5 1 to 5 Present study C. cutleri 1-89 1 to 4 Castle & Wright (1916) C. rufescens 1-35 Detlefson (1914) C. porcellus 2-1 1 to 5 Present study C. porcellus 2-58 1 to 8 Haines (1931) C. porcellus (MRC stock, Frant) 4-0 1 to 10 Rowlands (1949) C. porcellus (MRC stock, Pirbright) 4-3 Hoyland (1965) C. porcellus (MRC stock, Frant) 3-5 1 to 13* Sutherland (personal communication, 1969) C. aperea C. porcellus hybrid F1s 2-5 1 to 6 Present study

* The largest viable litter contained nine young. The most obvious reproductive character for which domestic guinea-pigs have been selectively bred is litter size. In 1931, Haines reported that the mean litter size in his large sample of guinea-pigs was 2-58 (range 1 to 8) while the Dunkin-Hartley, or MRC, strain which originally had a mean litter size of 3-0 (Dunkin, Hartley, Lewis-Fanning & Russell, 1930) was developed at Frant to produce 4-0 young per litter (Rowlands, 1949). Recently, the Pirbright strain of the MRC stock has been reported as having a mean litter size of 4-3 (Hoy- land, 1965). A comparison of the reported variations in litter size of different species and strains of Caviinae is shown in Table 3. The observation that, in the domestic guinea-pig, smaller litters were pro¬ duced after longer pregnancies was recorded by Rowlands (1949) and by Goy et al. (1957), and it is interesting to note a similar trend in C. aperea, although this was a small sample. Since domestic guinea-pigs have been bred to produce

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access Reproduction infernale wild guinea-pigs 407 larger litters, it is surprising that there has been no apparent shortening of gestation length. The pregnancies of other Cavia species are shorter than that of C. porcellus (see Table 4), and mating a female C. porcellus to a C. aperea male appeared to shorten pregnancy by several days. This has been found for all crosses of C. porcellus with wild cavies (Pictet & Ferrerò, 1951 ; Ubisch & Mello, 1940). It is noteworthy that Galea and Microcavia have gestation lengths (52-7 and 54-2 days respectively) which are not only shorter than those Cavia species but are the shortest known within the Hystricomorpha. The youngest animals to conceive in the present study were a 17-day-old Galea and a 27-day-old C. aperea. These values represent an extreme (see Table2), but Ubisch & Mello (1940) reported that C. rufescens can breed at 1 month of age and conceptions in domestic guinea-pigs less than 21 days old occur in large colonies (Sutherland, personal communication). The loss of a seasonal effect on the onset of reproduction in laboratory C. aperea and G. musteloides probably reflects the fact that the environmental conditions were more constant than

Table 4 reported gestation lengths in species of cavia

Gestation length (days) Species Mean Range Author C. aperea 62-4 59 to 74 Present study C. cutleri 63-3 56 to 69 Castle & Wright (1916) C. rufescens 62 to 65 Ubisch & Mello (1940) C. porcellus 680 59 to 72 Rowlands (1949) C. rufescens ¿* x C. porcellus Çj 63 to 67 Detlefson (1914) C. aperea

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access 408 J. P. Rood and Barbara J. Weir Ferrero's colony. However, C. aperea would seem to be a more probable ancestor of C. porcellus than, for example, C. rufescens, in which the male offspring of crosses with C. porcellus are sterile (Detlefson, 1914; Ubisch & Mello, 1940). Crosses of C. cutleri males with C. porcellus females have been reported to be fertile (Castle & Wright, 1916), and Hiickinghaus (1962) considers that C. porcellus, C. aperea and C. cutleri are all conspecific.

ACKNOWLEDGMENTS The field work in Argentina was supported by United States Public Health Service Fellowship Grant 5 F2 HD-28, 615-01 from the Institute of Child Health and Human Development, and a Training Grant GM-1751 from the Institute of General Medicine. Hazel Rood assisted in all phases ofthe investiga¬ tion and Dr John King, Michigan State University, also contributed greatly to the study through his co-operation and interest. We are also indebted to the owners of the estancias on which study areas were established, namely Sr G. Walter, Sr C. Olson and Sr R. Pearson, and to the Institute Nacional de Tecnologia Agropecuaria, Buenos Aires, for their donation of the domestic guinea-pigs used in the study. The laboratory work in England was financed by the Ford Foundation, and the expedition to Argentina during 1967 when the cavies were obtained was supported by the Royal Society, the Wellcome Trust and the Medical Research Council. It is impossible to thank individually the many kind people who assisted the expedition, but B.J.W. would like to thank particularly the Rumboll family for their help in Argentina and Dr I. W. Rowlands for his constant encouragement of the project.

REFERENCES Asdell, S. A. (1964) Patterns of mammalian reproduction, 2nd edn. Cornell University Press, Ithaca, New York. Blandau, R. J. & Young, W. C. (1939) The effects of delayed fertilization on the development of the guinea-pig ovum. Am. J. Anat. 64, 303. Cabrera, A. (1954) Los roedores argentinos de la famile a 'Caviidae'. Publnes Esc. Vet. Univ. B. Aires, 6, 93. Castle, W. E. & Wright, S. (1916) Studies of inheritance in guinea-pigs and rats. Pubis Carnegie Insin, 241, 1. Detlefson, J. A. (1914) Genetic studies on a cavy species cross. Pubis Carnegie Instn, 205, 1. Dunkin, G. W., Hartley, P., Lewis-Fanning, E. & Russell, W. T. (1930) A comparative biometrie study of albino and coloured guinea-pigs from the point of view of their suitability for experi¬ mental use. ,7. Hyg., Camb. 30, 311. Goy, R. W., Hoar, R. M. & Young, W. C. (1957) Length of gestation in the guinea-pig with data on the frequency and time of abortion and stillbirth. Anat. Ree. 128, 747. Guyenot, E. & DuszYNSKA-WiETRZYKOwsKA, J. (1935) Sterilite et virilisme chez des femelles de cobayes issues d'un croisement interspécifique. Revue suisse Z°°l· 42, 341. Haines, G. (1931) A statistical study of the relation between various expressions of fertility and vigor in the guinea-pig. J. agrie. Res. 42, 123. Hoyland, F. (1965) Further observations on the build up of a large guinea-pig colony (1964). J. Anim. Techns Ass. XVI, 59. HÜCKINGHAUS, F. (1962) Vergleichende Untersuchungen über die Formenmannig faltigkeit der Unterfamilie-Caviinae Murray 1886 (Ergebnisse de Südamerika-expedition Herre/Röhrs 1956-57). - "¿ - Zool. 166, 1. Ibsen, H. L. (1950) The guinea-pig. In: The Care and Breeding of Laboratory Animals. Ed. E. J. Farris. Wiley, New York.

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access Reproduction infernale wild guinea-pigs 409 Kleiman, D. G. (1969) The reproductive behaviour of the green acouchi, Myoprocta pratti. Ph.D. thesis, University of London. Marshall, F. H. A. (1952-66) Physiology of reproduction, 3rd edn. Ed. A. S. Parkes. Longmans, London. Nehring, A. (1893) Ueber Kreuzungen von Cavia aperea und Cavia cobaya. Sber. Ges. naturf. Freunde Beri. Part 10, 247. Pearson, O. P. (1951) Mammals in the highlands of Southern Peru. Bull. Mus. comp. Z°°l· Harv. 3, 117. Pictet, A. & Ferrerò, A. (1951) La descendance d'un croisement interspécifique de cobayes (Cavia aperea D'Az. x Cavia cobaya Marc.) analysée durant 25 années. Genetica, 25, 357. Reíd, M. E. (1958) The guinea-pig in research: biologi/, nutrition,physiology. Publication 557, Human Factors Research Bureau. National Press, Washington. Rood, J. P. (1970) Ecology and social behaviour of the desert cavy (Microcavia australis). Am. Midi. Nat. 83, 415. Rowlands, L W. (1949) Post-partum breeding in the guinea-pig. J. Hyg., Camb. 47, 281. Ubisch, G. & Mello, R. F. (1940) Genetic studies on a cavy species cross (Cavia rufescens Lund and Cavia porcellus Linné). J. Hered. 31, (9), 389. Weir, B. J. (1966) Aspects of reproduction in chinchilla. J. Reprod. Fert. 12, 410. Weir, . J. ( 196 7a) Aspects ofreproduction in some hystricomorph rodents. Ph .D. thesis, University ofCambridge. Weir, B. J. (1967b) The care and management of laboratory hystricomorph rodents. Lab. Anim. 1, 95. Weir, B. J. (1970) The management and breeding of some more hystricomorph rodents. Lab. Anim. 4, 83. Whitten, W. K. (1956) Modification of the oestrous cycle of the mouse by external stimuli associated with the male. J. Endocr. 13, 399. Wright, S. (1960) The genetics of vital characters of the guinea-pig. J. cell. comp. Physiol. 56, Suppl. 1, 123. Young, W. C, Dempsey, E. W., Hagquist, C. W. & Boling, J. L. (1939) Sexual behaviour and sexual receptivity in the female guinea-pig. J. comp. Psychol. 27, 49.

Downloaded from Bioscientifica.com at 10/10/2021 10:51:38AM via free access