3921.pdf ALCES VOL. 43, 2007 GAILLARD - ARE MOOSE ONLY A LARGE DEER? 1 ARE MOOSE ONLY A LARGE DEER?: SOME LIFE HISTORY CONSIDERATIONS Jean-Michel Gaillard Unité Mixte de Recherche N°5558 “Biométrie et Biologie Evolutive”, Université Claude Bernard Lyon 1, Bâtiment 711, 43 Boulevard du 11 novembre 1918, 69622 Villeurbanne Cedex, France ABSTRACT: Body mass generally accounts for a large part of variation in life history traits of ungu- lates. However, phylogeny and ecological features such as habitat or diet have been shown to cause differences in life history patterns among species of similar size. To assess the factors that shape life history traits of moose (Alces alces - with both traits expected from allometric equations and traits of similar-sized bovids. Both kinds of analyses led to the same results. While moose calves grow as expected from the size of their mothers, they start life at only about half the expected size. Moose populations have higher growth rates and shorter generation times as compared to similar-sized ungulates. Females reproduce earlier and have larger litters relative to their body size. The resulting faster than expected life cycle for moose can- population dynamics characterized by a low and variable juvenile survival as opposed to a high and constant survival of prime-age females. High reproductive output accounts for the fast life cycle of moose populations compared to other similar-sized ungulates. I propose that the high reproductive output has evolved in response to the unpredictable environmental conditions of early successional habitats preferred by moose. The evolutionary strategy of moose appears more similar to that of a very large roe deer (Capreolus capreolus) than that associated with larger deer in general. ALCES VOL. 43: 1-11 (2007) Key words: output, survival patterns, ungulates Since the pioneering work by Stearns (1976), the study of variation in life history traits has become a popular task among evo- lutionary ecologists. The analyses of varia- tion in life history traits can be performed at two different scales. First, the variation at the literature by using comparative analyses (sensu Harvey and Pagel 1991). Second, the existence of evolutionary trade-offs between history variation generated by differences in phenotypic quality are usually performed at - level, the variation in life history traits of vertebrates is mostly accounted for by three major structuring factors. Variation in body size generally accounts for more than half of the variation in most life history traits (Peters 1983, Calder 1984, Brown and West 2000 for reviews). In mammals, for instance, it is well- established that large mammals live longer, reproduce later, and produce fewer offspring per year than small ones (Stearns 1983, Gail- lard et al. 1989). However, for a given size, taxa often show marked differences in life history traits. For example, it is well known that bats outlive similar-sized rodents. Thus, ecological correlates of life history traits also occur. Differences in diet and differences in habitat quality have been shown to generate ARE MOOSE ONLY A LARGE DEER? - GAILLARD ALCES VOL. 43, 2007 2 differences in life history traits (Sæther and Gordon 1994 for ungulates, Fisher et al. 2001 for marsupials). Moose (Alces alces) are the largest mem- bers of the Cervidae family (from 200 to 825 kg, Novak 1993). Therefore, I expect that its large body size may have markedly shaped life history traits currently observed in moose populations. From comparative analyses of maternal care and demographic patterns reported in populations of moose and related ungulate species, I assessed whether moose life history can simply be accounted for by large size (i.e., moose are only large deer), or independent of their size relative to other deer (i.e., moose are different than a large deer). METHODS To assess whether moose are simply large deer, I performed two types of analyses on life history traits related to maternal care (birth mass and early growth rate) and population dynamics (population growth rate, generation I used allometric analyses in a three-step - ships without including moose for studied life history traits among ungulate species for which published information was available. Although species do not represent independent - ting usual linear models without accounting for phylogenetic relationships among species. My approach was based upon: (1) the similar results obtained from analyses on raw data (as performed here) and analyses including cor- rections for phylogenetic dependence (such as independent contrasts, see Garland et al. 1992) often reported (e.g., Fisher and Owens 2000); and (2) criticisms of the usefulness of phylogenetic methods such as independent contrasts (Ricklefs and Starck 1996, Björk- lund 1997, Price 1997), mainly based on the strong assumptions made by such methods on evolutionary changes of traits (Harvey and Rambaut 2001). Then I used the allometric equation to obtain the predicted value of the traits for a cervid with the same size as moose. Lastly, I compared predicted trait values with those reported in literature for moose popula- tions. The second type of analyses consisted of comparing life history traits observed in moose with those observed in similar-sized bovids. to the life history traits (birth mass and early size (polytocous and monotocous species). Indeed, individual offspring of polytocous ungulates that produce 2 offspring per breeding attempt might be lighter at birth than single offspring of monotocous ungulates (Roff 1992). Moreover, birth mass of singletons is often higher than birth mass of twins in polytocous species (e.g., moose, Schwartz and Hundertmark 1993). I found data for 38 (birth mass) and 22 (growth rate) monotocous ungulates and for 8 (birth mass) and 6 (growth rate) polytocous ungulates. To assess demographic patterns of moose as well as of other ungulate populations, I (i.e., the mean age of mothers at the time of birth, TB, in a given population, Leslie 1966) and population growth rate, r (i.e., the Malthusian parameter, Fisher 1930) from de- mographic data collected from the literature. To do that, I considered the following female of ungulate populations: the juvenile survival from birth to 1 year of age, the yearling sur- vival between 1 and 2 years of age, the annual survival of prime-age females between 2 and 7 years of age (or 10 depending on the size, Gaillard et al. 2000), the annual survival of females from 7 (or 10) years of age onwards, ALCES VOL. 43, 2007 GAILLARD - ARE MOOSE ONLY A LARGE DEER? 3 into Leslie matrix models and estimated both r and TB (see Caswell 2000 for further details). For species in which I obtained data from sev- populations belonging to 22 species including 6 moose populations (in South-Central Alaska, Ballard et al. 1991; in South Coast Barrens of Newfoundland, Albright and Keith 1987; in Northwest Territories (Canada), Stenhouse et al. 1995; and 3 populations in Northern Norway, Stubsjoen et al. 2000). To assess whether observed survival patterns account for the relatively rapid life cycle observed in moose populations, I used published estimates of both adult and juvenile survival in ungulate species. From a previ- ous literature review (Gaillard et al. 2000), I found data on adult survival in 61 populations belonging to 25 species (including 9 popula- tions of moose) and on juvenile survival in 53 populations belonging to 25 species (including 7 populations of moose). Because very low between-year variation in survival could also contribute to the higher than expected popu- lation growth rate of moose (see Tuljapurkar 1989 for a discussion of the changes in popula- tion growth generated by environmental varia- tion), I also compared the magnitude of annual variation of both juvenile and adult survival of annual estimates) in moose populations with the variation reported in other ungulate spe- cies. To account for the expected increase in survival with increasing body size (see Peters 1983, Calder 1984 for reviews), I regressed both mean survival and CV of survival for juveniles and adult females (measured as the a given species) on adult body mass. To assess whether observed reproduc- tive patterns account for the relatively rapid life cycle observed in moose populations, I collected data for ungulate species on two and litter size), as well as on body mass and generation time (see above). I found data for differences in reproductive traits according to adult body mass with 1-way ANOVAs using reproductive traits as factors (i.e., three classes age; and two classes of litter size: 1 or 2) and the log-transformed adult body mass as the dependent variable. I then compared the adult body mass of moose with the mean mass ex- pected from species with similar reproductive traits. In a second step, I performed the same kind of analysis by using the log-transformed generation time instead of adult body mass. I assumed that once variation in adult body mass is taken into account, differences in reproductive traits between moose and other ungulates account for the relatively faster life cycle of moose compared to other ungulates, and moose should reproduce earlier and more frequently relative to their size but perform as expected from their generation time. RESULTS AND DISCUSSION Patterns of Maternal Care in Moose: Birth Mass and Early Growth Rate As expected, a strong positive relation- ship occurred between birth mass (BW) and adult body mass (ABW) in both monotocous (Ln (BW) = -1.366 + 0.902 Ln (ABW); r = 0.977, P < 0.0001) and polytocous (Ln (BW) = -3.274 + 1.059 Ln (ABW); r = 0.892, P = 0.0029) species. There was no difference between slopes according to litter size (F = 0.903; df = 1, 42; P = 0.347). How- ever, for a given adult body mass, birth mass was larger in monotocous than in polytocous species (difference in intercepts of 0.295 (SE = 0.107); F = 7.630; df = 1, 43; P = 0.008). Similarly, early growth rate (GR) was allo- metrically related to adult body mass (ABW) in both monotocous (Ln (GR) = -2.689 + 0.733 Ln (ABW); r = 0.968, P < 0.0001) and polytocous (Ln (GR) = -0.691 + 0.561 Ln (ABW); r = 0.744, P = 0.0090) species. However, litter - ARE MOOSE ONLY A LARGE DEER? - GAILLARD ALCES VOL. 43, 2007 4 ship between early growth rate and adult body mass (differences in slope: F = 0.574; df = 1, 24; P = 0.456; differences in intercept: F = 2.770; df = 1, 25; P = 0.109). Using such allometric relationships to estimate expected values for moose, I obtained birth mass of 30.12 kg and 34.99 kg and early growth rates of 777.11 g/d and 639.61 g/d from the equa- tions of monotocous and polytocous species, respectively. Observed birth mass was only about half the expected values: 16.2 kg for monotocous moose and 13.5 kg for polytocous moose (Schwartz and Hundertmark 1993). On the other hand, an observed early growth rate of 785 g/d (Reese and Robbins 1994) was very similar to the expected values from allometric equations. Comparison of moose to similar-sized bovids led to the same conclusions. Moose had a much lighter birth mass than similar-sized species (Table 1). Birth mass in moose was similar to the birth mass of wildebeest (Con- nochaetes taurinus) whose adult body size is only half that of moose. On the other hand, early growth rates in moose were mid-range to those measured in similar-sized bovids. Comparative analyses of maternal care show that moose produce small newborns in relation to their size (about half the newborn size expected from other cervid species and similar-sized bovids). On the other hand, relative to their size, newborn moose grow at the same rate as other cervids and similar- sized bovids. I can thus also conclude that moose allocate energy to maternal care as a monotocous species during the gestation period but as a polytocous species during the lactation period. Demographic Patterns of Moose Popula- tions As expected, a marked positive allomet- ric relationship occurred between TB and adult body mass (ABW) among the 21 un- gulate species other than moose (Ln (TB) = 0.967 + 0.247 Ln (ABW); r = 0.646, P = 0.0016; Fig. 1). The allometric exponent was very close to that expected for a measure of bio- logical time such as generation time (0.25; Calder 1984), indicating that populations of large ungulate species have relatively slower life cycles than populations of small ungulate species. From such a relationship, TB of moose would be expected to be 11.76 years. From the 6 moose populations for which I found published information, the estimated TB was consistently shorter (from 4.57 to 10.66 years) than the expected value (Table LHT Wildebeest (Connachaetes taurinus) Cattle (Bos taurus) Eland (Taurotragus derbianus) Moose (Alces alces) Buffalo (Syncerus caffer) Adult mass (kg) 165 309 363 340-450 536 Birth mass (kg) 16.5 24 31.5 13-16 37.2 Growth rate (kg/d) 0.29 0.64 1.11 0.79 1.47 Litter size 1 1 1 1-2 1 Table 1. Comparison of life history traits (LHT) among moose to similar-sized bovids as related to maternal care. 765432 1,0 1,5 2,0 2,5 3,0 Ln (A dult Body M ass) L n (G en er at io n T im e) Fig. 1. Allometric relationship between generation - Observed generation time of moose (as measured by the median value from 6 populations) cor- responds to the open square. ALCES VOL. 43, 2007 GAILLARD - ARE MOOSE ONLY A LARGE DEER? 5 2), meaning that moose have a relatively short TB for their size. The median values observed for moose in the allometric relationship link- ing TB and adult body mass led to the largest negative residual. Likewise, according to previous work on a large range of taxa (e.g., Blueweiss et (e.g., Sinclair 1996, 1997), a negative allo- metric relationship tended to occur between r and adult body mass (ABW) among 17 un- gulate species other than moose that showed a positive r (i.e., increasing populations: Ln (r) = -1.402 - 0.303 Ln (ABW); r = 0.395, P = 0.117; Fig. 2). The slope was close to the theoretical expectation of -0.25 (Calder 1984), meaning that the product between r and TB is a dimensionless number (life history invariant sensu Charnov 1993). From such a relationship, r would be expected to be 0.039 for increasing populations of moose. From the 5 increasing moose populations for which I found published information, the estimated r was consistently higher (from 0.077 to 0.344; Table 2) than the expected value, meaning that moose populations have a high growth rate relative to female body size. The median value observed for moose on the allometric relationship between r and adult body mass led to one of the two largest positive residuals with a colonizing population of bison (Bison bison) (Van Vuren and Bray 1986). Such allometric analyses suggest that overall demographic patterns of moose popu- lations are more similar to those of small- or medium-sized ungulates than to those of sim- ilar-sized species. From expectations based on their body size alone, moose populations increase faster and the turnover of individuals is faster. Such overall demographic features can have three explanations: (1) survival of juveniles and/or adult female moose is much lower than that of similar-sized ungulates; (2) reproductive output of moose is much higher than that of similar-sized ungulates; or (3) both lower survival and higher reproductive output occur simultaneously in moose populations relative to similar-sized ungulates. Do Observed Survival Patterns Account for the Relatively Rapid Life Cycle Observed in Moose Populations? Contrary to expectation, the logit of female adult survival (LAS, which cor- responds to the log-transformed adult life expectancy) did not increase with increasing adult body mass among ungulate species (LAS = 2.205 + 0.040 Ln (ABW); r = 0.054, P = 0.801; Fig. 3). Female adult survival was high irrespective of body mass (from 0.710 in topi (Damaliscus lunatus) to 0.978 in pronghorn (Antilocapra americana); mean of 0.903, SE = 0.012). Female survival varied r TB Reference 0.10 8.89 Ballard et al. 1991 0.26 6.14 Stubsjoen et al. 2000 0.34 4.57 Stubsjoen et al. 2000 0.27 6.02 Stubsjoen et al. 2000 -0.03 10.66 Albright and Keith 1987 0.08 6.91 Stenhouse et al. 1995 Table 2. Population growth rate (r) and genera- tion time (TB) estimated for 6 moose popula- literature. 765432 -4 -3 -2 -1 Ln (A dult Body M ass) L n (P o p u la ti on G ro w th ) Fig. 2. Allometric relationship between population data collected from 17 ungulate species with population growth rate of moose (as measured by the median value from 5 increasing popula- tions) corresponds to the open square. ARE MOOSE ONLY A LARGE DEER? - GAILLARD ALCES VOL. 43, 2007 6 from 0.780 to 0.976 and averaged 0.907 (± 0.022) among the 9 moose populations for which data were available. Therefore, we can conclude that female adult survival of moose is similar to adult survival reported for other female ungulates. Likewise, there was no effect of adult body mass on CV of adult survival in female ungulates (CV = 0.069 + 0.001 Ln (ABW); r = 0.021, P = 0.927; Fig. 4). CV of female adult survival was low, irre- spective of body mass (from 0.017 in reindeer (Rangifer tarandus Ovis gmelini); mean of 0.073, SE = 0.007). CV of adult survival of females varied from 0.009 to 0.051 and averaged 0.035 (± 0.005) among the 7 populations of moose for which data were available. Such between-year variation appears to be a little lower than that observed in other ungulates. As expected, the logit of juvenile survival (LSJ) tended to increase with increasing adult body mass among ungulate species (LSJ = -1.648 + 0.409 Ln (ABW); r = 0.314, P = 0.135; Fig. 5). Expected juvenile survival indeed increased from 0.40 for an ungulate weighing 20 kg to 0.69 for an ungulate weighing 400 kg. Juvenile survival varied from 0.235 to 0.835 and averaged 0.640 (± 0.088) among the 7 moose populations from which I found data. Therefore, I can conclude that juvenile survival of moose is similar to juvenile survival reported for similar-sized ungulates. Likewise, there was a trend in the CV of juvenile survival to decrease with increasing adult body mass among ungulates (CV = 0.648 – 0.072 Ln (ABW); r = 0.353, P = 0.099; Fig. 6). Expected CV in juvenile survival decreased from 0.430 for an ungu- late weighing 20 kg to 0.220 for an ungulate weighing 400 kg. CV of juvenile survival of moose varied from 0.126 to 0.710 (average of 0.332 (± 0.130), median of 0.245) among the 765432 0 1 2 3 4 L n (A dult Body M ass) L o g it (f em al e ad u lt su rv iv a l) Fig. 3. Allometric relationship between adult survival of females (after logistic transforma- - Observed female adult survival of moose in 9 populations corresponds to the open squares. 765432 0,00 0,05 0,10 0,15 L n (Adult Body M ass) C V (f em al e ad u lt su rv iv al ) Fig. 4. Allometric relationship between temporal variation in adult survival of females (as mea- temporal variation in female adult survival of moose in 7 populations corresponds to the open squares. 765432 -2 -1 0 1 2 3 4 L n (Adult Body M ass) L o g it (J u v en il e su rv iv al ) Fig. 5. Allometric relationship between juvenile survival (after logistic transformation) and adult - survival of moose in 7 populations corresponds to the open squares. ALCES VOL. 43, 2007 GAILLARD - ARE MOOSE ONLY A LARGE DEER? 7 4 populations of moose with available data. Such between-year variation was similar to that observed in other similar-sized ungulates. Comparison of moose survival to sur- vival in similar-sized bovids led to the same conclusions (Table 3). Both survival esti- mates and temporal variation in survival of juvenile and adult female moose were very close to the values for similar-sized bovids. - eral survival pattern of ungulates, especially among smaller species, characterized by both a high and constant survival of adult females, irrespective of the species considered, and a low juvenile survival with high variability among years (see Gaillard et al. 1998, 2000 for similar conclusions). Because survival found in other ungulate species, survival pat- terns cannot be an explanation for the relatively rapid life cycle in moose. Do Observed Reproductive Patterns Ac- count for the Relatively Rapid Life Cycle Observed in Moose Populations? As expected, the mean body mass of ungulates differed according to the observed F = 4.808; df = 2, 18; P = 0.021), increasing from species that give birth at 1 year of age (20 kg; n = 1) to species n = 10). Ungulates that usually start to give birth at 2 years of age had an intermediate mean from that of ungulates starting to give birth at 3 years of age or older (Fisher’s LSD test, P = 0.028). Moose often give birth at 2 years of age when they are faced with favorable envi- ronmental conditions (Schwartz 1992). With a female adult body mass usually between 350 and 450 kg, moose do not belong to the size distribution of ungulates that normally start to reproduce at 2 years of age but belong to the size distribution of ungulates that normally do not start to reproduce before 3 years of age. Likewise, the mean body mass of un- gulates differed according to the observed litter size (F = 13.57; df = 1, 19; P = 0.002), decreasing from species that give birth to 765432 0,0 0,2 0,4 0,6 0,8 1,0 L n (Adult Body M ass) C V (J u v en il e su rv iv al ) Fig. 6. Allometric relationship between temporal variation in juvenile survival (as measured by - tion in juvenile survival of moose in 4 populations corresponds to the open squares. Traits Wildebeest Kudu Moose Bison Buffalo Adult mass (kg) 165 170 340-450 450 536 JS1 0.58 0.45 0.71 0.97 0.45 CV (JS)2 0.37 0.52 0.25 0.04 0.24 AS3 0.88 0.91 0.91 0.95 0.93 CV (AS)4 0.12 0.05 0.03 0.05 0.04 Table 3. Comparison of traits related to survival patterns among moose and similar-sized bovids. 1Juvenile survival. 2 3Adult survival. 4 ARE MOOSE ONLY A LARGE DEER? - GAILLARD ALCES VOL. 43, 2007 8 single offspring (139; n = 10) to species that give birth to twins (37 kg; n = 6). Moose often give birth to twins when they are faced with favorable environmental conditions (Boer 1992). With a female adult body mass usually between 350 and 450 kg, moose do not belong to the size distribution of ungulates that are expected to produce twins but belong to the size distribution of ungulates that normally produce single offspring. Looking now at the relationship between I found that as expected, generation time - tion (F = 5.89; df = 2, 18; P = 0.011) from yearlings to 10 years for those giving birth for of age had an intermediate generation time (Fisher’s LSD test, P = 0.013). The generation times observed in moose populations (from 4.57 to 10.66 years) match the distribution of generation times of ungulates that reproduce generation time decreased as expected with increasing litter size (F = 11.27; df = 1, 19; P = 0.003) from 9.3 years for ungulates that normally produce single offspring to 5.7 years for those that can produce twins. The generation times observed in moose popula- tions (from 4.57 to 10.66 years) match the distribution of generation times of ungulates that can produce twins. Comparison of moose reproductive pat- terns with those of similar-sized bovids led to the same conclusions (Table 4). The age at of smaller wildebeest and kudu (Tragelaphus strepsiceros) than that of larger bison and buffalo (Syncerus caffer). Moreover, both the proportion of 2 year-old females that give birth and the litter size were greater in moose than in smaller wildebeest and kudu. I can therefore conclude that female moose reproduce earlier (often giving birth at 2 years of age instead of 3 years of age for similar- sized ungulates) and have larger litters (often producing twins instead of single offspring as in similar-sized ungulates) than expected from their size. High reproductive output accounts for the rapid life cycle of moose populations compared to populations of other, similar- sized ungulates. Indeed, the distribution of generation times reported in moose popula- expected for ungulate populations that give often produce twins. Conclusions: Are Moose a Large Roe Deer or a Very Large Deer? Although only a few comparative analyses have reported clear ecological correlates of life history strategy, there is general agree- ment among evolutionary ecologists that Traits Wildebeest Kudu Moose Bison Buffalo Adult mass (kg) 165 170 340-450 450 536 Age of primiparity 2 2 2 3 3 % 21 0.27 0.1 0.4 0 0 % M2 0.9 0.9 0.81 0.62 0.7 Mean litter size 1 1 1.32 1 1 Table 4. Comparison of reproductive traits among moose to similar-sized bovids. 1Proportion of 2 year-old females giving birth in a given year in a population. 2Proportion of multiparous females giving birth in a given year in a population. ALCES VOL. 43, 2007 GAILLARD - ARE MOOSE ONLY A LARGE DEER? 9 among-species differences in habitat and diet should lead to differences in life history traits (Stearns 1992), especially in mammals (Saether and Gordon 1994 for ungulates, Gef- for marsupials). I therefore may ask whether for the relatively high reproductive output of moose? Contrary to other large cervids such as red deer (Cervus elaphus), moose appear to select early successional vegetation stages as a preferred habitat and are concentrate selectors (browsers) rather than grazers or mixed-feeders (Hofmann 1989). From these features, moose are much closer to roe deer (Capreolus capreolus), with which they occur often in sympatry, than to larger deer. Like all cervids, roe deer and moose both patterns of other ungulates. Strong selective pressures might have been operating during the evolutionary history of ungulates in response to predation (see Byers 1997). The canalization of adult survival (Gaillard and Yoccoz 2003) and the production of large and fast growing offspring within allometic constraints might ungulate evolution, leading these life history traits to vary little among ungulate species as ecological conditions vary. On the other hand, both moose and roe deer have a relatively high reproductive output, maybe in response to unpredictable environmental conditions in early successional habitats (as proposed by Liberg and Wahlström 1995). Female moose (weighing about 400 kg) cannot pro- duce offspring as fast and as often as roe deer (weighing about 25 kg) because of allometric constraints (Peters 1983, Calder 1984, Brown and West 2000). Thus, most female roe deer under a large range of environmental condi- tions, while only about half of female moose in the most productive populations (e.g., Vega Island in Norway where most females produce twins, Solberg, personal communication) do the same (Schwartz and Hundertmark 1993). Litter size of roe deer can be 3 offspring in very good conditions, while moose litter size is commonly 2 in the same situation. Lastly, because of their large size, moose cannot be a true polytocous species. To reach their high reproductive output, moose have to trade qual- ity of offspring (moose offspring are half the size of other ungulates’ offspring) for a higher quantity of offspring (female moose produce twins as soon as environmental conditions allow). This comparative analysis of moose life history traits suggests that moose are large roe deer rather than simply large deer, and sup- ports current theory on life history evolution that species occupying unpredictable habitats live at a faster rate than species living in more predictable habitats (Yodzis 1989). ACKNOWLEDGEMENTS This paper was prepared for the 5th Inter- national Moose Symposium held at Lilleham- mer, Norway, August 2002. I thank Erling Solberg and Bernt-Erik Saether for inviting me to the Symposium. REFERENCES ALBRIGHT, C. A., and L. B. KEITH. 1987. Population dynamics of moose, Alces alces, on the south-coast barrens of New- foundland. The Canadian Field-Naturalist 101:373-387. BALLARD, W. B., J. S. WHITMAN, and D. J. REED. 1991. Population dynamics of moose in South-central Alaska. Wildlife Monographs 114. BJÖRKLUND, M. 1997. Are ‘comparative methods’ always necessary? Oikos 80:607-612. BLUEWEISS, L., H. FOX, V. KUDZMA, D. NAKASHIMA, R. PETERS, and S. SAMS. 1978. Relationship between body size and some life history parameters. Oecologia 37:257-272. BOER, A. H. 1992. Fecundity of North Amer- ARE MOOSE ONLY A LARGE DEER? - GAILLARD ALCES VOL. 43, 2007 10 ican moose (Alces alces): a review. Alces Supplement 1:1-10. BROWN, J. H., and G. B. WEST. 2000. Scaling In Biology. Oxford University Press, New York, New York, USA. BYERS, J. A. 1997. American Pronghorn: Social Adaptations and the Ghosts of Predators Past. The University of Chicago Press, Chicago, Illinois, USA. CALDER, W. A. III. 1984. Size, Function and Life History. Harvard University Press, Harvard, Cambridge, Massachusetts, USA. CASWELL, H. 2000. Matrix Population Models. Construction, Analysis and Interpretation. Sinauer Associates, Sunderland, Massa- chuetts, USA. CHARNOV, E. L. 1993. Life History Invari- ants. Some Explorations of Symmetry in Evolutionary Ecology. Oxford University Press, Oxford, U.K. FISHER, D. A., and I. F. P. OWENS. 2000. Female home range size and the evolu- tion of social organization in macropod marsupials. Journal of Animal Ecology 69:1083-1098. _____, _____, and C. N. JOHNSON. 2001. The ecological basis of life history variation in marsupials. Ecology 82:3531-3540. FISHER, R. A. 1930. The Genetical Theory of Natural Selection. Oxford University Press, Oxford, U.K. GAILLARD, J.-M., M. FESTA-BIANCHET, and N. G. YOCCOZ. 1998. Population dynamics of large herbivores: variable recruitment with constant adult survival. Trends in Ecology and Evolution 13:58-63. _____, _____, _____, A. LOISON, and C. TOIGO. 2000. Temporal variation in fitness com- ponents and population dynamics of large herbivores. Annual Review of Ecology and Systematics 31:367-393. _____, D. PONTIER, D. ALLAINE, J. D. LEBRETON, J. TROUVILLIEZ, and J. CLOBERT. 1989. An analysis of demographic tactics in birds and mammals. Oikos 56:59-76. _____, and N. G. YOCCOZ. 2003. Temporal variation in survival of mammals: a case of environmental canalization?. Ecology 84:3294–3306. GARLAND, T., P. H. HARVEY, and A. R. IVES. 1992. Procedures for the analysis of comparative data using phylogenetically independent contrasts. Systematic Bio-Bio- logy 41:18-32. GEFFEN, E., M. E. GOMPPER, J. L. GITTLEMAN, H. K. KUH, D. MACDONALD, and R. K. WAYNE. 1996. Size, life history traits, and social organization in the Canidae: a reevaluation. American Naturalist 147:140-160. HARVEY, P. H., and M. D. PAGEL. 1991. The Comparative Method in Evolutionary Biology. Oxford University Press, Ox- ford, U.K. _____, and A. RAMBAUT. 2001. Comparative analyses for adaptive radiations. Philo- sophical Transactions of the Royal Society of London Series B 355:1599-1605. HOFMANN, R. R. 1989. Evolutionary steps of ecophysiological adaptation and di- versification of ruminants: a comparative view of their digestive system. Oecologia 78:449-457. LESLIE, P. H. 1966. The intrinsic rate of increase and the overlap of successive generations in a population of guillemot (Uria aalge Pont). Journal of Animal Ecology 35:291-301. LIBERG, O., and K. WAHLSTROM. 1995. Habitat stability and litter size in the Cervidae; a comparative analysis. Pages 1-60 in K. Wahlström. Natal Dispersal in Roe Deer. An Evolutionary Perspective. Unpublished Ph.D. Thesis, University of Stockholm, Sweden. NOVAK, R. M. 1993. Walker’s Mammals of the World. Fifth Edition, Volume II. The John Hopkins University Press, Baltimore and London, U.K. PETERS, R. H. 1983. The Ecological Implica- tion of Body Size. Cambridge University ALCES VOL. 43, 2007 GAILLARD - ARE MOOSE ONLY A LARGE DEER? 11 Press, Cambridge, U.K. PRICE, T. 1997. Correlated evolution and independent contrasts. Philosophical Transactions of the Royal Society of Lon- don Series B 352:519-529. REESE, E. O., and C. T. ROBBINS. 1994. Char- acteristics of moose lactation and neonatal growth. Canadian Journal of Zoology 72:953-957. RICKLEFS, R. E., and J. M. STARCK. 1996. Ap- plications of phylogenetically independ- ent contrasts: a mixed progress report. Oikos 77:167-172. ROFF, D. A. 1992. The Evolution of Life Histories. Chapman and Hall, London, U.K. SAETHER, B. E., and I. J. GORDON. 1994. The adaptive significance of reproductive strategies in ungulates. Proceedings of the Royal Society of London Series B 256:263-268. SCHWARTZ, C. C. 1992. Reproductive biol- ogy of North American moose. Alces 28:165-173. _____, and K. J. HUNDERTMARK. 1993. Re- productive characteristics of Alaskan moose. Journal of Wildlife Management 57:454-468. SINCLAIR, A. R. E. 1996. Mammal popula- tions: fluctuation, regulation, life history theory and their implications for conser- vation. Pages 127-154 in R. B. Floyd, A. W. Sheppard, and P. J. De Barro, editors. Frontiers of Population Ecology. CSIRO Publishing, Melbourne, Australia. _____, 1997. Fertility control of mammal pests and the conservation of endangered marsupials. Reproduction, Fertility and Development 9:1-16. STEARNS, S. C. 1976. Life-history tactics: a review of the ideas. Quarterly Review of Biology 51:3-47. _____, 1983. The influence of size and phy- logeny on patterns of covariation among life history traits in the mammals. Oikos 41:173-187. _____, 1992. The Evolution of Life Histories. Oxford University Press, Oxford, U.K. STENHOUSE, G. B., P. B. LATOUR, L. KUTNY, N. MACLEAN, and G. GLOVER. 1995. Productivity, survival, and movements of female moose in a low-density population, Northwest Territories, Canada. Arctic 48:57-62. STUBSJOEN, T., B. E. SAETHER, E. J. SOLBERG, M. HEIM, and C. M. ROLANDSEN. 2000. Moose (Alces alces) survival in three populations in northern Norway. Canadian Journal of Zoology 78:1822-1830. TULJAPURKAR, S. D. 1989. An uncertain life: demography in random environment. Theoretical Population Biology 21:141- 165. VAN VUREN, D., and M. P. BRAY. 1986. Popu- lation dynamics of bison in the Henry Mountains, Utah. Journal of Mammalogy 67:503-511. YODZIS, P. 1989. Introduction to Theoretical Ecology. Harper & Row, New York, New York, USA.