JOURNAL OF EXPERIMENTAL ZOOLOGY 293:171–178 (2002)

Inbreeding in Littledale’s Whistling Rat Parotomys littledalei NEVILLE PILLAY Ecophysiological Studies Research Group, School of , Plant and Environmental Sciences, University of the Witwatersrand, WITS 2050, South Africa

ABSTRACT Despite its rarity in nature, inbreeding is sometimes evident in occupying ephemeral, unpredictable habitats, and which occur at low densities. One such species is Littledale’s whistling rat, Parotomys littledalei, a murid endemic to the south-west arid region of South Africa. Using a captive population of P. littledalei, I studied mate choice for kin and nonkin, and the reproductive performance of inbred and outbred pairs. In choice tests, estrous females presented with either odors or actual males showed a preference for siblings or half-siblings to unrelated males. Males did not discriminate between the odor of estrous kin and nonkin. In breeding studies, inbred (mother-son; brother-sister) and outbred (proven female and an unrelated young male and nonsiblings) pairs had a similar reproductive output, although the sex ratio favored males in inbred litters. The development of inbred young was indistinguishable from outbred young. The results indicate that female P. littledalei prefer to inbreed, but there are no apparent advantages to inbreeding over outbreeding. J. Exp. Zool. 293:171–178, 2002. r 2002 Wiley-Liss, Inc.

Inbreeding is generally regarded as being disruption of gene complexes (Partridge, ’83; disadvantageous, particularly when mating occurs Shields, ’93). between close relatives, such as between parents Ecologically, inbreeding is apparently more and offspring and siblings (Clutton-Brock and common in populations occupying ephemeral, Harvey, ’76). The detrimental effects of inbreeding patchy, or unpredictable microhabitats than in are well known, and they are assumed to be the those occupying habitats that are permanent and result of increased homozygosity of deleterious predictable (Batzli et al., ’77; Anderson, ’89). In recessive alleles (Charlesworth and Charlesworth, ephemeral-patchy microhabitats, the costs of dis- ’87), affecting, for example, the pre- and post- persal may favor philopatry and mating with kin zygotic viability of inbred offspring (Lacy et al., (Smith and Ivins, ’83; Bollinger et al., ’91), ’96) and the reproductive success of inbred adults particularly at low population densities. For (Margulis, ’98). Many have argued and shown that example, the high level of inbreeding in the the detrimental effects of inbreeding lead to meadow vole Microtus pennsylvanicus may be outbreeding mechanisms through behavioral and explained by the high costs of dispersal between physiological adaptations. widely spaced grassland patches (Pugh and Ta- Despite the wide acceptance of inbreeding marin, ’88). avoidance, it is known that close inbreeding may Littledale’s whistling rat Parotomys littledalei is occur when alternative mates are not available. a medium-sized (adult mass range 107–120 g) For example, mating occurs in sibling groups in murid rodent endemic to the south-west arid some insects upon emergence but females will biome of southern Africa. The known distribution later outbreed (Hamilton, ’93; Chapman and pattern of P. littledalei includes the Nama Karoo, Stewart, ’96). Inbreeding may also occur when the costs of avoiding inbreeding outweigh the Grant sponsor: National Research Foundation; Grant sponsor: benefits (Partridge, ’83; Pugh and Tamarin, ’88; University of the Witwatersrand. Shields, ’93). For example, the costs of dispersing Correspondence to: Neville Pillay, Ecophysiological Studies Re- search Group, School of Animal, Plant and Environmental Sciences, or delaying reproduction may be higher than those University of the Witwatersrand, Private Bag 3, WITS 2050, South of mating with close relatives (Rowley et al., ’86). Africa. E-mail: [email protected] Received 9 November 2001; Accepted 28 February 2002 In addition, inbreeding may maintain adaptations Published online in Wiley InterScience (www.interscience.wiley. to local conditions (Partridge, ’83), and prevent com). DOI: 10.1002/jez.10107 r 2002 WILEY-LISS, INC. 172 N. PILLAY succulent Karoo, the southern Kalahari regions in different sites that were unlikely to be closely the South Africa-Namibia border, and along the related. The F1 offspring were outbred and their Namib and Kunene rivers (Coetzee and Jackson, offspring (F2) were used in experiments. ’99). This distribution covers some of the most were singly maintained in Lab-o-tec arid regions of southern Africa where rainfall may cages (400 250 120 mm) or as pairs in glass be less than 200 mm per annum. aquaria (900 300 400 mm). Environmental The presence of P. littledalei in an area is conditions were partially controlled (22–251C; 30– usually indicated by a complex warren system 70% rH; light regime of 15L:9D, lights on at 0500 under bushes (Coetzee and Jackson, ’99; Jackson, hr). Coarse wood shavings were provided as litter 2000). Apart from providing protective cover and hay was provided as nesting material. Cages against predators and extreme environmental were washed and the shavings were replaced conditions, bushes serve as a source of food; P. weekly. Food was comprised of ad libitum spinach, littledalei is strictly herbivorous, feeding on the vegetable tops (e.g., carrot and beetroot), and Epol leaves of a limited number of succulent and rabbit pellets. Water was provided at all times. nonsucculent vegetation (Coetzee and Jackson, ’99). Field studies indicate that the taxon appar- Mate preference ently does not readily venture beyond the warren to feed or forage (Coetzee and Jackson, ’99), and The preference by individual females and males Jackson (2000) maintains that such burrowing for the odor of related and unrelated opposite sex and foraging habits of P. littledalei effectively individuals was tested in two combinations: (i) restrict its distribution within its geographical littermate sibling vs. unrelated individual; and (ii) range. half-sibling (same father) vs. unrelated individual. Coetzee and Jackson (’99) suggest that locating For each combination, I tested the responses of 15 a suitable bush, a limiting resource in the arid estrous females and 14 males; males were tested environments, will influence dispersal in P. lit- with the bedding of estrous females. The estrous tledalei, and cohabitation has been observed in state of all females (test or stimuli) was deter- large bushes in parts of its distributional range. mined using vaginal smears. Whether such a situation leads to inbreeding in Tests were conducted in a neutral arena, nature is unknown, but I observed inbreeding consisting of a square perspex apparatus (400 during an earlier breeding study of captive P. 400 200 mm) connected to three clear, plastic littledalei (Pillay, 2002). Similarly, I showed in two choice chambers (150 150 150 mm) by PVC mesic relatives of P. littledalei that inbreeding is pipes (60 mm internal diameter) at the middle of common in the species ( angoniensis) each side. Each choice chamber contained a petri occupying patchy ephemeral habitats and absent dish (55 mm diameter) with soiled bedding from in the species (O. irroratus) inhabiting dense donor individuals or clean bedding (control). The habitatsFthe latter displaying behavioral in- entrance to each choice chamber was guarded with breeding avoidance (Pillay, ’98). a removable, opaque partition. Prior to each test, The objectives of the present study were to soiled bedding was randomly assigned to one of establish whether male and female P. littledalei the three nest boxes. A test individual was placed discriminate between kin and nonkin in mate in the middle of the test arena. After an acclima- choice tests and to ascertain the reproductive tion period of 10 min, partitions to the choice performance of inbred and outbred pairs. chambers were removed and the time spent within each of the three chambers by test subjects was MATERIALS AND METHODS video-recorded for 15 min under fluorescent lights during the light phase of the light-dark cycleF P. Study animals were live-trapped at Springbok littledalei is a diurnal rodent. After tests, the (291400S, 171530E) in the Northern Cape Province, choice apparatus was thoroughly cleaned with South Africa. Thirty adults were collected from warm soapy water and alcohol, dried, and several sites within the Springbok area. The the floor was covered with a thin layer of wood minimum distance between sites was 400 m, and shavings. although dispersal distances of P. littledalei are Soiled bedding used in choice tests was collected unknown, they are unlikely to exceed 100 m from estrous stimuli females, sealed in air-tight (Jackson, personal communication). In captivity, bags, and stored at 151C until use (after Heth pairings were established between individuals of et al., ’98). It was necessary to freeze samples since INBREEDING IN WHISTLING RATS 173 female P. littledalei do not display synchronous the pair had access to each other and the entire oestrus, making it unlikely that all the appropriate cage. Aggression was minimal or absent at this stimuli were available on one day. For consistency, time, with pairs either nesting together or sepa- bedding from donor males and clean bedding were rately. All females were very aggressive toward also frozen. Frozen samples were thawed out at their mates a week before parturition, so males room temperature just prior to tests. No bedding were removed from the breeding tank at this time was frozen for more than a week. and housed elsewhere. Pairs that had not pro- All test subjects were sexually mature (i.e., 4 duced a litter after 120 days were separated. This 100 days old; Pillay, 2002) virgins that were used time period allowed for 30-day old individuals to once in experiments. None made contact with half- attain sexual maturity (approximately 90 days) siblings or unrelated individuals prior to tests, and and complete gestation (41 days; Pillay, 2002). they were separated from siblings for at least two I compared the reproductive performance of months (i.e., they were housed in different rooms). inbred and unrelated pairs by recording the The effect of relatedness upon female preference proportion of pairs that reproduced, the litter for males, as opposed to odors, was examined in a size, and sex ratio. In addition, reproductive effort separate experiment by comparing individual of females was calculated using a modified version 0.75 0.75 1 preferences of estrous females for a brother and of Millar’s (’77) formula: Re ¼ NW (m ) , an unrelated male, and for a half-brother and an where N ¼ litter size, W ¼ mass of litters at unrelated male; 12 replicates were conducted per weaning (16 days old), and m ¼ mass of post- combination. The protocol of these tests was partum adult female. identical to that of the odor preference tests The viability of inbred relative to outbred except that instead of containing soiled bedding, offspring was ascertained by studying their post- the nest boxes were divided in two using a wire- natal ontogeny. Young were weighed to the mesh partition. In two of these nest boxes, males nearest 0.1g every second day from birth to 20 were placed behind the partition, and the third days of age, and weekly thereafter to 98 days of nest box contained clean bedding behind the age. Using the growth data from those young that partition. When making a choice, test females survived to 98 days of age, the increase in body entered a nest box but could not make physical mass of both males and females was fitted with the contact with males or wood shavings. Again, the Gompertz equation (Zullinger et al., ’84): W ¼ time that a test female spent within each of the A.exp [exp (K x(tti))], where W ¼ predicted three chambers was recorded. In addition to mass at time t; A ¼ asymptotic mass; exp ¼ recording the duration of visits, I also recorded exponential function; K ¼ growth rate constant; ti the number of aggressive-retreat and sexual acts ¼ a parameter indicating the inflection time. The that females displayed toward males across the Gompertz model generated the growth rate and wire mesh barrier. Typical aggressive-retreat inflection point, while the asymptotic value was (hereafter referred to as aggressive) behaviour fixed to the mean body mass at 98 days of age included biting wire mesh, pilo-erection, and when the last weighing was made and mass gain upright sparring. Sexual behaviour included pre- had leveled off (see Pillay, 2002). senting and/or lordotic behaviour. During postnatal development, mass and age data at several developmental stages were of particular biological interest, including litter mass Breeding biology at birth and weaning, and age and mass at sexual Two types of inbred and outbred (control) maturity for individual males and females. Re- pairings were made: (i) mother and son; (ii) productive maturity in males was indicated by the brother and sisterF siblings removed from the presence of scrotal testes that were of adult size mother at 30 days old and housed elsewhere; (iii) (i.e., 2 cm) and in females by a perforate vagina unrelated female and younger maleF a sexually- and completion of at least one estrous cycle (as experienced female and a virgin, 100-day old male; determined by vaginal smears). and (iv) nonsiblingsF both were approximately 100 days of age at pairing. Data analysis In outbred pairs, females displayed extreme aggression to males, necessitating separation of Before using parametric tests, I tested data sets the pair with a wire mesh screen. Following a for departure from normality and for homogenous period of two weeks, the barrier was removed, and variances. If the assumptions of normality were 174 N. PILLAY not met, nonparametric tests were used (Zar, ’96). brothers (2.08 7 0.22). A similar pattern was Data from the choice tests were analyzed using a observed in tests of half-brothers (2.53 7 0.36) Friedman ANOVA to compare the time spent with and unrelated males (4.30 7 0.33; G ¼ 5.96; P o the different types of bedding or males; specific 0.05). In contrast, females directed significantly differences were identified using a nonparametric more sexual acts toward related than unrelated post-hoc test and a Wilcoxon paired ranked test. A males: brothers (3.02 7 0.40) vs. unrelated males G-test was used to analyze the number of (1.13 7 0.12; G ¼ 5.96; P o 0.01); half-brothers aggressive and sexual behaviors displayed by (2.61 7 0.28) vs. unrelated males (0.95 7 0.16; G females to stimulus males. In breeding studies, ¼ 9.92; P o 0.01). statistical analyses used included w2 contingency Like females, males preferred the bedding of analyses and one-way ANOVA and Kruskall- females to unsoiled bedding (Fig. 2). However, Wallis tests. For comparisons of birth mass and males did not discriminate between related and weaning mass of litters, I used the ANCOVA test unrelated females, spending equal time with the to control for the influences of mother’s mass and bedding of sisters and unrelated females (z ¼ 0.53; litter size. All tests were two-tailed and a was set P 4 0.05), and with bedding of half sisters and at 0.05. unrelated females (z ¼ 0.47; P 4 0.05).

RESULTS Breeding biology Inbred and outbred pairs had similar high Mate preference reproductive success and reproductive output in Figure 1 gives the duration of time that estrous terms of litter size (Table 1). However, the sex females spent with the soiled bedding of males or ratio of inbred litters significantly favored males the males themselves. In all cases, females spent while parity was achieved in outbred litters. Post- significantly more time with the stimuli (bedding partum reproductive investment was similar in or males) and showed little interest in unsoiled inbred and outbred litters. bedding (Friedman ANOVA and nonparametric After statistically accounting for litter size and post tests). Therefore, comparisons were made maternal mass, the mass of inbred litters was between the times spent with stimuli using a similar to that of outbred litters at birth and Wilcoxon matched pairs test. In tests with odor weaning (Table 2). Male P. littledalei are signifi- (soiled bedding), females preferred brothers to cantly heavier than females, but only after wean- unrelated males (z ¼ 1.95; P o 0.05) and half- ing (Pillay, 2002), so that the skewed sex ratio of brothers to unrelated males (z ¼ 2.95; P o 0.01). inbred litters did not affect litter masses at birth Similarly, females tested with actual males spent and weaning. Evidence of male-biased sexual significantly more time with brothers than un- dimorphism was provided statistically using an related males (z ¼ 3.06; P o 0.01) and with half- ANOVA for repeated measures (after Ferron and brothers than unrelated males (z ¼ 2.90; P o Ouellet, ’91) with sex as the grouping variable and 0.01). In addition, females displayed significantly weekly readings from day 0 to 98 as the dependent higher (G ¼ 4.82; P o 0.05) levels of aggression variables. From 28 days of age onward, inbred toward unrelated males (3.55 7 0.32) than toward males were significantly heavier than inbred

Fig. 1. Mean time spent by estrous P. littledalei with the odor (soiled bedding) of related and unrelated males, the Fig. 2. Mean time spent by male P. littledalei with the males themselves, and controls. Vertical bars ¼ 1 SE above the soiled bedding of estrous related and unrelated females or mean. controls. Vertical bars ¼ 1 SE above the mean. INBREEDING IN WHISTLING RATS 175

TABLE 1. Reproductive data for inbred and outbred P. littledalei paris1 Unrelated female-younger Parameters Mother-son Brother-sister male Non-siblings Statistics

Proportion of pairs reproducing 11/13 12/12 14/16 12/12 w2=5.27 NS Litter size 2.6 (0.15) 2.3 (0.19) 2.4 (0.13) 2.4 (0.19) E3,45=0.64 NS Sex ratio (M:F) 23:8 21:7 14:19 13:16 w2=11.53; Po0.01 Reproductive e¡ort 2.4 (0.21) 2.3 (0.27) 2.0 (0.18) 1.9 (0.21) H3,49=5.60 NS

1Litter size and reproductive e¡ort values given as mean (7SE).

females (F1,48 ¼ 101.07; P o 0.001), and outbred brothers and half-brothers to unrelated males. In males were heavier than outbred females (F1,46 ¼ addition, females displayed less aggression and 31.69; P o 0.001). more sexual behaviour toward brothers and half- There were no differences between inbred and brothers than unrelated males. The similarity of outbred young in the age and mass at sexual responses by estrous females to males or their maturity (Table 2). Sex-specific growth rate did odors highlights the importance of male odor, and not differ between inbred and outbred young. No in particular deposited chemosignals, in mate significant differences were observed in the num- choice in P. littledalei. ber of young of all pairs that survived to 100 days Females distinguished between half-brothers of age. and unrelated males, despite the fact that they never made contact with these males prior to tests. DISCUSSION Therefore, it appears that prior association is not necessary for the establishment of kin recognition Mate preference and mate preference in P. littledalei, as is the case Female P. littledalei can distinguish between in several other (see Batzli et al., ’77; related and unrelated males. In odor and actual Gavish et al., ’84; Boyd and Blaustein, ’85; Pillay, male preference experiments, females preferred ’98). Thus, it is possible that P. littledalei can use

TABLE 2. Postnatal characteristics of o¡spring resulting from inbred and outbred P. littledalei pairs1 Unrelated female-younger Parameters Mother-son Brother-sister male Non-siblings Statistics

2 Litter mass at birth (g) 30.7 (1.23) 30.5 (1.32) 27.7 (1.75) 27.0 (1.72) F3, 43=1.31 NS n=11 n=12 n=14 n=12 2 Litter mass at weaning (g) 111.8 (6.80) 97.4 (5.03) 97.0 (6.04) 89.8 (5.32) F3, 43=2.20 NS n=11 n=12 n=14 n=11

Male age at sexual maturity (weeks) 7.0 (0.16) 6.6 (0.22) 6.5 (0.3) 6.6 (0.33) H3, 57=2.56 NS n=19 n=18 n=12 n=10

Female age at sexual maturity (weeks) 6.3 (0.36) 6.3 (0.49) 6.7 (0.33) 6.5 (0.31) H3, 39=0.56 NS n=7 n=6 n=13 n=13

Male mass at sexual maturity (g) 92.8 (2.38) 95.1 (2.30) 91.4 (3.32) 88.4 (3.44) F3, 53=0.95 NS n=19 n=18 n=12 n=10

Female mass at sexual maturity (g) 74.7 (2.90) 74.3 (4.02) 72.5 (2.71) 72.7 (2.38) F3,35=1.32 NS n=7 n=6 n=13 n=13 1 Male growth rate: K (day ) 0.045 (0.001) 0.046 (0.001) 0.046 (0.001) 0.045 (0.001) H3, 59=0.19 NS n=19 n=18 n=12 n=10 1 Female growth rate: K (day ) 0.051 (0.004) 0.042 (0.003) 0.045 (0.003) 0.048 (0.003) H3, 39=4.95 NS n=7 n=6 n=13 n=13 Proportion surviving to 100 days of age 26/31 28/29 25/33 23/29 w2=5.27 NS

1All values except proportion of surviving young given as mean (7SE). Sample size=number of litters (birth and weaning mass and growth rates) and number of individuals (age and mass at sexual maturity). 2F values=ANCOVAtest analysis, with maternal mass and litter size as co-variates. 176 N. PILLAY other mechanisms, such as phenotypic matching affect the probability of producing offspring or the (Hepper, ’91), to distinguish between kin and viability of inbred offspring. nonkin. In species where there is sexual Unlike females, males failed to distinguish dimorphism in body size, maternal condition between related and unrelated females. This lack may influence the sex ratio of progeny (Trivers of preference should be interpreted with caution, and Willard, ’73; Myers, ’78). The larger sex is since such a situation could imply either an usually more costly to produce and females will inability to discriminate between the scents or produce fewer of this sex when stressed (Fisher, an equal interest in all scents. It is unlikely that ’30). Using this premise, Lacy and Horner (’97) males require other stimuli in addition to odor for argued that since inbreeding can be stressful mate preference, since females were able to use (because of inbreeding depression) and is known odor for mate choice. During breeding studies, to cause differential survival between male and males demonstrated equal interest in related and female offspring, females may respond by adjust- unrelated estrous females (Pillay, personal obser- ing the sex ratio of offspring in response to vation), suggesting that male P. littledalei do not inbreeding. Although empirical evidence for this discriminate between kin and nonkin. prediction is lacking, a correlation between litter Sexual asymmetry in mate choice for kin and size and sex ratio has been demonstrated in inbred nonkin has also been demonstrated in house mice, Australian rats Rattus villosissimus (Lacy and but the patterns of preference are not always Horner, ’97). These rats have male-biased sexual predictable. Females preferred nonkin to kin in dimorphism, and inbreeding results in either large some laboratory mouse strains (Yanai and litters with more females or small litters with McClearn, ’72). In other strains, females preferred more males. If inbreeding leads to more male close kin to unrelated males, while males pre- young in P. littledalei, does this imply that ferred distantly related females (Barnard and inbreeding is not stressful in this species? An Fitzsimons, ’88). Females of wild house mice did important consideration is that gender differences not discriminate between kin and nonkin but in body mass only occurred after weaning, such males preferred unrelated females (Krackow and that relative investment in inbred and outbred Matuschak, ’91). young (reproductive effort) may not have been affected by the skewed sex ratio. Breeding biology Conclusion Although females preferred brothers and half- brothers in mate preference tests, they bred with Breeding studies revealed no indication of out- unfamiliar males. However, female P. littledalei breeding depression, suggesting that outbreeding displayed high levels of aggression toward unre- would be favored. It is intriguing, therefore, that lated males in choice tests and initially during females preferred full and half siblings in choice breeding experiments. In the latter case, a tests, suggesting that there is some fitness familiarization period was required before females advantage to inbreeding, although such advan- accepted and mated with unrelated males. tages were not observed, since inbred and outbred Captive breeding studies of laboratory and free- pairs had similar reproductive performance. There living rodents have shown that inbreeding can may be other advantages to inbreeding in P. have various negative effects on reproductive littledalei, such as promoting a high degree of success and output. Some studies have shown that adaptation to local conditions (Wilson, ’75; May- inbreeding affects the probability of successful nard Smith, ’78), but the asymmetric preferences reproduction, but not any other measure of in choice tests suggest that there may be sex- reproductive performance (Blaustein et al., ’87; specific fitness benefits in mate preferences. Smith Dewsbury, ’88; Krackow and Matuschak, ’91). (’79) proposed an inclusive fitness argument to Other studies have shown otherwise, and inbreed- explain sex differences in preference for inbreed- ing is known to decrease litter size (e.g., Hill, ’74; ing in polygynous species. Assuming that females Barnard and Fitzsimons, ’89; Margulis, ’98; Pillay, are philopatric and males emigrate from natal ’98) or detrimentally affect offspring growth demes, incestuous females increase the fitness of a (Laird and Howard, ’67; Keane, ’90; Pillay, ’98) male relative (e.g., father, brother) as well as their and survival (Keane, ’90; Haigh, ’83; Pillay, ’98). own inclusive fitness (Maynard Smith, ’78; Smith Inbreeding in P. littledalei does not appear to ’79). In contrast, dominant males mate with INBREEDING IN WHISTLING RATS 177 relatives within a deme, whereas emigrating Blaustein AR, Bekoff M, Daniels TJ. 1987. Kin recognition males may have to outbreed (Smith, ’79). Little in vertebrates (excluding primates): empirical evidence. is known about the social biology of P. littledalei, In: Fletcher DJC, Michener CD, editors. Kin recognition in animals. Chichester: John Wiley. p 287–331. so it is not possible to fully evaluate Smith’s Bollinger EK, Harper SJ, Kramer JM, Barrett GW. 1991. prediction. Avoidance of inbreeding in the meadow vole (Microtus Inbreeding depression occurs because of an pennsylvanicus). J Mamm 72:419–421. increase in homozygosity due to the accumulation Boyd SK, Blaustein AR. 1985. Familiarity and inbreeding of deleterious genes or genetic load (Charlesworth avoidance in the gray-tailed vole (Microtus canicaudus). J and Charlesworth, ’87). Since genetic load is Mamm 66:348–352. Campbell RB. 1986. The interdependence of mating cumulative over generations (Waller, ’93), the structure and inbreeding depression. Theor Popul Biol genetic consequences of inbreeding depend on 30:232–244. the history of parental inbreeding (Campbell, Chapman TW, Stewart SC. 1996. Extremely high levels of ’86). If this is the case in P. littledalei, the inbreeding in a natural population of the free-living wasp apparent lack of inbreeding depression could have Anscistrocerus antilope (Hymenoptera: Vespidae: Eumeni- nae). Heredity 76:65–69. been because the founder population was unre- Charlesworth D, Charlesworth B. 1987. Inbreeding depression lated and initial matings were outbred. In addi- and its evolutionary consequences. Annu Rev Ecol Syst tion, Charlesworth and Charlesworth (’87) 18:237–268. predicted that outcrossing in inbred species could Clutton-Brock TH, Harvey PH. 1976. Evolutionary rules significantly improve fitness. Future studies and primate societies. In: Bateson PPG, Hinde RA, editors. Growing points in ethology. Cambridge: Cambridge should therefore consider the genealogical effects University Press. p 195–237. of inbreeding. Coetzee CG, Jackson TP. 1999. The comparative behaviour I predicted that the ecological circumstances of and ecology of two species of Parotomys (Mammalia, P. littledalei would lead to inbreeding as in the Rodentia, Otomyinae) found in the arid areas of southern meadow vole Microtus pennsylvanicus (Pugh and Africa. J Nam Sci Soc 47:87–106. Tamarin, ’88). P. littledalei did not avoid inbreed- Dewsbury DA. 1988. Kin discrimination and reproductive behavior in muroid rodents. Behav Genet 18:525–536. ing in captivity, and inbreeding did not influence Ferron J, Ouellet J-P. 1991. Physical and behavioral postnatal reproductive performance. However, the relation- development of woodchucks (Marmota monax). Can J Zool ship between dispersal and inbreeding, and the 69:1040–1047. genetic composition of free-living P. littledalei Fisher RA. 1930. The genetical theory of natural selection. populations, should be studied to verify the Oxford: Oxford University Press. Gavish LH, Hoffman JE, Getz LL. 1984. Sibling recognition findings made here. in the prairie vole, Microtus ochrogaster. Anim Behav 32:362–366. ACKNOWLEDGMENTS Haigh GR. 1983. Effects of inbreeding and social factors on the reproduction of young female Peromyscus maniculatus I thank D. Padiachy, J. Lancaster, A. Reuther, baidii. J Mamm 64:48–54. Hamilton WD. 1993. Inbreeding in Egypt and in the book: a U. Schwaibold, C. Schradin, and staff of the childish perspective. In: Thornhill NW, editor. The natural Central Animal Services for technical assistance, history of inbreeding and outbreeding. Chicago: University and Northern Cape Nature Conservation for of Chicago Press. p 429–450. permission to collect animals. Animal ethics Hepper PG. 1991. Recognizing kin: ontogeny and classifica- clearance number: AESC 99/08/2a. tion. In: Hepper PG, editor. Kin recognition. Cambridge: Cambridge University Press. p 259–288. Heth G, Todrank J, Johnston RE. 1998. Kin recognition in LITERATURE CITED golden hamsters: evidence for phenotypic matching. Anim Behav 56:409–417. Anderson PK. 1989. Dispersal in rodents: a resident fitness Hill JL. 1974. Peromyscus: effect of early pairing on reproduc- hypothesis. Special publication No. 9. The American Society tion. Science 186:1042–1044. of Mammalogists. Jackson TP. 2000. Adaptation to living in an open Barnard CJ, Fitzsimons J. 1988. Kin recognition and mate arid environment; lessons from the burrow structure choice in mice: the effects of kinship, familiarity and social of the two southern African whistling rats, Parotomys interference on intersexual interaction. Anim Behav brantsii and P. littledalei. J Arid Environ 46: 36:1078–1090. 345–355. Barnard CJ, Fitzsimons J. 1989. Kin recognition and mate Keane B. 1990. Dispersal and inbreeding avoidance in the choice in mice: fitness consequences of mating with kin. white-footed mouse, Peromyscus leucopus. Anim Behav Anim Behav 38:35–40. 40:143–152. Batzli GO, Getz LL, Hurley SS. 1977. Suppression of growth Krackow S, Matuschak B. 1991. Mate choice for non-siblings and reproduction of microtine rodents by social factors. in wild house mice: evidence from a choice test and a J Mamm 58:583–591. reproductive test. Ethology 88:99–108. 178 N. PILLAY

Lacy RC, Horner BE. 1997. Effects of inbreeding on the Pugh SR, Tamarin RH. 1988. Inbreeding in a population of reproduction and sex ratio of Rattus villosissimus. J Mamm meadow voles, Microtus pennsylvanicus. Can J Zool 78:877–887. 66:1831–1834. Lacy RC, Alaks G, Walsh A. 1996. Hierarchical analysis of Rowley I, Russell EM, Brooker MG. 1986. Inbreeding: benefits inbreeding depression in Peromyscus polionotus. Evolution may outweigh costs. Anim Behav 34:939–941. 50:2187–2200. Shields WM. 1993. The natural unnatural history of inbreed- Laird AK, Howard A. 1967. Growth curves in inbred mice. ing and outbreeding. In: Thornhill NW, editor. The natural Nature 213:786–788. history of inbreeding and outbreeding. Chicago: University Margulis SW. 1998. Differential effects of inbreeding of Chicago Press. p 143–169. at juvenile and adult life-history stages in Peromyscus Smith AT, Ivins BL. 1983. Colonization in a pika popula- polionotus. J Mamm 79:326–336. tion: dispersal vs philopatry. Behav Ecol Sociobiol 13:37–47. Maynard Smith J. 1978. The sex habit of plants and animals. Smith RH. 1979. On selection for inbreeding in polygynous In: Christiansen FB, Fenchel TM, editors. Measuring animals. Heredity 43:205–211. selection in natural populations. Berlin: Springer-Verlag. Trivers RL, Willard DE. 1973. Natural selection of parental p 315–331. ability to vary the sex ratio of offspring. Science 179:90–92. Millar JS. 1977. Adaptive features of mammalian reproduc- Waller DM. 1993. The statics and dynamics of mating system tion. Evolution 31:370–386. evolution. In: Thornhill NW, editor. The natural history of Myers JH. 1978. Sex ratio adjustment under food stress: inbreeding and outbreeding. Chicago: University of Chicago maximization of quality or numbers of offspring. Am Nat Press. p 97–117. 112:381–388. Wilson EO. 1975. Sociobiology: the new synthesis. Cambridge, Partridge L. 1983. Non-random mating and offspring fitness. MA: Harvard University Press. In: Bateson P, editor. Mate choice. Cambridge: Cambridge Yanai J, McClearn GE. 1972. Assortative mating in mice and University Press. p 227–255. the incest taboo. Nature 238:281–282. Pillay N. 1998. The effects of inbreeding in two otomyine Zar JH. 1996. Biostatistical analysis, 3rd edition. London: rodents. Mammalia 62:521–528. Prentice Hall, Inc. Pillay N. 2002. Reproduction and postnatal development in Zullinger EM, Robert ER, Kent HR, Mace GM. 1984. Fitting Littledale’s whistling rat Parotomys littledalei. Mammalia sigmoidal equations to mammalian growth curves. J Mamm 66:71–82. 65:607–636.