birth-site selection by alaskan moose: maternal strategies for coping ...

BIRTH-SITE SELECTION BY ALASKAN MOOSE: MATERNAL STRATEGIES FOR COPING WITH A RISKY ENVIRONMENT

R. TERRY BOWYER, VICTOR V AN BALLENBERGHE, JOHN G. KIE, AND JULIE A. K. MAIER

Institute of Arctic Biology, University of Alaska Fairbanks, Fairbanks, AK 99775 (RTB, JAKM) United States Forest Service, Pacific Northwest Research Station, 3301 C Street, Suite 200,

Anchorage, AK 99503 (WB) United States Forest Service, Pacific Northwest Research Station, 1401 Gekeler Lane, La Grande,

OR 97850 (JGK)

We studied birth-site selection in Alaskan moose (Alces alces gigas) from 1990 to 1994 in Denali National Park and Preserve in interior Alaska. Twenty percent of preparturient females made extensive movements (2:::5 km) immediately before giving birth. Females selected (use was greater than availability) sites for giving birth (n = 39) that were on southerly exposures with low soil moisture and high variability in overstory cover. Moose selected birth sites based on micro-site characteristics rather than on broad types of habitat, which were used in proportion to their availability. Spatial distribution of birth sites did not differ significantly from random locations. We hypothesize that such unpredictable behavior by females is a strategy to avoid predators. Parturient females also selected sites with high visibility that were located at high elevation, which ostensibly allowed them to see and then hide from approaching predators. We rejected the hypothesis, however, that moose in this population spaced themselves away from predators or avoided habitat types favored by large carnivores. Likewise, we rejected the hypothesis that moose gave birth close to human developments to avoid predators; random' sites were > 100 m closer to human developments than were birth sites. Cover of forage, especially willows (Salix), was more than twice as abundant at birth sites than random sites. Forage quality, as indexed by nitrogen content and in vitro dry matter digestibility, was slightly but significantly higher at birth sites. An inverse relationship between visibility and availability of forage indicated that female moose made tradeoffs between risk of predation and food in selecting sites to give birth. Thus, maternal females coped with a risky environment; they gave birth at sites that helped them minimize risk of predation but exhibited risk-averse behavior with respect to the forage necessary to support the high cost of lactation. We hypothesize that risk of predation prevented moose from seeking birth sites with more forage and, hence, a greater nutritional reward, which reduced the variance in forage availability at birth sites.

Key words: Alces alces, Alaskan moose, maternal strategies, birth-site selection, riskaverse foraging, risk of predation, tradeoffs, interior Alaska

Female mammals bear costs of both gestation and lactation (Millar, 1977; Pond, 1977). Moreover, among many polygynous mammals, males contribute little more than genes to their offspring; consequently, the burden of rearing young in such species rests entirely with females (Clutton-Brock, 1991). Ungulates, which are among the most polygynous and sexually dimorphic mammals (Ralls, 1977; Weckerly, 1998),

Journal of Mammalogy, 80(4):1070-1083, 1999 1070

often follow this differential pattern of parental investment (Clutton-Brock, 1991). Only recently, however, have effects of maternal behavior on performance of offspring become of interest to evolutionary biologists (Bernardo, 1996), although that topic has fascinated those studying the biology of moose (Alces alces) for many years (Altmann, 1958, 1963; Peterson, 1955).

Female ungulates often encounter severe

Downloaded from https://academic.oup.com/jmammal/article-abstract/80/4/1070/851824by gueston 12 February 2018

November 1999 SPECIAL FEATURE-UNGULATE LIFE-HISTORY STRATEGIES 1071

environmental constraints on their ability to conceive, gestate, provision, and rear offspring successfully (Bowyer, 1991; Rachlow and Bowyer, 1991, 1994). Most fetal growth occurs in the last one-third of gestation (Schwartz and Hundertmark, 1993), but females in northern environments must incur costs of' sustaining such growth near the end of winter when forage is relatively unpalatable (Weixelman et al., 1998) and body reserves are at yearly minima (Schwartz, 1998). In addition, costs of lactation for large herbivores are enormous compared with other aspects of maternal investment (White and Luick, 1984). Females in poor physical condition may give birth to young that are underweight and exhibit low survivorship (Byers and Hogg, 1995; Clutton-Brock et aI., 1987; Festa-Bianchet and Jorgenson, 1998). Moose with low body mass at birth may not exhibit compensatory growth and remain among the smallest individuals in their cohort (Keech et al., 1999). Indeed, females obtaining a poor diet may reduce investment in neonates (Rachlow and Bowyer, 1994) or curtail it altogether (Langenau and Lerg, 1976). Undernourished mothers also may fail to defend their young adequately from predators (Smith, 1987). Losses of young to predators can be substantial, and this pressure has helped shape adaptations of ungulates for coping with the environments they inhabit (Bleich, 1999; Bowyer, 1987; Bowyer et al., 1998a; Hirth, 1977; Kie, 1999; Van Ballenberghe and Ballard, 1994). Indeed, the need to balance the requirement of obtaining a nutritious diet against risk of predation has been well documented for ungulates (Berger, 1991; Bleich et al., 1997; Kohlmann et aI., 1996; Nicholson et al., 1997).

In environments with a positive relationship between quality and abundance of forage and risk of predation (i.e., where ungulates seek out areas with ample forage, and predators concentrate their hunting in such areas), a tradeoff between those environmental factors may occur (Bowyer et al.,

1998b). Thus, female ungulates may be forced to tradeoff adequate forage to support lactation against risk of predation for them to rear young successfully (Bowyer et al., 1998b; Kohlmann et al., 1996; Nicholson et al., 1997; Rachlow and Bowyer, 1998). Understanding the nature of that tradeoff is necessary to comprehend how maternal females cope with their environment while attempting to rear young.

Because nutritional requirements of female ungulates and risk of predation on their neonates reach maxima during and shortly after partUrition (Bowyer et al., 1998a, 1998b; Rachlow and Bowyer, 1998), we chose that period to study maternal tradeoffs in Alaskan moose (A. a. gigas). We selected moose for our analysis because they remain at or near the birth site for several weeks following parturition (Addison et al., 1990); therefore, measurements of forage biomass and quality are simplified compared with situations where neonates follow their mothers (Bowyer et aI., 1998b). In addition, moose in Denali National Park and Preserve, Alaska, where we conducted our study, contend with a full array of natural predators, including wolves (Canis lupus) and grizzly bears (Ursus arctos-Bowyer et aI., 1998a; Miquelle et al., 1992).

We tested if forage quality and abundance or risk of predation was more important in determining selection of birth sites by moose. We determined if maternal females avoided predators by spacing their birth sites away from habitats used by predators, or located their birth sites near human developments. We also tested for a tradeoff between forage abundance and risk of predation by examining variation in forage available to moose at birth and random sites to infer if they followed a risk-prone or risk-averse strategy (sensu Stephens and Krebs, 1986). Finally, we discuss selection of birth sites and the role of this behavior in shaping patterns of maternal tradeoffs in moose.


1072 JOURNAL OF MAMMALOGY Vol. 80, No.4

I' Ra.-._ • BrthSitas DSIudy--ri-

FIG. I.-The eastern end of Denali National Park and Preserve, Alaska, showing the three intensive areas where we sampled random (n = 61) and birth (n = 39) sites of Alaskan moose from 1990 to 1994. Contour intervals are 100 m.

MATERIALS AND METHODS

Study area.-We studied birth-site selection by moose in the eastern part of Denali National Park and Preserve in interior Alaska during each spring from 1990 to 1994. The study area extended from Highway 3 westward along the Denali Park Road to the Sanctuary River and included ca. 300 km2 of wilderness (Fig. 1).

Moose were distributed in a broad valley with elevations ranging from 650 to 1,200 m. Rugged foothills bounded that area to the south, and the terrain of the Alaskan Range rose precipitously to the north. Vegetation was dominated by brushy tundra characterized by resin birch (Betula glandulosa) intermixed with stands of spruce (Picea glauca and P. mariana) often with a willow (Salix) understory. Herbaceous tundra occurred at higher elevations, with low-lying areas dominated by meandering creeks and dry washes with stringers of willow. Trembling aspen (Populus tremuloides), poplar (P. balsamifera), and alder (Alnus) were more common in the eastern part of our study area, although isolated stands occurred throughout the area (Fig. 2). More complete descriptions of topography and vegetation of this area were provided elsewhere (Bowyer et aI., 1998a; Miquelle et al., 1992; Molvar and Bowyer, 1994; Molvar et aI., 1993). Our classification of habitat types was modified from Viereck et aI. (1992); we recognized eight broad types of habitat that occurred in the Park (Fig. 2).

Summers in the Park were short and cool, and winters were long, cold, and often severe. Average temperature ranged from -17°C in January to 12°C in July; yearly snowfall averaged 190 cm, with snow sometimes persisting for 9 months. Depth of snow was above average (41 cm) during the 5 years of our study, with winter 1991-1992 producing exceedingly deep snow (>90 cm-Bowyer et aI., 1998a). Climatic conditions were highly variable among years (Bowyer et aI., 1998a; Rachlow and Bowyer, 1991, 1994, 1998).

About 150 moose were present during our study. The population of moose was typical of others in interior Alaska that were held below carrying capacity by heavy predation (Gasaway et al., 1992; Van Ballenberghe and Ballard, 1994). Low survivorship of young (ca. 0.20 by 20 days old) indicated that the population likely was declining (Bowyer et al., 1998a). Moose appeared to be in excellent physical condition and exhibited high rates of twinning (32-64% of births-Bowyer et aI., 1998a).

The Park contained relatively high densities of wolves and grizzly bears, the primary predators of moose (Albert and Bowyer, 1991; Mech et aI., 1998; Miquelle et al., 1992). Grizzly bears were responsible for most (53%) mortality of young moose from 1990 to 1994 (Bowyer et al., 1998a) and also killed many young caribou (Rangifer tarandus-Adams et aI., 1995a, 1995b).

Like most polygynous ungulates, moose sexually segregate at the time of parturition (Bleich et al., 1997; Bowyer 1984; Bowyer et aI, 1996; 1997; Kie and Bowyer, 1999; Miller and Litvaitis, 1992; Miquelle et al., 1992). Maternal females become solitary in early spring and seek secluded areas for giving birth (Cederlund et al., 1987; MacCraken et al., 1997; Molvar and Bowyer, 1994).

Sampling procedures.-We located birth sites of moose by tracking adult females fitted with radiotelemetry collars (Telonics, Mesa, AZ) from early May to mid-June 1990-1994. Mean date of birth for moose in the Park was 25 May, and births were highly synchronized (95% of births in 16 days-Bowyer et al., 1998a). During our 5-year study, 11-18 females wore telemetry collars each year. We attempted to locate females twice each day by driving westward for 40 km and then returning east along the park road. When two to three sequential telemetry



40'

o 10 20 Kilometers N

A ~~~~iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii

c==l Herbaceous tundra Low shrub .. Forest c=J Alpine tundra

Dwarf shrub .. Tall shrub River bar c=J Snow & Ice

FrG. 2.-Habitat types in the eastern end of Denali National Park and Preserve, Alaska. Moose did not give birth in the steep, precipitous terrain of alpine habitat.

fixes indicated a female was relatively stationary, we followed that signal to locate the female and potentially her birth site. Some parturient females also moved away from their previous location immediately before they gave birth, which also helped us to identify such individuals and locate their birth sites. We recorded the proportion of moose that made such movements and the linear distances they moved in 1993 and 1994. If a female had given birth, we took care not to approach too closely . or to disturb the mother or her young. All aspects of this research were approved by the Institutional Animal Care and Use Committee at the University of Alaska Fairbanks.

We also located birth sites opportunistically while driving along the park road by observing lone females closely with binoculars or a spotting scope to determine if a neonate was present. We also examined areas with trees or shrubs that had their bark stripped recently by moose. Such areas were obvious before leaf-out in early

spring. Bark stripping occurred around birth sites because the female seldom ventured > 100 m from her young and rapidly depleted forage around the site (Miquelle and Van Ballenberghe, 1989). We also searched for birth sites for ca. 3h using a fixed-winged aircraft. We mapped locations of all birth sites we discovered and often placed flagging near the site (> 100 m) so we could relocate it later. We never sampled a birth site, however, until the female" and her young had departed.

Birth sites of moose were concentrated in three areas (X = 2,058 ha ± 1,424 SD) in the eastern end of the Park (Fig. 1). Consequently, we distributed our random samples in those same three areas so that we included only habitat variables that were available to calving moose. More random sites were sampled than birth sites because habitat characteristics were more variable at random locations and a larger sample was required to describe those sites adequately. Three areas were selected for intensive study be-


1074 JOURNAL OF MAMMALOGY Vol. 80. No.4

cause our previous observations, and those of others (Miquelle et al., 1992), indicated that they were used traditionally by moose for calving. We sampled three birth sites, however, that were located outside of our areas of intensive study. We were unable to sample one birth site and one random site because we could not cross the rising water in Riley Creek safely to reach those areas. We sampled 39 birth sites and 70 random sites, but because of some missing data, fewer random sites were used in particular analyses. We attempted to distribute samples of birth and random sites throughout the spring so that a preponderance of samples did not occur in a particular week. Our desire to avoid birth sites until moose were no longer using them, however, sometimes caused us to modify that sampling design.

At each birth or random site, we recorded the location (2-5 m accuracy) using a global positioning system (GPS). We measured concealment cover at each site with a cover pole (Griffith and Youtie, 1988) that was 2 m in height and divided into 20-cm segments. The pole was observed from 10 m at each cardinal direction, and percent cover was determined by noting the proportion of segments on the pole that were obscured (2':50% of each segment) by vegetation or topographic features. We also recorded overstory cover at the center of a birth site or random location using a spherical densiometer (Lemmon, 1957) that was read from each of the four cardinal directions. We determined gravimetric moisture of soil using a soil core that was 2 cm in diameter and driven 10 cm into the substrate. We also recorded wind speed at sites with a windgauge by noting the maximum value obtained over a I-min interval. We noted if snow was present at birth or random sites ($;2 m from the center of the site).

We sampled woody vegetation (browse) available to moose at each birth and random site by aligning a 50-m transect, which was centered on the GPS location, in a random direction. Percent cover of forage within the reach of moose (2.5 min height-Weixelman et aI., 1998) was sampled with the line-intercept method (Canfield, 1941). We recorded cumulative crown cover of forage as an index of available forage to moose but considered small gaps «4 cm) in crowns of small trees and shrubs as continuous cover. We also recorded the relative amount of foraging by moose on each species of browse by ranking

each contiguous area of cover along the transect as high (>50%), moderate (26-50%), or low ($;25%). Those ranks corresponded to the percentage of leaders of current annual growth that was browsed by moose. Such rankings of use have been used as an index to browsing intensity for many years (Aldous, 1944). We created a mean index to browsing for each transect by weighting the rank for browsing intensity by the percent cover of a particular contiguous grouping of forage. We did not sample herbaceous vegetation because it was uncommon in early spring, and moose in the Park eat mostly browse at that time of year (Van Ballenberghe et aI., 1989).

We also clipped samples of current annual growth for forage species if they were available at birth or random sites. We collected a minimum of 15 leaders of current annual growth from willows at each site, which were composited into a single sample for analysis. Those samples were placed in plastic bags and stored frozen until they could be analyzed for forage quality. Before analysis, samples were dried to a constant weight at 50°C, and ground with' a Wiley mill so fragments would pass through a 1-mm mesh screen. Forage quality was indexed by determining in vitro dry matter digestibility (IVDMD-Van Soest, 1982) with rumen liquor from a caribou fed a diet that included willows. Percent nitrogen (N) also was determined using standard techniques (Van Soest, 1982) at the Institute of Arctic Biology of the University of Alaska Fairbanks.

We used the Geographic Information System (GIS) ARCIINFO (Environmental Systems Research Institute, Redlands, CA) to derive several variables that were indicative of a broader scale than data collected at birth and random sites, including distance of random and birth sites to streams, forest, and human developments. Human developments included campgrounds, the park road, and other Park facilities. We used a LANDSAT-TM scene that was classified to determine eight broad habitat types. We used the GRIDS module of ARCIINFO with a cell size of 80 m to determine slope, aspect, and steepness around sites. Aspect was transformed to the sine and cosine of the direction (in degrees) of the slope face. We also obtained elevation from a digital elevation model (United States Geological Survey, scale = 1:250,000). We determined terrain ruggedness by multiplying the angular



deviation of aspect by the SD of slope steepness (Nicholson et al., 1997). We also calculated the "viewshed," which was the area a standing moose (2 m in height) might view to a maximum distance of 300 m based on topography around a birth or random site. That value was corrected for cover of vegetation by multiplying the proportional cover in the top 40 cm of the 2-m tall cover pole by the viewshed. Our index to visibility increased with the unobstructed view from the site.

Statistical analysis.-We used stepwise (n to enter and remain = 0.15) logistic regression to identify variables important in discriminating birth (coded 1) from random sites (coded 0-Agresti, 1990). Because moose modify surrounding vegetation and substrate at most birth sites, we were confident that no random sites were used by moose for parturition. We eliminated one of any pair of variables with a r 2:

0.5 to control for multicollinearity. We assured that our logistic model was apt by examining a Homser-Lemeshow test for goodness-of-fit. Because year (P > 0.15) failed to enter our logistic regression, we pooled years to compare birth with random sites. For descriptive purposes only, we present selection of habitat features as use minus availability. We also performed twotailed t-tests for unequal variances (Zar, 1984) on some of the individual variables. We used multivariate analysis of variance (MANOV A) to test for differences in forage use between birth and random sites for the four most common species of browse eaten by moose. Quality of forage (IVDMD, N) was tested with multivariate analysis of covariance (MANCOV A) with site (birth or random) as the main effect and Julian date and year as covariates. That approach was necessary because plant phenology and, hence, quality of forage in the Park varied among years (Bowyer et aI., 1998a). Female moose remained at some birth sites longer than at others thereby determining dates on which sites could be sampled. We use the SAS statistical package for those analyses (SAS Institute Inc., 1988). We examined differences in use of habitat types for birth sites compared with their relative availability with a G-test (Zar, 1984).

We used multi-response permutation procedures (MRPP) to test for differences between locations of birth sites of moose and random locations within and across years (Slauson et aI., 1991; Zimmerman et al., 1985). We subset our

random sites for this analysis so they equaled the number of birth sites in our three areas of intensive sampling.

RESULTS

Movements of preparturient females were sampled only in 1993-1994; 20% of 20 females made unusual movements immediately before giving birth (7.3 km ± 2.3 SD). We also examined spatial distribution of birth and random sites within years (1990-1994) and strata (Fig. 1; three outlying birth sites withheld from analysis). Nearest-neighbor distance for birth sites was 1.1 ± 0.6 km, whereas the distance between random sites was 0.9 ± 0.5 km. MRPP analysis indicated no significant difference (P > 0.6) between the spatial arrangement of birth and random sites for within-year data pooled, or for the 3 individual years for which we had sufficient data to allow analyses: 1992 (P > 0.9), 1993 (P > 0.4), and 1994 (P > 0.3).

Birth sites of moose were typically small areas (1-3 m across) in which the ground had been pawed thereby exposing fresh soil. Hair from the molting female often was scattered across the site, and feces of the female and her offspring were present. Heavy use of forage around birth sites was obvious with willows being most consumed (Table 1). Females remained near birth sites (:=:; 100 m) if undisturbed in all but one instance (n = 39). On that occasion, a female moved> 125 m to a nearby hillside to eat snow during the unusually warm spring of 1993 eX = 10.4°C for May-June). Females that did not lose young to predators (n = 5) remained at the birth site for 3-4 weeks, but most young moose (78%) were killed by predators and did not survive >20 days of age (Bowyer et al., 1998a). Female moose never used the same birth site twice.

Female moose did not select broad habitat types for giving birth; proportional occurrence of birth sites and random ones were distributed equally among those habitats (Fig. 3). Additionally, maternal females did not position themselves closer to



TABLE 1.-Relative use (proportional cover x index of forage use) of browse by Alaskan moose at birth and random sites in Denali National Park and Preserve, Alaska, in spring 1990-1994. Multivariate analysis of variance (MANOVA) revealed an overall difference in use between birth and random sites (F = 15.63, d.f. = 4, 104, P < 0.0001); P-values are from planned contrasts following MANOVA.

Birth sites Random sites

Browse (n = 39) (n = 70)a

species X SD X SD P-value

Willow 1.72 0.76 0.86 0.53 <0.0001 Poplar 0.23 0.59 0.14 0.35 0.32 Alder 0.10 0.38 0.06 0.23 0.44 Aspen 0.13 0.34 0.04 0.20 0.10

, Sample size differs slightly for random sites from that use in logistic regression (Fig. 4) because of missing variables in the regression analysis.

human developments such as campgrounds, the Visitor Center, or the park road to avoid predators; random locations were on average > 100 m closer to human developments than were birth sites (t = 9.53, dj = 80, P < 0.001; Table 2).

Selection of birth sites may have been based on habitat characteristics' that occurred at a much smaller scale than broad habitat types (Fig. 2). Thus, we examined a suite of micro-site characteristics for birth and random sites that were related to the geographic, topographic, and climatic conditions in the Park, forage availability, and risk of predation (Table 2). From that list, stepwise logistic regression identified three variables that discriminated birth from random sites: forage, aspect, and visibility (Fig. 4). Female moose selected sites to give birth with more forage (especially willow), better visibility, and southeasterly exposures (Table 2, Fig. 4). Selection of southerly aspects likely related to climatic conditions in spring. Some snow was on the ground in all years when females were giving birth, but differences in snow cover between birth and random sites were small (Table 2). Soil moisture, however, was 50% lower at birth than random sites (t = 16.8,

en w t: en u.. o w <!l <C fZ W U a: w 0..

G=1.37,d.f. =5,P= .

FOREST

• BI~:~ ~~~ES

~ R~~D,,?~ )SITES

TUNDRA

HABITAT TYPES

FIa. 3.-A comparison of broad habitat types available to (random sites) and used (birth sites) by Alaskan moose in Denali National Park and Preserve, Alaska, from 1990 to 1994.

d.! = 93, P < 0.001; Table 2). Thus, aspect probably influenced whether neonates were likely to get wet. There also was greater CV in overstory cover at birth sites than at random sites (t = 8.37, d.! = 42, P < 0.001; Table 2).

Visibility, which also entered the logistic-regression model (Fig. 4), probably was related to the ability of parturient females to observe predators before those carnivores approached birth sites closely. On average, birth sites were 96 m higher in elevation than were random ones (t = 20.8, dj = 80, P < 0.001; Table 2), which likely contributed to a better view. Concealment cover, however, was similar at birth and random sites (Table 2). Finally, birth sites had more than twice the available forage than did random sites (Table 2). Willow was primarily responsible for that relationship (birth sites, 19.4% ± 21.1 SD; random sites, 8.1 % ± 9.8 SD); other species of browse were < 1 % on all sites. The CV of willow cover on birth sites (109%) was less than on random sites (121 %). There also was a weak but inverse correlation between visibility and percent cover of willows (r =

-0.15, P = 0.10) for random sites. Female moose selected sites with a high

er quality of forage (e.g., willows) to give birth. Forage at birth sites was slightly but



TABLE 2.-Summary statistics for habitat characteristics of random and birth sites of Alaskan moose in Denali National Park and Preserve, Alaska, during spring 1990-1994. Suite of variables was analyzed with stepwise logistic regression.

Birth sites (n = 39) Random sites (n = 61)

Habitat variables X

Geographical

Distance to streams (m) 362 Distance to forest (m) 159 Distance to human

developments (m) 559

Topographical

E-W aspect (radians) 0.250 N-S aspect (radians) -0.450 Slope (%) 3.9 Terrain ruggedness (index) 6.24 Elevation (m) 829

Climatic

Julian date 145 Windspeed (km/h) 8.0 Presence of snow (%) 7.7 Overstory cover (%) 12.3 CV overstory cover (%) 51.0 Soil moisture (%) 135

Forage

Cover of browse (%) 20.2

Risk of predation

Concealment cover (%) 49.5 CV concealment cover (%) 53.9 Visibility (index) 10.1

significantly higher in Nand IVDMD than at random sites (Table 3).

DISCUSSION

Although one variable (visibility) that related to predation entered our logistic-regression model (Table 2, Fig. 4), the hypothesis concerning the role of human developments in protecting young moose from predation was rejected. Visitors to the Park observed grizzly bears pursuing female moose and their offspring, (and sometimes killing them) especially in the eastern end of the Park where a large campground (Riley Creek), train station, the Visitor Center, and Park Headquarters concentrated human activities (Albert and Bowyer, 1991). Parturient moose were thought to select such areas for giving birth because high

SD X SD

277 383 255 196 109 147

396 453 357

0.534 0.163 0.625 0.686 -0.147 0.759 3.3 4.1 3.5 9.20 5.91 8.48

174 733 170

15 147 14 6.4 8.3 9.2

27.0 12.9 33.7 17.0 13.6 23.3 61.1 37.3 58.4

172 185 275

20.7 8.8 9.7

19.6 43.4 25.4 33.2 61.5 45.7 11.3 7.2 9.5

levels of human activity around those developments deterred bears. Random locations, however, were significantly closer to human developments than were the birth sites of moose (Table 2). In addition, moose selected some areas in the Park for giving birth that had few human developments except the park road. Finally, low survivorship of young moose (Bowyer et al., 1998a) and bears killing them adjacent to human developments (Albert and Bowyer, 1991) indicated that location of birth sites near such developments had little affect on reducing bear-moose encounters. Langley and Pletscher (1994) also observed no relationship between birth sites of moose and their distance to human habitation in northwestern Montana and southeastern British Columbia.



35

30 OVERALL LOGISTIC MODEL, P = 0.0001

P= 0.002

Z 25

0 20 72.7% CONCORDANT

f- PREDICTIONS FOR

() 15 39 BIRTH SITES AND W 61 RANDOM SITES ....J 10 w C/)

P=0.036 P=0.033

(INDEX) (SIN) (% COVER)

HABITAT VARIABLES

FIG. 4.-Selection (use minus available) of habitat variables associated with birth sites of Alaskan moose in Denali National Park and Preserve, Alaska, from 1990 to 1994. Stepwise logistic regression indicated moose selected (use > availability) birth sites with more forage and greater visibility but avoided (use < availability) north-facing slopes (Le., selected south-facing ones).

The most likely explanation for numerous visitors observing grizzly bears preying on young moose was that rates of people visiting the Park have increased dramatically in recent years, and most campgrounds and other facilities were constructed in prime habitat or along routes of travel for bears (Albert and Bowyer, 1991). Moreover, the eastern end of the Park, which is the most intensively developed, is a traditional calving area for moose (Miquelle et aI., 1992; Fig. 1). Thus, humans, female moose and their neonates, and grizzly bears co-occurred in the same area, and observations of bears preying on moose increased.

Risk of predation was related to birth-site selection in moose, as other authors have proposed (Addison et al., 1990; Bailey and Bangs, 1980; Langley and P1etscher, 1994; Leptich and Gilbert, 1986). Our index of visibility (Table 2, Fig. 4) likely related to the ability of a female moose to locate predators before the predator became aware of her presence. Female moose and their neonates attempt to hide at the birth site to

TABLE 3.-Forage quality of willow (Salix) available to Alaskan moose at birth and random sites in Denali National Park and Preserve, Alaska, 1992-1994. Least-square means corrected for Julian date and year are presented. The overall Multivariate analysis of covariance (MANCOVA) comparing differences between sites was significant (F = 2.679, d.f. = 68, 104, P = 0.0001).

Birth sites Random sites

Measures of (n = 16) (n = 20)

quality X SD X SD P-value

Nitrogen (%) 2.5 0.5 2.4 0.6 0.0001 In vitro dry matter digestibility (%) 39.9 8.5 34.9 6.0 0.0012

elude predators rather than to flee immediately (R.T. Bowyer, in litt.). Offspring may join their mother if she flees, or if too young and small, remain motionless at the birth site when predators approach (Bowyer et aI., 1998a). Female moose sometimes stand their ground and attempt to defend their young from predators but may be killed themselves in doing so (Bowyer et al., 1998a). Although concealment cover at birth sites was similar to that at random sites (Table 2), there likely was sufficient vegetative cover to conceal an adult female from view when she was lying down. Consequently, female moose that observed predators before they were observed would have the opportunity to hide. This pattern of maternal care and defense does not fit traditional concepts related to the hider-follower dichotomy proposed for ungulates by Lent (1974) and Walther (1984). Indeed, several authors have questioned the usefulness of this concept as an organizing principle for understanding mother-young relationships among ungulates (Bowyer et al., 1998b; Green and Rothstein, 1993).

Moose (Stephens and Peterson, 1984) and other cervids (Bergerud, 1985; Bergerud and Page, 1987) are thought to "space away" from predators at the time of parturition. Birth sites were located at higher elevations than random sites (Table 2), as



others have reported for moose (Wilton and Garner, 1991). Moreover, predators tend to be less abundant in the Park at high elevations (Adams et aI., 1995a; Mech et al., 1998), and bears and wolves occur infrequently in the alpine zone (Rachlow and Bowyer, 1991). Other cervids presumably seek high-elevation sites to space away froIP predators (Adams et aI., 1995a; Bergerud et al., 1984). Evidence we collected, however, did not support that hypothesis. Shrub tundra occurred at higher elevation than other habitats used for parturition (Fig. 2), yet moose did not select that high-elevation habitat for giving birth (Fig. 3). Likewise, moose did not use high-elevation slopes of rugged and steep alpine habitat (Fig. 2) for calving. Dry washes and river bars were used extensively by grizzly bears as routes of travel (Albert and Bowyer, 1991), but moose did not avoid those habitats for birth sites (Fig. 3). Thus, use of higher-elevation sites (Table 2) likely related to the better view such sites provided, and moose selected such sites independently of the broad habitat types in which they occurred (Figs. 3 and 4). Perhaps other female ungulates that spaced themselves away from predators to give birth also could have been selecting sites that provided a superior view of approaching predators. Moose also locate birth sites on islands presumably to avoid predators (Addison et aI., 1990), but lakes and ponds were too rare in the eastern part of the Park (Fig. 2) to test that hypothesis.

Although moose did not space away from predators to elude them, maternal females may have attempted to negate hunting tactics of bears and wolves by behaving unpredictably at the level of the landscape. Twenty percent of preparturient females made extensive movements before giving birth. Females also did not select broad habitat types for parturition but used such habitats in proportion to their availability. The spatial arrangement of birth sites we located did not differ from random locations. Those behaviors would prevent pred-

ators from keying on previous locations of some females to aid in locating their neonates. Likewise, predators could not focus their hunting activities profitably in particular habitats or localized areas. We hypothesize that such unpredictable behavior by maternal females is a strategy to thwart some hunting tactics by predators, especially grizzly bears, which are the primary cause of death for young moose.

Females remaining at or near the birth site may represent an anti-predator strategy. Female moose are nearby to defend their young from predators, although such defense is not always successful (Bowyer et aI., 1998a). We hypothesize that remaining near the birth site would reduce scent trails deposited by the female that would lead to the birth site, thereby making neonates more difficult to locate. Moose possess interdigital glands (Chapman, 1985), and grizzly bears are thought to locate prey by following their scent (Craighead and Mitchell, 1982).

Another variable that may relate to risk of predation is the variability (CV) in overstory cover at a birth site (Table 2), which also was proposed for birth-site selection in black-tailed deer (Odocoileus hemionusBowyer et al., 1998b). Such variability in crown cover from a tree or tall shrub would create broad patches of sun and shade at the birth site, and that contrast might help camouflage hiding neonates. Similarly, Eastland et al. (1989) proposed that the high contrast produced by a patchy cover of snow might help conceal young caribou from view. Another possibility is that the lower portion of the tree or shrub that produced variability in overstory cover also would help break up the silhouette of a female moose standing against a hill top or skyline, thereby making her more difficult for a predator to locate visually. A standing female moose that was readily visible might provide a cue to predators as to the location of her neonate. After grizzly bears located the general area of a birth site, young moose seldom survived (Bowyer et aI., 1998a).



Variability in overstory cover (Table 2) also may have been associated with the thermal environment of the birth site, as Bowyer et al. (1998b) hypothesized for sites used by neonatal black-tailed deer. Southeast-facing slopes entered our logistic-regression model (Fig. 4) and may have been correlated with other climatic variables. For instance, slopes with a more southerly exposure undoubtedly were warmer than other aspects, and birth sites had significantly lower soil moisture than random sites (Table 2). Perhaps variability in overstory cover allowed neonates to thermoregulate more effectively by providing patches of sun and shade at the birth site. We cannot discriminate between that hypothesis and one related to predation from our data.

The final variable that entered our model was availability of forage (Fig. 4); selection for that variable was driven by abundance of willows, which were an important component in the diet of moose in the Park (Molvar and Bowyer, 1994; Molvar et al., 1993; Van Ballenberghe et aI., 1989). Our conclusion is supported by the heavy use of willows at birth sites compared with random sites (Table 1). Moose also selected sites to give birth with slightly but significantly higher-quality forage than at random sites (Table 3). Maternal females experience tremendous nutritional demands associated with lactation (White and Luick, 1984). Even slight differences in quality of forage can be crucial in successfully provisioning young cervids (White, 1983); this especially holds for moose, which remain with the neonate at the birth site for several weeks. That some females stripped bark to feed rather than consuming current annual growth of browse species indicated they experienced nutritional stress. Bark stripping is associated with undernutrition in moose during winter (Miquelle and Van Ballenberghe, 1989). Thus, forage quality and quantity played a major role in determining selection of birth sites by Alaskan moose. Moreover, moose selected larger, more pal-

atable stems when foraging than stems they left behind (Bowyer and Bowyer, 1997; Vivas et al., 1991). Consequently, stems we sampled after moose already had foraged at birth sites may have underestimated forage quality of the stems eaten by moose.

Three characteristics of the environment were generally responsible for birth-site selection in moose: risk of predation, microclimate, and forage abundance and quality. Moose apparently dealt with the needs for rearing young under climatic conditions that were hospitable by selecting slopes with a southerly exposure, which were available across an array of broad habitat types (Fig. 3). Meeting nutritional needs of females while avoiding predators, however, was more complex. Visibility, which ostensibly varied inversely with risk of predation, also was related inversely to abundance of forage-risk of predation varied directly with abundance of food. Sites were not available that allowed females to maximize forage while minimizing risk of predation. Thus, parturient females made a tradeoff between those variables in selecting sites where they gave birth.

Risk of predation, as indexed by visibility, was an important component of habitat selection by parturient females (Fig. 4); young moose may experience high rates of predation (Ballard et al., 1981; Bowyer et al., 1998a; Franzmann et aI., 1980; Gasaway et aI., 1992). We believe, however, that too little attention has been given to nutritional needs of maternal females in understanding where they give birth and how this selection relates to survivorship of their young.

More research is needed to understand how changes in population density or climatic variability affect habitat selection by females. For example, female DalI's sheep (Ovis dalli) selected areas with steep terrain in a year with good growing conditions for forage, but selected areas with more food in a year when growth of forage was limited by cool weather (Rachlow and Bowyer, 1998). Models of habitat selection also will



not include variables that are in sufficient supply in the environment, but such variables may be essential for rearing young successfully. Studies seldom make that distinction, and interannual variation in important components of habitat may be necessary to identify those factors or understand their value.

Moose tried to cope with a risky environment in both meanings of that term. First, they attempted to minimize risk of predation by selecting birth sites that allowed them to detect approaching predators at a sufficient distance to elude those large carnivores. Second, females were risk averse (Stephens and Krebs, 1986) with respect to selecting sites that allowed them to meet high nutritional costs of lactation. Failure to satisfy both of those demands has huge implications for their reproductive success. Female moose attempted to find a balance between those variables and, in doing so, made important maternal tradeoffs. Indeed, the lower CV for willows at birth sites is likely the result of moose avoiding sites with too little forage to meet requirements for lactation. The positive relationship between abundance of willow and risk of predation (inverse of visibility) also indicates that there were areas with ample food that were too dangerous to be used for birthing. Thus, we hypothesize that predation risk and demands of lactation help to cause risk-averse foraging by female moose at birth sites.

Much emphasis has been placed on mother-infant relationships (Stringham, 1974) and the role of maternal care in determining survivorship of young (Byers and Hogg, 1995; Clutton-Brock et al., 1987; Festa-Bianchet and Jorgenson, 1998). We suggest that habitat selection related to birth sites may help regulate type and amount of care given (Rachlow and Bowyer, 1994, 1998), and that birth-site selection is especially important in survival of young, particularly in environments with effective predators. Finally, a growing body of evidence indicates the hider-follower dichoto-

my is too simplistic to explain complex patterns of maternal behavior exhibited by ungulates; environmental conditions may play a larger role in influencing maternal behavior than previously recognized.

ACKNOWLEDGMENTS

We thank the personnel of Denali National Park and Preserve for their assistance during our field research, especially their help in arranging for the necessary permits. We also thank E. Rexstad for a helpful discussion concerning this manuscript. We are grateful to V. Baxter for assisting with the field work. B. M. Pierce and G. L. Kirkland, Jr. provided helpful comments on this manuscript. This research was funded in part by the Institute of Arctic Biology at the University of Alaska Fairbanks, and the United States Forest Service.

LITERATURE CITED

ADAMS, L. G., B. W. DALE, AND L. D. MECH. 1995a. Wolf predation on caribou calves in Denali National Park, Alaska. pp. 245-260, in Ecology and conservation of wolves in a changing world (L. N. Carbyn, S. H. Fritts, and D. R. Seip, eds.). Canadian Circumpolar Institute, Occasional Publication, 35:1-620.

ADAMS, L. G., F. J. SINGER, AND B. W. DALE. I 995b. Caribou calf mortality in Denali National Park, Alaska. The Journal of Wildlife Management, 59: 584-594.

ADDISON, E. M., W. L. WILTON, R. F. McLAUGHLIN, AND M. E. Buss. 1990. Calving sites of moose in Central Ontario. Alces, 26: 142-153.

AGRESTI, A. 1990. Categorical data analysis. John Wiley & Sons, New York.

ALBERT, D. M., AND R. T. BOWYER. 1991. Factors related to grizzly bear-human interactions in Denali National Park. Wildlife Society Bulletin, 19:339-349.

ALDOUS, S. E. 1944. A deer browse survey method. Journal of Mammalogy, 25:130-136.

ALTMANN, M. 1958. Social integration of the moose calf. Animal Behaviour, 6:155-159.

---. 1963. Naturalistic studies of maternal care in moose and elk. Pp. 233-253, in Maternal behavior in mammals (H. L. Rheingold, ed.). John Wiley & Sons, New York.

BAILEY, T. N., AND E. E. BANGS. 1980. Moose calving areas and use on the Kenai National Wildlife Refuge, Alaska. Proceedings of the North American Moose Conference and Workshop, 16:289-313.

BALLARD, W. B., T. H. SPRAKER, AND K. P. TAYLOR. 1981. Causes of neonatal moose calf mortality in southcentral Alaska. The Journal of Wildlife Management, 45:335-342.

BERGER, J. 1991. Pregnancy incentives, predation constraints and habitat shifts: experimental and field ev-



idence for wild bighorn sheep. Animal Behaviour, 41:61-77.

BERGERUD, A. T. 1985. Antipredator strategies of caribou: dispersion along shorelines. Canadian Journal of Zoology, 63:1324-1329.

BERGERUD, A. T., AND R. E. PAGE. 1987. Displacement and dispersion of parturient caribou at calving as antipredator tactics. Canadian Journal of Zoology, 65:1597-1606.

BERGERUD, A. T., H. E. BUTLER, AND D. R. MILLER. 1984. Antipredator tactics of calving caribou: dispersion in mountains. Canadian Journal of Zoology, 62:1566-1575.

BERNARDO, J. 1996. Maternal effects in animal ecology. American Zoologist, 36:83-105.

BLEICH, V. C. 1999. Mountain sheep and coyotes: patterns of predator evasion in a mountain ungulate.

. Journal of Mammalogy, 80:283-289. BLEICH, V. c., R. T. BOWYER, AND J. D. WEHAUSEN.

1997. Sexual segregation in mountain sheep: resources or predation? Wildlife Monographs, 134:1-50.

BOWYER, J. W., AND R. T. BOWYER. 1997. Effects of previous browsing on the selection of willow stems by Alaskan moose. Alces, 33: 11-18.

BOWYER, R. T. 1984. Sexual segregation in southern mule deer. Journal of Mammalogy, 65:410-417.

---. 1987. Coyote group size relative to predation on mule deer. Mammalia, 51:515-526.

---. 1991. Timing and synchrony of parturition and lactation in southern mule deer. Journal of Mammalogy, 72:138-145.

BOWYER, R. T., J. G. KIE, AND V. VAN BALLENBERGHE. 1996. Sexual segregation in black-tailed deer: effects of scale. The Journal of Wildlife Management, 60: 10-17.

---. 1998b. Habitat selection by neonatal blacktailed deer: climate, forage, or risk of predation? Journal of Mammalogy, 70:415-425.

BOWYER, R. T., V. VAN BALLENBERGHE, AND J. G. KIE. 1997. The role of moose in landscape processes: effects of biogeography, population dynamics, and predation. Pp. 265-287, in Wildlife and landscape ecology: effects of pattern and scale (J. A. Bissonette, ed.). Springer-Verlag, New York.

---. 1998a. Timing and synchrony of parturition in Alaskan moose: long-term versus proximal effects of climate. Journal of Mammalogy, 79:1332-1344.

BYERS, J. A., AND J. T. HOGG. 1995. Environmental effects on prenatal growth rate in pronghorn and bighorn: further evidence for energy constraint on sexbiased maternal expenditure. Behavioral Ecology, 6: 451-457.

CANFIELD, R. H. 1941. Application of the line intercept method in sampling range vegetation. Journal of Forestry, 39:388-394.

CEDERLUND, G., R. SANDEGREN, AND K. LARSSON. 1987. Summer movements of female moose and dispersal of their offspring. The Journal of Wildlife Management, 51:342-352.

CHAPMAN, D. M. 1985. Histology of the moose (Alces alees) interdigital glands and associated green hairs. Canadian Journal of Zoology, 63:899-911.

CLUTION-BROCK, T. H. 1991. The evolution of parental

care. Princeton University Press, Princeton, New Jersey.

CLUTION-BROCK, T. H., M. MAJOR, S. D. ALBON, AND F. E. GUINNESS. 1987. Early development and population dynamics in red deer. I. Demographic consequences of density-dependent changes in birth weight and date. The Journal of Animal Ecology, 56:53-67.

CRAIGHEAD, J. J., AND J. A. MITCHELL. 1982. Grizzly bear. Pp. 515-556, in Wild mammals of North America: biology, management and economics (J. A. Chapman and G. A. Feldhammer, eds.). The Johns Hopkins University Press, Baltimore, Maryland.

EASTLAND, W. G., R. T. BOWYER, AND S. G. FANCY. 1989. Effects of snow cover on selection of calving sites by caribou. Journal of Mammalogy, 70:824-828 .

FESTA-BlANCHET, M., AND J. T. JORGENSON. 1998. Selfish mothers: reproductive expenditure and resource availability in bighorn ewes. Behavioral Ecology, 9: 144-150.

FRANZMANN, A. W., C. C. SCHWARTZ, AND R. O. PETERSON. 1980. Moose calf mortality in summer on the Kenai Peninsula, Alaska. The Journal of Wildlife Management, 44:764-768.

GASAWAY, W. c., R. D. BOERTJE, D. V. GRANGAARD, D. G. KELLYHOUSE, R. O. STEPHENSON, AND D. G. LARSEN. 1992. The role of predation in limiting moose at low densities in Alaska and Yukon and implication for conservation. Wildlife Monographs, 120:1-59.

GREEN, W. C., AND A. ROTHSTEIN. 1993. Asynchronous parturition in bison: implications for the hider-follower dichotomy. Journal of Mammalogy, 74:920-925.

GRIFFITH, D. B., AND B. A. YOUTIE. 1988. Two devices for estimating foliage density and hiding cover. Wildlife Society Bulletin, 16:206-210.

HIRTH, D. H. 1977. Social behavior of white-tailed deer in relation to habitat. Wildlife Monographs, 53: 1-55.

KEECH, M. A., R. D. BOERTJE, R. T. BOWYER, AND B. W. DALE. 1999. Effects of birth weight on growth of young moose: do low-weight neonates compensate? Alces, 35:51-57.

KIE, J. G. 1999. Optimal foraging and risk of predation: effects on behavior and social structure in ungulates. Journal of Mammalogy, 80:1114-1129.

KIE, J. G., AND R. T. BOWYER. 1999. Sexual segregation in white-tailed deer: density-dependent changes in use of space, habitat selection, and dietary niche. Journal of Mammalogy, 80:1004-1020.

KOHLMANN, S. G., D. M. MULLER, AND P. U. ALKON. 1996. Antipredator constraints on lactating Nubian ibexes. Journal of Mammalogy, 77:1122-1131.

LANGELY, M. A., AND D. H. PLETSCHER. 1994. Calving areas of moose in northwestern Montana and southeastern British Columbia. Alces, 30:127-135.

LANGENAU, E. E., JR., AND J. M. LERG. 1976. The effects of winter nutritional stress on maternal and neonatal behavior in penned white-tailed deer. Applied Animal Ethology, 2:207-223.

LEMMON, P. E. 1957. A new instrument for estimating overstory density. Journal of Forestry, 55:667-669.



LENT, P. C. 1974. Mother-infant relationships in ungulates. Pp. 14-53, in The behaviour of ungulates and its relation to management (V. Geist and F. Walther, eds.). International Union for Conservation of Nature and Natural Resources, Publications New Series,24:1-940.

LEPTICH, D. J., AND J. R. GILBERT. 1986. Characteristics of moose calving sites in northern Maine: a preliminary investigation. Alces, 22:69-81.

MACCRACKEN, J. G., V. VAN BALLENBERGHE, AND J. M. PEEK. 1997. Habitat relationships of moose on the Copper River Delta in coastal south-central Alaska. Wildlife Monographs, 136:1-52.

MECH, L. D., L. G. ADAMS, T. J. MEIER, J. W. BURCH, AND B. W. DALE. 1998. The wolves of Denali. University of Minnesota Press, Minneapolis.

MILLAR, J. S. 1977. Adaptive features of mammalian reproduction. Evolution, 31 :370-386.

MILLER, B. K., AND J. A. LITVAITIS. 1992. Habitat segregation by moose in a boreal forest ecotone. Acta Theriologica,37:41-50.

MIQUELLE, D. G., AND V. VAN BALLENBERGHE. 1989. Impact of bark stripping by moose on aspen-spruce communities. The Journal of Wildlife Management, 53:577-586.

MIQUELLE, D. M., J. M. PEEK, AND V. V AN BALLENBERGHE. 1992. Sexual segregation in Alaskan moose. Wildlife Monographs, 122:1-57.

MOLVAR, E. M., AND R. T. BOWYER. 1994. Costs and· benefits of group living in a recently social ungulate: the Alaskan moose. Journal of Mammalogy, 75: 621-630.

MOLVAR, E. M., R. T. BOWYER, AND V. VAN BALLENBERGHE. 1993. Moose herbivory, browse quality, and nutrient cycling in an Alaskan treeline community. Oecologia, 94:472-479.

NICHOLSON, M. C., R. T. BOWYER, AND J. G. KIE. 1997. Habitat selection and survival of mule deer: tradeoffs associated with migration. Journal of Mammalogy, 78:483-504.

PETERSON, R. L. 1995. North American moose. University of Toronto Press, Toronto, Ontario, Canada.

POND, C. M. 1977. The significance of lactation in the evolution of mammals. Evolution, 31:177-199.

RACHLOW, J. L., AND R. T. BOWYER. 1991. Interannual variation in timing and synchrony of parturition in Dall's sheep. Journal of Mammalogy, 72:487-492.

---. 1994. Variability in maternal behavior by Dall's sheep: environmental tracking or adaptive strategy? Journal of Mammalogy, 75:328-337.

---. 1998. Habitat selection by Dall's sheep (Ovis dalli): maternal trade-offs. Journal of Zoology (London), 245:457-465.

RALLS, K. 1977. Sexual dimorphism in mammals: avian models and unanswered questions. The American Naturalist, 122:917-938.

SAS Institute Inc. 1988. SAS/STAT user's guide release 6.03. SAS Institute Inc., Cary, North Carolina.

SCHWARTZ, C. C. 1998. Reproduction, natality, and growth. pp. 141-171, in Ecology and management of North American moose (A. W. Franzmann and C. C. Schwartz, eds.). Smithsonian Institution Press, Washington, D.C.

SCHWARTZ, C. c., AND K. J. HUNDERTMARK. 1993. Re-

productive characteristics of Alaskan moose. The Journal of Wildlife Management, 57:454-468.

SLAUSON, W. L., B. S. CADE, AND J. D. RICHARDS. 1991. Users manual for BLOSSOM statistical software. United States Fish and Wildlife Service, National Research Center, Fort Collins, Colorado.

SMITH, W. P. 1987. Maternal defense in Columbian white-tailed deer: when is it worth it? The American Naturalist, 130:310-316.

STEPHENS, D. W., AND J. R. KREBS. 1986. Foraging theory. Princeton University Press, Princeton, New Jersey.

STEPHENS, P. W., AND R. O. PETERSON. 1984. Wolfavoidance strategies of moose. Holarctic Ecology, 7: 239-244.

STRINGHAM, S. F. 1974. Mother-infant relations in moose. Naturaliste canadien, 101 :325-369.

VAN BALLENBERGHE, V., AND W. B. BALLARD. 1994. Limitations and regulation of moose populations: the role of predation. Canadian Journal of Zoology, 72: 2071-2077.

VAN BALLENBERGE, v., D. G. MIQUELLE, AND J. G. MCCRACKEN. 1989. Heavy utilization of woody plants by moose during summer in Denali National Park, Alaska. Alces, 25:31-35.

VAN SOEST, P. J. 1982. Nutritional ecology of the ruminant. 0 & B Books, Corvallis, Oregon.

VIERECK, L. A., C. T. DYRNESS, A. R. BAITEN, AND K. J. WENZLICK. 1992. The Alaskan vegetation classification. United States Forest Service General Technical Report, PNW-GTR-286: 1-278.

VIVAS, H. J., B. E. SAETHER, AND R. ANDERSON. 1991. Optimal twig-size selection of a generalist herbivore, the moose Alces alces: implications for plant-herbivore interactions. The Journal of Animal Ecology, 60:395-408.

WALTHER, F. R. 1984. Communication and expression in hooved mammals. Indiana University Press, Bloomington.

WECKERLY, F. W. 1998. Sexual size dimorphism: influence of mass and mating systems in the most dimorphic mammals. Journal of Mammalogy, 79:33-52.

WEIXELMAN, D. A., R. T. BOWYER, AND V. VAN BALLENBERGHE. 1998. Diet selection by Alaskan moose during winter: effects of fire and forest succession. Alces, 34:213-238.

WHITE, R. G. 1983. Foraging patterns and their multiplier effects on productivity of northern ungulates. Oikos, 40:377-384.

WHITE, R. G., AND J. R. LUICK. 1984. Plasticity and constraints in the lactational strategy of reindeer and caribou. Symposia of the Zoological Society of London, 51:215-232.

WILTON, M. L., AND D. L. GARNER. 1991. Preliminary findings regarding elevation as a major factor in moose calving site selection in South Central Ontario, Canada. Alces, 27: 111-117.

ZAR, J. H. 1984. Biostatistical analysis. Second ed. Prentice-Hall, Inc., Englewood Cliffs, New Jersey.

ZIMMERMAN, G. M., H. GoETZ, AND W. P. MIELKE. 1985. Use of an improved statistical method for group comparisons to study effects of prairie fire. Ecology, 66:606-611.


birth-site selection by alaskan moose: maternal strategies for coping ...

Documents

Transcript of birth-site selection by alaskan moose: maternal strategies for coping ...