NATIONAL PARK SERVICE Research Report GRTE-N-1 The Elk of Grand Teton and Southern Yellowstone National Parks

This section attempts to integrate study findings into an overall account of the relationships between the elk population and its environment. Reconstructions of probable past relationships for comparison purposes were aided by reference literature on ecological principles (Elton, 1927; Allee et al., 1949; and Daubenmire, 1968), Errington's (1946) reviews on vertebrate predation, and by the author's preliminary findings from a study (since June 1967) of a naturally regulated elk population within the west central portion of Yellowstone Park. Other cited literature also aided interpretations on population ecology or predation. Subsections on environmental influences, behavior, and habitat relationships led to final considerations of elk population regulation.

Environmental Influences

Climate and Weather

Climate had long-term influences through its role in directing relatively slow plant succession processes which changed habitat and food conditions. As periodic severe winter weather, it had pronounced short-term effects on the elk population. Deep snow and/or particular crust conditions limited the quantity and/or quality of winter food available to the animals by restricting their movements and foraging actions. These conditions, in combination with low temperatures, caused energy stresses which influenced population mortality and reproductive rates.

Winter Food and Plant Succession

Winter food represented the most limited source of energy for the studied elk population. Other environmental influences and the distribution and size of the elk population determined the extent to which winter food was utilized or in limited supply.

Elk that freeranged on ecologically complete winter habitats that were complexes of bottomland, swale, and slope areas (or ecological equivalents) seemed unable to progressively reduce snow-covered food sources by their foraging activities alone. The animals maintained what could be considered natural biotic disclimaxes or zootic climaxes (Daubenmire, 1968) on limited snow-free ridgetops and upper slopes. Elk that were concentrated by artificial food sources or were otherwise restricted from using ecologically essential units of winter habitat maintained disclimax conditions that probably would not have occurred with freeranging animals.

Plant succession since the retreat of the last glaciers has undoubtedly increased winter food sources for elk. This process still continues on wintering areas where pioneer and some disclimax stands (from past summer livestock grazing) of big sagebrush are being slowly replaced by more preferred grasses and shrubs.

The replacement of seral willow and aspen stands on elk wintering areas with deep snows has and will continue to represent some reduction in available food. This was partially offset by the reestablishment of successionally young stands, but long range trends seemed to be toward progressively less willow and aspen. The replacement of willow stands in areas with lesser snow depths resulted in quantitative gains in herbaceous forage, These gains were partially offset by herbaceous vegetation being less available than willow during most winters. The replacement of aspen by conifers or herbaceous vegetation on areas with lesser snow depths has proceeded to the point where existing remnant stands do not represent a quantitatively significant winter food source for the majority of the elk population. Aspens' role during earlier times with geologically young substrates, an absence of competing vegetation, more frequent fires, and possibly a different climate may have been more significant. Elk and other plant-eating animals hastened the replacement of seral willow and aspen after succession reached advanced stages or stands were reduced to remnant status.

Predators and Scavengers

These roles are not separated because many native meat eaters were both predators and scavengers. Also, relationships appear to exist where scavenging required more efficient predators to make additional kills or contend with food limitations set by scavengers. Original predator and scavenger populations may have periodically dampened and extended the interval between elk population peaks (see Population Regulation). Reductions in the numbers and/or kinds of predator and scavenger fauna have undoubtedly changed these relationships.

Organized predator control programs during the early 1900's contributed to the near extermination of the mountain lion (Felis concolor) and grey wolf (Canis lupus) from the Grand Teton and Yellowstone regions. Mountain lions have been sighted in both parks in recent years, but seem unable to increase their numbers. A known 134 wolves were killed within Yellowstone Park between 1916 and 1926. The animals were believed to have been eliminated. However, park records show fairly consistent sightings of single wolves and groups of up to four animals have been reported within Yellowstone Park since 1932. Present low numbers and distributions, almost entirely within Yellowstone Park, precluded the wolf from having a significant influence on the studied elk population. The black bear (Ursus americanus) and grizzly bear (Ursus horribilis) were also precluded as significant predators by their low numbers or distributions.

The coyote was protected within Yellowstone Park after 1935. Unrestricted shooting or control programs continue to the present on the lands outside national park and elk refuge boundaries. The coyote seemed to have sufficiently secure habitats to remain abundant and in some dynamic relationships with the studied elk population.

Accounts of predators killing elk and other wild animals from Murie (1940), other unpublished records from Yellowstone Park files, and the author's observations of attempted or circumstantially successful (examinations of carcasses, tracks in snow, etc.) predation between 1962 and 1968 suggested that past and present predation on elk did not depart from general principles summarized by Errington (1946). These were that predation was largely limited to some portion of the annual production of young, to animals predisposed to being preyed upon by accidents, sickness and old age, or to animals attempting to inhabit low security level habitats where environmental conditions made them vulnerable.

Elk carrion may have represented a seasonally important food source for golden (Aguila chrysoetos) and bald eagles (Haliaeetus leucocephalus), ravens (Corvus corax) and magpies (Pica pica). It, in addition to predisposed adult elk, appeared to represent a food source which could influence coyote and grizzly bear numbers. Murie's (1940) study in Yellowstone Park suggested that direct relationships exist between winter and spring carrion sources and coyote numbers. Studies by Jonkel (1967) suggest that, following summer or fall seasons with poor berry or whitebark pine nut crops, the survival of subadult bears could depend upon their finding carrion or having predisposed animals to prey on after they emerged from hibernation.

Coyote predation on adult elk seemed limited to weak animals unable to stay within groups or very near death. Other elk actively kept coyotes at the margins of larger groups and away from predisposed individuals. Grizzly bears appeared to be more efficient predators through their apparent ability to course and pull down adult elk. Their predation, however, seemed to be largely restricted to newborn young and predisposed or vulnerable animals.

Most early accounts of wolf predation in the Grand Teton and Yellowstone region reflect the predator control attitudes which prevailed into the mid-1930's. Park records since this period (to 1969) are limited to eight reports of one to four wolves feeding on elk within Yellowstone Park and one sighting of a single wolf on the offal from a hunter-killed elk about 3 miles north of Grand Teton. Reports by Olson (1938), Murie (1944), Cowan (1947), Stenlund (1955), Burkholder (1959), Mech (1966), and Pimlott (1967) suggest that wolves could be a more efficient predator on large ungulates than the grizzly. This, however, might only be to the extent that the animals occurred in packs of some optimum size.

Parasites and Disease

The original ecological role of indigenous parasites and diseases probably represented a form of selective "predation" on weak or aged elk and assistance (predisposing) to predation that hastened the deaths of such animals. Such relationships appeared to still partially exist, but at a relatively inefficient level. Without selective culling by larger predators, predisposed elk (by disease and other agents) commonly persisted over winter or into early spring before dying. Infestations of the winter tick (Dermacentor albipictus), and scabies from a psoroptic mite (Psoroptes sp,) were the most apparent manifestations of disease in the studied elk population.

Other Animals

Present day relationships between elk, other wild animals, and their food sources are illustrated on a portion of a food web diagram (Figure 15). The vertical lines between herbivores and representative forage plants within different habitats illustrate food sources which could, in combination with other environmental influences, limit population size. Dashed sloping lines show dual or multiple use on food sources between different herbivores. Solid lines from herbivores to predators, scavengers, and parasites may represent either specific or general food links.

Fig. 15. Food relationships between animals and vegetation types on valley winter ranges. (click on image for an enlargement in a new window)

Significant intensities of dual or multiple use appeared to be limited to areas where the different habitat types or main food sources for wintering elk, moose, or mule deer adjoined each other. Suggested relationships were that differences in food habits and/or environmental conditions influenced distributions and prevented biologically significant competition where one species could progressively displace the other. Varying degrees of "exclusion" competition occurred on portions of areas or habitats with the greatest overlapping use by different species. Such interspecific competition seemed an essential relationship which prevented one species from appropriating the habitat or food niche of another.

Deep snow (24 or more inches) appeared to allow moose to "outcompete" and ultimately preclude elk from using significant portions of their main willow food sources. Lesser snow depths permitted elk to utilize willow in conjunction with their more abundant herbaceous food sources on or adjacent to bottomlands and ultimately "outcompete" moose. Wintering mule deer contributed to overlapping use, but appeared to persist because of abundant shrub food sources on slope areas and their ability to utilize big sagebrush as a main food item. Snow conditions variably influenced the availability of shrub food sources on slopes and the extent of overlapping use by different herbivores.

Small numbers of pronghorn antelope that were a remnant of a larger population that summered within Grand Teton Park at the turn of the century. Some animals occasionally attempted to winter within the park. Deep snow made them vulnerable to the substantial coyote population which appeared to be largely sustained by elk carrion.

Small bands of bighorn sheep (Ovis canadensis) stayed within the Teton Mountain range yearlong. Another group wintered on the east side of the National Elk Refuge. These animals mainly wintered on and adjacent to rock outcrop areas. Such habitat areas were undoubtedly more extensive when the region was geologically younger and/or the climate was sufficiently warm to permit the winter use of sites which are now within deep snow zones and/or tree covered.

A warmer and drier climate occurred between 9,000 and 4,500 years ago (Kind, 1965). It seems probable that climatic change and ecological succession, involving elk as well as other faunal species, since this period have contributed to the present relict status of the bighorn. Archeological excavations bordering Yellowstone Park show a race of mountain dwelling Indians may have subsisted mainly on bighorn sheep at least 4,500 years ago (Wedel et al., 1968). Modern man's influences may have additionally contributed to the bighorn's present status by eliminating more vulnerable groups from accessible habitat areas and grazing domestic stock on their winter ranges.

Elk relationships to predators, scavengers, parasites and diseases have been discussed. In addition to their use of the elk as a host, certain insects strongly influenced the distributions and movements of elk and thereby their use of food sources (see Elk Habits and Habitat Use).

Man

It seems unlikely that the summer visits of Indians, early trappers, or even the first few settlers in the Jackson Hole region up to about 1910 had significant impacts on the large elk population or changed environmental conditions for the animals. This was not the case after human settlement reached levels where the elk's historical wintering areas were almost completely appropriated for domestic stock grazing or hay raising. Figure 16 illustrates the extent to which modern man has become part of the elk's environment. Human developments, agriculture and hunting, in combination with the practice of winter feeding the animals, have become major influences on present day elk populations.

Fig. 16. Elk population regulation. (click on image for an enlargement in a new window)

Developments and Winter Ranges

The use of valley lands south and north of the town of Jackson for hayfields, livestock pastures, ranch or home sites, and more recently, a golf course has led to the present elk herds being restricted to wintering on about 60 percent of the land described as their winter range in 1911 (Figure 6). No lands were specifically set aside for wintering elk until 1913 when 2,800 acres were designated as the National Elk Refuge. About 1,760 acres were privately donated in 1927. Purchases and administrative actions between 1935 and 1950 led to the refuge reaching its present 22,700 acre size.

Refuge, adjoining national forest, and enclosed state lands presently make up a 37,500 acre unit (Houston, 1968a). This, a 18,700 acre unit within Grand Teton south of Ditch Creek, and an adjoining 4,000 acre slope area to Horsetail Creek total about 60,000 acres and make up the largest single block of publicly owned historical winter range available to the elk in the Jackson Role valley. Additional wintering areas within Grand Teton Park north of Ditch Creek and the adjoining Buffalo River valley total over 7,000 acres.

Preble (1911) described the extensive "marsh" bottomlands along Flat Creek above the town of Jackson as a "favorite haunt" for wintering elk. He reports other bottomland or slope areas as wintering 2,000 or 3,000 elk, harboring good sized herds, a few hundred, or a few animals. Greater numbers of elk appeared to winter where the animals had extensive bottomlands in addition to upland slopes. The Flat Creek bottomlands north of the town of Jackson occur within present refuge boundaries. Other comparable bottomland areas that Preble mapped as elk winter range lie west of present refuge boundaries and south of the town of Jackson. These have become almost entirely privately owned and wintering elk are either restricted to using small State owned feed grounds or slope areas bordering the main valley. Less extensive bottomland areas, formally mapped as winter range, occur in Grand Teton's Spread Creek-Uhl Hill area, the adjoining Buffalo River valley, within the Gros Ventre River flood plain along the north boundary of the refuge, and scattered along the Snake River in and outside Grand Teton Park. These bottomlands were or still are variously used for hay raising and/or livestock grazing.

Livestock Grazing

Approximately 11,000 a.u.m.'s (one animal unit grazing one month) of grazing by over 4,000 cattle and horses over a May 1 to November 15 period were acquired when national monument lands were added to Grand Teton Park in 1950. Progressive expirations of lifetime leases and transfers of land could result in the elimination of large scale livestock grazing within the park some time after the year 2,000. Extensive open range grazing in the park has been progressively transferred to fenced pastures on the east side of the Snake River since 1958. Some pastures included important wintering areas for elk, mule, deer, and moose.

Figure 15 shows cattle had the capacity to compete with elk, mule deer, or moose. This appeared to reach significant levels in the park's Spread Creek-Uhl Hill areas. Here, 1963-1966 measurements showed late summer and fall cattle grazing reduced main winter food sources for elk and mule deer (see Effects on Habitats). Elk shifts to use adjoining moose habitat and food sources were pronounced in 1964. Rancher cooperation was partially secured after 1964 to move cattle to irrigated bottomland pastures before they started to use elk and mule deer winter food sources. Cattle and horse grazing on pastures within the southern portion of Grand Teton Park would only cause conflicts if large numbers of elk again used the historical winter ranges between Ditch Creek and the north boundary of the refuge.

Hunting

Man's hunting directly influenced elk population death rates and distributions (Figure 16). Indirect influences resulted from hunting other wildlife that might occur in some relationship with the elk as a competitor, scavenger, or predator. Elk distributions were probably influenced by removing particular animals from herd groups, conditioning survivors, and by simple avoidance responses to hunting disturbances which caused animals to move to superior escape habitats or areas where they were undisturbed. Martinka (1969) reported on the 1964 movements of marked elk on and across Grand Teton and the National Elk Refuge and concluded hunting encouraged movements to and discouraged movements from areas closed to hunting. The results from the purposeful use of hunting to halt early habitual movements to the refuge and its unintended effects on the historical distributions and migrations of other elk groups are discussed in the Fall Migration section.

Artificial Feeding

Differences between high pregnancy rates and calves at heel, the high spring mortality of previous year calves following severe winters, and the comparative appearance of animals on and off refuge feed grounds suggested that artificially fed hay was either a nutritionally inadequate diet for subadults and pregnant females or the feeding process itself increased overwinter energy stresses to levels which reduced the net quality of the diet.

The greater exposure of animals on open feed grounds to weather influences may have been partially involved. Moen (1968) showed wind in combination with freezing temperatures could progressively reduce the effectiveness of both high quality and maintenance diets in balancing energy losses in white-tailed deer. Smaller animals were most affected. This basic surface to mass relationship suggests that the reduced opportunity for calf elk on open feed grounds to obtain shelter from wind could partially account for their usual less thrifty appearance (by late winter) than calves in free-ranging groups.

The establishment of energy-conserving social relationships (peck orders) between different aged animals or dominant individuals appeared impossible when large numbers of elk were concentrated on feed grounds. The subtle and overt agonistic behavior that occurred with the daily crowding or intraspecific competition of 1,000 to 4,000 elk on feed lines could be expected to expend energy that reduced the effectiveness of a diet. The consequences of crowding were most apparent in the rapid breakdown of female-calf associations. Large groups of calves that were obviously not at heel concentrated together on feed grounds. Calves in smaller free-ranging groups off feed grounds appeared to be associated with maternal females through the winter.

Refuge records (Table 26) indicate that approximately equal winter herds have been maintained from 1912, when large groups of elk (7,250 animals) were first fed, to recent years. Periodic low numbers up to about 1945 probably resulted from initially "high" mortality occurring both on, and after elk left, the refuge as well as the presence of herd segments that did not use feed grounds consistently. Undetected mortality on spring ranges apparently substituted for mortality on feed grounds after 1945. Elk also used feed grounds more consistently. After 1945, higher hunting removals, periodic "high" mortality off feed grounds and compensating reproductive increases apparently maintained winter herd numbers between about 6,000 and 11,000 animals (7,600 average).

Table 26 also shows artificial feeding for 3 to 4 months on relatively small areas has had the end result of partially substituting hay diets for natural foods on extensive areas previously used by the animals. This was undoubtedly necessary because of the limited wintering areas set aside for the animals before 1935, and a subsequent state damage law that tended to make artificial feeding less costly and controversial than continuous investigations and payments for elk damage.

The particular conditions that may have contributed to the refuge winter herd maintaining its numbers for over 50 years are:

¹ From National Elk Refuge records.

^* No feeding one year within period.

^** No feeding two years within period.

1. Initial high mortality of calves could have led to some selection for a population more adapted to feed ground conditions.

2. Abundant natural food sources became available to the animals as lands were added to the refuge after 1935.

3. Artificial feeding for 3 to 4 months provided a relatively consistent subsistence diet that could be supplemented by abundant natural foods adjacent to feed grounds.

4. Feed grounds were located on an extensive 6.5-mile long by 3-mile wide fertile bottomland flat where trampling and close cropping of vegetation was accommodated without causing destructive overuse of upland sites away from their immediate vicinity.

Man's continued routine feeding of the refuge elk herd after extensive historical wintering areas became available for their use appears to be partly tradition and partly the very real prospect of damage claims from the relatively few ranches scattered within or bordering historical wintering areas inside Grand Teton Park. A possible solution to the damage claim problem on park lands could involve securing scenic easements until present leases expire.

Behavior Relationships

Behavior that appeared to be caused by external environmental influences or relationships within the elk population itself is listed in Table 27. Spring dispersals off feed grounds, November migrations that occurred with particular snow conditions, and aggregations of animals into large groups on refuge feed grounds appeared to be food-linked responses.

Movements off the refuge to Grand Teton spring ranges and into Yellowstone Park seemed initially to be a response to reduced environmental resistance; by mid-May, some overriding attempts by females to reach calving areas. These, the dispersals off feed grounds and November migrations, tend to show behavior that resulted from a progressive relaxing and return of severe weather conditions that influenced the availability of food or restricted movements along a general elevational gradient (the Snake River drainage). It appeared that elk would remain in scattered distributions on summer home ranges in the absence of weather conditions that reduced the availability of their food. Weather severity along elevational gradients resulted in the animals moving to and concentrating on lower elevation areas where essential members of the population survived the severest winters. Moderating weather allowed reverse movements and a return to scattered distributions on summer home ranges.

Delayed movements of a portion of the elk population onto high elevation ranges, August dispersals into forest types, and the formation of harem groups represented behavior linked to reproduction. Newborn calves delayed the movements of maternal females and other associated animals to the most distant Yellowstone summer ranges and to high elevations until July. Subtle attempts by previously segregated adult male elk to associate with groups of females and calves (observed as early as August 9) contributed to breaking up summer group associations and caused dispersals into forest types. Overt displays by adult males and attempts to collect females in harems (rutting behavior) were observed as early as August 13 and commonly after this date. Field records showed all male elk older than yearlings had started or completed removing the velvet from their antlers by August 25. The process was observed to start as early as August 13. Attempts to hold females in harem groups appeared to be most successful between mid-September and mid-October. Over the mid-August through October period, the size of elk groups in mountain areas averaged about six animals (N = 4999). Yearling males were usually excluded from harem associations of females and calves that were attended by an adult male.

Classifications obtained by Martinka (1965) showed late August through October group sizes in valley areas were much larger, averaging about 30 animals (N = 12,936). An average group size of 11 animals was observed during the mid-September to mid-October peak of breeding activity. These higher group sizes probably resulted from fewer adult males being initially present in valley areas and a greater tendency for harem groups to aggregate on extensive fall range areas with limited forest cover.

October migrations of Grand Teton elk to refuge winter ranges appeared to represent behavior responses to viewing by park visitors and some illegal hunting disturbances on large elk groups (200 to 500 animals) within low security level habitats (outwash plain with limited forest cover). After 1964, old roads within the western portions of Grand Teton were closed to provide blocks of fall range where large elk groups could be seen from vista points, but not disturbed by too close an approach. These roads had either penetrated major fall range areas for large elk groups or allowed untolerated approaches between foraging animals and their forest cover. The road closures, in combination with limited permit hunting on eastern portions of Grand Teton and the National Elk Refuge, greatly reduced early October migrations from the western portions of the park.

As discussed in the Habitat Use section, aggregations of large numbers of elk at high elevations in mountain areas were caused by molesting insects. The equivalent behavior response for elk in Grand Teton valley areas or in groups scattered at low elevations within mountain areas was to retire into forest types or bed in dense herbaceous vegetation in meadows. This would suggest that high elevations were not a deciding or critical factor in the animals' establishment of summer home ranges. Brazda (1953) sampled molesting insect densities which indicated elk would obtain greater relief at high than low elevations. An additional relationship may have been that large aggregations of elk afforded relief. Allee et al. (1959) cites an account that aggregations of at least 300 to 400 reindeer (Rangifer sp.) permitted herds to remain intact under warble fly attacks.

The suggested relationship was that elk largely influenced their own scattered distributions when environmental stresses were minimal. These were variably maintained by matriarchal care-dependency and leader-follower associations with a dominant female elk; by mutual avoidance, subtle and overt agonistic behavior that maintained dominance subordination relationships within and between associations; and by sexual relationships between adult males and females. The latter involved either mutual or female avoidance behavior during periods other than the breeding season. Terminology follows Etkin (1964).

Aggregations of elk from scattered summer distributions and fall migrations occurred in relation to overriding environmental influences and coincident subjugations of some dominant females into leader-follower relationships. Subjugations may have resulted from some lessening of female dominance apart from home range "territories" used for the care of young. Aggregations of female and subadult elk on wintering areas apart from feed grounds probably re-formed as variably sized matriarchal associations with both care-dependency and leader-follower relationships. These re-established social order with energy conserving dominance-subordination (peck order) relations. Off feed grounds, adult males usually wintered in loosely organized herds socially apart from matriarchal associations. Such segregations of adult male elk apart from groups of females, calves, and yearling males were apparently the rule in early day elk populations (Preble, 1911).

Large groups of female, subadult, and adult male elk on feed grounds appeared to represent aggregations where social relationships progressively deteriorated. This may have resulted from daily occurrences of agonistic behavior between large numbers of variably dominant and subordinate elk on feed lines. The establishment of energy conserving peck orders was precluded and led to an early dissolving of maternal care-dependency relations (see Artificial Feeding). Moderating environmental conditions in spring, in combination with the re-establishment of maternal-care relationships, led to the re-establishment of scattered distributions on summer home ranges.

Aggregations of social groups did occur from foraging encounters on other than wintering areas. These were usually temporary or occurred under conditions where close crowding to obtain food was not necessary. Aggregations from escape encounters occurred in response to varying intensities of human disturbances and molesting insects. Male sexual behavior appeared to cause dispersals from aggregations and either temporary or lasting disruptions of summer matriarchal associations.

Habitat Relationships

Elk have apparently persisted for thousands of years in the Grand Teton and Yellowstone regions over a wide range of environmental changes which are still occurring. Vegetation changes have been short term and cyclic from fire, biotic influences, and variable growing conditions; or directional from developing soils, stream cutting, and climatic change. Selection pressures for the "most fit" plant and animal species have undoubtedly occurred and will continue. It seems unlikely that elk would have persisted if the animals were able to progressively deplete their main food sources which, in combination with other influences, determined their numbers (i.e., had population consequences).

Winter habitats that were interspersions of different physiographic sites and/or vegetation types provided increased opportunities for an elk population to remain in some dynamic balance with its food sources (homeostasis). These ecologically complete habitats had carrying capacity relationships where "the whole was greater than the sum of its parts." The elk's variable use of different habitat units, general food habit, protection from snow, and the capacity of native plants to withstand periodic heavy use appeared to preclude free-ranging animals from progressively depleting their main winter food sources. The density-influenced mortality of animals with low energy reserves also helped to maintain elk populations in balance with their food sources (see Population Regulations).

As biotic agents, elk influenced the rate at which late stages of seral vegetation were replaced, maintained relatively stable biotic disclimaxes on limited sites where their effects were either without population consequences, or were incidental to the use of food sources that had population consequences. They also occurred in some dynamic relationship with other native herbivores through "exclusion" or interspecific competition that retained a mixed species fauna within different food or habitat niches.

Exceptions to these apparently natural relationships occurred on upland areas adjacent to refuge feed grounds and wildlife wintering areas additionally grazed by domestic stock. Here, animal concentrations and/or consistent heavy or dual use of vegetation appeared to intensify disclimax conditions or cause seral vegetation to be replaced at a faster than "normal" rate.

These interpretations of free-ranging elk relationships to their winter habitats may only apply to other areas with equally variable and rigorous winter weather. They would have limited to almost no application where human influences restricted or precluded elk from using portions of a winter habitat (e.g., bottomlands, slopes, etc.) which were essential to homeostasis. What may be shown is that interpretations of elk habitat relationships require considerations of natural successional processes, the ecological completeness of winter habitats, and distinctions between food sources which do or do not have population consequences. Natural biotic effects or sucessuccessionalges would not require corrective management within a national park.

Population Regulation

The logistic curve relationship between population growth and environmental resistance may have been first expressed by Verhulst in 1883 (Allee, et al. 1949). Accumulated knowledge since this date further establishes that animal populations occur in some equilibrium (mean numerical stability) in the absence of environmental changes that consistently cause more or less resistance to population growth. Environmental changes which consistently offer less environmental resistance permit upward trends in population numbers. Consistently more environmental resistance results in downward population trends.

A regulating influence was considered to directly or indirectly cause deaths, or change population reproduction, or survival rates. The complementing influences from some animals emigrating from or immigrating into a population is recognized. The probable regulatory process for past as well as present populations is presented for comparison purposes.

Past Populations

Accumulated knowledge on the organization of life in natural communities tends to assure that past elk populations were regulated to the extent that they could not, by themselves, progressively deplete food sources which limited their numbers. This study suggests that the animals could have temporarily reduced the amount or quality of their own food sources as part of a natural regulatory process, reduced or maintained some food sources that did not limit their numbers as natural biotic disclimaxes, and accelerated late stages of plant succession to either increase or decrease their total food supply.

Intraspecific competition for available food and environmental influences from winter weather, predators, scavengers, and diseases probably interacted to lower the numbers in early-day elk populations in the following manner: When populations were at upper levels in relation to their available winter food, intraspecific competition intensified energy stresses. These stresses directly or indirectly caused the deaths of elk with the lowest energy reserves and sometimes lowered the subsequent year's reproductive success. The deaths of diseased and other energy-stressed animals were hastened by the combined effects of predators and scavengers.

Severe winter weather per se periodically caused higher than usual deaths, or what could be considered additional density-independent mortality, by increasing intraspecific competition, energy stress, and the efficiency of predators. These additional deaths were also animals with low energy reserves. Subsequent winters with less severe weather or intraspecific competition permitted elk populations to compensate for the deaths of animals with low energy reserves and return to higher numbers. Compensations resulted from increased reproductive success and survival or the process Errington (1946) calls "compensatory trends."

The reports of "high" winter mortality in the Jackson Hole herd at 4 to 6-year intervals during 1882, 1887, 1891, 1897, and 1911 (Preble, 1911; Anonymous, 1915; Sheldon, 1927; Brown, 1947) suggest that early day predator populations did not prevent the elk from contending with the regulatory influences from intraspecific competition for food or periodic severe winters. This should not be interpreted that predation on elk populations was without ecological significance. Original predator populations may have reduced the intensity of intraspecific competition within an elk population during more severe winters. The extent to which this occurred would have extended the interval between and dampened elk population fluctuations. The compensatory trend process (Errington, 1946) could be expected to compensate for periodically higher than usual mortality from predation.

Present Populations

Intraspecific competition for food and environmental influences from man, winter weather, a limited predator-scavenger fauna, and disease acted to lower numbers in present elk populations (Figure 16). The regulatory process mainly differed from that on past populations to the extent that man increased or decreased the intensity of natural regulatory influences. This resulted from his artificially feeding the animals, displacing the original predator-scavenger fauna, and changing herd distributions so as to reduce total food sources.

When elk on winter feed grounds were at upper levels (apparently a wide range) in relation to the available energy from their artificial diets and adjacent food sources or the effects of periodic severe weather, intraspecific competition increased energy stresses. These stresses directly or indirectly caused the deaths of subadults and older adults with the lowest energy reserves and sometimes lowered the subsequent year's reproduction. The deaths of diseased or energy-stressed animals were not significantly hastened by the remnant predator fauna which was mainly restricted to preying on newborn elk. Population increases back to higher levels, by compensating reproduction and/or survival, were influenced by hunting removals and other human influences.

The density-influenced or periodically higher density-independent deaths of animals with low energy reserves would not represent a loss of biologically essential population members and would be predestined to occur to the extent that elk populations were self-regulated (intraspecific competition) in relation to their available winter food. Such mortality would not occur to the extent that hunting removals could substitute for density-influenced deaths.

From 1962 through 1967 about 700 to 800 elk were estimated to consistently winter without using artificial food sources. These animals occurred in scattered distributions on historical winter areas on the refuge, Grand Teton Park, and adjoining national forest lands. Their numbers were variously regulated by interspecific competition, weather influences, hunting, and competition with domestic livestock. Small elk groups, that remained within Grand Teton areas closed to hunting or arrived on park winter ranges after hunting seasons were closed, were largely self-regulated by intraspecific competition for food and weather influences.

Generally high hunting kills from 1940 through the late 1950's coincided with progressive declines in the numbers of elk that freeranged off refuge feed grounds. This reduction in animals, which were partly or wholly on different food sources, may preclude presently distributed winter herds from reaching 1955-56 and earlier year population peaks of 10,000 to 11,000 animals. Sustained hunting removals that approximate herd increase rates and yearly artificial feeding may additionally restrict winter herd numbers from fluctuating outside the general 6,000 to 8,000 range that has prevailed over severe, average, and mild winters since 1961. Herds could be expected to occasionally fluctuate to higher levels with lower or less consistent hunting removals. The extent to which artificial feeding prevented subadults and/or other population members from freeranging in scattered distributions and obtaining more adequate diets could also restrict fluctuations to higher levels.

Man's actions in restricting elk from freeranging were not always unintentional when the animals' historical winter range was largely privately owned and used for livestock grazing or hay raising. Transfers of land, purchases, and administrative withdrawals have progressed to where the refuge herd could be allowed to freerange over a 60,000-acre block of this historical winter range. This could conceivably replace all artificial feeding.

Consistent artificial feeding of the refuge winter herd may result in a more apparent than real lowering of overwinter mortality rates by maintaining low proportions of vulnerable subadults in the population and only deferring until spring what Preble (1911) and others have considered "high" mortality from severe winters (about 15 to 20 percent of herd numbers, or a large portion of the calves). Comparable mortality of subadults and other elk with limited energy reserves appears to have occurred periodically over a wide range of population size up to the most recent severe winter of 1961-62. Such partially density-independent mortality would preclude maintaining highly stable elk populations.

If the present refuge herd was largely self-regulated as a result of low hunting removals and was not artificially fed, its winter numbers could conceivably fluctuate within a 5,000 to 9,000 range. If artificial feeding supplied more energy to subadults and pregnant females than they could obtain by freeranging (this needs to be demonstrated), the population might fluctuate within a 6,000 to 9,000 range; if it did not, a 5,000 to 9,000 range. The latter approximates the usual range of winter herd numbers that occurred before the higher hunting removals of the 1950's (Table 23 and Figure 8).

With hunting removals that attempted to reduce intraspecific competition by approximating average population increase rates, the winter numbers of a free-ranging herd might fluctuate between 5,000 and 8,000 animals. This compares with 6,000 to 8,000 fluctuations between 1961 and 1967 which may or may not have been maintained at higher levels by artificial feeding. These figures should be considered approximations which are mainly used to illustrate suggested relationships. Fluctuation ranges could vary with unusual sequences of winter weather, variable hunting removals, or changes in artificial feeding practices.

Man's hunting since 1955 appears to have become somewhat more efficient than original predator-scavenger complexes in preventing extreme fluctuations in elk numbers. It has not prevented the elk population from being additionally regulated by periodic severe weather and intraspecific competition. Complete substitutions of hunting for all natural mortality do not appear possible because severe weather influences on the availability of food and on subadult elk or other animals with low energy reserves were not completely density-dependent within the full range of population numbers accommodated by variation in the winter environment.

Man was obviously less successful than the original predator fauna in allowing the elk population to maintain its numbers and distributions in relation to suitable habitats and food sources. This resulted from his more efficient and less restricted (to predisposed and vulnerable elk) hunting reducing elk population groups that used particular habitat areas and forage sources. Conditioned avoidance behavior appeared to additionally restrict elk from using extensive wintering areas with abundant forage sources.