Activity Patterns, Foraging Ecology, and Summer Range Carrying Capacity of Moose (Alces Alces Shirasi) in Rocky Mountain National Park, Colorado

ALCES VOL. 46, 2010 DUNGAN ET AL. - MOOSE SUMMER RANGE CARRYING CAPACITY

ACTIVITY PATTERNS, FORAGING ECOLOGY, AND SUMMER RANGE CARRYING CAPACITY OF MOOSE (ALCES ALCES SHIRASI) IN ROCKY MOUNTAIN NATIONAL PARK, COLORADO

Jason D. Dungan1, Lisa A. Shipley2, and R. Gerald Wright3

1Department of Fish and Wildlife Resources, University of Idaho, P.O. Box 441136, Moscow, ID 83844-1136, USA; 2Department of Natural Resource Sciences, Washington State University, 105 Johnson Hall, Pullman, WA 99164-6410, USA; 3USGS Idaho Cooperative Fish and Wildlife Research Unit, Department of Fish and Wildlife Resources, University of Idaho, P.O. Box 441136, Moscow, ID 83844-1136, USA

ABSTRACT: To ensure sustainable populations of native animals and plants, managers of protected areas must understand carrying capacity of large wild herbivores. Estimates of carrying capacity and how large herbivores may influence native vegetation require knowledge of their activity and foraging patterns. Therefore, we examined activity patterns and foraging behavior of adult male moose (Alces alces shirasi) in Rocky Mountain National Park, Colorado by following individual animals and counting bites during the summer and fall of 2003 and 2004. Mean active time per day peaked in late June at 11.3 h and declined to 8.9 h in early fall preceding the breeding season. Moose averaged 6.7 feeding periods/d, each lasting 79 min; feeding bouts were longer around sunrise and sunset and were shorter midday presumably because of high ambient temperature. Activities associated with feeding and resting constituted 94.0% of daily time budgets. Feeding declined and social behavior and movement increased in fall with the onset of the breeding season. Food consumption increased steadily through early summer peaking at 126.8 g/kg BW0.75 in early August, followed by a sharp decline to a low of 69.1 g/kg BW0.75 in early September. Daily digestible energy intake was estimated at 1191 kJ/ kg BW0.75/d. Maximum rates of instantaneous intake were recorded in early August at 22.3 g/min. Because intake rates of willow (Salix spp.) increased from June-August, but nutritional quality peaked in mid-June, increases in daily and instantaneous intake rates during summer seemed more related to forage availability than protein and energy content of willow leaves. The nutritional carrying capacity of summer range in Rocky Mountain National Park in 2004 was estimated from the range supply of metabolizable energy, digestible energy, and available nitrogen. Based on the digestible energy intake and energy requirements of a 344 kg male moose, the summer range carrying capacity was estimated at 0.21 moose/km2. Nitrogen based estimates were considerably higher at 0.32 moose/km2.

ALCES VOL. 46: 71-87 (2010)

Key words: activity patterns, Alces alces, behavior, carrying capacity, consumption, energy, foraging, moose, summer.

Because hunting is prohibited and large greatest extent possible to regulate ungulate carnivores are rare in many protected areas in populations (NPS 2001). However, the policy the world, resource managers are concerned is flexible resulting in varied management ap- about the potential effects of large populations proaches in different situations. Where natural of wild herbivores on sensitive plant species controls have been altered by human activity, (Schreiner et al. 1996, Baker et al. 1997, Au- unnatural concentrations of ungulates may be gustine and McNaughton 1998). The United managed by park staff (Huff and Varley 1999). States National Park Service (NPS) states that The number of Rocky Mountain elk (Cervus natural processes should be relied upon to the elaphus) in the eastern portion of the Rocky

1Present address: Wallowa Whitman National Forest, Whitman Ranger District, 3285 11th St., Baker City, OR 97814, USA 71 MOOSE SUMMER RANGE CARRYING CAPACITY- DUNGAN ET AL. ALCES VOL. 46, 2010

Mountain National Park (RMNP) in Colorado of northern ungulates (Crete 1989, Kufeld was artificially reduced in 1943-1968 in an and Steinert 1990, MacCracken et al. 1997, attempt to maintain the herd at about 500 Zheleznov-Chukotsky and Votiashova 1998), (Wright 1992). Since 1968, elk have no longer especially moose (LeResche and Davis 1973). been controlled within RMNP and the popu- However, ecologists are increasingly focusing lation has increased substantially and been their attention on forage availability and K of implicated in the decline of the abundance, summer and autumn ranges (Hett et al. 1978, distribution, and stature of riparian willow Beck and Peek 2001, Beck et al. 2006). Sum- (Salix spp.) and upland shrub communities mer is a key period in which moose recover in the eastern portion of RMNP (Singer and and build fat reserves for fall breeding and Zeigenfuss 2002). More recently, managers winter survival (Schwartz et al. 1988b), and have expressed concern for the health of ripar- body condition in fall has been linked to preg- ian willow communities in the western portion nancy rates (Testa and Adams 1998). Moose as a consequence of a presumed increase in enter winter with an estimated 20-26% body moose (Alces alces shirasi). fat (Schwartz et al. 1988b) which is critical Understanding how many herbivores because winter diets are insufficient to meet a given ecosystem can sustain and maintain nutrient requirements and moose enter energy ecological integrity (i.e., carrying capacity, deficit. Although summer diets of moose are K) is an important, yet often elusive and dif- typically 1.5-3 times more nutritious than ficult goal for resource managers (Caughley winter diets (Schwartz 1992), warm summer 1979, Hobbs and Swift 1985, Regelin et al. temperatures may restrict the time large-bod- 1987, Boyce 1989). For wild ungulates that ied ungulates forage (Taylor and Maloiy 1970, rely on plants that vary greatly in quality and Van Soest 1982, Hudson and White 1985, abundance both temporally and spatially, Owen-Smith 1998); for example, Renecker carrying capacity based on nutritional needs and Hudson (1986a) measured reduced intake is generally measured as the ratio of nutrient rate and loss of body weight during warm sum- supply of the range to the nutrient requirement mer periods. Summer habitat of poor quality of the animal (Wallmo et al. 1977, Hobbs et and/or increasing summer temperature due to al. 1982, Potvin and Huot 1983), and often global climate change may represent limiting integrates information on food habits, daily factors in how moose prepare for and survive intake, and the forage production of a given winter regardless of availability and quality area (Crete 1989, Kufeld and Steinert 1990, of winter forage. Zheleznov-Chukotsky and Votiashova 1998). Despite the importance of summer nutri- However, these methods require accurate tion for moose populations, and the influence measures of seasonal changes in intake, diet of high ambient temperatures on moose forag- selection, and forage production. Previous ing behavior, few studies have examined the studies with moose have examined differences foraging ecology, activity patterns, and K of an in seasonal dry matter intake (DMI) among area for moose in summer. The objectives of seasons (Schwartz et al. 1984, Renecker and this study were to 1) examine activity patterns Hudson 1985, Hubbert 1987), but the dynamics and foraging behavior of free-ranging male within seasons has received less attention. moose during summer and fall in RMNP, 2) Most estimates of K for ungulate popula- investigate the relationship between ambient tions have focused on winter ranges (Hobbs et temperature and daily activity budgets, and 3) al. 1982, Potvin and Huot 1983, Crete 1989, use these data with nutritional and metabolic MacCracken et al. 1997) because winter information from the literature to estimate the food supply is considered the limiting factor nutritional carrying capacity of summer range

72 ALCES VOL. 46, 2010 DUNGAN ET AL. - MOOSE SUMMER RANGE CARRYING CAPACITY for moose in RMNP. attempt to record complete 24 h observation periods, 1 June-30 September 2003 and STUDY AREA 2004. Individual moose were identified based This study was conducted during summer- on antler configuration, and an attempt was early fall in 2003 and 2004 in RMNP that is made to sample as many different moose as located in north-central Colorado, covering possible. Because individual moose were 1,075 km2 at 2,389-4,345 m elevation; RMNP not marked, and an attempt to mark moose lies astride the continental divide with differ- was unsuccessful, the same moose may have ent climates on the west and east sides. We been sampled both years. Observer safety conducted this study on the west side within was assessed from the initial reaction of the the Colorado River Drainage, an area covering moose to the observer and continued behav- 390.8 km2. Annual precipitation ranges from ior over several hours before an observation 37.6-51.7 cm with a temperature range of 24o shift began. A moose showing aggressive or C in July-August and -17o C in December- annoyed behavior toward observers was not February (Monello and Johnson 2003). followed. We recorded locations of moose Lower elevation riparian meadows are with hand-held global positioning systems characterized by large stands of geyer willow (GPS) units. At night we used headlamps to (Salix geyeriana), mountain willow (S. monti- record categorical behavior, but were unable cola), drummond willow (S. drummondiana), to make detailed observations of feeding plane-leaf willow (S. planifolia), and smaller behavior (e.g., bite counts). Observations of stands of whiplash willow (S. lasiandra) and different activities were recorded in journals, wolf willow (S. wolfii). Other common species and we used hand-held voice recorders to are beaked sedge (Carex utriculata), bog birch document feeding behavior; observers were (Betula glandulosa), mountain alder (Alnus replaced every 8-10 h. incana), marsh reed grass (Calamagrotis canadensis), western dock (Rumex aquati- Activity Budgets cus), white clover (Trifolium repens), and Activity patterns and time budgets were strawberry (Fragaria ovalis). These areas are estimated by continuous time sampling over surrounded by stands of ponderosa pine (Pinus 24-h observation periods. Activities were ponderosa), Douglas fir (Pseudotsuga men- categorized as resting, feeding, standing, ziesii), quaking aspen (Populus tremuloides), extended movement, engaged in social inter- and narrowleaf cottonwood (P. angustifolia). actions, and other (e.g., drinking, defecating, Higher elevation meadows are characterized grooming). We recorded all major activities by large stands of plane-leaf willow, wolf to the nearest minute. Extended movement willow, and bog birch; surrounding trees in- was considered any movement lasting >5 clude quaking aspen, lodgepole pine (Pinus min. We used only complete activity bouts contorta), and subalpine fir (Abies lasiocarpa) lasting >15 min in our analyses of duration (Beidleman et al. 2000). and number of feeding and resting bouts per day. Activity and resting bouts lasting ≤15 min METHODS were generally associated with disturbances We directly observed and measured activ- caused by the observer or other animals. All ity patterns and foraging behavior of moose activity data were used in estimating amount at close range (5-20 m) because moose in of time active per day and time budgets; data RMNP are habituated to people and tolerate were grouped into biweekly periods. close observation. We observed free-ranging Following Risenhoover (1986), we com- adult male moose as long as possible in an pared the amount of time spent active (feed-

73 MOOSE SUMMER RANGE CARRYING CAPACITY- DUNGAN ET AL. ALCES VOL. 46, 2010 ing, moving, social, other) during the day and were bagged, oven-dried at 60 oC for 48 h, and night using the normalized day:night ratio (R) weighed as described by Renecker and Hudson calculated by the equation: (1985). Bite size (g/bite) of each consumed plant species was estimated separately.

Da / Dt We calculated intake rate as the product R = of mean bite size and mean bite rate, both on Na / Nt an instantaneous and daily basis. Daily DMI (g/day) was the product of species-specific where Da is the number of daylight min active, intake rates and foraging time per day for each

Dt is the total of daylight min, Na is the number species in the diet. Foraging time per day was of night min active, and Nt is the total number the product of the average number of feeding of night min. We estimated the length of day periods per day and mean continuous feeding and night periods using sunset and sunrise within bouts. Estimates of overall summer data for Denver, Colorado obtained from the consumption were the product of daily DMI, National Climatic Data Center, U.S. Depart- number of days per period, and number of ment of Commerce, Washington, D.C. periods per summer. We calculated digestible energy intake Foraging Ecology (DEI) by multiplying estimates of summer Feeding data were grouped into individual consumption of the principle forage (willow) feeding bouts. We defined feeding bouts as any by its dry matter digestibility (DMD) and gross span of feeding ≥15 min. Feeding time within energy (GE) estimates in the literature. The bouts was the sum of seconds spent biting, DMD of willow in RMNP during summer chewing, swallowing (but not ruminating), 2004 was 60.2% (Stumph 2005), and the GE and moving between bites. We stopped voice estimate of willow was 20.38 kJ/g (Hjeljord et recorders during non-feeding activities within al. 1982). Metabolizable energy (ME) intake bouts to calculate continuous feeding within was assumed to be 88.6% of DEI (Schwartz bouts. Time was recorded at the end of each et al. 1985). bout to estimate average bout length, and total feeding time within bouts was the proportion Nutritional Carrying Capacity of the time spent continuously feeding within We estimated nutritional K using data a given feeding bout. from the literature and from our feeding trials. We counted individual bites taken, and First, our estimates of forage biomass was these were classified by plant species when based solely on willow forage because 1) wil- possible. Bite rates (bites/min) were deter- low habitat was most used by all moose in all mined by continuous time sampling over the seasons in North Park, Colorado in 1991-1995 duration of the bout. Bite rate for each plant (Kufeld and Bowden 1996) and in summer in species was estimated during feeding periods, RMNP (Dungan 2007), and 2) 6 willow species and mean bite rates were grouped according comprise 91.3% of the summer diet in RMNP to species and bout. (June through mid-September) (Dungan and During feeding periods, the sizes of all Wright 2005). Estimates of annual biomass bites were recorded to estimate intake rates, production (269,974 kg), available dry matter and bites were classified as small (1-5 leaves), (162,524 kg), and available protein (22,012 medium (5-10 leaves), or large (>10 leaves). kg) of willow on the western side of RMNP After each observation, simulated moose bites were obtained from Stumph (2005). Because were collected by clipping 10-20 samples/ 100% use of forage is clearly not sustainable, species in each of the 3 sizes. These samples we reduced total available biomass to reflect

74 ALCES VOL. 46, 2010 DUNGAN ET AL. - MOOSE SUMMER RANGE CARRYING CAPACITY sustainable utilization by moose. We based Colorado were also unavailable. Because most this reduction on the “allowable use criteria” of our foraging data were recorded from adult recommended by Singer and Zeigenfuss bulls (5.5+ years), we used an estimate of the (2002); we assumed that ≤21% utilization of body mass of male Shiras moose from Lander, riparian willow by elk in RMNP would al- Wyoming during winter (344 kg; Hanna et low riparian willow communities to recover. al. 1989). Daily energy requirements for Because more biomass is available during maintenance were calculated from this body summer, we reduced total available biomass mass using 1) activity budgets measured from available to moose for foraging (R) by 75% our moose, and 2) the energy expenditure of (Frank and McNaughton 1992). This conser- 2 free-ranging, non-pregnant female moose vative reduction reduced our values of annual weighing 320 ± 5 kg during July in central biomass production to 67,494 kg, available dry Alberta, Canada (Renecker and Hudson mass to 40,631 kg, and available protein to 1989a). The time spent per day in various 5503 kg. The range supply (RS) of available activities during summer was multiplied by nitrogen (N) was calculated by multiplying the the energetic costs of those activities, and the available protein estimates by 0.16 (Robbins costs summed over 24 h (Table 1). The nitro- 2001). Available dry matter was calculated by gen requirement for winter maintenance was multiplying estimates of annual biomass pro- estimated as 0.627 g/kg BW0.75/d (Schwartz et duction by the digestibility coefficient (60.2%; al. 1987b). Because no estimate of nitrogen Stumph 2005). The RS of digestible energy requirement for summer maintenance exists, (DE) was calculated by multiplying estimates this estimate was increased 33% (0.835 g/kg of available dry matter of willow by the GE BW0.75/d) to match the % increase of activity content of willow (20.38 kJ/g; Hjeljord et al. level during summer (VanBallenberghe and 1982). Finally, the RS of ME was the product Miquelle 1990). of DE x 0.89 (Schwartz et al. 1985). Using data on energy and nitrogen pro- Next, we calculated energy and nitrogen duction at RMNP and energy and nitrogen requirements of moose from the literature. requirements of moose, we estimated K for Unfortunately we were unable to weigh our male moose on summer range in RMNP dur- study moose and summer weights of moose in ing the 2004 growing season with 3 different Table 1. Calculations of daily energy requirements for a 344 kg moose during summer in Rocky Mountain National Park, Colorado. Activity Daily time spent per activity Energy costa per activity Daily energy cost (hr) (kJ/kg BW0.75/hr) (kJ/kg BW0.75/day) Feeding 8.91 39.7 353.7 Resting 4.53 26.9 121.9 Resting/Ruminatingb 9.17 30.0 275.1 Moving 0.55 71.1 39.1 Standing 0.50 51.3 25.7 Social Interactionsc 0.34 57.4 19.5 Total 835.0 a Energy cost estimates are the average expenditure costs for various activities in May and July; taken from Renecker and Hudson (1989a). bTime spent resting/ruminating is based on 67% time spent resting (Risenhoover 1986). cEnergy cost of social interaction was based on estimates of grooming (Renecker and Hudson 1989a). 75 MOOSE SUMMER RANGE CARRYING CAPACITY- DUNGAN ET AL. ALCES VOL. 46, 2010 models. The first model was based on esti- gression to analyze relationships between bite mates of DEI across the summer and the RS sizes and bite rates. We compared differences of DE. The second model was based on the in the duration of feeding bouts and mean RS of ME and the daily energy requirements temperatures within bouts using a one-way of an adult bull moose. The third model was parametric ANOVA. For this analysis, we based on the RS of available N and daily N only used data from feeding bouts recorded requirements of moose. during midday, and from late June-August, The estimate of K on summer range in because of significant differences in feeding RMNP was calculated in Model 1 by divid- bout length during other periods. We obtained ing the RS of DE by DEI during the summer, temperature data every 15 min from a mini- in Model 2 by dividing the RS of ME by the weather station located within the study area product of daily energy requirements multi- (Universal Transverse Mercator zone 13, plied by number of days of summer, and in Northing 4470423.01, Easting 427322.43, Model 3 by dividing the RS of available N by elevation 2719.10 m), and used those data the daily N requirements during summer. We to estimate temperature during resting and assumed that summer was 106 days (1 June-15 feeding bouts. September) based on observations from May- All pairwise differences were located us- December, 2002-2004 (Dungan 2007). ing the Tukey HSD procedure. Differences were considered significant at alpha <0.05 Statistical Analyses and all statistics were performed using SAS We compared biweekly differences in ac- 8.3 (SAS Institute Inc., Cary, North Carolina) tive time per day, number of feeding periods statistical software. Results are reported as per day, duration of feeding and resting bouts, mean ± standard deviation (SD). continuous feeding within feeding bouts, day:night ratios, species-specific bite sizes, RESULTS species-specific bite rates, and rate of move- Activity Budgets ment within feeding bouts using a two-way Active time per day ─ We recorded 208 analysis of variance (ANOVA) (Sokal and individual free-ranging male moose during Rohlf 1987), with time periods and years as summer 2003 and 2004; 55 individual males the main effects. We compared differences were identified based on antler configuration in percent time engaged in different activities and photographs with 29 observed on mul- (time budgets) using a two-way parametric tiple occasions. Direct observations resulted ANOVA on arcsine-transformed data to in 1456 h of data collected June-September, correct for non-normal distributions (Sokal 2003 and 2004. Data for 37 complete 24-h and Rohlf 1987). Experimental units were observation periods were obtained from 17 complete 24-h observation periods for these individual adult males. Data on duration of analyses. Following Van Ballenberghe and activity and resting bouts were obtained from Miquelle (1990), changes in the duration of both complete and incomplete 24-h observa- feeding and resting bouts during the 24-h day tion periods. Activity budgets did not differ were tested by dividing the day into eight 3-h between years (F = 0.11, df = 1, 28, P >0.05); periods and comparing means. therefore, data were pooled across years for We used linear regression to separately biweekly comparisons of activity. Daily ac- analyze the relationship between mean tem- tivity fluctuated greatly across months (F = perature and the duration of feeding and resting 2.54, df = 5, 28, P = 0.05) (Table 2). bouts, and between movement and bite rates Time budgets ─ Moose averaged 10.1 h/ within feeding bouts. We used non-linear re- day active and 13.9 h/day inactive during sum-

76 ALCES VOL. 46, 2010 DUNGAN ET AL. - MOOSE SUMMER RANGE CARRYING CAPACITY

Table 2. Mean (X ± SD) biweekly activity patterns of moose in Rocky Mountain National Park, Colorado, 16 June-15 September, 2003 and 2004. Letters indicate significant (P ≤0.05) differences in activity patterns between periods. Day:night ratio refers to the time spent active during the day compared to night. Period n Active hr/day Feeding periods/day Day:night ratio (:1) June 16-30 2 11.4 ± 0.1 A 6.9 ± 0.1 AB 1.23 ± 0.13 AB July Jan-15 3 10.3 ± 0.8 AB 6.3 ± 0.3 AB 0.79 ± 0.14 B 16-31 6 10.8 ± 0.2 A 7.3 ± 0.5 B 1.27 ± 0.44 AB Aug Jan-15 10 9.0 ± 1.1 B 6.9 ± 0.6 B 0.98 ± 0.21 B 16-31 4 8.9 ± 0.9 B 7.2 ± 0.5 B 0.89 ± 0.14 B Sept Jan-15 10 9.8 ± 1.7 AB 5.5 ± 0.7 A 1.64 ± 0.46 A mer and fall. Activities associated with feeding 2a). The mean duration of feeding bouts was and resting constituted 94.0% of daily time 79.4 ± 29.9 min (n = 277, range = 19.3-277.7). budgets. Other activities included extended 45 A. FEEDING movement (2.3%), standing (2.1%), and social B B 40 B behavior (1.4%). The dominant social behav- B B B ior was sparring (41.3 %), followed by fraying 35 and thrashing of trees (32.1%), smelling and 30 following of females (15.2%), and pawing 25 A 20 and urinating in pit holes (6.1%). JUNE 16-30 JULY 1-15 JULY 16-31 AUG 1-15 AUG 16-31 SEPT 1-15SEPT 16-30

14 The percent time engaged in activities B. EXTENDED MOVEMENT A differed among biweekly time periods for 12 10 A feeding (F = 3.29, df = 6, 28, P = 0.01), 8 extended movement (F = 6.82, df = 6, 28, 6 4 P = 0.0002), and social behavior (F = 6.24, B B

ACTIVITIES(%) B 2 B B df = 6, 28, P = 0.0003). Feeding was lower 0 in late September, larger extended move- JUNE 16-30 JULY 1-15 JULY 16-31 AUG 1-15 AUG 16-31 SEPT 1-15SEPT 16-30 ments occurred in late June and September, 5 C. SOCIAL INTERACTION A and social interaction was higher in mid-late 4 September (Fig. 1). The average day:night 3 2 A ratio for time spent active (active day:active 1 B B B B B night) was 1.13:1. Average biweekly day:night 0 ratios varied monthly (F = 9.19, df = 5, 28, JUNE 16-30 JULY 1-15 JULY 16-31 AUG 1-15 AUG 16-31 SEPT 1-15SEPT 16-30 P <0.0001) peaking in late July and early PERIOD September (Table 2). Fig. 1. Percentage of feeding (A), extended move- Feeding and resting bouts ─ Moose ment (B), and social interaction (C) activities of engaged in more feeding bouts per day in late male moose during 2-week observation periods June-August than early September (F = 11.97, in Rocky Mountain National Park, Colorado, df = 5, 28, P <0.0001) (Table 2). Duration of 16 June-30 September, 2003 and 2004. Letters feeding bouts differed among bi-weekly time indicate significant (P ≤0.05) differences in periods (F = 4.98, df = 5, 270, P = 0.0002) (Fig. activity levels between periods. 77 MOOSE SUMMER RANGE CARRYING CAPACITY- DUNGAN ET AL. ALCES VOL. 46, 2010

140 A tion was shortest those same periods (Fig. 3). A. FEEDING BOUTS

120 The duration of feeding bouts declined with AB B increasing mean temperature over 24-h period AB B 100 2 B (r = 0.032, P = 0.01), and between dawn and 2 80 dusk (r = 0.21, P <0.001) (Fig. 4); length of resting bouts was not related to daily tempera- 60 ture (r2 <0.001, P= 0.86). During midday in

40 June-August moose fed for shorter periods June 16-30 July 1-15 July 16-31 Aug 1-15 Aug 16-31 Sept 1-15 when temperatures were >20o C (F = 28.74, df = 1, 111, P <0.0001). 240

220 B. RESTING BOUTS A

200

DURATIONBOUTSOF (min) AB Foraging Ecology 180 AB AB AB Bite size, bite rate, and search effort ─ 160 B 140 Bite size differed among plant species con- 120 sumed (F = 22.77, df = 13, 461, P <0.0001) 100 and 2-week periods (F = 7.56, df = 5, 461, P 80

60 <0.0001), but not among years (F = 1.27, df

40 June 16-30 July 1-15 July 16-31 Aug 1-15 Aug 16-31 Sept 1-15 = 1, 461, P = 0.26). Moose took the largest PERIOD bites from quaking aspen (2.7 ± 0.1 g/bite) in Fig. 2. Biweekly feeding (A) and resting (B) bouts early July; this was a result of stripping bark of male moose in Rocky Mountain National Park, rather than solely cropping leaves. During Colorado, 16 June-15 September, 2003 and 2004. summer the largest average bite size was taken Vertical bars indicate standard deviation (SD), from western dock (2.5 ± 0.2 g/bite), and the and letters indicate significant (P ≤0.05) differ- largest bite size of a species dominant in the ences in bout duration between periods. diet was drummond willow (1.6 ± 0.4 g/bite) Continuous feeding within bouts also differed (Table 3). The smallest average bite size was (F = 4.62, df = 5, 212, P = 0.0005) among from wolf willow (0.2 ± 0.02 g/bite) in early bi-weekly periods; it was longest in early July July. The mean size of simulated moose bites

(73.7 ± 29.7 min, n = 19) and early September 160 Aa FEEDING BOUTS (86.2 ± 32.4 min, n = 70), and lowest in late 140 Aa aA RESTING BOUTS aA July (67.2 ± 26.3 min, n = 66). Continu- Aa 120 aA ous feeding within bouts averaged 12.2 min 100 B A b bB A (range = 4.0-27.8) less than the mean bout B 80 B B BB BB length, with the greatest difference oc- B B curring in early September (27.8 min). 60 The duration of resting bouts was also 40 different among bi-weekly time periods BOUTDURATION (min) 20

(F = 3.93, df = 5, 275, P = 0.002) (Fig. 2b); 0 the mean was 125.9 ± 58.6 min (n = 282, range = 19.1-410.3). 23:00-01:59 02:00-04:59 05:00-07:59 08:00-10:59 11:00-13:59 14:00-16:59 17:00-19:59 20:00-22:59 TIME OF DAY The duration of feeding (F = 3.61, df = 7, Fig. 3. Mean duration of activity and resting 198, P = 0.0011) and resting (F = 5.49, df = bouts in 3-h periods during diel observations of 7, 197, P <0.0001) bouts differed across the male moose in Rocky Mountain National Park, day. Feeding bout duration was longer around Colorado, 16 June-15 September, 2003 and 2004. dawn and shortly after dusk than at midday Letters indicate significant (P ≤0.05) differences and night; conversely, the resting bout dura- in the duration of bouts between periods.

78 ALCES VOL. 46, 2010 DUNGAN ET AL. - MOOSE SUMMER RANGE CARRYING CAPACITY

160 2 Y = 111 - 2.41X feeding bouts (r = 0.04, P = 0.44). R2 = .207 140 P < .001 Dry matter intake ─ Daily DMI of male 120 moose varied throughout the summer and au-

100 tumn. The estimated mean daily consumption

80 showed a steady increase in early summer 0.75 60 peaking at 10,133 g/d (126.8 g/kg BW ) in early August, followed by a sharp decline to 40 Y = 89.3 - .675X a low of 5,522 g/d in early September (Fig. BOUTDURATION (min) 20 R2 = .032 P = .011 6). The highest estimated intake rate (22.3 g/ 0 -5 0 5 10 15 20 25 30 min) was recorded in early August; the low TEMPERATURE (C) was 10.2 g/min in late June. Intake rates of Fig. 4. The relationship between the duration of willow (all species) increased from a season feeding bouts of male moose and temperature low of 11.2 g/min in late June to a peak of 18.7 in Rocky Mountain National Park, Colorado, 16 g/min in early August, and declined to 11.7 June-31 August, 2003 and 2004. Filled circles represent data collected during the 24-h day, and g/min in early September. Species-specific open circles between dawn and dusk only. intake rates during summer were highest for western dock and drummond willow, and (all species) increased from 1.1 ± 0.6 g/bite in lowest for bog birch (Table 3). Dry matter late June, peaked at 1.4 ± 0.2 g/bite in early consumption was estimated in 2-week inter- August, and declined to 0.9 ± 0.1 g/bite in vals from 1 June-15 September (106 days) early September. The mean size of a moose with total summer consumption by male moose bite during summer was 1.1 ± 0.2 g/bite. estimated at 822.8 kg. As a mechanical consequence of con- Daily DEI increased in early summer suming different bite sizes, bite rates de- peaking at 1555 kJ/kg BW0.75 (124,257 kJ)/d clined curvilinearly with increasing bite in early August, and declined to a season low 2 size (r = 0.94, P <0.0001) (Fig. 5). Bite of 847 kJ/kg BW0.75 (67,714 kJ)/d in early rate differed among plant species (F = 6.85, September. Daily DEI was estimated as 1191 df = 13, 461, P <0.0001) (Table 3) and periods kJ/kg BW0.75 (95,182 kJ)/d, and overall sum- (F = 16.75, df = 5, 461, P <0.0001), but did mer DEI was estimated at 1008.9 x 104 kJ. not differ between years (F = 0.59, df = 1, Daily ME intake was estimated as 1055 kJ/ 461, P = 0.44). We recorded the highest bite 0.75 kg BW /d. rates for wolf willow (21.5 ± 2.9 bites/min) in early August, and the lowest for quaking aspen 30 Y = 23.0 (6.3 ± 3.4 bites/min) in late July. 25 1.02 + X R2 = .94

The rate of movement (search effort) P > .0001 during feeding bouts differed among periods 20 (F = 2.79, df = 5, 145, P = 0.01) and between years (F = 14.21, df = 1, 145, P = 0.0002). 15 We measured the highest average rates of 10 movement in late June (2.5 ± 1.7 steps/min) BITERATE (bites/min) and late August (2.4 ± 1.0 steps/min), and 5 the lowest in late July (1.9 ± 0.9 steps/min). 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 BITE SIZE (g/bite) The average rate of movement during feed- Fig. 5. The relationship between bite size and ing bouts was higher in 2003 (2.3 ± 1.5 steps/ bite rate of moose in Rocky Mountain National min) than 2004 (1.6 ± 0.8 steps/min). Rate of Park, Colorado, 16 June-15 September, 2003 movement was not related to bite rate during and 2004.

79 MOOSE SUMMER RANGE CARRYING CAPACITY- DUNGAN ET AL. ALCES VOL. 46, 2010 a 8.84 11.04 11.22 18.96 23.98 15.43 20.42 19.64 37.02 19.50 22.12 19.26 13.12 23.36 Harvesting rate (g/min) SD 1.86 4.50 4.88 4.36 7.46 4.05 4.81 4.81 7.39 0.66 4.18 6.52 2.35 1.96 B B B B A A B AB AB AB AB AB AB AB x 12.29 18.70 20.51 15.38 14.29 15.70 14.88 16.75 16.39 13.97 10.72 15.24 14.93 12.16 Bite rate (bites/min) n 8 8 7 3 2 3 36 24 15 40 77 10 119 128 SD 0.35 0.12 0.16 0.05 0.02 0.23 0.22 0.04 0.26 0.31 0.17 0.03 0.19 0.03 x 1.56 1.08 0.66 0.60 0.54 1.30 1.38 0.64 1.32 1.80 2.18 1.28 2.48 1.56 Bite size (g/bite) n 6 3 6 3 2 8 3 11 28 52 20 45 23 61 ) ) ) ) ) ) ) ) ) ) Senecio triangularis ) Salix drummondiana ) Salix planifolia Salix monticola Populus tremuloides ( Alnus incana Salix geyeriana Cirsium spp. ) Colorado. Different letters denote significant differences (P ≤0.05) among bite rates. (P letters denote significant differences Colorado. Different Intake rate is the product of mean bite size and rate. bog birch ( Betula glandulosa drummond willow ( geyer willow ( grasses mountain willow ( plane-leaf willow ( thistle ( trembling aspen ( whiplash willow ( Salix lasiandra wolf willow ( Salix wolfi ( Nuphar lutea ssp. Polysepalum yellow water-lily mountain alder western dock ( Rumex aquaticus Table 3. Table Bite size, bite rate, and intake rate (i.e., the product of bite size and bite rate) by moose during summer 2003-2004 in Rocky Mountain Park, Species a arrow-leaved groundsel (

80 ALCES VOL. 46, 2010 DUNGAN ET AL. - MOOSE SUMMER RANGE CARRYING CAPACITY

11000 range in North America.

10000 Activity Budgets

9000 The activity of male moose in RMNP fluctuated during summer through fall, and 8000 these fluctuations are most likely attributable to changes in behavior preceding the breed- 7000 ing season. Moose were most active in early

6000 summer, feeding for 82% of the 10 h/d they DRYMATTER INTAKE (g/day) spent active. Similarly, free-ranging moose 5000 June 16-30 July 1-15 July 16-31 Aug 1-15 Aug 16-31 Sept 1-15 in Denali National Park (DNP), Alaska spent PERIOD 13 h/d active during early summer, feeding Fig. 6. Dry matter intake (g/day) of free-ranging for about 75% of their active time (VanBal- male moose in Rocky Mountain National Park, lenberghe and Miquelle 1990). As breeding Colorado, 16 June-15 September, 2003 and 2004. season approached in early fall, activity level Estimates are based on bite-count techniques. of moose in RMNP declined to <9 h/d. Ac- Nutritional Carrying Capacity tivity levels increased with the onset of the The total RS of willow habitat represented breeding season in fall, but the proportion of 6 6 3310.4 x 10 kJ of DE, 2946.2 x 10 kJ of time spent feeding declined to 52% of active ME, and 3521.9 kg nitrogen during summer time. Moose spent more time moving and in 2004. The estimated available biomass that social activities related to breeding; sparring would provide sustainable moose foraging and thrashing trees began around 1 Septem- (i.e., available dry matter reduced by 75%) ber with smelling and following females was 67,494 kg (dry weight). The estimates towards the end of September. A reduction of K with Model 1 and 2 were similar; Model in feeding and increase in social behavior 2 1 estimated 82 moose or 0.21 moose/km and during breeding season was also observed in Model 2 estimated 83 moose and 0.21 moose/ Alberta, Canada (Best et al. 1978, Renecker 2 km . Model 3 predicted a K of 125 moose and Hudson 1989b). Schwartz et al. (1984) 2 and 0.32 moose/km , estimates 34% higher noted complete fasting by males as long as 18 than those of Models 1 and 2. days during the breeding season. Little is know about the nocturnal activ- DISCUSSION ity of moose (Klassen and Rea 2008), and Our observations of summer activity to the best of our knowledge this is the first patterns and foraging ecology of moose in study to record activity patterns of free- RMNP at the southernmost extent of their ranging moose for extended periods at night range fills a key gap in the understanding of (between dusk and dawn). Although moose environmental interactions between moose were 13% more active during day than night, and willow habitats. It complements and these observations were influenced by early provides a novel ecological comparison with season movements spent searching for suitable the extensive research of activity patterns forage and late season mating behavior, both (Risenhoover 1986, Renecker and Hudson done mostly during the day. During July and 1989b, Bevins et al. 1990, VanBallenberghe August moose were up to 20% more active at and Miquelle 1990) and foraging ecology night than during the day (Table 2). Likewise, (Miquelle and Jordan 1979, Schwartz et al. Renecker and Hudson (1989b) found penned 1984, Renecker and Hudson 1985, Renecker moose in Alberta, Canada were more active at and Hudson 1986b) of moose in their northern night during spring and summer than in fall

81 MOOSE SUMMER RANGE CARRYING CAPACITY- DUNGAN ET AL. ALCES VOL. 46, 2010 and winter. Feeding and resting/ruminating Temperatures regularly rose above 20oC dur- dominated nocturnal activity of our moose ing the afternoon in mid-June to mid-August throughout summer, and extended movement in RMNP. When temperature was >20oC, the (>5 min) was rarely observed at night. Pre- length of foraging bouts declined, and moose dation of moose within RMNP is considered were rarely observed feeding for extended rare, and hunting is not allowed within RMNP, periods during midday, and typically bedded thus predator avoidance at night may be less in shade or water. However, they did not shift influential and contribute to low movement to nocturnal feeding suggesting that tempera- at night. Ericsson and Wallin (1996) found ture in RMNP was not high enough to cause increased diurnal movement during hunting major behavioral change in foraging activity. season in northern Sweden. Kelsall and Telfer (1974) suggested that areas Although moose had fewer feeding bouts/d with prolonged temperature >27° C are unsuit- as fall breeding approached (Table 2), mean able for moose because of thermal stress, and feeding bout duration in early September was temperature rarely reached that level. longer than in summer because moose spent Peak feeding times for moose were around more time socializing during feeding bouts; dawn and dusk (Fig. 3) as indicated in previous continuous feeding within activity bouts was studies (Cederlund et al. 1989, Renecker and also highest at that time resulting in similar Hudson 1989b, MacCracken et al. 1997); the time spent foraging during summer and early timing of feeding shifted with photoperiod. fall. Renecker and Hudson (1989b) found that However, summer studies in Alaska where the longest foraging bouts of captive moose days are longer indicated less synchrony of occurred in spring and fall, and the number feeding at dawn and dusk (Bevins et al. 1990, of activity bouts/d did not vary between sea- VanBallenberghe and Miquelle 1990). Best sons. The number of activity bouts/d varied et al. (1978) concluded that moose activity between and within seasons, as also reported in Alberta, Canada was controlled by light by Bevins et al. (1990), VanBallenberghe and triggered daily by sunrise and sunset; and Miquelle (1990), and MacCracken et al. our data are supportive of this. If activity is (1997). The variation among studies may be triggered by photoperiod, continuous day- attributed to the differences in behavior of light during Alaskan summers would lead to free-ranging and captive moose. Free-ranging less synchrony of moose activity. It follows moose most likely endure more environmental that studies at different latitudes and seasons constraints than captive moose. The aver- should not necessarily yield similar results. age number of feeding bouts/d was similar We estimated the average daily energy ex- to that measured in Michigan (Miquelle and penditure of male moose in summer at 835 Jordan 1979), Alberta, Canada (Bevins et al. kJ/kg BW0.75/d, similar to the mean for 2 1990), and DNP, Alaska (VanBallenberghe free-ranging moose cows in May and July and Miquelle 1990). (890 kJ/kg BW0.75/d; Renecker and Hudson The length of feeding bouts during the 1989a). Energy expenditure during summer is day declined as ambient temperature increased considerably higher than in winter (Schwartz (Fig. 4), similar to that reported for moose in et al. 1987a, Schwartz et al. 1988a, Renecker other studies (Ackerman 1987, Bevins et al. and Hudson 1989a), most likely because of 1990, VanBallenberghe and Miquelle 1990). increased metabolic rate (Renecker and Hud- Thermal stress in moose begins at 14-20oC, and son 1986a), increased activity (Renecker and when temperatures exceed 20oC, metabolic Hudson 1989b, VanBallenberghe and Miquelle rates of moose increase at a rate of 0.7 kJ/kg 1990), and higher ambient temperature (Re- BW 0.75/h/oC (Renecker and Hudson 1986a). necker and Hudson 1989a).

82 ALCES VOL. 46, 2010 DUNGAN ET AL. - MOOSE SUMMER RANGE CARRYING CAPACITY

Foraging Behavior that the available biomass of willow is likely Moose had the highest harvesting rates more important than its protein content. on plants from which they could obtain the largest bites. Bites >2 g were measured for Carrying Capacity Estimates 2 aquatic plants, western dock and Rocky The estimate of nutritional K from the Mountain cow-lily. The harvest rate of wil- nitrogen-based model (3) was about 34% low averaged 17.1 g/min, with the largest bite higher than those from the energy-based size and fastest harvest on drummond willow models (1 and 2). We suspect that Model 3 and the smallest bite and slowest harvest on overestimated K and attribute the larger esti- wolf willow. Because larger bites require mate to our assumptions and use of the 33% more time to chew (Risenhoover 1989), bite multiplier to estimate nitrogen requirements. rates were lowest on the plant species from The same estimates were more similar for elk which moose took the largest bites. Bite size, during winter in RMNP (Hobbs et al. 1982) bite rate, and harvest rate of willow were however, estimates of energy requirements are similar to those reported by Renecker and less complex in winter than summer. Hudson (1986b) for cow moose in summer Because >90% of the summer and fall in Alberta, Canada, and moose in DNP (Van diet of moose in RMNP is willow (Dungan Ballenberghe, unpublished data). The mean and Wright 2005), modeling carrying capacity bite size increased early in summer, peaked was much simpler than in situations where in early August at the height of the growing herbivores consume a wide range of plants season, and declined in early September with that vary substantially in nutritional quality the senescence of available forage. and concentrations of plant secondary me- Daily DMI varied in the study, slowly tabolites (Hobbs and Swift 1985, Beck et al. increasing from May to August, followed by 2006). Although moose consumed 6 species a sharp decline in fall (Fig. 6). Consump- of willow that differ somewhat in nutritional tion likely mirrored the changes in willow quality (0.99-1.43% available N, 51.7-59.9% biomass that peaked in late summer (Stumph available dry matter; Stumph 2005), and a 2005). Decreases in DMI in fall are usually minor proportion of 14 other species or cat- associated with the onset of breeding. Peak egories of plants (Dungan and Wright 2005), DMI in early August and DMI estimates were we believe that basing estimates of K on total similar to those reported for moose in Alaska willow biomass, average willow quality, and and Alberta, Canada (Schwartz et al. 1984, diet selection, activity patterns, and intake rates Renecker and Hudson 1985, Miquelle and Jor- of moose provided a reasonable estimate of dan 1979), and predictions from a simulation the number of moose that can be supported by model (Hubbert 1987). Estimates of daily ME the riparian communities in RMNP in summer. intake across the summer were well above the However, direct inference to other locations estimated ME requirements for free ranging or seasons should be cautious. moose (584.7 kJ/kg BW0.75/d; Renecker and Because 100% use of forage by ungulates Hudson 1985), indicating that moose were able in most ecosystems reduces plant productivity to obtain adequate energy for maintenance and and is not sustainable (McInnes et al. 1992, replenishing fat reserves. Peak DEI did not Pastor et al. 1993), and ungulates are not coincide with peak nutritional content of wil- capable of consuming forage at that level, low. Crude protein of willow peaked around researchers have reduced estimates of for- late June in 2003, and mid-June in 2004 for all age availability based on snow depth (Potvin species (Stumph 2005); however, intake rates and Huot 1983, Stephenson et al. 2006), diet of willow peaked in early August suggesting choice (Wallmo et al. 1977, Hobbs et al. 1982,

83 MOOSE SUMMER RANGE CARRYING CAPACITY- DUNGAN ET AL. ALCES VOL. 46, 2010

Stephenson et al. 2006), and habitat selec- ACKNOWLEDGEMENTS tion (Beck et al. 2006). We conservatively We want to acknowledge field technicians reduced available browse estimates by 75% J. Skinner, B. Iannone, and M. Cluck. We (25% sustainable use), although some studies want to thank B. Stumph and C. Westbrook have considered consumption rates of 42-50% for allowing us to use their data in construct- sustainable for shrub communities (Wolfe et ing this manuscript. Funding for this project al. 1983, Bergstrom and Danell 1987, Singer was provided by Rocky Mountain National et al. 1994). The reduction of available bio- Park through the USGS Idaho Cooperative mass greatly reduced the estimates of K for Fish and Wildlife Research Unit. moose on the western side of RMNP, and higher consumption rates of riparian willow REFERENCES communities may indeed be sustainable. Ac k e r m a n , T. 1987. Moose response to heat on Isle Royale. M.S. Thesis, Michigan MANAGEMENT IMPLICATIONS Technological University, Houghton, Willow communities in the portion of Michigan, USA. RMNP west of the Continental Divide have Au g u s t i n e , D. J., and S. J. McNa u g h t o n . experienced degradation over the last 100 1998. Ungulate effects on the functional years. Two historic influences were 1) water species composition of plant communities: diversions from the upper Colorado River to herbivore selectivity and plant tolerance. eastern Colorado put in place prior to estab- Journal of Wildlife Management 62: lishment of RMNP, and 2) intensive trapping 1165-1183. of a still-depleted beaver population in the Ba k e r , W. L., J. A. Mu n r o e , and A. E. He s s l . 19th and early 20th centuries (Hess1993). 1997. The effects of elk on aspen in the More recently drought, climate change, and winter range in Rocky Mountain National an expanding elk population have likely Park. Ecography 20: 155-165. added stress to native plant communities, Be c k , J. L., and J. M. Pe e k . 2001. Jarbidge particularly riparian willow habitat. This cooperative elk herd carrying capacity study has not only provided valuable informa- study. Idaho Bureau of Land Manage- tion to evaluate the moose population and its ment. Technical Bulletin No. 01-3. sustainability in the western part of RMNP, _____, _____, and E. K. St r a n d . 2006. Esti- but it also aids in addressing the larger issue mates of elk summer nutritional carrying of willow communities. Our study estimated capacity constrained by probabilities that 82-125 moose could be sustained in this of habitat selection. Journal of Wildlife 391 km2 area west of the Continental Divide, Management 70: 283-294. which is 30-72% more than the 37-59 moose Be i d l e m a n , L. H., R. G. Be i d l e m a n , and B. estimated as resident in summer, 2003-2004 E. Wi l l a r d . 2000. Plants of Rocky (Dungan 2007); thus, moose at present pose Mountain National Park. Falcon Press, no measurable ecological threat in the study Helena, Montana, USA. area. Importantly, our data should be useful Be r g s t r o m , R. and K. Da n e l l . 1987. Effects to estimate K for moose in willow communi- of simulated winter browsing by moose on ties on the east of the Continental Divide that morphology and biomass of two birch spe- are more degraded and heavily used by elk cies. Journal of Ecology 75: 533-544. (Singer and Zeigenfuss 2002), should moose Be s t , D. A., G. M. Ly n c h , and O. J. Ro n g - ever disperse to this area of RMNP. s t a d . 1978. Seasonal activity patterns of moose in the Swan Hills, Alberta. Alces 14: 109-125.

84 ALCES VOL. 46, 2010 DUNGAN ET AL. - MOOSE SUMMER RANGE CARRYING CAPACITY

Be v i n s , J. S., C. C. Sc h w a r t z , and A. W. tain National Park. University Press of Fr a n z m a n n . 1990. Seasonal activity Colorado, Boulder, Colorado, USA. patterns of moose on the Kenai Peninsula, He t t , J. M., R. Ta b e r , J. Lo n g , and J. Sc h o e n . Alaska. Alces 26: 14-23. 1978. Forest management policies and Bo y c e , M. S. 1989. The Jackson Elk Herd: elk summer carrying capacity in the Abies Intensive Wildlife Management in North Amabilis Forest, western Washington En- America. Cambridge University Press, vironmental Management 2: 561-566. Cambridge, Massachusetts, USA. Hj e l j o r d , O., F. Su n d s t o l , and H. H a a g e n r u d . Ca u g h l e y , G. 1979. What is this thing called 1982. The nutritional value of browse to carrying capacity? Pages 2-8 in M. S. moose. Journal of Wildlife Management Boyce and L. D. Hayden Wing, editors. 46: 333-343. North American Elk: Ecology, Behavior, Ho bb s , N. T., D. L. Ba k e r , J. E. El l i s , D. M. and Management. University of Wyoming Sw i f t , and R. A. Gr e e n . 1982. Energy Press, Laramie, Wyoming, USA. and nitrogen based estimates of elk win- Cr e t e , M. 1989. Approximation of K car- ter range carrying capacity. Journal of rying capacity for moose in eastern Wildlife Management 46: 12-21. Quebec. Canadian Journal of Zoology _____, and D. M. Sw i f t . 1985. Estimates of 67: 373-380. habitat carrying capacity incorporating Ce d e r l u n d , G., Be r g s t r o m , R., and F. Sa n - nutritional constraints. Journal of Wildlife d e g r e n . 1989. Winter activity patterns Management 49: 814-822. of females in two moose populations. Hu bb e r t , M. E. 1987. The effects of diet Canadian Journal of Zoology 67: 1516- on energy partitioning in moose. Ph.D. 1522. Thesis, University of Alaska, Fairbanks, Du n g a n , J. D. 2007. Activity patterns, forag- Alaska, USA. ing ecology, and summer range carrying Hu d s o n , R. J., and R. G. Wh i t e . 1985. Bioen- capacity of moose in Rocky Mountain ergetics of Wild Herbivores. CRC Press, National Park, Colorado. M.S. Thesis. Boca Raton, Florida, USA. University of Idaho, Moscow, Idaho, Hu ff , D. E., and S. D. Va r l e y . 1999. Natural USA. Regulation in Yellowstone National Park’s _____, and R. G. Wr i g h t . 2005. Summer Northern Range. Ecological Society of diet composition of moose in Rocky America, Washington D.C., USA. Mountain National Park, Colorado. Alces Ke l s a l l , J. P., and E. S. Te l f e r . 1974. 41: 139-146. Biogeography of moose with particular Er i c s s o n , G. and K. Wa l l i n . 1996. The reference to western North America. impact of hunting on moose movements. Naturaliste Canadian 101: 117-130. Alces 32: 31-40. Kl a s s e n , N. A., and R. V. Re a . 2008. What Fr a n k , D. A., and S. J. McNa u g h t o n . 1992. do we know about nocturnal activity of The ecology of plants, large mammalian moose? Alces 44: 101-109. herbivores, and drought in Yellowstone Ku f e l d , R. C., and S. F. St e i n e r t . 1990. An National Park. Ecology 73: 2034-2058. estimate of moose carrying capacity in Ha n n a , J. S., S. Be n s o n , J. Em m e r i c h , and R. willow habitat in North Park, Colorado. O. Ol s o n . 1989. Lander moose herd unit Colorado Division of Wildlife. Unpub- winter range study. Wyoming Game and lished Report. Fish Department. Enhancement Project ____, and D. C. Bo w d e n . 1996. Movement BFGFSN3511, Final Report. and habitat selection of Shiras moose He s s , K. 1993. Rocky Times in Rocky Moun- (Alces alces shirasi) in Colorado. Alces

85 MOOSE SUMMER RANGE CARRYING CAPACITY- DUNGAN ET AL. ALCES VOL. 46, 2010

32: 85-99. Estimation of dry matter intake of free- Le r e s c h e , R. E., and J. L. Da v i s . 1973. Im- ranging moose. Journal of Wildlife portance of nonbrowse foods to moose on Management 49: 785-792. the Kenai Peninsula, Alaska. Journal of _____, and _____. 1986a. Seasonal energy Wildlife Management 37: 279-287. expenditures and thermoregulatory re- Ma c Cr a c k e n , J. G., V. Va n Ba l l e n b e r g h e , sponses of moose. Canadian Journal of and J. M. Pe e k . 1997. Habitat relation- Zoology 64: 322-327. ships of moose on the Copper River delta _____, and _____. 1986b. Seasonal foraging in coastal south-central Alaska. Wildlife rates of free-ranging moose. Journal of Monographs 136: 1-52. Wildlife Management 50: 143-147. McIn n e s , P. F., R. J. Na i m a n , J. Pa s t o r , and Y. _____, and _____. 1989a. Ecological me- Co h e n . 1992. Effects of moose browsing tabolism of moose in aspen-dominated on vegetation and litter of the boreal for- boreal forest, central Alberta. Canadian est, Isle Royale, Michigan, USA. Ecology Journal of Zoology 67: 1923-1928. 73: 2059-2075. _____, and _____. 1989b. Seasonal activity Mi q u e l l e , D. G., and P. A. Jo r d a n . 1979. budgets of moose in aspen-dominated The importance of diversity in the diet boreal forest. Journal of Wildlife Man- of moose. Proceedings North American agement 53: 296-302. Moose Conference Workshop 15: 54- Ri s e n h o o v e r , K. L. 1986. Winter activity pat- 79. terns of moose in interior Alaska. Journal Mo n e l l o , R. and T. Jo h n s o n . 2003. The elk of Wildlife Management 50: 727-734. herd in Rocky Mountain National Park. _____. 1989. Composition and quality of Rocky Mountain National Park. Unpub- moose winter diets in interior Alaska. lished Report. Journal of Wildlife Management 53: Na t i o n a l Pa r k Se r v i c e . 2001. Management 568-577. Policies. U.S. Government Printing Of- Ro bb i n s , C. T. 2001. Wildlife Feeding and fice, Washington, D.C., USA. Nutrition. 2nd edition. Academic Press, Ow e n -Sm i t h , N. 1998. How high ambient San Diego, California, USA. temperature affects the daily activity and Sc h r e i n e r , E. G., K. A. Kr u e g e r , P. J. Ha pp e , foraging time of a sub-tropical ungulate, and D. B. Ho u s t o n . 1996. Understory the greater Kudu (Tragelaphus strepsice- patch dynamics and ungulate herbivory ros). Journal of Zoology 246: 183-192. in old-growth forest of Olympic National Pa s t o r , J., B. De w e y , R. J. Na i m a n , P. F. Park, Washington. Canadian Journal of McIn n e s , and Y. Co h e n . 1993. Moose Forest Research 26: 255-265. browsing and soil fertility in the boreal Sc h w a r t z , C. C. 1992. Physiological and forest of Isle Royale National Park. Ecol- nutritional adaptations of moose to north- ogy 74: 467-480. ern environments. Alces Supplement 1: Po t v i n , F. and J. Hu o t . 1983. Estimating 139-155. carrying capacity of a white-tailed deer _____, A. W. Fr a n z m a n N, and D. C. Jo h n s o n . wintering area in Quebec. Journal of 1988b. Wildlife Research and Manage- Wildlife Management 47: 463-475. ment Moose Nutrition and Physiology Re g El i n , W. L., M. E. Hu bb e r t , C. C. Studies. Alaskan Department of Fish Sc h w a r t z , and D. J. Re e d . 1987. Field and Game. test of a moose carrying capacity model. _____, M. E. Hu bb e r t , and A. W. Fr a n z m a n N. Alces 23: 243-284. 1988a. Energy requirements of adult Re n e c k e r , L. A., and R. J. Hu d s o n . 1985. moose for winter maintenance. Journal

86 ALCES VOL. 46, 2010 DUNGAN ET AL. - MOOSE SUMMER RANGE CARRYING CAPACITY

of Wildlife Management 52: 26-33. Ta y l o r , C. R., and G. M. Ma l o i y . 1970. The _____, W. L. Re g e l i n , and A. W. Fr a n z m a n n . effects of dehydration and heat stress on 1984. Seasonal dynamics of food intake intake and digestion of food in some east in moose. Alces 20: 223-241. African bovids. International Congress _____, _____, and _____. 1985. Suitability of Game Biology 8: 324-325. of a formulated ration for moose. Journal Te s t a , J. W., and G. P. Ad a m s . 1998. Body of Wildlife Management 49: 137-141. condition and adjustment to reproductive _____, _____, and _____. 1987a. Nutritional effort in female moose (Alces alces). Jour- energetics of moose. Swedish Wildlife nal of Mammalogy 79: 1345-1354. Research Supplement 1: 265-280. Va n Ba l l e n b e r g h e , V. D., and D. G. Mi q u e l l e . _____, _____, and _____. 1987b. Protein 1990. Activity of moose during spring digestion in moose. Journal of Wildlife and summer in interior Alaska. Journal of Management 51: 352-357. Wildlife Management 54: 391-396. Si n g e r , F. J., L. Ma c k , and R. G. Ca t e s . 1994. Va n So e s t , P. J. 1982. Nutritional Ecology Ungulate herbivory of willow on Yellow- of the Ruminant. O & B Books, Inc., stone’s northern winter range. Journal of Corvallis, Oregon, USA. Range Management 47: 435-443. Wa l l m o , O. C., L. H. Ca r p e n t e r , W. L. Re g E- _____, and L. C. Ze i g e n f u s s , compilers. 2002. l i n , R. B. Gi l l , and D. L. Ba k e r . 1977. Ecological evaluation of the abundance Evaluation of deer habitat on a nutritional and effects of elk herbivory in Rocky basis. Journal of Range Management Mountain Nation Park, Colorado, 1994- 30: 122-127. 1999. U.S. Geological Survey. Open File Wo l f e , M. L., W. H. Ba b c o c k , and R. M. Report 02-208. Fort Collins, Colorado, We l c h . 1983. Effects of simulated USA. browsing on Drummond’s willow. Alces So k a l , R. R., and F. J. Ro h l f . 1987. Intro- 19: 14-35. duction to Biostatistics. Freeman and Wr i g h t , R. G. 1992. Wildlife Research Company, New York, New York, USA. and Management in the National Parks. St e p h e n s o n , T. R., V. Va n Ba l l e n b e r g h e , J. University of Illinois Press. Urbana, Il- M. Pe e k , and J. G. Ma c Cr a c k e n . 2006. linois, USA. Spatio-temporal constraints on moose Zh e l e z n o v -Ch u k o t s k y , N. K., and E. S. Vo- habitat and carrying capacity in coastal t i a s h o v a . 1998. Comparative analysis Alaska: vegetation succession and cli- of moose nutrition of the Anadyrsky and mate. Rangeland Ecology and Manage- Omolonsky populations (far north east) in ment 59: 359-372. different seasons. Alces 34: 445-451. St u m p h , B. P. 2005. The summer forage quality of willow communities and its influence on moose forage ecology in Rocky Mountain National Park, Colorado. M. S. Thesis. University of Idaho, Moscow, Idaho, USA.