933

Habitat use patterns in relation to escape terrain: are alpine ungulate females trading off better foraging sites for safety? S. Hamel and S.D. Coˆte´

Abstract: Predation risk often forces prey to trade off good foraging sites for safety, and compromises are expected to be greater in females with vulnerable offspring than in barren females. To determine whether adult females of large herbivores traded off forage for safety, we assessed habitat use and estimated vegetation abundance and quality in relation to distance to escape terrain in marked mountain goats (Oreamnos americanus de Blainville, 1816). We found that all females spent more time foraging near escape terrain than away from them. Females with young foraged on average 20 m closer to escape terrain than barren females in June, a time when offspring were particularly vulnerable to predation. Plant biomass did not vary with distance to escape terrain in June, but was lower closer than away from escape terrain during all other months. The abundance of forbs and shrubs increased with distance to escape terrain, but their quality did not vary. For grasses and sedges, plant digestible content decreased closer to escape terrain, but interestingly proteins increased. Our results suggest that females traded off forage abundance, and to a lesser extent forage quality, for safety. Compared with barren females, females with offspring may face a trade-off in plant digestible content by foraging in safer areas than barren females. Re´sume´ : Le risque de pre´dation oblige souvent les proies a` faire des compromis entre de bons sites alimentaires et leur se´curite´ et on s’attend a` ce que ces compromis soient plus importants chez les femelles accompagne´es d’un jeune vulne´rables que chez les femelles sans jeune. Afin de de´terminer si les femelles adultes chez les grands herbivores font des compromis entre la nourriture et la se´curite´, nous avons e´value´ l’utilisation de l’habitat et estime´ l’abondance et la qualite´ de la ve´ge´tation en relation avec la distance au terrain de fuite chez la che`vre de montagne (Oreamnos americanus de Blainville, 1816). Toutes les femelles passent plus de temps a` s’alimenter a` proximite´ que loin des terrains de fuite. En juin, un moment ou` les jeunes sont particulie`rement vulne´rables a` la pre´dation, les femelles avec un jeune s’alimentent en moyenne 20 m plus pre`s des terrains de fuite que les femelles sans jeune. En juin, la biomasse des plantes ne varie pas en fonction de la distance au terrain de fuite, mais durant tous les autres mois, elle est plus faible pre`s des terrains de fuite qu’a` une certaine distance. L’abondance des plantes arbustives et des plantes herbace´es augmente en fonction de la distance au terrain de fuite, mais la qualite´ de ces plantes ne varie pas. Dans le cas des gramine´es et des cype´race´es, le contenu digestible des plantes diminue a` proximite´ des terrains de fuite, mais curieusement leur contenu en prote´ines augmente. Nos re´sultats indiquent que les femelles font un compromis entre l’abondance, et dans une moindre mesure la qualite´, de leur nourriture et leur se´curite´. En s’alimentant dans des sites plus se´curitaires que les femelles sans jeune, les femelles accompagne´es d’un jeune semblent faire un compromis sur le contenu digestible des plantes. [Traduit par la Re´daction]

Introduction Predation risk has shaped the foraging behaviour of prey (Brown et al. 1999), leading to evolutionary trade-offs between using safe versus productive habitats (Sih 1980; Brown 1999). In temperate regions, mammalian herbivores must replenish their body reserves during the snow-free period, when forage is abundant and of high quality. The best foraging sites, however, are often located in risky areas (Lima and Dill 1990; Berger 1991; Rachlow and Bowyer 1998), generating a fundamental trade-off between food and Received 25 January 2007. Accepted 12 July 2007. Published on the NRC Research Press Web site at cjz.nrc.ca on 2 October 2007. S. Hamel1 and S.D. Coˆte´. De´partement de Biologie and Centre d’E´tudes Nordiques, Universite´ Laval, Que´bec, QC G1K 7P4, Canada. 1Corresponding

author (e-mail: [email protected]).

Can. J. Zool. 85: 933–943 (2007)

protection. Balancing habitat selection between safe and productive sites, however, is likely more crucial for lactating than nonlactating females, since they have to protect their offspring from predators and to compensate for the high energetic costs of lactation (Hanwell and Peaker 1977; Loudon and Kay 1984). Lactating females must also maintain a good body condition to produce abundant and high-quality milk (Landete-Castillejos et al. 2003). Additionally, we can expect this trade-off to be more important in early lactation when offspring are most vulnerable to predation and depend almost entirely on their mothers’ milk (Barten et al. 2001). Indeed, when female caribou (Rangifer tarandus L., 1758) were accompanied by a calf, Barten et al. (2001) showed that they used higher elevation sites than when alone, trading off abundant vegetation for a lower risk of predation. When calves were older, females moved back to lower elevations alongside nonlactating females. In many species, lactating females increase time spent in vigilance to reduce predation risk (Toı¨go 1999; Laundre´ et

doi:10.1139/Z07-080

#

2007 NRC Canada

934

al. 2001; Childress and Lung 2003). In other species, however, females with young do not spend more time vigilant than nonlactating females (Ruckstuhl and Festa-Bianchet 1998; Neuhaus and Ruckstuhl 2002). As scanning is incompatible with feeding (Illius and FitzGibbon 1994; Cowlishaw et al. 2004), lactating females in many species may be unable to augment time spent vigilant without concomitantly reducing time spent foraging. An alternative for these females is to forage in safer areas, so that they could maintain time spent foraging while simultaneously reducing predation risk. Evidence for this strategy has been described in moose (Alces alces L., 1758), where females spent more time foraging in close proximity to safe areas when they had a young at heel than when alone (White and Berger 2001). Foraging in a safer area, however, often means foraging in a different habitat, which may be of lower quality than the riskier area. For alpine ungulates, safe habitats are either high-elevation sites or areas located close to cliffs (FestaBianchet 1988; Gross et al. 1996). High-altitude sites are less commonly used by predators (Barten et al. 2001), and alpine ungulates can escape predators in cliffs. Because of these advantages, areas close to escape terrain are often heavily used for foraging compared with areas located further away from them (Pfitsch and Bliss 1985). Since intensive grazing or browsing can greatly reduce plant abundance and quality (Coˆte´ et al. 2004; Schoenecker et al. 2004), areas near escape terrain may be of lower quality than areas far from them. Interestingly, however, light to moderate grazing can also enhance plant abundance or quality because some species can show overcompensatory growth after defoliation (Archer and Tieszen 1980; Chapin 1980; McNaughton 1983). Regrown plants often increase in quality because they are at a younger stage than older, ungrazed plants (Bryant et al. 1983; Moser and Schu¨tz 2006). To our knowledge, the relationships among plant quality, plant abundance, and distance to escape terrain have never been evaluated and compared with habitat use in a wild herbivore. Prior studies have described the importance of escape terrain in habitat use of mountain goats (Oreamnos americanus de Blainville, 1816; Haynes 1992; Gross et al. 2002), and some have proposed that goats traded off forage and safety (McFetridge 1977; von Elsner-Schack 1986); however, none have demonstrated this trade-off and quantified its importance. Furthermore, we have previously shown that lactating mountain goats spent 3% more time foraging but had similar vigilance time than nonlactating females during summer (Hamel and Coˆte´, In press). We therefore hypothesized that lactating females would increase offspring protection by foraging in safer areas than nonlactating females as opposed to increasing time spent vigilant. Our objectives were (i) to determine if lactating mountain goats forage closer to escape terrain than nonlactating females and (ii) to assess whether plant quality and abundance vary according to distance from escape terrain to determine if females are facing a trade-off between foraging in safe versus productive habitats. We predicted that all females would spend a greater proportion of time foraging closer to escape terrain than away from them, but that lactating females would be found closer to escape terrain than nonlactating females. We expected that this would result in a fundamental

Can. J. Zool. Vol. 85, 2007

trade-off in habitat use, because vegetation abundance and quality should be lower near escape terrain.

Materials and methods Study area We studied mountain goats at Caw Ridge (548N, 1198W), west central Alberta, Canada, in the front range of the Rocky Mountains. The climate is subarctic–arctic, characterized by long winters and short, cool summers, where snowfall can occur during any month of the year. Goats use 28 km2 of alpine tundra and subalpine open forest of Engelmann spruce (Picea engelmannii Parry ex Engelm.) between 1750 and 2170 m of elevation. Landscape includes gently rolling hills and steep grassy slopes, as well as rockslides and a few abrupt cliff faces that are crucial escape terrain. Predation is the principal cause of mortality (Coˆte´ and Festa-Bianchet 2003). The main predators are gray wolves (Canis lupus L., 1758), grizzly bears (Ursus arctos L., 1758), and cougars (Felis concolor L., 1771), but also potentially include black bears (Ursus americanus Pallas, 1780), coyotes (Canis latrans Say, 1823), wolverines (Gulo gulo L., 1758), and golden eagles (Aquila chrysaetos L., 1758) (Coˆte´ and Festa-Bianchet 2003). During summer, females and juveniles seldom travel far from escape terrain or below the treeline. Mountain goats are generalist herbivores and eat mainly grass (>50%), forb (30%), and browse (15%) species growing in alpine meadows at close proximity to cliffs or rocky ledges (Coˆte´ and Festa-Bianchet 2003). At Caw Ridge, the main grass and sedge species include Agropyron sp. Gaertn., Carex spp. L., Festuca spp. L., Kobresia spp. Willd., Poa spp. L., and Phleum sp. L. The most important consumed forbs are Aconitum delphiniifolium DC., Anemone drummondii S. Wats., Astragalus spp. L., Campanula lasiocarpa Cham., Castilleja spp. Mutis ex L. f., Epilobium spp. L., Gentiana spp. L., Myosotis alpestris alpestris F.W. Schmidt, Oxyria spp. Hill, Oxytropis spp. DC., Pedicularis spp. L., Polemonium sp. L., Polygonum viviparum L., and Potentilla spp. L. Palatable shrubs consist mainly of Betula glandulosa Michx., Salix spp. L., and Vaccinium spp. L. Behavioural observations We captured goats from 1988 to 2005 in remotely controlled box traps and self-tripping Clover traps baited with salt. We immobilized goats with xylazine hydrochloride, and we reversed the effects of xylazine with injection of idazoxan (Haviernick et al. 1998). We marked adult females with plastic ear tags and canvas collars. Since 1993, 98% of goats 1 year and older were marked. We aged adult goats not marked as juveniles (12% of females used in this study) by counting their horn annuli, a technique reliable up to 7 years of age (Stevens and Houston 1989). From mid-May to late September 2002–2005, we recorded daily behavioural observations on adult females (i.e., 3 years of age and older) using spotting scopes (15
–45
). The number of adult females in each year ranged between 50 and 58, for a total of 75 different females. Total population size ranged between 147 and 159 individuals. In all groups observed, we noted the identity of each individual, and we determined the reproductive status of females from #

2007 NRC Canada

Hamel and Coˆte´

observations of nursing behaviour. Since milk production decreases rapidly after the death of offspring (Hanwell and Peaker 1977), we considered a female as nonlactating 10 days after she had lost her kid (n = 16 cases). We used scan sampling (Altmann 1974) to compare time spent near escape terrain while foraging between lactating and nonlactating females. For each group observed (n = 62), we recorded the distance to the closest escape terrain for each female that was foraging every 30 min (n = 3250 observations). We used a large scan interval to ensure independence of consecutive data. The mean (±SE) number of adult females per group was 16 ± 2, ranging from 1 to 55. All distances were estimated by a single observer (S.H.), using known distances between objects in the field and ‘‘number of goat lengths’’ between the female and the escape terrain. We recorded data between 0700 and 2130. Because social rank may influence how females use space around escape terrain, we recorded agonistic encounters among females using ad-libitum sampling (n = 5803 interactions from 2002 to 2005; Altmann 1974; Coˆte´ 2000). Because dominance matrices were significantly linear for all years (all h’ ‡ 0.2, all P < 0.001), we ordered adult females in annual hierarchies according to the methodology of de Vries (1998) using Matman 1.0 for Windows (Noldus Information Technology 1998). The number of adult females varied annually, so we transformed social ranks according to the formula (1 – rank)/Ni, where Ni was the number of adult females during year i (Coˆte´ 2000). Hence, social rank varied from 0 to 1 (i.e., from subordinate to dominant). Experimental design We sampled vegetation to estimate plant abundance and quality in relation to distance to escape terrain. Our experimental setup consisted of four transects placed perpendicular to a cliff. Transects were 20 m apart. Along each transect, we placed a quadrat (400 cm2) at 20, 60, 100, and 140 m from the cliff. We replicated this setup at three sites and sampled vegetation in the second week of each month between June and September. We sampled vegetation during 3 years for plant biomass (2002–2004) and during 2 years for plant quality (2002–2003). For each quadrat (n = 576 in 3 years), we estimated the percent cover of plants in 5% classes as well as plant height (cm). We averaged height among plant clusters when it was heterogeneous. This was done for each plant category: live grasses and sedges, live forbs, live shrubs, and dead plants. In June, shrubs consisted of spring buds. At each site, we randomly selected a different transect each month to clip all plant categories separately at 1 cm above ground. Vegetation was first air-dried in the field and then oven-dried at 45 8C for 48 h to determine aboveground biomass. For each plant category, we used a regression analysis to calculate the relationship between percent cover and height, and aboveground biomass (Table A1; Bonham 1989). We then used these equations to estimate plant biomass for quadrats that we did not clip. Lastly, we calculated the mean biomass for each distance to escape terrain and plant category at each site (n = 144 biomass values for each plant category over 3 years). To assess vegetation quality, we estimated digestible plant content (percent NDS (the fraction of the plant cell that dis-

935

solves in neutral detergent) and percent ADS (the fraction that dissolves in acidic detergent); Van Soest 1994) and protein content (macro-Kjeldhal acid digestion technique; Association of Official Analytical Chemists 1984) of all plant categories found in each clipped quadrat in 2002 and 2003 (n = 96). Because these analyses required about 8 g of dried vegetation per sample and quadrats never contained that much, we collected additional plants within 1 m of the quadrat. To determine the proportion of faecal pellets deposited at 20, 60, 100, and 140 m from escape terrain, we randomly sampled 10 sites and counted the number of goat faecal pellets in circular quadrats of 4 m2 positioned at each of the four distances to escape terrain (n = 40 quadrats). Statistical analyses Distance to escape terrain We used two different analyses to quantify the distance to escape terrain of lactating and nonlactating females while foraging using the 30 min scans. First, we compared the mean distance to escape terrain between both types of females. We used a linear mixed model (LMM) to assess the influence of reproductive status, month, age, social rank, and their interactions on distances to escape terrain (n = 3250 observations on 75 different females). We normalized distances using square-root transformations. We included both year and female identity as random effects to control for stochastic between-year variation and pseudoreplication (Machlis et al. 1985). Because age is highly correlated with social rank in female mountain goats (r > 0.9; Coˆte´ 2000), we used the residuals of the regression of social rank on age (hereafter age-specific social rank) in all analyses. Second, we standardized values to obtain the relative distance of females to escape terrain compared with the rest of the group. For each scan, we computed the mean distance to escape terrain for all females and then standardized distances around this mean. This procedure allowed us to determine if lactating females foraged on average closer to escape terrain than nonlactating females, but also if they foraged relatively closer than nonlactating females within the same group. Standardized distances (hereafter relative distances) were normally distributed, and we used a similar LMM as above. To compare the proportional amount of time that lactating and nonlactating females spent at different distances from escape terrain, we separated data by female identity, month, and year to compute the proportion of time each female spent foraging at 0–40, 41–80, 81–120, 121–160, and ‡161 m from escape terrain. We used these categories because the mean distance of the first four categories were sampled for plant abundance and quality. This procedure led to 358 vectors, each containing the proportion of time spent in each of the five distance categories for a single female in a specific month and year. To control for the nonindependence of proportions, we used a MANOVA (Littell et al. 2002) to assess the influence of reproductive status, month, age, social rank, and their interactions on these vectors. Because it is not possible to include random factors in a MANOVA, we forced the variable ‘‘year’’ in the model to account for between-year variation prior to assessing the influence of the other variables. Pseudoreplication was con#

2007 NRC Canada

936

Can. J. Zool. Vol. 85, 2007 Table 1. Linear mixed models on the mean distance adult females foraged from escape terrain (ET) and the relative distance to escape terrain that a female foraged compared with other females in a given group (i.e., standardized distance to escape terrain) in mountain goats at Caw Ridge, Alberta (2002–2005). Dependent variable Distance to ET

Distance to ET within a group (standardized distance)

Independent variable Reproductive status Month Reproductive status
month Reproductive status Month Reproductive status
month

df 1,67 3,67 3,67 1,67 3,67 3,67

F 3.4 10.7 3.8 4.2 0.3 5.2

P 0.070 <0.001 0.015 0.045 0.851 0.003

V (%)a 20

18

Note: n = 3250 observations on 75 different females. Percent variance explained by the model.

a

trolled by using a single vector per female, month, and year for every female in the population (see Leger and Didrichsons 1994). Vegetation abundance and quality We first performed simple ANOVAs to assess the overall difference in biomass, percent NDS, percent ADS, and percent protein among plant categories. When significant (P £ 0.05), we compared differences among groups using LSD tests (Sokal and Rohlf 1981). Then, for each plant category, we assessed the influence of distance to escape terrain, month, and their interaction on plant biomass by performing an ANOVA according to a split-split-plot model (Littell et al. 2002), which allowed us to control for variability among years and sites. We used year as repeated blocks (n = 3), site as the first-level plot (n = 3), distance to escape terrain as the second-level plot (n = 4), and month as the third-level plot (n = 4). Sample sizes for plant quality were lower than for plant biomass (grasses and sedges, n = 58; forbs, n = 29; shrubs, n = 14) because we could not collect vegetation when biomass was null. Thus, we could not use a split-split-plot design, and we performed simpler linear models using year and site as repeated measures. We forced year followed by site in the models to account for between-year and betweensite variations prior to assessing the influence of month, distance to escape terrain, and their interaction. Finally, we performed an LMM to assess the relationship between the number of faecal pellets and distance to escape terrain, including site as a random effect. We performed all analyses with SAS software (Littell et al. 2002, 2006). We used a significance level of 0.05. We computed the observed variance of LMM models using the following (Xu 2003): R2 ¼ 1  ðSSR=SSTOÞ where SSR ¼

X ½ðYobserved  Ypredicted Þ2 

and SSTO ¼

X ½ðYobserved  Y
observed Þ2 

All results are presented as mean ± SE.

Results Distance to escape terrain Age-specific social rank, age, and their interaction did not influence distance to escape terrain for foraging females in all analyses (all P > 0.2). Therefore, we removed these variables from the final models. In June, lactating females foraged about 20 m closer to escape terrain than nonlactating females, whereas there was no difference in the other months (Table 1; Fig. 1a). Using relative distances, we also found that lactating females foraged about 10 m closer to escape terrain than nonlactating females of the same group, but again only in June (Table 1; Fig. 1b). Interestingly, kid mortality during summer occurred mainly during the first month of life; 54% of mortalities occurred from birth in late May to the end of June (n = 23 mortalities in 43 days), while 46% occurred from July to mid-September (n = 18 mortalities in 78 days; w2 = 15.9, P < 0.001, df = 1, n = 166 kids born from 2002 to 2006). In June, lactating females spent about 15% more time foraging within 40 m of escape terrain than nonlactating females, thereby reducing time spent foraging at distances of >120 m (month: F[4,348] = 27.8, P < 0.001; reproductive status: F[4,346] = 0.5, P = 0.8; month
reproductive status: F[4,348] = 3.8, P = 0.004; Fig. 2). During the rest of the summer, time spent foraging at various distances from escape terrain was not influenced by females’ reproductive status (Fig. 2). Overall, all females spent more time within 40 m of escape terrain than further away (Fig. 2). Similarly, faecal pellets of goats were more abundant closer than further away from escape terrain (F[3,27] = 4.4, P = 0.01, n = 40, R2 = 46%; 20 m = 94 ± 4 pellets; 60 m = 23 ± 7 pellets; 100 m = 14 ± 6 pellets; 140 m = 9 ± 4 pellets). Vegetation abundance The average biomass of plants for the whole summer was greater for grasses and sedges than for forbs and shrubs, which had a similar biomass (F[2,429] = 76.2, P <0.001, n = 432, R2 = 26%; grasses and sedges: 0.40 ± 0.04 g/400 cm2; forbs: 0.23 ± 0.03 g/400 cm2; shrubs: 0.19 ± 0.03 g/ 400 cm2). The same pattern was observed each month and at each distance from escape terrain (Fig. 3). Plant biomass was affected by distance to escape terrain, month, and their interaction for all plant categories, except for the biomass of shrubs, where the interaction was not sig#

2007 NRC Canada

Hamel and Coˆte´

937

60

both were greater than in grasses and sedges (NDS: F[2,99] = 512.2, P <0.001, n = 102, R2 = 91%; grasses and sedges = 49.7% ± 0.6%; forbs = 78.2% ± 0.7%; and shrubs = 70.6% ± 1.1%; ADS: F[2,99] = 83.3, P < 0.001, n = 102, R2 = 63%; grasses and sedges = 77.2% ± 0.4%; forbs = 85.4% ± 0.4%; and shrubs = 79.6% ± 0.9%; Fig. 4). Plant proteins in forbs and shrubs did not differ, but both were slightly higher than in grasses and sedges (F[2,99] = 10.6, P <0.001, n = 102, R2 = 18%; grasses and sedges = 12.1% ± 0.4%; forbs = 15.5% ± 0.7%; and shrubs = 14.3% ± 0.9%; Fig. 4). Plant digestible content only varied in grasses and sedges, as percent NDS increased with distance to escape terrain and was higher in September than in August (Table 3; Fig. 4). Percent ADS was similar across months and distances to escape terrain for all species (Table 3; Fig. 4). In grasses and sedges, proteins were higher closer (20 and 60 m) than away (100 and 140 m) from escape terrain, while it did not vary with distance to escape terrain for forbs and shrubs (Table 3; Fig. 4). Percent protein content decreased considerably over the summer in all species (Table 3; Fig. 4).

50

Discussion

40

During summer, females spent more than 60% of their time foraging within 80 m of escape terrain. Although plant biomass did not vary with distance to escape terrain in June, it was much lower within 80 m from escape terrain during all other months, suggesting that all females were facing a trade-off between forage abundance and safety. Regarding plant quality, digestible content of the most abundant forage, grasses and sedges, increased with increasing distances to escape terrain, whereas protein content decreased. This finding was surprising, since a positive correlation between plant digestible content and proteins usually occurs in plants (Minson 1990; Robbins 1993). Interestingly, some studies have shown that faecal and urinary depositions were positively influencing protein concentration in plants (McKendrick et al. 1980; Bryant et al. 1983; Hik et al. 1991; Hobbs 2006). Since we found that goats intensively used areas close to escape terrain and that faecal pellets were more abundant in these areas than elsewhere, the high protein content of plants close to escape terrain may be the result of more frequent faecal and urinary depositions in these areas. Plant digestible content (Berteaux et al. 1998; Bowyer et al. 1998) and proteins (Pellew 1984) have been shown to influence ungulates’ selectivity and appear to be the most limiting currencies in large foraging herbivores (Owen-Smith and Novellie 1982; Robbins 1993). Consequently, it is difficult to disentangle the effects of these two plant components on females’ diet and habitat choice (Wilmshurst and Fryxell 1995). Our results suggest that by foraging in safer areas, females experienced a cost in terms of the reduced digestible content of grasses and sedges, which may have been counterbalanced, however, by higher protein conten. Importantly, however, even if forbs and shrubs were less abundant than grasses and sedges, they were of higher quality and found in much greater abundance further than closer to escape terrain, thereby increasing the magnitude of the trade-off. Thus, predation threat appears to force females to remain close to escape terrain, as it can enhance their chances of escaping predators (Berger 1991; Coˆte´ and Beaudoin 1997;

Fig. 1. Monthly comparisons of (a) the mean distance (+SE) adult females foraged from escape terrain and (b) the relative distance (+SE) to escape terrain that a female foraged in comparison with other females of a given group (i.e., standardized distance to escape terrain) in mountain goats at Caw Ridge, Alberta (2002–2005). For standardized distances to escape terrain, positive values represent individuals that were located further away from escape terrain than the average, and negative values represent individuals that were located closer to escape terrain than the average. n = 3250; significant comparisons are indicated by an asterisk (*). Distance to escape terrain (m)

110

(a)

100

*

90 80 70

June 10 8 Standardized distance to escape terrain (m)

Lactating Nonlactating

(b)

July

August

*

September

Lactating Nonlactating

6 4 2 0 -2 -4 -6 -8 June

July

August

September

nificant (Table 2). For all plant categories, biomass was fairly low in June, increased in July and August, and decreased in September (Fig. 3). Plant biomass in June was similar at all distances to escape terrain, but the biomass of grasses and sedges was about five times greater than for other species (Fig. 3). In July and August, plant abundance was generally higher at distances >100 m from escape terrain than near escape terrain (Fig. 3). Plant biomass, however, varied less importantly among distances in July compared with August, where fewer plants were found closer (20 and 60 m) than away (100 and 140 m) from escape terrain (Fig. 3). Overall, grasses and sedges were relatively abundant at all distances, whereas forbs and shrubs were, respectively, most abundant at 100 and 140 m from escape terrain (Fig. 3). Vegetation quality Overall, plant digestible contents (percent NDS and percent ADS) were significantly greater in forbs than in shrubs, and

#

2007 NRC Canada

938

Can. J. Zool. Vol. 85, 2007

Fig. 2. Monthly comparisons of the percent time (+SE) adult females spent foraging at various distances to escape terrain (A, 0–40 m; B, 41–80 m; C, 81–120 m; D, 121–160 m; E, ‡161 m) in relation to female reproductive status (lactating, solid bars; nonlactating, shaded bars) in mountain goats at Caw Ridge, Alberta (2002–2005). n = 358; significant comparisons are indicated by an asterisk (*).

60

July

August

A B C D E

A B C D E

June

50

September

*

Percent time

40 30

*

20

*

10 0 A B C D E

A B C D E

Distance to escape terrain Fig. 3. Monthly least-squares means comparisons of plant biomass (+SE) in relation to distance to escape terrain (ET). Vegetation was sampled (n = 576 quadrats in three different sites) at Caw Ridge, Alberta (2002–2004). Letters contrast significant differences (P £ 0.05) within month.

Dry biomass (g/400 cm2)

1.0

c

(a) grasses and sedges

Distance ET

20m 60m 100m 140m

c

0.8 0.6 0.4

ab

a

ab

a

b

0.8 b

0.6

b a

a

0.2

a

a

a

0.2

0.0

b

a

a

a

a

a

ab a

0.0 June

Dry biomass (g/400 cm2)

a

a

0.4

a a

1.0

20m 60m 100m 140m

ab a

a

Distance ET

(c) shrubs

c

b

bc

1.0

July

August

Distance ET

(b) forbs

June

September

20m 60m 100m 140m

0.8

2.5

July

2.0

a

a a a a

a

a a a

0.0

a

a

b

a

a

a

a

1.0

b

0.2

a

b

c

a

20m 60m 100m 140m

b

ac

1.5 a

a

Distance ET

c

b

c

0.4

September

b

(d) total plant biomass

b

0.6

August

a

a

0.5

0.0 June

July

August

September

June

July

August

September

#

2007 NRC Canada

Hamel and Coˆte´

939 Table 2. Split-split-plot models on biomass of grasses and sedges, forbs, and shrubs in relation to distance to escape terrain (ET) and month, controlling for year and site (see text). Dependent variable Biomass of grasses and sedges

Biomass of forbs

Biomass of shrubs

Biomass of all species

Independent variable ET Month ET
month ET Month ET
month ET Month ET
month ET Month ET
month

df 3,18 3,18 9,18 3,18 3,18 9,18 3,18 3,18 9,18 3,18 3,18 9,18

F 13.8 19.3 2.6 26.1 33.3 3.4 14.1 9.9 1.3 28.8 15.0 3.1

P <0.001 <0.001 0.011 <0.001 <0.001 0.001 <0.001 <0.001 0.264 <0.001 <0.001 0.004

V (%)a 87

89

77

89

Note: n = 144. Data were collected at Caw Ridge, Alberta (2002–2004). Percent variance explained by the model.

a

Fig. 4. Least-squares means comparisons of (a) percent NDS (the fraction of the plant cell that dissolves in neutral detergent) (+SE), (b) percent ADS (the fraction that dissolves in acidic detergent) (+SE), and (c) percent protein (+SE) in grasses and sedges, forbs, and shrubs, in relation to month (left to right: J, June; J, July; A, August; and S, September; solid bars) and distance to escape terrain (in metres; open bars). Samples were collected at Caw Ridge, Alberta (2002–2003). Grasses and sedges: n = 58; forbs: n = 29; shrubs: n = 14. Letters contrast significant differences (P £ 0.05) among months or distances to escape terrain (ET) (Table 3). SHRUBS GRASSES AND SEDGES FORBS Percent NDS

80

a

a

a

a

a

a

a a

70

a

a

a

a

a

a

a

60

ab ab

50

40 95 90

Percent ADS

a

J

J

a

A

b a

S

20

ab

b

b

60 100 140

J

J

A

b a

85

a

80

a

a

a

S

20

a

a

60 100 140

a

a

J

J

a

a

a

A

S

20

60 100 140

S

20

a a

a

A

a

a

a a

a

a

60 100 140

a

a

75 70 65

Percent protein

25

20

J

J

J

J

A

S

20

60 100 140

J

A

a

ab

S

20

60 100 140

c a

a

a a

a

b

15

c 10

5

J

J

J

A

Month

a

b

a

b

b

a

a

a

a

a

a b

c

S

a

20

60 100 140

Distance ET

J

J

A

Month

Gross et al. 2002), but at the expenses of reduced plant abundance and, to a lesser extent, vegetation quality. This suggests that predation risk is the primary factor that has shaped habitat use in female mountain goats, similarly to

S

20

60 100 140

Distance ET

J

J

A

Month

S

20

60 100 140

Distance ET

what has been suggested in other ungulates (Bergerud et al. 1984; Pe´rez-Barberı´a and Nores 1994; Kohlmann et al. 1996; Bowyer et al. 1999). By spending much time foraging close to escape terrain, #

2007 NRC Canada

940

Can. J. Zool. Vol. 85, 2007

Table 3. Linear models on plant quality (percent NDS, percent ADS, and percent protein) in relation to month and distance to escape terrain (ET), controlling for year and site (see text). % NDSa Plant category Grasses and sedges (n = 58) Forbsc (n = 29)

Shrubsc (n = 14)

Variable Month ET ET
month Month ET ET
month Month ET

df 3,39 3,39 9,39 2,14 3,14 6,14 3,5 3,5

F 3.5 3.4 0.5 0.9 0.9 0.2 4.0 0.4

P 0.024 0.027 0.884 0.438 0.475 0.958 0.084 0.790

% ADSa V 55

47

89

(%)b

F 1.7 0.5 1.7 2.9 0.8 0.7 1.5 1.4

P 0.180 0.668 0.130 0.092 0.516 0.668 0.330 0.337

% Protein V (%)b 71

53

75

F 33.0 3.8 1.1 15.5 0.1 0.5 3.4 1.7

P <0.001 0.018 0.403 <0.001 0.981 0.828 0.111 0.289

V (%)b 78

72

82 .

Note: Data were collected at Caw Ridge, Alberta (2002–2003). NDS is the fraction of the plant cell that dissolves in neutral detergent; ADS is the fraction that dissolves in acidic detergent (Van Soest 1994). Percent variance explained by the model. c Only 3 months were used in the analysis of forb quality, as there were no samples in June. Similarly, sample size for shrubs was too small to test the interaction ET  month. a b

females created a high grazing pressure on plants in these habitats. Light to moderate grazing has been shown to enhance vegetation production because of plant overcompensatory growth after defoliation (Archer and Tieszen 1980; Bryant et al. 1983; McNaughton 1983), whereas intense grazing pressure can reduce vegetation abundance (Coˆte´ et al. 2004; Schoenecker et al. 2004). A compensatory growth response has been described mainly in grass and sedge species that are morphologically adapted to rapidly grow new leaves to replace those lost to herbivores (Archer and Tieszen 1980; Bryant et al. 1983; Jefferies et al. 1994). For forbs and shrubs, however, regrowth within the same season is limited and seems to occur only in some species (Archer and Tieszen 1980; Bryant et al. 1983; Jefferies et al. 1994). In our study, plant abundance was lower for all plant categories near escape terrain during all months, except in June when forage abundance was low everywhere. Plants, therefore, did not appear to fully compensate for increased defoliation in safer areas. Furthermore, light to moderate grazing can also increase forage quality, since compensatory growth maintains plants at a younger stage (Chapin 1980; Bryant et al. 1983). We only found variations in plant quality for grasses and sedges, where plant digestible content was lower at closer distances to escape terrain than further away, but protein content was higher. This indicates that grasses and sedges increased fibre content following herbivory while concomitantly increasing protein investment in leaves. The latter, however, may also be the result of high nitrogen levels available in the ground because of important goat urine and faecal depositions (Hobbs 2006). Therefore, we cannot strongly conclude that grazing pressure in safe areas influenced vegetation quality positively or negatively, but it was definitely too high to enhance forage production. Females accompanied by a young goat are expected to behave more safely than females without young to ensure the survival of their offspring that are more vulnerable to predation than adults. Because lactating female mountain goats increased time spent foraging but not time spent vigilant (Hamel and Coˆte´, in press), they were constrained to forage in safer areas in June, when their offspring were younger and more vulnerable to predation. At that time of year, kids rarely venture more than a few metres away from their mothers (S. Hamel and S.D. Coˆte´, unpublished data). On average,

individuals in nursery groups were observed foraging at 66.9 ± 3.1 m (range 1 to 250 m) from escape terrain. Females with kids, therefore, increased their chance of escaping predators by foraging 20 m closer to escape terrain than nonlactating females. Similarly, female Cantabrian chamois (Rupicapra pyrenaica parva Cabrera, 1910) were found to remain in close proximity to escape terrain until their offspring were able to escape predators (Pe´rez-Barberı´a and Nores 1994). In female Stone’s sheep (Ovis dalli stonei Allen, 1897), nursery groups were found closer to escape terrain and in habitats in proximity to larger escape terrain than nonmaternal groups (Walker et al. 2006). In June, vegetation was just starting to grow and it was rare everywhere. By foraging in safer areas then, females with young offspring did not have to face a trade-off between safety and forage abundance. Lactating females foraging near escape terrain in June had access to plants with lower digestible content but greater protein content. Both plant proteins and digestible content, which are related to energy availability (Van Soest 1994), are crucial for lactating females, as they can greatly affect milk quality (Loudon and Kay 1984; Minson 1990). For mountain goats, however, if safe areas were of better quality than areas away from escape terrain, there would be no reason for nonlactating females not to use the same areas as lactating females. This suggests that plant digestible content was more important than protein content in determining foraging patch choice of nonlactating females in early summer. At this time of year, protein content of the growing vegetation was at its peak, around 18%, and likely higher than the minimum protein requirements of mountain goats anywhere on the ridge (Owens and Zinn 1993). These results suggest that lactating females most likely trade off access to plants of higher digestible content by foraging in safer areas. In other species, females with young offspring have often been found to trade off vegetation quality and abundance for safety (bighorn sheep (Ovis canadensis Shaw, 1804; Festa-Bianchet 1988; Berger 1991), moose (Bowyer et al. 1999), caribou (Bergerud et al. 1984; Barten et al. 2001), and Cantabrian chamois (Pe´rez-Barberı´a and Nores 1994)). Most of these species increase offspring safety in spring by moving to higher altitudes, where forage abundance is much reduced compared with lower elevations. In comparison with these ungulates, our results suggest that the trade-off be#

2007 NRC Canada

Hamel and Coˆte´

tween forage and safety is probably lower in mountain goats, since lactating females had access to similar plant abundance as nonlactating females in early summer. We demonstrated the influence of escape terrain in the evolution of foraging strategies and summer habitat use of female mountain goats. As predation risk seems to push females to spend most of their time foraging close to escape terrain, this grazing pressure led to an important reduction in plant abundance and a slight diminution in digestible content of grasses and sedges near escape terrain. Because vegetation abundance was lower closer than away from escape terrain, females experienced an important trade-off between forage abundance and safety. These patterns of habitat use by foraging female mountain goats demonstrate the evolutionary importance of escape terrain and suggest that predation pressure is an important factor determining habitat use in this alpine ungulate. Future studies should investigate the impacts of the foraging constraints associated with remaining close to escape terrain on the body condition of individuals, such as variations in summer mass gain between lactating and nonlactating females.

Acknowledgements Our research was financed by the Alberta Conservation Association (ACA); the Natural Sciences and Engineering Research Council of Canada (NSERC); the Challenge Grant in Biodiversity of ACA; and the Alberta Sport, Recreation, Parks and Wildlife Foundation. S.H. received scholarships from NSERC and the Fonds Que´be´cois de la Recherche sur la Nature et les Technologies. The Alberta Fish and Wildlife Division provided logistical support. We are grateful to C. Cameron, E. Cardinal, G. Coˆte´, E. Drouin, C.A. Gagnon, M. Houle, S. Popp, S. Rioux, and V. Viera for help with field work and to A. Brousseau, M. Gravel, L. L’Italien, and G. Picard for help with laboratory work. We especially thank M. Festa-Bianchet and K.G. Smith for fruitful discussions. Valuable comments from J.-P. Tremblay, R.B. Weladji, and one anonymous reviewer improved earlier versions of this manuscript. This research project was approved by the Canadian Council for Animal Care committee of Universite´ Laval.

References Altmann, J. 1974. Observational study of behavior: sampling methods. Behaviour, 49: 227–267. PMID:4597405. Archer, S., and Tieszen, L.L. 1980. Growth and physiological responses of tundra plants to defoliation. Arct. Alp. Res. 12: 531– 552. doi:10.2307/1550499. Association of Official Analytical Chemists. 1984. Official methods of analysis. Association of Official Analytical Chemists, Washington, D.C. Barten, N.L., Bowyer, R.T., and Jenkins, K.J. 2001. Habitat use by female caribou: tradeoffs associated with parturition. J. Wildl. Manage. 65: 77–92. doi:10.2307/3803279. Berger, J. 1991. Pregnancy incentives, predation constraints and habitat shifts: experimental and field evidence for wild bighorn sheep. Anim. Behav. 41: 61–77. doi:10.1016/S0003-3472(05)80503-2. Bergerud, A.T., Butler, H.E., and Miller, D.R. 1984. Antipredator tactics of calving caribou: dispersion in mountains. Can. J. Zool. 62: 1566–1575. Berteaux, D., Creˆte, M., Huot, J., Maltais, J., and Ouellet, J.-P. 1998. Food choice by white-tailed deer in relation to protein

941 and energy content of the diet: a field experiment. Oecologia (Berl.), 115: 84–92. doi:10.1007/s004420050494. Bonham, C.D. 1989. Measurements for terrestrial vegetation. John Wiley and Sons, New York. Bowyer, R.T., Kie, J.G., and Ballenberghe, V.V. 1998. Habitat selection by neonatal black-tailed deer: climate, forage, or risk of predation? J. Mammal. 79: 415–425. doi:10.2307/1382972. Bowyer, R.T., Ballenberghe, V.V., Kie, J.G., and Maier, J.A.K. 1999. Birth-site selection by Alaskan moose: maternal strategies for coping with a risky environment. J. Mammal. 80: 1070– 1083. doi:10.2307/1383161. Brown, J.S. 1999. Vigilance, patch use and habitat selection: foraging under predation risk. Evol. Ecol. Res. 1: 49–71. Brown, J.S., Laundre´, J.W., and Gurung, M. 1999. The ecology of fear: optimal foraging, game theory, and trophic interactions. J. Mammal. 80: 385–399. doi:10.2307/1383287. Bryant, J.P., Chapin, F.S., III, and Klein, D.R. 1983. Carbon/ nutrient balance of boreal plants in relation to vertebrate herbivory. Oikos, 40: 357–368. doi:10.2307/3544308. Chapin, F.S., III. 1980. Nutrient allocation and responses to defoliation in tundra plants. Arct. Alp. Res. 12: 553–563. doi:10. 2307/1550500. Childress, M.J., and Lung, M.A. 2003. Predation risk, gender and the group size effect: does elk vigilance depend upon the behaviour of conspecifics? Anim. Behav. 66: 389–398. doi:10.1006/ anbe.2003.2217. Coˆte´, S.D. 2000. Dominance hierarchies in female mountain goats: stability, aggressiveness and determinants of rank. Behaviour, 137: 1541–1566. doi:10.1163/156853900502718. Coˆte´, S.D., and Beaudoin, C. 1997. Grizzly bear (Ursus arctos) attacks and nanny–kid separation on mountain goats (Oreamnos americanus). Mammalia, 61: 614–617. Coˆte´, S.D., and Festa-Bianchet, M. 2003. Mountain goat. In Wild mammals of North America: biology, management and conservation. Edited by G.A. Feldhamer, B. Thompson, and J. Chapman. The John Hopkins University Press, Baltimore, Md. pp. 1061–1075. Coˆte´, S.D., Rooney, T.P., Tremblay, J.-P., Dussault, C., and Waller, D.M. 2004. Ecological impacts of deer overabundance. Annu. Rev. Ecol. Evol. Syst. 35: 113–147. doi:10.1146/annurev. ecolsys.35.021103.105725. Cowlishaw, G., Lawes, M.J., Lightbody, M., Martin, A., Pettifor, R., and Rowcliffe, J.M. 2004. A simple rule for the costs of vigilance: empirical evidence from a social forager. Proc. R. Soc. Lond. B Biol. Sci. 271: 27–33. doi:10.1098/rspb.2003.2522. de Vries, H. 1998. Finding a dominance order most consistent with a linear hierarchy: a new procedure and review. Anim. Behav. 55: 827–843. doi:10.1006/anbe.1997.0708. PMID:9632471. Festa-Bianchet, M. 1988. Seasonal range selection in bighorn sheep: conflicts between forage quality, forage quantity, and predator avoidance. Oecologia (Berl.), 75: 580–586. doi:10. 1007/BF00776423. Gross, J.E., Alkon, P.U., and Demment, M.W. 1996. Nutritional ecology of dimorphic herbivores: digestion and passage rates in Nubian ibex. Oecologia (Berl.), 107: 170–178. doi:10.1007/ BF00327900. Gross, J.E., Kneeland, M.C., Reed, D.F., and Reich, R.M. 2002. GISbased habitat models for mountain goats. J. Mammal. 83: 218–228. doi:10.1644/1545-1542(2002)083<0218:GBHMFM>2.0.CO;2. Hamel, S., and Coˆte´, S.D. In press. Trade-offs in activity budget in an alpine ungulate: contrasting lactating and nonlactating females. Anim. Behav. doi:10.1016/j.anbehav.2007.04.028. Hanwell, A., and Peaker, M. 1977. Physiological effects of lactation on the mother. Symp. Zool. Soc. Lond. 41: 297–312. Haviernick, M., Coˆte´, S.D., and Festa-Bianchet, M. 1998. Immobi#

2007 NRC Canada

942 lization of mountain goats with xylazine and reversal with idazoxan. J. Wildl. Dis. 34: 342–347. PMID:9577782. Haynes, L.A. 1992. Mountain goat habitat of Wyoming’s Beartooth Plateau: implications for management. Biennial Symposium of the Northern Wild Sheep and Goat Council, 8: 325–339. Hik, D.S., Sadul, H.A., and Jefferies, R.L. 1991. Effects of the timing of multiple grazings by geese on net above-ground primary production of swards of Puccinellia phryganodes. J. Ecol. 79: 715–730. doi:10.2307/2260663. Hobbs, N.T. 2006. Large herbivores as sources of disturbance in ecosystems. In Large herbivore ecology, ecosystem dynamics and conservation. Edited by K. Danell, P. Duncan, R. Bergstro¨m, and J. Pastor. Cambridge University Press, Cambridge. pp. 261– 288. Illius, A.W., and FitzGibbon, C. 1994. Costs of vigilance in foraging ungulates. Anim. Behav. 47: 481–484. doi:10.1006/anbe.1994.1067. Jefferies, R.L., Klein, D.R., and Shaver, G.R. 1994. Vertebrate herbivores and northern plant communities: reciprocal influences and responses. Oikos, 71: 193–206. doi:10.2307/3546267. Kohlmann, S.G., Mu¨ller, D.M., and Alkon, P.U. 1996. Antipredator constraints on lactating Nubian ibexes. J. Mammal. 77: 1122– 1131. doi:10.2307/1382794. Landete-Castillejos, T., Garcı´a, A., Gomez, J.A., and Gallego, L. 2003. Lactation under food constraints in Iberian red deer Cervus elaphus hispanicus. Wildl. Biol. 9: 99–107. Laundre´, J.W., Herna`ndez, L., and Altendorf, K.B. 2001. Wolves, elk, and bison: reestablishing the ‘‘landscape of fear’’ in Yellowstone National Park, U.S.A. Can. J. Zool. 79: 1401–1409. doi:10.1139/cjz-79-8-1401. Leger, D.W., and Didrichsons, I.A. 1994. An assessment of data pooling and some alternatives. Anim. Behav. 48: 823–832. doi:10.1006/anbe.1994.1306. Lima, S.L., and Dill, L.M. 1990. Behavioral decisions made under the risk of predation: a review and prospectus. Can. J. Zool. 68: 619–640. Littell, R.C., Stroup, W.W., and Freund, R.J. 2002. SAS for linear models. SAS Institute Inc., Cary, N.C. Littell, R.C., Milliken, G.A., Stroup, W.W., Wolfinger, R.D., and Schabenberger, O. 2006. SAS for mixed models. SAS Institute Inc., Cary, N.C. Loudon, A.S.I., and Kay, R.N.B. 1984. Lactational constraints on a seasonally breeding mammal: the red deer. Symp. Zool. Soc. Lond. 51: 233–252. Machlis, L., Dodd, P.W., and Fentress, J.C. 1985. The pooling fallacy: problems arising when individuals contribute more than one observation to the data set. Z. Tierpsychol. 68: 201–214. McFetridge, R.J. 1977. Strategy of resource use by mountain goat nursery groups. In Proceedings of the 1st International Mountain Goat Symposium, 19 February 1977, Kalispell, Montana. Edited by W. Samuel and W.G. Macgregor. Province of British Columbia Ministry of Recreation and Conservation, Fish and Wildlife Branch, Victoria, B.C. pp. 169–173. McKendrick, J.D., Batzli, G.O., Everett, K.R., and Swanson, J.C. 1980. Some effects of mammalian herbivores and fertilisation on tundra soils and vegetation. Arct. Alp. Res. 12: 565–578. doi:10.2307/1550501. McNaughton, S.J. 1983. Compensatory plant growth as a response to herbivory. Oikos, 40: 329–336. doi:10.2307/3544305. Minson, D.J. 1990. Forage in ruminant nutrition. Academic Press, San Diego, Calif. Moser, B., and Schu¨tz, M. 2006. Tolerance of understory plants subject to herbivory by roe deer. Oikos, 114: 311–321. doi:10. 1111/j.2006.0030-1299.14386.x. Neuhaus, P., and Ruckstuhl, K.E. 2002. The link between sexual

Can. J. Zool. Vol. 85, 2007 dimorphism, activity budgets, and group cohesion: the case of the plains zebra (Equus burchelli). Can. J. Zool. 80: 1437–1441. Noldus Information Technology. 1998. Matman, reference manual. Version 1.0 for Windows. Noldus Information Technology, Wageningen, The Netherlands. Owens, F.N., and Zinn, R. 1993. Protein metabolism of ruminant animals. In The ruminant animal: digestive physiology and nutrition. Edited by D.C. Church. Waveland Press, Inc., Long Grove, Ill. pp. 227–249. Owen-Smith, N., and Novellie, P. 1982. What should a clever ungulate eat? Am. Nat. 119: 151–178. doi:10.1086/283902. Pellew, R.A. 1984. The feeding ecology of a selective browser, the giraffe (Giraffa camelopardalis tippelskirchi). J. Zool. 202: 57–81. Pe´rez-Barberı´a, F.J., and Nores, C. 1994. Seasonal variation in group size of Cantabrian chamois in relation to escape terrain and food. Acta Theriol. 39: 295–305. Pfitsch, W.A., and Bliss, L.C. 1985. Seasonal forage availability and potential vegetation limitations to a mountain goat population, Olympic National Park. Am. Midl. Nat. 113: 109–121. doi:10.2307/2425353. Rachlow, J.L., and Bowyer, R.T. 1998. Habitat selection by Dall’s sheep (Ovis dalli): maternal trade-offs. J. Zool. 245: 457–465. doi:10.1111/j.1469-7998.1998.tb00120.x. Robbins, C.T. 1993. Wildlife feeding and nutrition. Academic Press, New York. Ruckstuhl, K.E., and Festa-Bianchet, M. 1998. Do reproductive status and lamb gender affect the foraging behavior of bighorn ewes? Ethology, 104: 941–954. Schoenecker, K.A., Singer, F.J., Zeigenfuss, L.C., Binkley, D., and Menezes, R.S.C. 2004. Effects of elk herbivory on vegetation and nitrogen processes. J. Wildl. Manag. 68: 837–849. doi:10. 2193/0022-541X(2004)068[0837:EOEHOV]2.0.CO;2. Sih, A. 1980. Optimal behavior: can foragers balance two conflicting demands? Science (Washington, D.C.), 210: 1041–1043. doi:10.1126/science.210.4473.1041. Sokal, R.R., and Rohlf, F.J. 1981. Biometry: the principles and practice of statistics in biological research. W.H. Freeman and Co., San Francisco, Calif. Stevens, V., and Houston, D.B. 1989. Reliability of age determination of mountain goats. Wildl. Soc. Bull. 17: 72–74. Toı¨go, C. 1999. Vigilance behavior in lactating female Alpine ibex. Can. J. Zool. 77: 1060–1063. doi:10.1139/cjz-77-7-1060. Van Soest, P.J. 1994. Nutritional ecology of the ruminant. Cornell University Press, New York. von Elsner-Schack, I. 1986. Habitat use by mountain goats, Oreamnos americanus, on the eastern slopes region of the Rocky Mountains at Mount Hamell, Alberta. Can. Field Nat. 100: 319–324. Walker, A.B.D., Parker, K.L., and Gillingham, M.P. 2006. Behaviour, habitat associations, and intrasexual differences of female Stone’s sheep. Can. J. Zool. 84: 1187–1201. doi:10.1139/Z06-109. White, K.S., and Berger, J. 2001. Antipredator strategies of Alaskan moose: are maternal trade-offs influenced by offspring activity? Can. J. Zool. 79: 2055–2062. doi:10.1139/cjz-79-11-2055. Wilmshurst, J.F., and Fryxell, J.M. 1995. Patch selection by red deer in relation to energy and protein intake: a re-evaluation of Langvatn and Hanley’s (1993) results. Oecologia (Berl.), 104: 297–300. doi:10.1007/BF00328364. Xu, R. 2003. Measuring explained variation in linear mixed effects models. Stat. Med. 22: 3527–3541. doi:10.1002/sim.1572. PMID:14601017.

Appendix A Appendix appears on the following page. #

2007 NRC Canada

Hamel and Coˆte´

943

Table A1. Regression equations from linear models used to predict dry plant biomass according to visual estimations of plant height and percent cover based on vegetation collected at Caw Ridge, Alberta (2002–2004). Plant category

Regression equation: ln(biomass + 0.1) =

Grasses and sedges

–2.4680 + [ln(cover + 0.1)
0.3659] + [ln(height + {[ln(cover + 0.1)
ln(height + 0.1)]
0.0889} –2.3379 + [ln(cover + 0.1)
0.5379] + [ln(height + –2.0780 + [ln(cover + 0.1)
0.5810] + [ln(height + –2.5329 + [ln(cover + 0.1)
0.2127] + [ln(height + {[ln(cover + 0.1)
ln(height + 0.1)]
0.1360}

Forbs Shrubs Dead plants

n

df

F

P

Adjusted r2

0.1)
0.5243] +

271

3,267

310.4

<0.0001

0.77

0.1)
0.5646] 0.1)
0.6458] 0.1)
0.5639] +

157 86 188

2,154 2,83 3,184

367.0 161.2 164.3

<0.0001 <0.0001 <0.0001

0.82 0.79 0.72

Note: Cover: percent plant cover in 5% classes; height: plant height (cm), which was averaged among plant clusters when it was heterogeneous. We determined final models by removing variables and interactions that did not affect (P > 0.1) plant biomass. To reach normality, we transformed all variables to the logarithmic scale after adding a value of 0.1 (Sokal and Rohlf 1981. Biometry: the principles and practice of statistics in biological research. W.H. Freeman and Co., San Francisco, Calif.).

#

2007 NRC Canada