NATIONAL PARK SERVICE Mountain Goats in Olympic National Park: Biology and Management of an Introduced Species

Concern has been expressed about interspecific competition between goats and the native fauna of the Olympic Peninsula (National Park Service 1987) because the distribution of mountain goats and other herbivores overlap seasonally. Data bearing directly on the issue of competitive interactions, however, are scant.

Herbivores

Ungulates

Roosevelt elk (Cervus elaphus roosevelti) and black-tailed deer show two life-history patterns in the park: year-round residence at low elevations and seasonal migrations to high elevation summer ranges (Houston et al. 1990). Migratory elk, deer, and goats were broadly associated on subalpine summer ranges; for example, 244 elk, 109 deer, and 396 goats were tallied in counting units during the July 1983 goat census (see Chapter 4). Similar results occurred during the 1990 count; 273 elk and 99 deer occurred in the units with 172 goats. The extensive survey of herbivory on subalpine and alpine summer ranges also showed broad overlap in use of plant communities among goats, elk, deer, and marmots (see Chapter 9). That survey and a radiotelemetry study (Schroer 1987) suggested that elk tended to select highly productive meadows in cirques rather than the sparsely vegetated ridges occupied by goats.

Microhistological examination of fecal pellets showed that elk diets from subalpine summer ranges contained 30 plant taxa; 32 taxa occurred in deer diets. Diets of both cervids were composed mainly of forbs and grasses, but forbs were relatively more important in deer diets (Table 21). Thus, summer diets of elk and deer in subalpine areas were dominated by high quality herbaceous forage. Several plant taxa were common to diets of the three ungulates in these limited samples (Carex spp., Luzula spp., and Vaccinium spp.), and it seems likely that, parkwide, many more taxa would be consumed in common. However, limitations on either the quantity or the quality of food resources for ungulates are least likely during summer in the Olympic Mountains (see Chapter 6). Mere overlap in distribution or diets among species is insufficient to demonstrate interspecific competition.

Table 21. Summary of Roosevelt elk (Cervus elaphus roosevelti) and black-tailed deer (Odocoileus hemionus) diets from subalpine ranges in Olympic National Park, 1989.

Observations of elk, deer, and goat distributions during winter, when resources are most likely limited, suggested little overlap between goats and the other ungulates. Elk and deer occupied valley bottoms and slopes below about 500 m (Leslie and Starkey 1982; Jenkins and Starkey 1984; Leslie et al. 1984; Houston et al. 1987). Although wintering goats were found both above and below 500 m (see Chapter 6), their affinity for cliffs and rock outcrops suggested a limited potential for interspecific competition, even during winter.

The occurrence and strength of interspecific relations are best demonstrated by experiments where population densities change or are manipulated (Houston 1982; Sinclair and Norton-Griffiths 1982). An average of 15.9 ± 10.50 SD (range = 3-32) deer were tallied during 14 aerial censuses of Klahhane Ridge from 1981 to 1990 as goat density was reduced by more than 98% (see Chapter 4). A regression of deer numbers observed on the estimated goat population showed no significant association (P > 0.10, b = 6.03, r² = 0.67), and interspecific competition for resources could not be inferred.

Resource competition between goats and other ungulates has not been demonstrated elsewhere in Washington. However, there is concern that clearcut logging of low-elevation forests has increased elk populations locally and that elk have potential for competing with goats (Johnson 1983).

Small Mammals

The relations between introduced mountain goats and endemic mammals were of interest. The Olympic marmot (Marmota olympus) occurs in colonies throughout the subalpine areas of the Olympic Mountains. Marmots hibernate for 7-8 months annually, occupy restricted home ranges while active during summer, and may be food-limited (Barash 1970, 1974; Wood 1973). They forage heavily on forbs and grasses, preferring high-quality new growth of leaves and inflorescences. Plants favored by goats in the Klahhane Ridge area also were utilized by marmots, including Festuca idahoensis and Carex spp. (Wood 1973). The plot data showed widespread effects on soils from marmot herbivory in subalpine areas and suggested several factors important in den selection (see Chapter 9). Marmot foraging and digging were powerful disturbance forces shaping the vegetation of subalpine meadows (del Moral 1984). Heavy marmot grazing around colonies reduced graminoids and palatable forbs and increased ruderal and unpalatable species (e.g., Achillea millefolium), not unlike effects documented for mountain goats (see Chapter 10).

Marmot occupancy of selected colonies has been sampled occasionally over the past 32 years as part of other field studies. Occupancy of 30 colonies was examined in 1989 in relation to mountain goat occurrence by initially considering the Hurricane Ridge and Royal Basin colonies to be in high-density goat range and the Obstruction Point and the Deer Park colonies to be in low-density range (Table 22). Eleven of 15 colonies occupied in either 1957 or 1967-69 on Hurricane Ridge and vicinity were still occupied in 1989 (73%). Nine of these colonies were also known to be occupied in the 1970's. Two colonies in Royal Basin were occupied in 1972 and 1989. For the low-density goat areas, 8 of 12 colonies occupied in either 1957 or 1967-69 were occupied in 1989 (67%), 3 were unoccupied, and occupancy could not be determined for 1. If these colonies were assumed to be representative samples, then the proportion occupied over time did not differ between high and low goat density areas (X² test, P > 0.10).

Table 22. Occupancy of 30 marmot (Marmota olympus) colonies, Olympic National Park, 1957-1989.^a

At this broad level of resolution, high densities of goats by the 1970's did not result in competitive exclusion of marmots from the sampled areas. Unfortunately, however, marmot dens surveyed at Hurricane Ridge were virtually all on the periphery of the most intensively used goat areas, so they did not provide a good measure of potential interspecific effects at high goat densities. Although marmots occurred throughout the Hurricane Ridge area and seem to have persisted during the period of high goat density, we cannot discount possible effects on colonies that may have involved reduced densities, reduced recruitment, or elimination.

Distribution of the endemic Olympic pocket gopher (Thomomys mazama melanops) seems to have declined from that reported historically (Johnson 1977, and personal communication, 1992). Surveys in 1949, 1951, and 1976 showed that animals were absent from a site at the headwaters of the Soleduck River and at Canyon Creek divide near Bogachiel Peak, areas where they occurred in 1897 and 1931. The decline may have resulted from habitat fragmentation because subalpine meadows have been invaded by forest during the past century. Interspecific competition with goats seems—to us and to M. L. Johnson—unlikely to have been the force driving the decline in gophers.

At least 16 species of insectivores, rodents, and lagomorphs occur in alpine areas of the Olympic Mountains (Johnson and Johnson 1952; Sheehan 1978; Reichel 1984). Dragavon and Weisbrod (1978) conducted a preliminary survey of small mammals on the south slopes of Klahhane Ridge. They trapped one 50- x 50-m grid in each of two plant communities (Phlox—Festuca scree slope and Artemisia—Festuca—Elymus) during 10-15 September 1978. Fifty-one individuals were trapped, mostly deer mice (Peromyscus spp.; 53%), yellow-pine chipmunk (Tamias amoenus; 18%), and Oregon voles (Microtus oregoni; 14%). Although they speculated that modification of the plant communities (species richness, physiognomy) by mountain goats might affect distribution of certain small mammals, they also recognized the limited nature of the survey, including the likelihood that differences in species trapped between grids reflected only habitat preferences of the small mammals.

The Oregon vole occupies a wide variety of herbaceous communities in the park. Sheehan (1978) studied density, demography, and diet of Oregon voles in three herbaceous plant communities (Carex spectabilis, Valeriana sitchensis, and Saussurea americana) in the northeastern Olympics. Vole densities and recruitment were greatest in the Carex community. Twenty-five plant species were identified in vole diets, including several taxa also consumed by mountain goats (e.g., Luzula spp. and Carex spp.). Goats are known to alter the species composition of at least the Carex spp. community, but we do not know how this may affect voles.

On a broader geographic scale, examinations of the population densities, structures, and individual body weights of small mammals suggested that high-elevation areas in the Pacific Northwest generally represented marginal habitats for many species. Sites often functioned as dispersal sinks for populations found at lower elevation sites; that is, populations in alpine or high subalpine areas seemed to be maintained by the influx of individuals from below (Reichel 1984). Effects of goats on small mammals in these apparently marginal habitats—by altering plant cover, morphology, or composition—are unknown.

Carnivores

Predators in Olympic National Park that are known or suspected of taking mountain goats (either from observations in the park or elsewhere) include cougars (Felis concolor), bald and golden eagles (Haliaeetus leucocephalus and Aquila chrysaetos), black bear (Ursus americanus), and coyote (Brandborg 1955; Rideout and Hoffman 1975; Johnson 1983). We consider both cougars and black bears to be abundant in the park. For example, the 56 and 26 bears counted incidentally during the July 1983 and 1990 goat censuses represented respective densities of ~0.25 and 0.11 bear/km². However, evidence of actual or attempted predation in the park is scarce and consists of cougar tracks at a goat carcass (Stevens 1980), a cougar observed stalking goat kids (D. B. Houston, personal observation), and three attempts by a golden eagle to take a kid (Hutchins 1984). Predation on goats is considered a comparatively minor source of goat deaths in the park. We have more records of deaths by falls than predators. Observations of goats dead from winter mortality, falls, or other causes suggested that carcasses were consumed rapidly by an array of scavengers including ravens (Corvus corax), crows (C. brachyrhynchos), gray jays (Perisoreus canadensis), coyotes, and bears.

Interpretation

Relations between goats and the established native fauna in the Olympic Mountains are obscure. Goats clearly coexist in the nearby Cascade Mountains and elsewhere in western North America with the same species (or near relatives) as those in the Olympics. This suggests that interspecific competition for resources, involving either interference or exploitation as components of the competitive relation (Miller 1967), may be mitigated through resource division.

Still, it seems unlikely to us that herbivores as large as mountain goats could be introduced into the Olympic Mountains, modify the vegetation to the extent documented, reach local densities of 0.08 animal/ha, and not affect the distribution or abundance of the pre-existing fauna. Goats seem most likely to have affected populations of small mammals (and perhaps ground-nesting birds?) associated with rock outcrops, cliffs, and similar sparsely vegetated sites because of their modification of the vegetation—but even this cannot be demonstrated empirically. In addition, goat deaths have likely increased food supplies for carnivores, particularly at high elevations.