This post introduces Jonathan James Farr and Clifford A. White’s recently published Diversity article, “Buffalo on the Edge: Factors Affecting Historical Distribution and Restoration of Bison bison in the Western Cordillera, North America.”
verb (used with object), buf·fa·loed, buf·fa·lo·ing. Informal.
-to puzzle or baffle; confuse; mystify: He was buffaloed by the problem.
-to impress or intimidate by a display of power, importance, etc.: The older boys buffaloed him.
The North American bison, commonly called “the buffalo” is an iconic species to many cultures. In Canada, the ongoing restoration of bison to Banff National Park is being followed by conservationists, First Nations, the tourism industry, among others. In this essay, I describe the past few decades of bison restoration in the mountain west, and my long involvement in an endeavor that seeks to link the thinking of Indigenous peoples, ecologists, historians, and many other research disciplines.
A Story of Four Maps
Map One: The First Map of Bison Extinction (Allen, 1876)
This research endeavor can be boiled down into understanding the background behind four maps. The first is a famous 1876 map by Joel Allen, a meticulous American scientist, showing the process of one of the greatest near-extinction ever-documented—the disappearance of wild bison from the North American continent. Allen consulted hundreds of historical texts and then-living observers to show the maximum extent of the buffalo across the continent, and the species retreat against the European onslaught. In the forward to his autobiography, the great biologist Henry Fairfield Osborne writes:
“The life and writings of Joel Asaph Allen have exerted so great an influence on the progress of ornithology and mammalogy in America that all who have the interests of these branches of science at heart, both in this country and abroad, will welcome this biographical and bibliographical volume… The biographical sketch was especially desired… because it sets forth so clearly the broad groundwork of travel, of field observation and of field record which has established a model for all modern American work on the birds and the mammals.”
These words ring true. Allen travelled with Louis Agassiz across Brazil, with George Custer up the Yellowstone, and, on horseback, he followed the tracks one of the last small herds of bison as they followed along an ancient movement route from the Great Plains, up the South Platte River, and through the Colorado Rockies to South Park in the center of the cordillera. With regards to his bison map of the buffalo’s shrinking range, Allen explains that the distribution of bison when Europeans first visited America “is still a matter of uncertainty, yet reliable data are sufficiently abundant to establish the boundaries of its habitat at that time with tolerable exactness.” For decades, Allen’s work stood the test of time. In the 1950s, Canadian Frank Roe, in a detailed volume of over 600 pages, painstakingly re-documented the record of early European bison observations. Roe’s map differs in only minor ways from Allen’s earlier plotting of historical bison range.
Map Two: Updating Allen’s Map
The second map is modern science’s most recent contribution updating that updates Allen’s analysis, a product of the International Union for the Conservation of Nature (IUCN) scientific working group. Instead of bison’s historic range being mainly limited to the Great Plains, the most recent work presents a map that suggests that at first European contact, bison could be found much more broadly across the continent. This paper uses IUCN guidelines to delineate an Extent of Occurrence for bison, and by including some scant evidence of bison bones in Alaska it suggests that in the 1500s the shaggy beast may have ranged from the Gulf of Mexico, right across to the Pacific and Arctic oceans. The story behind this science is convoluted, but ironically it would, through bison, link two of the most remote ecological research laboratories in North America—one at the windswept headquarters of Yellowstone National Park in Mammoth, Wyoming, and another on a wave-pounded point at Bamfield, British Columbia on the west coast of Vancouver Island.
Briefly told, this research on bison abundance and distribution had its roots in Yellowstone National Park’s infamous “natural regulation hypothesis.” (Huff and Varley 1999, Wagner 2006). This hypothesis arose in the 1960s from political and social pressure to stop elk and bison culls in Yellowstone, even as Park Rangers grappled with overabundant herbivore populations due to the previous removal of the ecosystem’s top predators (Indigenous peoples and wolves). Applied to Yellowstone, the natural regulation hypothesis proposed that the long-term ecological role of humans, wolves, and other large predators was irrelevant. These species “were non-essential adjuncts,” and the park’s ecosystem would “self-regulate” in their absence.
The untested concept provided the framework for bold new management directions in the world’s first national park. By the 1990s, the role of wolves in reducing elk numbers would be recognized, but restoring, or even understanding, the role of humans in the park’s ecosystem remained sacrosanct. Bison and elk numbers began to increase unhindered, and the only good fire was a non-human ignition caused lightning despite historical evidence of extensive seasonal burning from Indigenous people across North America. The park’s forage was eaten to the roots. The massive mid-summer burns of 1988 scorched the forests, but ironically didn’t have the fuel to carry across meadows once frequently burned by Indigenous residents. In short, Yellowstone’s ecosystem began to evolve well outside its long-term range of variability.
For bison, Yellowstone’s biologists had once recognized Allen’s range edge in rugged mountains. Long-time Yellowstone bison researcher Mary Meagher believed that “Prehistorically, YNP bison ranges were probably the “tips of the fingers” of seasonal migration from large source populations associated with expansive grasslands lying to the north, west and southwest around the Yellowstone Plateau (Interview with Mary Meagher, July 15, 2004, p. 79 in Gates et al. 2005). In contrast, motivated to support the “natural regulation hypothesis,” Yellowstone National Park historians Schullery and Whittelsey (2006) reviewed bison abundance in the mountains and plains of the Greater Yellowstone area. Re-interpreting Allen’s and other’s findings about the rarity of bison in the mountains concluded that “what the historical record does tell us is that bison were here, they were all over the place, they were abundant.”
To provide further scientific veneer to the new paradigm of natural regulation in the absence of human and other species influences, researchers began to estimate what numbers of bison were theoretically possible in the absence of human or other species influences before starvation or migration might limit populations. For Yellowstone, scientists predicted that the park could support at least 6000 bison before starvation or migration might limit the population. Moreover, some scientists believed that, at least for bison, the natural regulation hypothesis might have general application across western North America, and bison might have a had a much greater recent historical distribution and numbers in some areas of the west than what Allen had documented. The natural regulation hypothesis explicitly ignores the role of Indigenous people in shaping the vegetation, fire regimes, and large mammal populations of North America, including their impacts on bison.
Map Three: Challenging the Myth of Natural Regulation
The result was a new map of historic extent of occurrence recently published in September 2022 in the esteemed journal Ecology (Martin et al. 2022). Using liberal mapping extrapolations, but specifically recognizing that the “bison is not a marine mammal” the researchers inflated bison’s historic range to possibly including much of the Pacific and Arctic ocean’s coastline. In essence, an obscure and politically driven hypothesis formulated in a research lab at Yellowstone might now even influence scientific thinking across North America. Amusingly, the paradigm of “natural regulation” had even brought the probability of historic bison presence to Bamfield, one of Canada’s premier ecological research institutes lying on the remote west coast of Vancouver Island.
As one could surmise, this recent revolution in scientific thought and the challenge of the “natural regulation” theory to previous work on bison distribution by such greats as Allen and Roe, did not go unnoticed. In the 1980s, Charles Kay, a doctoral researcher at Utah State University began to compile the archaeological and historical records for wildlife for the Yellowstone region, and to measure the impacts of the now rapidly growing, “naturally regulated” elk and bison populations on plants (Kay 1994).
In the 1970s, I had taken an undergraduate field trip to Yellowstone and was entranced with the concept as “the way to go” for Canadian national parks. However, by 1985, I was the vegetation and fire management warden for Banff National Park. My fire history research in Banff revealed the importance of Indigenous burning. Moreover, it was clear that restoring the fire, in combination with intense wildlife grazing and browsing could eliminate many plant communities. Fire would top-kill the tall aspen and willow stems, and the flush of sprouts from their roots would be then browsed off by over-abundant herbivores. This ongoing field work in Yellowstone and Banff both contradicted predictions of “natural regulation” which, by this time had evolved from being an untested theory to now being a paradigm for park managers.
In 1994, now Banff’s ecosystem research manager, I contracted Charles Kay to extend his field work north into Canada. By then Kay was becoming infamous. His doctoral work had provided the ecological science basis for Alston Chase’s best-selling 1986 book “Playing God in Yellowstone: The Destruction of America’s First National Park.” Here Chase provided the political origin’s of “natural regulation” and in layman’s terms explained Kay’s results on how assuming away human’s long-term effects was an ecological fallacy with potential serious impacts. For Canada’s Rocky Mountain national parks, Kay’s research revealed similar evidence for the importance of indigenous peoples as ecosystem managers. In Canada, these findings were not as controversial as in the United States. The Canada Parks Act mandated that maintaining or restoring “ecological integrity,” or the characteristic long-term ecosystems was the “first priority” for national parks. This legislation did not discriminate between natural and cultural processes. Both were important, and inter-related.
An additional outcome of this ongoing research was to remind Parks Canada that bison had historically been a significant species present along the eastern slopes of the Rockies. However, small herds of bison here were “sink populations” at the edge of their range and would require an understanding of the cultural processes creating these conditions in the past for successful restoration today. By 2016, Parks Canada was acting on these recommendations and proposing to restore bison to Banff. But not surprisingly for a document prepared solely by ecologists, the environmental assessment documents for the restoration made no mention of the role of Indigenous people’s role as key parts of Banff’s ecosystem. Instead, tenets from the natural regulation paradigm were invoked, and restoration targets for over 500 bison were suggested based upon forage availability.
By then I had retired, and this scientific background for Parks Canada’s bison proposal struck me as highly questionable. However, having once been entranced myself by natural regulation, I could understand why Banff’s new generation of ecologists would be drawn to it. In addition, from a manager’s perspective, natural regulation means observation of ecological change with minimal management. This adheres to the standard principles of bureaucracy: 1) “Don’t do anything and you won’t do anything wrong”, and 2) Nobody moves, nobody get’s hurt” – good rules for bureaucrats but often not-so-good for maintaining or restoring native species.
For me there was no option but to it but to go back to work—this time as an independent researcher—to provide further contrary evidence to the myth of natural regulation.
For me there was no option but to it but to go back to work—this time as an independent researcher—to provide further contrary evidence to the myth of natural regulation. In 2007, Kay published another piece of fine work analysing the detailed journals of Lewis and Clark. He had mapped and quantified all 585 days of wildlife observations made during their expedition up the Missouri River, across the continental divide, out to the Pacific Coast, and then returning. Using his methodology, expanded to include plant or fish resource use and an index of Indigenous human abundance, I re-entered the Lewis and Clark journals into the latest Google Earth and Excel database format, and then, like Forest Gump, having crossed most the continent twice, I just kept going, following the routes of early European travellers, from east to west, and north to south, from Lake Winnipeg to Bamfield, and the Gulf of Mexico to Point Barrow.
By the spring of 2022 this culminated in Map 3 for the northwestern area of the continent showing the locations of over 10,000 daily or multi-day entries from the expeditions of such notables such as William Henry Kelsey (1692), La Verendrye (1738-43), Henday (1754-55), Samuel Hearne (1771-1774), Sir Alexander Mackenzie (1789-93), Peter Fidler (1791-1800), George Vancouver (1792-93), David Thompson (1796-1812) and Peter Ogden (1824-1828). Each location on the map is linked to the Western North America Journal Observations database with information numerous wildlife, plant and fish resources used by Indigenous peoples and early European travellers. Indeed, the dataset was so comprehensive that it was even used by the IUCN working group to plot Map 2, although this team then added additional datapoints that did not come detailed analysis of historical journals.
The transcontinental historical journal dataset provided unique opportunities to examine historic bison distributions, especially in relation to the role of Indigenous people. One hypothesis that this dataset provided the opportunity to test, initially proposed by zoologist William T. Hornaday in 1889, is that bison populations west of the Rocky Mountains were limited in part by heavy hunting pressure from Indigenous peoples.
“It is probable that had the bison remained unmolested by man and uninfluenced by him he would have crossed the Sierra Nevada and the Coast Range and taken his abode in the fertile valleys of the Pacific Slope”William T Hornaday, 1889
Map Four: The Latest Map of Historical Bison Range
Over the past several decades, numerous scientists have attempted to attribute the apparent western edge of bison distribution to hunting pressure, deep snow, lower quality vegetation, and discontinuous habitat. Fortunately, the latest quantitative techniques for “big data” provide opportunities to test these hypotheses, and, as a bonus, to generate the Map 4 which shows the latest map of historic bison range.
For this work, Jonathan Farr from the University of Alberta took the lead, and, in November 2022, our paper summarizing this work was published in the journal Diversity. We recognized that finding the extent of occurrence for a species, as attempted by the IUCN working group is an important first step for understanding where species may have historically occurred. However, world leading experts in species distribution mapping have recently pushed for species conservation efforts to focus on identifying areas of habitat rather than simply using the IUCN’s current protocols for identifying extent of occurrences or area of occupancy. Therefore, we first built an area of habitat map for bison to provide an idea of historic bison range that would be more informative than simply drawing a straight-line polygon around potential bison occurrence points. With a process called Maximum Entropy Modelling (MaxEnt), we used satellite collected data on climate, land cover, and topographic variables with the historical journal observation dataset to estimate bison habitat suitability across Northwest North America. This method is more robust to outlier data points and is generally considered the gold standard for species distribution modelling. The method provides a more complete picture of where bison were likely to have occurred prior to their near extinction.
With a process called Maximum Entropy Modelling (MaxEnt), we used satellite collected data on climate, land cover, and topographic variables with the historical journal observation dataset to estimate bison habitat suitability across Northwest North America.
We were reassured that our Map 4—generated from modern ecological modelling techniques—could independently replicate Allen’s research from nearly 150 years ago and provide a much more believable picture of historic bison range. However, we also recognized in our paper that the rapid bison population growth from ongoing restoration projects in areas that we identify as marginal habitat clearly indicates that it is not bison habitat per se that likely limited bison range. More likely, the variables that we measured are just correlated with historic bison abundance, as their real influence was probably how, in the past, these variables influenced the abundance and hunting effectiveness of bison’s primary predator—humans. Archeological studies in Southwest Alberta and Northwest Montana have shown that Indigenous groups like the Blackfoot were especially effective at hunting bison in the rugged terrain with dozens of harvesting sites identified to date. In the narrow valleys, deep snows, and forested areas west of the Rockies, which we characterized as having low bison habitat suitability, bison range limitation was likely compounded by the fact that that people were easily able to limit bison abundance through hunting.
We were partially able to test this idea because the historical journal dataset included information on whether journalists had encountered people along their journeys, which provides some idea of the historic distribution of people in the area (Map 3). By using statistical modelling to relate this to bison probability of occurrence, we found evidence that people and bison were inversely correlated. High numbers of people were encountered in areas with low bison habitat suitability, especially west of the Rocky Mountains where rich salmon streams sustained many Indigenous peoples for millennia.
The IUCN bison working group’s mapping of potential historic range of bison to include Bamfield on the Pacific Coast is indeed a bold attempt to extend the application of the natural regulation hypothesis to its limits of believability. To evaluate its application to the area, we are fortunate to have extensive traditional knowledge from numerous First Nations, the historical observations of dozens of early European travellers, and, through archaeology, the ability to verify and extend these sources far back into the past. From just 50 km north of Bamfield, we even have Jewitt’s account of living with the Nuu-chah-nulth of Nootka Sound from 1803 to 1805, and specifics on their annual round of fishing, hunting, aquaculture, and use of plants.
This type of information is consistent up and down the whole west coast and inland along salmon-bearing streams in the northwest, or streams with irrigation agriculture potential in the southwest. These are descriptions of dense human populations, well-adapted to their local environment, and with regional trade routes extending well into the interior of the continent. For large ungulates, this trade between coastal, riverine, and interior peoples limited their density, or even their occurrence as they were hunted for robes, hides, and fat. Bison, as prime source of these resources was especially susceptible. This has long been obvious to bison historians:
“Even before the advance of the white men into the Rocky Mountain territory, the westward advance of the buffalo must have been much impeded by the ‘economic pressure’ of the Indian tribes beyond the actual buffalo range. For many Indians journeyed through the passes to procure bison meat and hides, either by hostile forays or by trade. This is attested by the earliest (European) observers and by many others and was clearly a long-established process.” (Frank Roe 1972, p. 259, emphasis added).
From an ecological perspective, determining the degree that Indigenous homelands should now be “naturally regulated” is a critical scientific and management question. National parks like Banff and Yellowstone protect some of the world’s most treasured and diverse ecosystems. As shown by images of the Madison River near West Yellowstone, ongoing high densities browsers and grazers, first elk, and now bison can irrevocably alter the habitat for hundreds of species, from grass to trees, fish to songbirds. The Society for Range Management has routinely criticized Yellowstone’s managers for maintaining high numbers of bovids—in this case bison, not cattle–in constrained high elevation valleys throughout the year (Mosely et al. 2018). This is a no-no for ranchers who preserve their forage productivity by moving their cattle to low elevations in winter, and even fencing stream-side zones. Banff now faces a similar issue. Under natural regulation management, the park’s recently restored bison herd could number over 400 animals within a decade, well more than the small bison herds sighted by historical travellers. For both parks, potential invasions by exotic species tolerant of high grazing pressure and altered stream processes due to loss of beaver could potentially damage unique ecosystems beyond ecological recovery.
Streamside willow, poplar and aspen along the Madison River in Yellowstone National Park on September 30, 1938 (from an early Kodachrome slide, Cushman Collection, Indiana University Archives P15797), and a mixed bison herd (bulls, cows, calves) along the river in the fall of ca. 2000 (US National Parks Service YELL-16356). Hunting by Indigenous peoples once kept elk and bison densities low here, but in recent decades high numbers of these animals have greatly altered this long-evolved, high biodiversity ecosystem. Beaver, once important here are now very rare.
The Upshot on Being “Buffaloed”
With a growing body of evidence suggesting that Indigenous people played a key role in shaping North America’s ecosystems, perhaps ecologists will eventually accept that the “natural regulation” paradigm is tragically flawed. This will be a difficult transition. For over a century, North American ecological researchers have sought out “pristine ecosystems” where they assume long-term human influences are negligible and can be ignored or assumed away. Accepting that human hunting, gathering, and culturing practices have influenced even the continent’s most remote landscapes will not come easy.
For over a century, North American ecological researchers have sought out “pristine ecosystems” where they assume long-term human influences are negligible and can be ignored or assumed away. Accepting that human hunting, gathering, and culturing practices have influenced even the continent’s most remote landscapes will not come easy.
As a sad, and pressing example, our generation is witnessing the demise of caribou, which maybe the northern equivalent to bison in its ability to herd in great numbers, or space-out in low densities depending on ecological and cultural conditions. Scientists researching the demise of caribou generally approach this from a natural regulation perspective, and rarely consider the hypothesis that long-term Indigenous cultural practices sustained caribou, and their removal is the primary cause of its demise. So while our modern society focuses on the usual culprits—timber harvesting, oil and gas development—the reality that caribou often disappear first in national parks, even large national parks, remains an enigma. Ultimately, the most unfortunate outcome of promoting the natural regulation paradigm is to the Indigenous cultures themselves.
Its difficult to believe that members of First Nations near Bamfield, or, for that matter, all along the western edge of the continent will do more than have a deep laugh if they see a map that shows that bison historically walked in the Pacific Ocean. But what about those peoples further inland? Their ancestors likely played a critical role in delimiting bison range for millennium. Historians are searching for the real story behind bison, even as Indigenous nations rekindle traditional knowledge and find their footing in political, public, and scientific arenas that affect wildlife management. One needs to look no further than the signatories of the Buffalo Treaty, an Indigenous-led effort that is pushing to restore free-roaming bison across their historic range. Of over 40 signatories, all but three are located within Allen’s historic map of bison occurrence. Indigenous nations seem to understand where bison were, and where they weren’t, and even the quirk of popular ecological science won’t leave them “buffaloed” by an image or map suggesting that bison historically walked on the rocky Pacific shores at Bamfield.
My latest technical report describing the edge of bison (and caribou) range is available HERE.
The historical journal observation Excel database and Google Earth map is updated regularly, and is available HERE.
Allen, Joel Asaph. 1876. The American Bisons, Living and Extinct. Vol. 10. Harvard College Memoirs of the Museum of Comparative Zoology 4. Cambridge, MA: University Press, Cambridge.
Chase, A. 1986. Playing God in Yellowstone: The Destruction of America’s First National Park. San Diego: Harcourt Brace.
Farr, Jonathan James, and Clifford A. White. 2022. “Buffalo on the Edge: Factors Affecting Historical Distribution and Restoration of Bison bison in the Western Cordillera, North America.” Diversity 14 (11): 937. https://doi.org/10.3390/d14110937.
Gates, C. Cormack, Brad Stelfox, Tyler Muhly, Thomas Chowns, and Robert J. Hudson. 2005. The Ecology of Bison Movements and Distribution in and Beyond Yellowstone National Park. Calgary, AB: Faculty of Environmental Design, University of Calgary.
Hornaday, William T. 1889. “The Extermination of the American Bison, with a Sketch of Its Discovery Life History.” In Report of the United States National Museum under the Direction of the Smithsonian Institution, 1887, 369–548. Washington DC.
Huff, Dan E., and John D. Varley. 1999. “Natural Regulation in Yellowstone National Park’s Northern Range.” Ecological Applications 9 (1): 17–29.
Kay, Charles E. 1994. “Aboriginal Overkill: The Role of Native Americans in Structuring Western Ecosystems.” Human Nature 5: 359–98.
Kay, Charles E. 2007. “Were Native People Keystone Predators? A Continuous-Time Analysis of Wildlife Observations Made by Lewis and Clark.” Canadian Field Naturalist 121: 1–16.
Martin, Jeff M., Rachel A. Short, Glenn E. Plumb, Lauren Markewicz, Dirk H. Van Vuren, Bradly Wehus-Tow, Erik Otarola-Castillo, and Matthew E. Hill Jr. 2022. “Integrated Evidence-Based Extent of Occurrence for North American Bison (Bison bison) since 1500 CE and Before.” Ecology, e3864–e3864.
Mosley, Jeffrey C., Joseph Fidel, Harold E. Hunter, Peter O. Husby, Charles E. Kay, John G. Mundinger, and Ryan M. Yonk. 2018. “An Ecological Assessment of the Northern Yellowstone Range: Introduction to the Special Issue.” Rangelands 40 (6): 173–76.
Roe, Frank G. 1972. The North American Buffalo. Toronto: University of Toronto Press.
Schullery, Paul, and Lee H. Whittlesey. 2006. “Greater Yellowstone Bison Distribution and Abundance in the Early Historic Period.” In Greater Yellowstone Public Lands. Vol. 8. Proceedings of the Eighth Biennial Scientific Conference on the Greater Yellowstone Ecosystem. Mammoth, Wyoming: Yellowstone National Park.
Wagner, Frederic H. 2006. Yellowstone’s Destabilized Ecosystem: Elk Effects, Science, and Policy Conflict. New York: Oxford University Press.