The Bison of Yellowstone National Park
NPS Scientific Monograph No. 1
NPS Logo


The Present Population


As stated previously, the Yellowstone bison of the present derive from two subspecies: plains bison from Montana (Pablo-Allard herd) and Texas (Goodnight herd), introduced in 1902, and a remnant of the original wild population of mountain bison. Skinner and Alcorn (1942-51) summarize the introduction, numbers, and subsequent management practices pertaining to the introduced herd at the Buffalo Ranch in Lamar. Population numbers are from that source and other official reports. Skinner also provides a resumé of official information concerning the wild bison, but does not attempt to evaluate the question of their survival. Information scattered in the diaries, reports, and correspondence of park personnel (Yellowstone National Park Archives) provides the basis for the following.

Before 1915, introduced bison of plains stock could not have escaped to form a wild group. The introduced herd was in a small fenced pasture at Mammoth from 1902 until moved to the Buffalo Ranch at Lamar in 1907. From 1907 until at least 1915, these animals were closely day-herded, and apparently put in a fenced pasture at night. Although one plains bull from the fenced herd was turned out in 1903, and one or two stray bulls were mentioned later in scout reports, these apparently never joined the wild bison. Twenty bulls from the Buffalo Ranch were driven 14 miles up the Lamar River in 1914; these all returned.

Members of two other small groups of semidomestic bison, those of the Yellowstone Lake Boat Co. and some from Henry's Lake west of the park, never mixed with the wild herd (Appendix III).

In spite of very low numbers and a pessimistic outlook, the original wild herd did persist, and gradually increased once protection from poaching was assured. The critical survival period extended from 1902 until about 1920; thereafter, groups of bison which did not frequent the Buffalo Ranch were more common. These probably contained escapees from the introduced herd as well as native animals. Table 2, from sources listed in Appendix III, shows dates, locations, and numbers for the wild herd from 1903-15. Later official estimates of wild herd numbers are not used, since there was some possibility of wild and introduced animals intermingling after 1915. The table shows clearly the presence of a remnant wild herd in Pelican Valley in winter, and on the Mirror Plateau in summer, as well as a few individuals elsewhere. There was a steady increase, indicated both by calves and by total bison seen. The known population more than doubled between 1903 and 1912. An actual count was difficult, as Nowlin (1912) of the Biological Survey found: "I have never seen buffaloes on the range so wary and difficult to locate as the wild ones in the Yellowstone Park."

Official population estimates of the time did not allow for more animals than were actually seen; they were undoubtedly conservative. By 1912, as Nowlin's classification of 35 animals shows, the survival of calves (8) and yearlings (7) was encouraging, and the potential for increase (13 females) was apparent. By 1915, the population may have been nearing 100. Known mortality is shown only in 1904. While winter loss probably occurred during other years, the death of many animals would surely have been observed and reported by Army patrols or scouts.

Present composition

An estimate of the contribution of the two subspecies to the present population gene pool is, at best, rough but is preferable to having none. Table 3 shows the sex and adult-calf composition of the fenced herd from 1902 through 1915. The addition to this herd of four calves captured by Army scouts (Fig. 13) from the wild herd for the purpose of adding a second bloodline is also shown. From these numbers, Table 4 was compiled to show the age classes according to sex. The bulls 4 years and older were assumed to have done the breeding in this fenced situation, although younger bulls may have been physically capable. Females were assumed to breed as 2-year-olds and to calve at 3. From these assumptions and the tables, the wild strain in the fenced group was estimated at a maximum of 10% by 1910. Further dilution of the wild strain in the fenced herd was assumed until perhaps 1917.

Buffalo Jones and Army Scout Holt
Fig. 13. Buffalo Jones and Army Scout Holt with a captured bison calf on a sled. Photo from Yellowstone National Park files.

TABLE 3. Composition of fenced herd, 1902-15

yrlg & other
yrlg & over
TotalMale calvesFemale calves Total calvesHerd totalShow herdb

1903218202 + 2 wild3727
19046202666 + 1 wild1339
190930a42a72a13 + 1 wild9239614
19147782159a20153519416 + 20

bShow herd is included in herd total.

Sometime between 1915 and 1920, intermingling of the introduced and wild animals began. At first this was probably gradual. Park records do not show the specific year, but after 1915 the close herding practices in use with the introduced herd were abandoned, and the animals were kept on open range all summer. Close account was kept of most of them for several more years, but there were some escapees. After 1921, with construction of a log drift fence across the Lamar Valley above Soda Butte Creek, deliberate efforts were made to keep the introduced herd on the higher summer ranges, where intermingling with the wild bison must have quickly increased.

After intermingling of wild and introduced animals began, several factors tended to increase the wild strain in the total population, although the plains type outnumbered the wild roughly 3:1 about 1917, and perhaps 4:1 in 1921. The number of males in the introduced herd was reduced by the yearly segregation of a show herd of bulls beginning in 1909. Additional bulls were removed by live shipment and slaughter. To further reduce the male surplus (from a ranch operation viewpoint), castration of bull calves averaged slightly over 50% from 1916 through 1931. As a result, the number of aggressive, dominant plains-type bulls with the intermingled groups would have been considerably decreased.

Table 5 shows the presumed numbers of freeranging males in various age classes for both wild and introduced herds in 1921. An estimated 40% of the bulls older than 5 years were of the mountain bison strain. Their contribution to the breeding activity may have been larger, as discussed above, than their numbers indicate.

The trend toward increased mountain bison strain would have continued during the 1920s. On this basis, a reasonable estimate of wild strain in the present bison population would seem to be 30-40%.

Numbers and distribution

Management practices for many years made little attempt to re-create a natural, wild bison population in the park. Efforts concentrated on ensuring bison in numbers sufficient to guarantee perpetuation. Through 1938, horseback riders rounded up as many bison as possible in late fall and drove them into the Lamar Valley for feeding and reductions. To cut population numbers to desired levels during these reductions (and in most later ones), cripples, aged animals, and those infected with brucellosis or otherwise considered undesirable were removed to improve the herd, in keeping with the ranching viewpoint (W.S. Chapman 1969 pers. comm.). In 1939, a hay-baiting operation was substituted for the roundup. Hay was fed to some extent every winter through 1952. Before 1936, most animals wintered in Lamar, with some in Pelican. During summer, bison concentrated on the Mirror Plateau and Upper Lamar, with scattered bulls and a few small groups to the north. In spite of very large populations wintering in Lamar, natural reestablishment of the population west beyond the Pelican area into Hayden Valley and the Firehole did not occur, although a 1946 file report contains a penciled notation of some bison in Hayden Valley in 1930-31.

TABLE 4. Presumed age and sex composition, fenced herd, 1903-10.


YearAge: Adult765432YrlgCalf Adult765432YrlgCalf



6-P 17-P


6-P3-P 17-P


6-P3-P9-P 17-P




6-P3-P9-P37 16-P



19 15-P3-P1-W

P — plains type; W — wild type.
aFor 1910, the number of females, yearlings and older, does not agree with table 3 because of discrepancies in available records.

TABLE 5. Presumed age classes of male bison on the range, 1921.

Age classesWild herdIntroduced herd

Total males 1 year and older51198a
Total population125 estimated526

An equal sex ratio is assumed.
aNumber includes 96 steers among the 1-5 year olds.

In 1936, animals were trucked to the Firehole and Hayden valleys for release. They were thought to have formed separate herds, but as numbers increased, some movement between the two valleys became obvious, and they were called the Mary Mountain herd. Two other herds were distinguished, on the basis of wintering areas, as the Lamar and Pelican. None of these herds is geographically isolated at all seasons of the year, but the names are still used to designate the wintering populations.

Population counts, estimates, and known losses (mainly reductions) are listed in Appendix IV by the three wintering populations and as park totals. Aerial counts were started for the four main wintering valleys (Lamar, Pelican, Hayden Valley, Firehole) in 1949. These counts were not made every year, nor were all main areas checked each time. Usually no attempt was made to count scattered animals in fringe areas, nor to check the Bechler Meadows before 1965 (Jim Stradley 1968 pers. comm.). When aerial counts were not available, winter ground counts were made. Estimates based on previous counts, reduction figures, and presumed increases were made by park personnel when counts were not available. Because of possible population shifts, these may be unreliable. Population numbers are for winter seasons, after reductions, but before calving.

Bison increased steadily after 1902 until, with a gradual change in policy about 1930 from one of ranching to one of preservation of bison in a natural state, the National Park Service decided to cut the numbers wintering in Lamar. The decision was based on the gradual elimination of artificial management practices and supported by information derived from a range-condition and carrying-capacity study (Rush 1932b). Lamar-wintering bison numbered over 1000 from 1929 through 1932, before Rush recommended a maximum of 1000. Later decisions lowered the maximum number until frequent reductions had cut Lamar herd numbers to a count of 143 in 1952. After an aerial count for all main wintering valleys totaled 1477 in January 1954, reductions were made on all population segments. An aerial count for the entire park of 397, made in March 1967, was considered very accurate. Thus, the park population of bison for most of the study period was lower than at any time since the early years of the introduced herd.

Winter and summer distributions for mixed herd groups and separate bulls at population levels of the study period are shown in Table 6. Divisions between major areas indicate geographic separation but do not imply population isolation. Bulls were found in all areas of herd use and also were scattered in places where herd groups were seldom or never seen. Past records indicate that mixed herd groups used some of these areas when population numbers were higher. In effect, places most frequented by mixed herd groups probably represent core areas or population centers from which bulls and mixed herd groups move into less-favored locations as the population increases.

Although Firehole and Hayden valleys are combined as the Mary Mountain area, winter distribution of total numbers and mixed herd groups favored Hayden Valley, according to available counts since the 1950s. During the reductions of 1964-65 and 1965-66, animals were herded from Hayden Valley to the Nez Perce trap on the Firehole side with helicopters, but a prereduction count in December 1964 showed 436 bison in Hayden Valley, and only 54 on the Firehole. Prereduction counts of 1965-66 also located most animals in Hayden Valley. Groups released from the trap usually remained on the west side the rest of the winter, but movements by them between the two wintering valleys were known for all seasons. Groups were seen only in the two main valleys and the intervening Nez Perce Creek drainage during the study period, but a group of 20-23 was seen at Beach or Dryad lakes, 5 miles south of Hayden Valley, the winter of 1955-56 (Jim Stradley 1968 pers. comm.).

The majority of the wintering mixed herd groups of the Mary Mountain area summered in Hayden Valley and to the south of it as far as Beach and Dryad lakes, although some animals from the same population summered to the west of the Firehole. Ranger reports indicate that infrequent mixed herd use of the Madison Plateau began in 1939, 3 years after the bison were released on the Firehole. By the early 1950s, use had become common, but apparently almost ceased after the 1955 reduction. But in 1963 as many as 50 ranged from the Little Firehole Meadows to the Pitchstone Plateau. In spite of more reductions, small mixed herd groups were seen during the summers of 1965, 1966, and 1967.

After the early 1900s, animals were not again reported wintering in the Bechler Meadows southwest of the Madison Plateau until February 1955, when three bulls were seen just outside the park. An occasional bull may have wintered in the meadows earlier, since a few animals again began to summer on the Madison Plateau in 1939. Observations and reports indicate a few animals in that area nearly every winter since the mid-1950s, in spite of considerable decrease in the Mary Mountain herd, from which these animals probably came. No mixed herd groups were reported wintering in the Bechler Meadows until 1962-63, but the area is seldom visited in winter. Periodic plane flights begun in March 1965 showed a small group there in 1964-65 and 1965-66.

Lamar and Pelican populations are isolated from each other during most of the winter except for occasional movements of a few hardy bulls. Table 6 shows the valleys as distinct wintering areas. Groups in Lamar ranged the valley from Soda Butte west to the Hellroaring Slopes. In Pelican, group use extended from the Mushpots-Mudkettles of Pelican Creek downstream to Vermilion Hot Springs and included the lower parts of Astringent Creek and adjacent hot-spring areas to the west. Although no groups frequented Ponuntpa Hot Springs (6 miles north of Pelican Valley) during the study, both historic records and reports of the mid-1950s (Jim Stradley 1969 pers. comm.) indicated some previous mixed herd use.

TABLE 6. Distribution of the present population, 1969.


Mixed herd groups Mixed herd groups
Mixed group areas
Round Prairie
Cache Cr.
Slough Cr.
Hellroaring-Tower Jct. area
Lamar Soda Butte area
aupper main Lamar Valley
lower main Lamar Valley
aHellroaring Slope area
aMt. Norris
Cache-Calfee ridge
Miller Cr. ridge
Canoe Lake-Hoodoos
aSaddlle Mt.
aLittle Saddle Mt.
head of Flint Cr. to heads of Pelican, Timothy, Raven Cr.
Upper Lamar and Mirror Plateau mixed group areas
widely scattered north of Lamar Valley and on the Mirror Plateau
Cache Cr.
Miller Cr.
along the upper Lamar R.
Mixed group areas
Ponuntpa Hot Sp.
Heads of Opal and Flint Cr.
White Lake
Astringent Cr.
Upper Pelican Valley
Mouth Pelican Cr.
Lakeshore from Steamboat Pt. to the Yellowstone R.
Pelican aMushpots-Mudkettles
main valley from Pelican Sp. to Vermilion Sp.
aTurbid Lake area
athermal areas west of Astringent Cr.
ahead of Opal Cr.
ahead of Buffalo Fork of Timothy Cr.

Mixed group areas
East side of Yellowstone R.
Mud Volcano
Entire main valley
Beach Lake
Hayden Valley west of Yellowstone River:
southwest corner
northeast to junction Alum-Violet Cr.
west and southwest part inc. Highland Hot Sp.
upper Alum Cr.
aupper Trout Cr.
acentral valley
aupper Nez Perce Cr.
upper Spruce Cr. to Beach Lake and area
Hayden Valley Area mixed group areas
Cygnet Lakes
lower Sour Cr.
Arnica Cr.
Mixed group areas
Entire Firehole area
from Old Faithful through the Lower Firehole
Geyser Basin
Madison Jct. area
Smokejumper Hot Sp.
Firehole aupper Nez Perce Cr.
lower Nez Perce Cr.
entire Lower Geyser Basin
aMidway Geyser Basin
aBiscuit Geyser Basin
from Little Firehole Mdws. south across Summit lake area to west end of Pitchstone Plateau
aupper Boundary Cr.
Madison and Pitchstone Plateaus widely scattered
Bechler Ford
Dunanda Hot Sp.
Bechler Bechler Ford

aLess used.

Bison from both Pelican and Lamar Valley wintering populations ranged widely during summers on the Mirror Plateau and the Upper Lamar. Neckband observations and aerial counts showed that the two populations, except scattered bulls, combined on the east boundary for several weeks in 1967. Main use areas in the Upper Lamar region during the study period were the Cache-Calfee and the Miller Creek ridges from the east boundary down, and the series of meadows and parks on the east rim of the Mirror plateau extending from the head of Flint Creek to the heads of Pelican-Timothy-Raven creeks. At higher population numbers, as reports of the 1930s and 1950s indicate, herd-group use included most of Specimen Ridge north and west of Flint Creek as well.

To the north of Lamar Valley, where at present only bulls summer, past reports indicate small mixed herd groups on the upper Slough Creek meadows and the Buffalo Plateau (Jim Stradley, Dave Pierson 1968 pers. comm.). In August 1943, a report was received of an estimated 150 bison near Lake Abundance, just outside the northeast corner of the park.

Two small areas, separated by distance and topography from the four main valleys, have had small wintering mixed herd groups before the study period, according to recent records. Approximately 4050 bison wintered north of the Madison River on the flats just southwest of the Cougar Creek patrol cabin in 1955-56. Some were also seen in 1959. In the Antelope Creek basin on the northeast flank of Mount Washburn (no date), 45-70 wintered one year (Jim Stradley 1968 pers. comm.). There were no summer reports of groups north of the Madison River, although the bison wintering there in the 1950s may also have summered there.

Bulls were not distributed proportionately among the four main wintering valleys, as shown in Table 7. Most of the Mary Mountain area bulls wintered in Hayden Valley. Lamar Valley had more bulls than did the Pelican area. In summer, bull distribution was widespread. Animals were so scattered that preference for certain general areas by a large percent of the bulls was not observed.

Recent and present population distributions generally resemble those described by historical sources (Fig. 11). Winter distributions within the park approximate those of early times, as do summer distributions on the Mirror-Upper Lamar. Other summer distributions have changed most. The herd which summered north of the Lamar-Yellowstone rivers in historic times is now gone. The large numbers which once ranged the west side (Madison-Pitchstone plateaus) in summer are reduced to comparatively few. The segments of these two historical summer herds which wintered beyond the park boundaries are also gone. The recent Hayden Valley summering population seems larger than the historic population. Bison use of the valley may actually have increased, compared to former times, as the population became reestablished. Or the increase may not be real, but may result from incomplete historical information.

TABLE 7. Comparative distribution of bulls wintering apart from mixed groups.

LamarPelican Hayden ValleyFirehole
Year No.

Although present winter distributions seem little altered (except in total numbers) compared to historic times, the Firehole population may be larger now. Changes in the summering distributions of Hayden Valley to the east and the Madison-Pitchstone plateaus to the west may have caused changes on the Firehole, or unrecorded poaching in that wintering valley may have resulted in apparent early low numbers.

<<< Previous <<< Contents >>> Next >>>

Last Updated: 24-Jan-2005