Ecological correlates of regional variation in life history of the moose Alces alces

Ecology, July, 1996 by Bernt-Erik Saether, Reidar Andersen, Olav Hjeljord, Morten Heim

INTRODUCTION

Regulation and limitation of populations through food shortage is a central paradigm in the population ecology of large herbivores (Caughley 1970, Sinclair 1977). Following Lack (1954), when resources are reduced seasonally, competition between the individuals for access to resources may increase, resulting in increased mortality (Sinclair 1977, Skogland 1985, 1990, Choquenot 1991) and decreased fecundity (McCullough 1979, Clutton-Brock et al. 1987, Fowler 1987). The fecundity of ungulates at high population densities may decrease through a resource-dependent reduction in body condition, often resulting in delayed maturity among young females (White 1983, Skogland 1990, Gaillard et al. 1992, Jorgenson et al. 1993) or reduction in the fecundity rates among adults (Franzmann and Schwartz 1985).

Regulation of population size (Sinclair 1989) through food limitation requires a rapid resource-dependent demographic response in the population (Maynard Smith 1974, Royama 1992). However, the establishment of such a feedback mechanism between resource level and population growth rate may be concealed by density-independent variation in demographic variables due to a stochastic environment. Both in tropical and temperate ungulates, environmental stochasticity may strongly influence population fluctuations (White 1983, Albon et al. 1987, Owen-Smith 1990). As in density-dependent cases, stochastic effects on fecundity often operate through body condition (Klein 1970, Saether 1985, Saether and Heim 1993, Langvatn et al., in press).

The present study compares patterns of covariation of life history characters in four different Norwegian moose populations living in very different range conditions (Hjeljord et al. 1994). We will examine whether differences in fecundity and mortality can be predicted from variation in resource availability. If food-limited regulation of moose populations is likely (cf. Messier 1994), we will expect, in the almost complete absence of large predators, a reduction in fecundity and an increase in mortality with reduced winter food availability.

STUDY AREAS

The study was conducted in one southern (59.9 [degrees] N), one interior (60.7 [degrees] N), one alpine (61.3 [degrees] N), and one northern (69.8 [degrees] N) study area. Data were collected during the years 1985-1987 in the southern, 1985-1990 in the interior, 1984-1987 in the alpine, and 1984-1990 in the northern study area. The southern and interior study areas were located in coniferous forests, subject to relatively intense management for commercial forestry. In both areas the major food resources for moose were located near or on clear-cut areas. In the southern study area, the moose had a variety of both winter and summer food plants available. In the interior, Scots pine (Pinus sylvestris) and birch (Betula pubescens) were the dominant winter browse species. During summer, fireweed (Chamaenerion angustifolium) and birch leaves were a large proportion of the diet. For further description of these two areas, see Hjeljord et al. (1994).

The alpine area was located above the timberline, 850-1000 m above sea level. Birch was the only winter food for moose (Saether and Andersen 1990); winter range conditions were very poor due to a combination of deep snow and low biomass of browse, but during summer, many food plants were available (Saether et al. 1992). During the winter moose in the northern study area concentrated in broad river valleys where river flats were covered with dense stands of Salix species. In summer, many animals spread out and were found in rich birch forest with high production of preferred food plants. For a further description of those two study areas, see Saether and Andersen (1990), Andersen and Saether (1992), and Saether and Heim (1993).

In both the alpine and northern study areas, winter ranges were covered by snow most of the period from December to May. In the interior study area the snow cover was 35-50 cm during the winters 1985-1988, whereas in the last two study years the ground was free from snow most of the winter. In the southern study area the ground was covered with snow only during January-March with a maximum snow depth of 35 cm.

MATERIAL AND METHODS

Data from radio-collared females

Data on the timing of calving, number of calves born, and juvenile survival were collected over several years from 11 adult radio-collared cows in the southern, 13 cows in the interior, 11 cows in the alpine, and 13 cows in the northern study areas. The capture procedures and the telemetry equipment are described in detail elsewhere (Saether and Andersen 1990, Andersen and Saether 1992).

Timing of calving was determined by locating and approaching the female on foot at regular intervals (usually 3-4 d) through the calving season from the middle of May to the end of June. Calving date was taken as the mid-date between the visits before and after calving. If a calf was found while still wet, calving was assumed to have occurred on the same day as the visit. All dates could be estimated within 3-4 d. Calf production was determined by the maximum number of calves recorded during the first 2 wk after calving. Because primiparous females calve later (Saether and Heim 1993) and have smaller litter size (Saether and Haagenrud 1983) than multiparous females (i.e., cows known to have calved previously), only data from multiparous cows were included in these analyses.