E CoCIEsNCE. On reproduction in lemmings 1. Introduction

Transcription

1 E CoCIEsNCE 8 (2) : (2001) On reproduction in lemmings 1 John S. MILLAR, Department of Zoology, University of Western Ontario, London, Ontario N6A 5B7, Canada, jsmillar@julian.uwo.ca Abstract: Many northern small mammals have restricted breeding seasons, which results in limited opportunities for reproduction by adults and delayed maturation until one year of age. Survival and longevity is enhanced under such conditions and the potential for population growth is low relative to small mammals in temperate environments. However, some northern small mammals, such as lemmings, are regarded as having high reproductive rates and a tremendous potential for population growth. Here I review what is known about lemming life histories. Available data indicates that lemmings have the potential for both summer and winter breeding seasons. During the summer, adults are generally restricted to only one or two litters during a short breeding season, and only the earliest-born young-of-the-year can mature in the summer of their birth. In most years, very few or no young mature in the summer of their birth, and in some years survival of nestlings is very poor. Survival to the winter breeding season is important for the persistence of populations under these conditions. Reproduction by lemmings in winter is poorly documented, but years with little or no winter breeding are not uncommon. Summers in which no young mature, or reproduction is unsuccessful, followed by winters in which no breeding occurs would favor lemmings that can successfully reproduce at an advanced age. A complete understanding of the role of age-specific reproduction in the dynamics of lemming populations awaits further study. Keywords : lemmings, reproduction, maturation, survival. Résumé : Plusieurs petits mammifères nordiques présentent des saisons de reproduction réduites qui limitent les opportunités de reproduction par les adultes et retardent la maturation sexuelle jusqu à l âge de 1 an. Sous ces conditions, la survie et la longévité sont augmentées et le potentiel de croissance des populations est relativement faible par rapport aux petits mammifères dans les environnements tempérés. Cependant, certains petits mammifères nordiques, comme les lemmings, sont considérés comme des espèces avec un taux de reproduction élevé et un énorme potentiel de croissance des populations. Ci-après, je passe en revue les connaissances sur le cycle biologique des lemmings. Les données disponibles indiquent que les lemmings ont des périodes de reproduction potentielles en été comme en hiver. Au cours de l été, les adultes sont généralement restreints à une ou deux portées pendant une courte saison de reproduction et seulement les jeunes de l année nés les plus tôt peuvent atteindre la maturité sexuelle lors de l été de leur naissance. La plupart des années, très peu ou pas de jeunes atteignent la maturité sexuelle lors de l été de leur naissance et, certaines années, la survie des jeunes au nid est très faible. La survie jusqu à la saison de reproduction hivernale est importante pour la persistance des populations soumises à ces conditions. La reproduction des lemmings en hiver est peu documentée, mais des années avec peu ou pas d effort reproducteur sont choses communes. Les lemmings capables de se reproduire à un âge avancé seraient favorisés lors d années où il n y a pas maturation sexuelle chez les jeunes de l année ou un succès de reproduction nul en été, suivi d un hiver sans effort de reproduction. Davantage d études sont nécessaires à la compréhension complète du rôle de la reproduction par classe d âge dans la dynamique des populations de lemmings. Mots-clé : lemmings, reproduction, maturation sexuelle, survie. 1 Rec ; acc Introduction Small northern rodents exhibit severe fluctuations in population density that may be related to the seasonality of their environment (Schaffer & Tamarin, 1973; Ostfeld & Tamarin, 1986). High reproductive rates and a tremendous potential for population growth have been noted during the increase phase of these fluctuations (Boonstra, Krebs & Stenseth, 1998; Yoccoz & Ims, 1999), implying that these species must be r-selected relative to most small mammals, with early maturation, large litters, frequent breeding, and short life spans. However, these conclusions were based primarily on assessments of population variability, and relatively few studies of life history parameters have been made on northern rodents. As noted by Chitty (1996) when discussing pioneering studies on vole populations, Lemmings and voles have the capacity to increase rapidly, so this phase, however spectacular, seemed understandable, and only vague hypotheses could be entertained about the relevance of reproduction to the cycle problem. This omission has been unfortunate because an understanding of the balance (or imbalance) between reproduction and mortality is fundamental to an understanding of population processes, and there is increasing evidence from life history studies that the growth potential of many northern rodent populations may not, in fact, be particularly high. Reproductive rates (the replacement rate per generation) are relevant to population dynamics only when they are considered in relation to generation time (the average age at which offspring are produced). Indeed, the instantaneous rate of increase for populations is defined as the log of reproductive rates divided by generation time. Reproductive rates per se depend on litter size and number of litters per lifetime, which may not be highly variable among populations. Numerous studies have produced no clear consensus on differences in litter size between northern and southern populations of small mammals. Some studies have suggested larger litters in northern populations than in

2 MILLAR: REPRODUCTION IN LEMMINGS southern populations, but others have found no trend or smaller litters in the north (see Millar, 1989 and Innes & Millar, 1994 for reviews on mice and voles). Similarly, number of litters per female per breeding season may not be highly variable. Post-partum breeding may be more common in northern environments than southern environments (Millar, 1984) but the lifetime production of litters appears to be approximately two in most populations. (Innes & Millar, 1987; Jacquot & Vessey, 1998). On the other hand, it has long been recognized that instantaneous rates of increase are greatly influenced by age at maturation because age at maturation dictates generation time (Cole, 1954). Age at maturation (and hence generation time) is highly variable within and among species because most small mammals are seasonal breeders, with shorter breeding seasons in northern environments than in southern environments (see Millar, 1984; 1989; Innes & Millar, 1994; Tkadlec & Zejda, 1998 for reviews). In temperate environments, many young have the potential to mature during their summer of birth and only the late-born young forego maturation until the next year. In short-season environments, only the earliest-born young mature during the summer of their birth, and most young mature the following spring. No (or very few) young mature in the summer of their birth in many small mammals with short breeding seasons (Gilbert & Krebs, 1991; Millar & Innes, 1985; Kostian, 1970; Wolff & Lidicker, 1980; Van Horne, 1981; 1982; Yoccoz & Ims, 1999), and this delayed maturation is offset by enhanced survival and longevity (Millar, 1984; 1994; Yoccoz & Mesnager, 1998). This effectively increases the generation time for the population, which drastically decreases the potential for population growth. Delayed maturation and enhanced longevity makes these small mammals K-selected relative to many southern small mammals. Lemmings are of special interest among northern small mammals because they are specifically adapted to arctic conditions, exhibit severe population fluctuations, and are known to breed both in summer and winter (see Stenseth & Ims, 1993a for a recent review). It is generally recognized that lemmings do not breed during snow-melt in the spring and that breeding ceases at the end of the summer, before snow accumulates, (Dunaeva, 1948; Mullen, 1968; Koshkina & Khalanski, 1962; Krebs, 1964; Koponen, 1970) so that they have a relatively short summer breeding season. They also do not breed during winter in all years, although winter breeding is thought to be important to their population dynamics (Hansson, 1984; Stenseth & Ims, 1993a; Krebs, 1993). However, many of their life history attributes remain unknown. For example, a recent review of reproductive parameters in lemmings (Stenseth & Ims, 1993b) indicates that almost nothing is known about age at maturation in the field. Among 10 species surveyed, age at maturation was listed as unknown for seven of them. Among the remaining three, the data for Dicrostonyx groenlandicus were from a laboratory colony (Hansen, 1957) the data for Synaptomys cooperi were collected in a temperate environment (Conner, 1959), and maturation in Lemmus lemmus was identified by cohort rather than age (Koponen, 1970). Some information about reproductive rates and patterns of maturation can be surmised if the time constraints on reproduction are known. Gestation and interbirth intervals in L. lemmus average 21 days in the laboratory (Semb- Johansson, Engh & Ostbye, 1993), which appears typical of lemmings in general (Stenseth & Ims, 1993b). Dunaeva (1948) recorded longer interbirth intervals in L. obensis = sibiricus (23-46 days), indicating that postpartum breeding may be delayed in some females in the field, but 21 days provides a likely minimum frequency at which litters can be born to wild dams. Several sources report about three weeks for maturation in the laboratory (reviewed by Stenseth & Ims, 1993b), so that another 21 days is likely the minimum age at which wild female lemmings mature under ideal conditions for growth in the nest. Therefore, an assumption of 21 days gestation and a further 21 days to maturation provides minimum estimates of the time constraints on reproduction by lemmings. These constraints can be applied to documented breeding phenologies to reveal probable breeding patterns. The purpose of this investigation was to review the available data on reproduction by lemmings in the wild in order to construct their probable life histories relative to other northern small mammals. Because lemmings can breed in summer or winter, referring to age categories as overwintered (OW) or young-of-the-year (YY) may be misleading. Therefore, for purposes of this review, OW refers to animals born in any previous breeding season and YY refers to animals born during the current breeding season. Reproduction during the summer Six studies (Dunaeva, 1948; Koshkina & Khalanski, 1962; Krebs, 1964; Mullen, 1968; Koponen, 1970; Erlinge et al., 2000) recorded sufficient data to be able to infer breeding patterns during the summer. Dunaeva (1948) documented reproduction in L. obensis and in D. torquatus (in 1941 and , respectively) on the Yamal Peninsula, U.S.S.R. The data were qualitative, but L. obensis began breeding in June 1942, with young emerging in early July. No pregnant females were caught after early August. Within these time constraints, OW females could produce 1-2 litters during the summer, which was confirmed from counts of placental scars in females collected in August. Only one of 25 adult females had three sets of placental scars at that time. However, some of the early-born YY females could have produced one litter before the end of the breeding season. A similar phenology was noted in D. torquatus, although only 10% of subadults were recorded as breeding in 1942 compared to 50% in 1941, indicating that very few YY females matured during the summer of Koshkina and Khalanski (1962) collected L. lemmus on the Kola Peninsula, U.S.S.R. during May-November, They recorded the first weaned young June 16 and the first pregnant YY on June 26. Many YY females (68%) had bred by late July, although immature females predominated in the population in August. No pregnant females were collected after August. Assuming July 5 to be a reasonable median weaning (and breeding) date for first-litter young, they would have been born to OW females who gave birth 146

3 ÉCOSCIENCE, VOL. 8 (2), 2001 approximately June 14. Survivors among the OW females could also give birth to second and third litters on July 5 and July 26, respectively. Young from the first cohort could give birth to one litter approximately July 26 and a few may have given birth to a second litter approximately 16 August. However, it is unlikely that their second litter young (born July 5 and weaned July 26) could have bred during the summer of their birth. Krebs (1964) studied L. trimucronatus and D. groenlandicus at Baker Lake, N.W.T. in The breeding phenology of both species showed high pregnancy rates in June, a decline in breeding in August, and breeding into September only in 1959 and Over the four years of the study, the average date of first parturitions by OW L. trimucronatus females was July 3. From this, weaning and maturation of first-litter YY occurred approximately July 25, with parturitions by YY occurring approximately August 15. Some early-born young could mature during the summer of their birth in 1959 and 1962, (as indicated by 61 and 58% pregnancy rates by first cohort young in those years, respectively) but no pregnancies were recorded after late July/early August in 1960 and Therefore, very few YY females matured in the summer of their birth in those years (as indicated by an 11% pregnancy rate among small females in 1960). The rate of maturation by YY in 1961 could not be determined conclusively because of a low rate of nestling/weanling survival (< 10%) in that year. However, three of four small females from the first cohort were pregnant during the summer of 1961, indicating a possible high rate of maturation of YY in that year. With slight differences in phenology, the same pattern of maturation and nest survival was evident in D. groenlandicus in all years. Mullen (1968) monitored breeding in L. trimucronatus in Alaska In 1960, the breeding season was very short (8 weeks). OW females could produce 1-2 litters in this time, but no YY females matured in the summer of their birth. In 1961, the breeding season was long (11 weeks), but very few females were lactating and very few weaned young were caught, indicating that failure in the nest may have been common. In 1962 and 1963, the breeding season was also long (10 and 12 weeks, respectively) and breeding by YY females was common. Given that OW females require six weeks from conception to weaning for their first litter, some early-born YY females could have produced a litter in these longer breeding seasons. Koponen (1970) studied L. lemmus at Kilpisjarvi, Finland in 1960, using age of young based on pelage characteristics in order to estimate their dates of parturition. He identified June 6-11 as the median parturition date of the first summer cohort born to OW females. Some YY females matured after June 21-30, but none did so after July 1-10, so that very few of the YY matured in the summer of their birth. However, breeding by OW females persisted into August. OW females giving birth to the first cohort (June 6-11) could have produced at least one subsequent litter during the summer breeding season. Finally, Erlinge et al., (2000) sampled populations of L. sibiricus in eastern Siberia during July-early August 1994 and recorded 6 and 100% breeding by YY females in peak and increasing populations, respectively. Reproduction during the winter Winter breeding is generally considered to be an important aspect of the biology of lemmings, and reproduction in the subnivean environment has been recorded for most species (Stenseth & Ims, 1993a). However, this physiological miracle has not been studied extensively (Krebs, 1993), and it is not known why they breed so extensively in winter (Stenseth & Ims, 1993b). Winter reproduction by northern small mammals is indeed a miracle given the nature of their biology and reproduction. First, small mammals are endotherms with high metabolic intensities, which gives them a fast physiological response time but little fasting endurance. Their metabolic costs increase with both decreasing temperature and increasing latitude (Speakman, 2000) They do not fare well in inclement weather, as indicated by records of nonpredator deaths (Dunaeva, 1948; Rausch, 1950; Krebs, 1964) and the impact of weather on their population dynamics (Scott, 1993; Reid & Krebs, 1996). Second, they are income breeders with few somatic reserves (Millar, 1987; Jonsson, 1997); energy to support reproduction must be obtained from the environment as needed. Third, they generally produce several young per litter, which makes reproduction energetically expensive, particularly during late lactation when the biomass of growing offspring can greatly exceed the biomass of the dam (Kaczmarski, 1966; Migula, 1969; Millar, 1975). Fourth, neonates are altricial and nestlings are ectothermic during early development (Hill, 1992); frequent nest attendance by the dam is critical to nestling survival. For these reasons, breeding success in small mammals is uncertain at all times. However, as income breeders, small mammals do not deplete somatic reserves to support reproduction, so that physiological trade-offs between current and future reproduction are minimal. Costs of reproduction are related primarily to the time commitment needed to complete pregnancy and lactation. This means that small mammals should breed whenever environmental conditions and nutritive resources are suitable. If the environment subsequently becomes unsuitable for reproduction, the process can be terminated and reinitiated in anticipation of better conditions in the future. Indeed, failure in the nest is common among small mammals, including lemmings. For example, Krebs (1964) estimated survival of nestling L. trimucronatus and D. groenlandicus (to 28 days) in the summer to never exceed 50%, Bondrup-Nielsen & Ims (1988) recorded an almost total failure of Myopus schisticolor in the nest during one of two summer breeding seasons, and Reid, Krebs & Kenney (1995) recorded nest mortality in D. kilangmiutak = groenlandicus to be 60, 70, and 43% during , respectively. If failures of this magnitude occur during the summer, the probability of successfully breeding under more stringent winter conditions may be quite low in most years. That lemmings breed during winter, at least in some years, is well documented. Much of this documentation is based on incidental records of breeding in winter (Krebs, 1964; Hansson, 1984; Kaikusalo & Tast, 1984), or inferred from samples from summer populations (Krebs, Boonstra & Kenney, 1995) but extensive winter sampling was conducted in a few studies. 147

4 MILLAR: REPRODUCTION IN LEMMINGS Dunaeva (1948) recorded the ratio of young to adult L. obensis recovered from the stomachs of harvested foxes during four winters ( , , , and ). No young were recorded in , indicating that breeding did not occur or was unsuccessful that winter. Young comprised only 37, 12, and 26% of the samples in , , and , respectively, indicating that breeding was of low intensity or not very successful in those winters. Koshkina and Kholanski (1962) collected L. lemmus in May 1958 and found evidence of previous subnivean reproduction in 47% of all females, which suggests that 53% of the females were immature. Among breeding females, counts of placental scars indicated that they had produced 1-2 litters (average 1.3) prior to snow melt. Litter size averaged 3.7 young, so that the total production of young can be estimated as 2.4 daughters per female, and the percentage of immature females in the population at birth can be estimated as 70%. With the population consisting of 53% immature females in May, survival of young in the subnivean environment must have been almost as high as that of adults. Mullen (1968) collected winter samples of L. trimucronatus during October-December 1960, and every month from December 1962 to April No breeding was recorded in In , 28% of females collected December-March were pregnant, and in , 13% were pregnant during the same months. Unfortunately, age classes of non-pregnant females were not recorded and it is not known if these females represent non-breeding adults or immature animals. MacLean, Fitzgerald and Pitelka (1974) also noted winter breeding in L. trimucronatus in Alaska, inferred from the remains of young lemmings in winter nests after snowmelt They recorded high levels of winter reproduction in 1969 and 1971, and moderate reproduction in 1972, but no breeding in 1970 and only low levels of reproduction in Koponen (1970) and Henttonen and Jarvinen (1981) estimated winter reproduction in L. lemmus in 1960 and 1978, respectively, from the age structure of samples collected after snowmelt. In both years, reproduction in the subnivean environment appeared to be intense during March to early April. Breeding females could produce 1-2 litters during this time, but their offspring would not likely mature until the summer breeding season. Finally, Krebs, Boonstra and Kenney (1995) inferred the presence, but not the intensity, of winter breeding in D. groenlandicus = kilangmiutak from the presence of young lemmings at the beginning of the summer breeding seasons Reid & Krebs (1996) also inferred winter breeding from population changes over winter in 1991, but suggested that little or no winter breeding occurred in Discussion The available data on lemming life histories in the wild is summarized in Table I. These data are qualitative and undoubtedly biased in favor of years when lemmings were abundant, but some generalizations can be surmised from these studies. TABLE I. Synopsis of reproduction in lemmings during winter and summer. Winter breeding refers to the winter prior to the summer breeding season. YY refers to young of the year born in that season. Source Species Year Winter breeding Summer breeding Dunaeva (1948) L. obensis 1937 Moderate intensity or success (37% YY) 1938 No breeding 1940 Low intensity or success (12% YY) 1941 Moderate intensity or success (26% YY) >1 litter by adults; 1 by some early YY D. torquatus 1941 >1 litter by adults; 1 by 50% of early YY 1942 >1 litter by adults; 1 by 10% of early YY Koshkina & Khalanski (1962) L. lemmus 1958 >1 litter in March; no breeding by YY >1 litter by adults; >1 by some early YY Krebs (1964) L. trimucronatus 1959 >1 litter by adults; 1 by 61% of early YY 1960 >1 litter by adults; 1 by 11% of early YY D.groenlandicus 1961 >1 litter by adults; <10% nest survival 1962 >1 litter by adults; 1 by 50% of early YY Mullen (1968) L. trimucronatus 1960 >1 litter by adults; none by YY 1961 No breeding >1 litter by adults; very low nest success 1962 >1 litter by adults; 1 by some early YY 1963 Moderate intensity or success (28% preg.) >1 litter by adults; 1 by some early YY 1964 Low intensity or success (13% preg.) MacLean et al. (1974) L. trimucronatus 1969 High intensity 1970 No breeding 1971 High intensity 1972 Moderate intensity 1973 Low intensity Koponen (1970) L. lemmus 1960 >1 litter in March; no breeding by YY >1 litter by adults; 1 by a few early YY Henttonen & Jarvenin (1981) L. lemmus 1978 >1 litter in March; no breeding by YY Krebs et al. (1995) D. groenlandicus 1988 some breeding 1989 some breeding 1990 some breeding Reid & Krebs (1996) D. kilangmiutak 1991 some breeding 1992 little or no breeding Erlinge et al. (2000) L. sibiricus % breeding by YY 148

5 ÉCOSCIENCE, VOL. 8 (2), 2001 The general pattern of summer reproduction is that OW females begin breeding following snowmelt (in June). OW females can produce two or more litters during the summer if they survive the breeding season. However, mortality during the breeding season likely erodes numbers to the extent that an average OW female produces only one or two litters during the summer. For example, Krebs (1964) estimated summer survival of OW L. trimucronatus and D. groenlandicus to average 0.38 and 0.33 per 28 days, respectively, so that the survival of OW females over a three-month breeding season would be only about 5%. Early-born YY appear able to reproduce in the summer of their birth only in some years (Table I). Among the 18 species-years documented, seven had very few or no young females maturing in the summer of their birth. Among the remaining speciesyears, only the earliest-born young matured during the summer, and very few of them produced more than one litter within the constraints of the breeding season. These limits to reproduction by YY were compounded by very low survival in the nest in three cases. Under these circumstances, populations could increase significantly during the summer only in years with successful breeding by both OW females and their YY daughters. Populations could increase very little during years with no maturation by YY females, and numbers would necessarily decline in those years with total or near-total failure in the nest. This pattern of summer reproduction is similar to that exhibited by other northern small rodents, although the potential for some early-born young to mature in the summer of their birth in some years gives lemmings the potential for higher population growth than northern mice and voles that do not mature in the summer of their birth. However, even successful summer reproduction is insufficient to produce a many-fold increase in numbers. Winter reproduction must be involved when high population growth is exhibited by lemmings. Winter reproduction in lemmings is poorly documented, but appears to be characterized by low to only moderate breeding intensity or success in most years, and no breeding in others (Table I). However, if breeding is successful during winter and the following summer, or during summer and the following winter, the potential for population growth is high because more than one generation can mature within a year. Peak populations in lemmings are found only every several years (Turchin et al., 2000) so that sequential, successful breeding seasons must be infrequent. Nevertheless, such occurrences are sufficient to give lemmings a reputation for having rapid population growth. Having said this, it must be recognized that, in most years, lemmings exhibit slow population growth because very few of them mature in the season of their birth. Most females born during the summer (all of them in some years), do not mature until the winter breeding season when they are several months of age. Similarly, those born in March to early April do not likely mature until the summer breeding season in June. Added to this pattern of delayed maturation is the fact that summer breeding fails in some years and reproduction does not occur (or is not successful) in some winters. If a summer with failed reproduction is followed by a winter with no reproduction, the persistence of the population must rely on successful breeding by females that exceed 15 months of age. If a summer with failed reproduction is both preceded and followed by winters with no reproduction, the persistence of the population must rely on successful breeding by females that exceed two years of age. Under these circumstances, selection for physiological longevity and successful reproduction at an advanced age would be very strong. The full extent to which longevity and delayed maturation is important to the biology of lemmings cannot be determined from available data because the frequencies of winter breeding, delayed maturation, and failed reproduction in lemmings are largely unknown. Without such data, demographic hypotheses to explain cyclic fluctuations in abundance (such as the senescence hypothesis [Boonstra, 1994] and the age at maturation hypothesis [Oli & Dobson, 1999]) cannot be tested. Such information requires commitment to longitudinal studies of lemming life histories, both summer and winter, over multiple years. Acknowledgements This study was supported by the Natural Sciences and Engineering Council of Canada. I thank R. Boonstra, A. Schulte- Hostedde, J. Wolff, and N. Yoccoz for comments on early drafts of the manuscript. Literature cited Bondrup-Nielsen, S. & R. A. Ims, Demography during a population crash of the wood lemming, Myopus schistocolor. Canadian Journal of Zoology, 66: Boonstra, R., Population cycles in microtines: The senescence hypothesis. Evolutionary Ecology, 8: Boonstra, R., C. J. Krebs & N. C. Stenseth, Population cycles in small mammals: The problem of explaining the low phase. Ecology, 79: Chitty, D., Do Lemmings Commit Suicide? Oxford University Press, New York. Cole, L. C., The population consequences of life history phenomena. Quarterly Review of Biology, 29: Conner, P. F., The bog lemming Synaptomys cooperi in southern New Jersey. Publication of the Museum, Michigan State University, Biological Series, 1: Dunaeva, T. N., Comparative survey of the ecology of the tundra voles of the Yamal, Peninsula. Trudy Institute of Demography, Academy of Sciences of USSR, XLI: (Translated by G. MacDonald, edited by W. A. Fuller.) Erlinge, S., D. Hasselquist, M. Svensson, P. Frodin & P. Nilsson, Reproductive behaviour of female Siberian lemmings during the increase and peak phase of the lemming cycle. Oecologia, 123: Gilbert, S. B. & C. J. Krebs, Population dynamics of Clethrionomys and Peromyscus in southwestern Yukon Holarctic Ecology, 14: Hansen, R., Development of young varying lemmings (Dicrostonyx). Arctic, 10: Hansson, L., Winter reproduction of small mammals in relation to food conditions and population dynamics. Carnegie Museum of Natural History Special Publication, 10:

6 MILLAR: REPRODUCTION IN LEMMINGS Henttonen, H. & A. Jarvinen, Lemmings in 1978 at Kilpisjarvi: Population characteristicsof a small peak. Memoranda Society Fauna Flora Fennica, 57: Hill, R. W., The altricial/precocial contrast in the thermal relations and energetics of small mammals. Pages in T. E. Tomasi & T. H. Horton (ed.). Mammalian Energetics. Cornell University Press, Ithaca, New York. Innes, D. G. L. & J. S. Millar, Life histories of Clethrionomys and Microtus (Microtinae). Mammal Review, 24: Innes, D. G. L. & J. S. Millar, The mean number of litters per breeding season in small mammal populations: A comparison of methods. Journal of Mammalogy, 68: Jacquot, J. J. & S. H. Vessey, Recruitment in white-footed mice (Peromyscus leucopus) as a function of litter size, parity, and season. Journal of Mammalogy, 79: Jonsson, K. I., Capital and income breeding as alternative tactics of resource use in reproduction. Oikos, 78: Kaczmarski, F., Bioenergetics of pregnancy and lactation in the bank vole. Acta Theriologica, 11: Kaikusalo, A. & J. Tast, Winter breeding of microtine rodents at Kilpisjarvi, Finnish Lapland. Carnegie Museum of Natural History Special Publication, 10: Koponen, T., Age structure in sedentary and migratory populations of the Norwegian lemming, Lemmus lemmus (L.), at Kilpisjarvi in Annales Zoologici Fennici, 7: Koshkina, T. V. & A. S. Khalanski, Reproduction of the Norway lemming (Lemmus lemmus L.) on the Kola Peninsula. Zoological Zhurnal, XLI: (Translated by W. A. Fuller) Kostian, E., Habitat requirements and breeding biology of the root vole, Microtus eoconomus (Pallas), on the shore meadows in the Gulf of Bothnia, Finland. Annales Zoologici Fennici, 7: Krebs, C. J., The lemming cycle at Baker Lake, Northwest Territories, during Arctic Institute of North America Technical Paper, 15: Krebs, C. J., Are lemmings large Microtus or small reindeer? A review of lemming cycles after 25 years and recommendations for future work. Pages in N. C. Stenseth & R. A. Ims (ed.). The Biology of Lemmings. Academic Press, London. Krebs, C. J., R. Boonstra & A. J. Kenney, Population dynamics of the collared lemming and the tundra vole at Pearce Point, Northwest Territories, Canada. Oecologia, 103: MacLean, S. F. Jr., B. M. Fitzgerald, & F. A. Pitelka, Population cycles in arctic lemmings: Winter reproduction and predation by weasels. Arctic and Alpine Research, 6: Migula, P., Bioenergetics of pregnancy and lactation in the European common vole. Acta Theriologica, 14: Millar, J. S., Tactics of energy partitioning in breeding Peromyscus. Canadian Journal of Zoology, 53: Millar, J. S., Reproduction and survival of Peromyscus in seasonal environments. Carnegie Museum of Natural History Special Publication, 10: Millar, J. S., Energy reserves in breeding small rodents. Zoological Society of London Symposia, 57: Millar, J. S., Reproduction and development. Pages in G.L. Kirkland, Jr. & J. N. Layne (ed.). Advances in the Study of Peromyscus (Rodentia). Texas Tech University Press, Lubbock, Texas. Millar, J. S., Senescence in a population of small mammals? Écoscience, 1: Millar, J. S. & D. G.-L. Innes, Breeding by Peromyscus over an elevational gradient. Canadian Journal of Zoology, 63: Mullen, D. A., Reproduction in brown lemmings (Lemmus trimucronatus) and its relevance to their cycle of abundance. University of California Publications in Zoology, 85: Oli, M. K. & F. S. Dobson, Population cycles in small mammals: The role of age at sexual maturity. Oikos, 86: Ostfeld, R. S. & R. H. Tamarin, The role of seasonality in vole cycles. Canadian Journal of Zoology, 64: Rausch, R., Observations of a cyclic decline of lemmings (Lemmus) on the arctic coast of Alaska during spring Arctic, 3: Reid, D. G. & C. J. Krebs, Limitation to collared lemming population growth in winter. Canadian Journal of Zoology, 74: Reid, D. G., C. J. Krebs, & A. Kenney, Limitation of collared lemming population growth at low densities by predation mortality. Oikos, 73: Schaffer, W. M. & R. H. Tamarin, Changing reproductive rates and population cycles in lemmings. Evolution, 27: Semb-Johansson, A., C. E. Engh, & E. Ostbye, Reproduction, litter size and survival in a laboratory strain of the Norwegian lemming (Lemmus lemmus). Pages in N. C. Stenseth & R. A. Ims (ed.). The Biology of Lemmings. Academic Press, London. Scott, P. A., Relationship between the onset of winter and collared lemming abundance at Churchill, Manitoba, Canada: Arctic, 46: Speakman, J. R., The cost of living: Field metabolic rates of small mammals. Advances in Ecological Research, 30: Stenseth, N. C. & R. A. Ims, 1993a. Population dynamics of lemmings: Temporal and spatial variation. An introduction. Pages in N.C. Stenseth & R.A. Ims. (ed.). The Biology of Lemmings. Academic Press, London. Stenseth, N. C. & R. A. Ims, 1993b. Food selection, individual growth and reproduction: An introduction. Pages in N. C. Stenseth & R. A. Ims. (ed.). The Biology of Lemmings. Academic Press, London. Tkadlec, E. & J. Zejda, Small rodent population fluctuations: The effects of age structure and seasonality. Evolutionary Ecology, 12: Turchin, P., L. Oksanen, P. Ekerholm, T. Oksanen & H. Henttonen, Are lemmings prey or predators? Nature, 405: Van Horne, B., Demography of Peromyscus maniculatus populations in seral stages of coastal coniferous forest in southeast Alaska. Canadian Journal of Zoology, 59: Van Horne, B., Demography of the longtail vole Microtus longicaudus in seral stages of coastal coniferous forest, southeast Alaska. Canadian Journal of Zoology, 60: Wolff, J. O. & W. Z. Lidicker Jr., Population ecology of the taiga vole, Microtus xanthognathus, in interior Alaska. Canadian Journal of Zoology, 58: Yoccoz, N. G. & R. A. Ims, Demography of small mammals in cold regions: The importance of environmental variability. Ecological Bulletins, 47: Yoccoz, N. G. & S. Mesnager, Are alpine voles larger and more sexually dimorphic because adults survive better? Oikos, 82: