VARIATION IN HUMP MORPHOLOGY BETWEEN PLAINS BISON (Bison bison bison) AND WOOD BISON (Bison bison athabascae).
Wes E. Olson and
______
The variation in hump morphology between plains bison (Bison bison bison Linnaeus 1758) and wood bison (Bison bison athabascae Rhoads 1897) was examinedin 13 populations and sub-populations from across North America,to determine whether hump characteristics were sufficiently different to support subspeciation. Plains bison from Elk Island National Park, AB, the National Bison Range, MT, Yellowstone National Park, WY, Grand Tetons National Park, WY, Alaska Delta Junction, AK, Badlands National Park, NE, Fort Niobrara Wildlife Refuge NE, Wind Cave National Park and Custer State Park SD were examined, as were wood bison from the sub-populations Lake Clair and Central in Wood Buffalo National Park,NWT, Mackenzie Bison Sanctuary, NWT, and Elk Island National Park, AB. Data were collected from photographs to produce size-independent ratios, measurements of surface area and degrees of slope for four hump components. Hump characteristics were found to different enough to consistently and reliably determine the subspecies of a bison of unknown origin.
Key words: allometry, bison, hump, morphology, subspecies,
INTRODUCTION
The classification of North American bison has long been a contentious topic (Geist 1971; Krumbeigal and Sehm 1989; Reynolds et al 2003) with two opposing points of view. Many authors support the view that two subspecies of bison currently inhabit North America – plains bison (Bison bison bison Linnaeus 1758) and wood bison (Bison bison athabascae Rhoads 1897) Attempts to justify subspeciation are numerous (Anthony 1928, Radford 1911, and Richardson 1829 in Soper 1941; CITES; Frick 1937; Gates et al 2001; Geist and Karsten 1971; Kurten and Anderson 1992; Mitchell and Gates 2002; Rhoads 1897; Seibert 1925; Seton 1929; Scott 1913; Soper 1941; van Camp 1989; van Zyll de Jong 1986; van Zyll de Jong et al 1995; Wilson and Strobeck 1999; Wilson, Zittlau and Nishi 2002; Wilson, Olson and Strobeck 2002; Wilson and Zittlau 2004; and Olson 2005). Other authors present the argument that one subspecies of bison resides in North America, and that any variation within or between populations is related to environmental and habitat factors rather than true subspeciation (Graham 1923, Mackenzie 1801, and Pike 1882 in Soper 1941; Geist 1991; Geist 1996; Halbert 2003).
The designation of bison as subspecies became an issue in 1990 when Agriculture Canada put forth a proposal to slaughter all resident bison in around WoodBuffaloNational Park in northern Alberta (Connelly et al 1990, Environmental Assessment Panel 1990). The proposal was the first step in an attempt to eradicate Mycobacterium bovis and Brucella abortus (Gates et al 1997) from this population. The proposal met with considerable opposition (Geist 1991, McCormack 1992) and the proposal was subsequently abandoned. The work by Geist (1991) was a catalyst in the debate about bison subspeciation. This paper was in direct contradiction to an earlier paper (Geist and Karsten 1977) that described the physical characteristics of both plains and wood bison, and which supported subspeciation of North American bison. Geist (1991) later suggested that plains and wood bison were a single subspecies, and that the pelage and morphometric differences he observed earlier were the result of ecotypic confinement effects. To validate this assumption he presented photographs which showed similarities in phenotypic expression between plains and wood bison. He then represented those photographs as simplified line drawings to clarify the differences / similarities between 16 individual bison. These 16 bison represented ages described as “young” to “very old”. The photographs were taken during early summer through to late winter. This confused ontogeny and season with phenotype. The combination of small sample size, wide geographic distribution (from Texas to Wood Buffalo and the Hamburg Zoo), different ages sampled and different seasons of sampling, led to research conducted and presented in van Zyll de Jong et al (1995). This research standardized the age of bison sampled to breeding age males and females, and season to the period of the rut (mid-July to mid-August). While that research supported the differences between plains and wood bison and concluded that they were indeed two distinct subspecies, it did not examine the variation within those populations. As Geist (1991) stated, “… subspecies are distinguished by at least one consistent taxonomic difference.”
The issue of subspeciation received further attention in 2004 when the Committee on the Status of Endangered Wildlife in Canada (COSEWIC) listed plains bison as “Threatened” in Canada. As a result of this proposal, the Canadian Bison Association expressed concern regarding the potential negative impacts of the Threatened designation on the bison ranching industry in Canada. COSEWIC agreed to a moratorium on SARA listing, until January 1 2011. Subsequent to that date, the CBA is to have identified at least 1000 breeding agebison of conservation value (see Appendix 5.2for a discussion on Conservation Herds) to satisfy the minimum population requirements of COSEWIC. An ancillary result of this decision is the need to be able to recognize wood bison from plains bison, and from their hybrid cross, the Parkland bison.
We developed a methodology that can assign a adult breeding aged male bison, of unknown subspecific origin, to the correct subspecies.
RESEARCH QUESTIONS
Geist (1991) postulated that plains and wood bison are ecotypes, rather than true subspecies. He also stated that for a species to be a“true” subspecies, they must be differentiated by at least one consistent taxonomic difference. How long, and under what circumstances, does it take for an isolated ecotype of a species to evolve characteristics of behavior, taxonomy, and genetics distinct enough for it to be scientifically accepted as a true subspecies? Where is the evolutionary dividing line that says, this animal has attained the required characteristics that make it unique from its parent type? Given mankind’s intervention in the evolutionary process of bison and the interruptions it inadvertently caused, is it still possible to determine whether plains and wood bison were evolving towards subspeciation or is it too late?
Limited research has been conducted on the differences in skeletal morphometry between plains and wood bison with the specific goal of identifying post-cranial skeletal variation between the two types. One of the largest studies was conducted by van Zyll de Jong in 1985. He conducted a wide variety of measurements on the skull and post-cranial skeleton of plains bison (B.b. bison), wood bison (B. b. athabascae)and the Wisent (B.. bonasus) and this research helped to explain why bison are shaped the way they are. This anatomical exploration is further enhanced by the work of Guthrie (1990) who describes the ecology of the hump, its role and function as it relates to combat, locomotion and behavior.
All extinct and extant bison have 7 cervical, 14 thoracic, 5 lumbar vertebrae (Guthrie 1990) and 8 caudal vertebrae (Figure 1). It is the placement and length of the vertebral neural spinal processes which provide the foundation upon which the various hump muscles (the rhomboideus, splenius and spinalis muscles for example) and tissues (such as the nucal ligament) are laid and supported (Guthrie 1990, pg 138). These in turn are overlain with a layer of fat and then the hide with its secondary sexual pelage characteristics. Together, these structural elements produce the dorsal contour of the bison hump that is measurable on photographs.
For a cursorial species like the plains bison (Guthrie 1989, Geist 1991), the ability to travel long distances during seasonal migrations (McHugh 1958, Meagher 1986, Schult and Haugen 1979) is enhanced when the centre of gravity is located over a central pivot point. Bison have an exceptionally long scapula that permits long stride lengths (Guthrie 1989). With the highest point and consequently greatest mass, placed over the front legs, the plains bison is superbly adapted to traveling great distances,. The centrally placed hump aids in offensive and defensive agonistic behavior between rival bulls during the rut, and permits fast and agile defence responses against the large wolf populations which existed on the great plains (Bergon 2003, pg 116).
Wood bison living in the boreal forests of northern Canada did not make the long distance movements typical of their southern cousins (Graham 1923, Soper 1941, van Camp 1975). While wolves remain a significant predator on wood bison (Carbyn 1981, 1987, Fuller 1951, Gates and Larter 1990, Joly and Messier 1990) they probably never existed in the numbers reported historically for plains bison (Bergon 2003, MacEwan 1995, McHugh 1958), and perhaps the frequency of attack by wolves was less for wood bison than for plains, due to lower wolf densities. Bergmann’s Rule (1847), suggests that mammals increase in body size with increases in latitude; an adaptation to colder climates. Olson (2006) found that wood bison in Elk Island National Park were an average of 17 % heavier than plains bison in Elk Island. Van Zyll de Jong (1986) measured skulls from extant, historical through to the Holocene and Pleistocene bison. Using his data, wood bison skulls were an average of 13% to 14% larger than those of plains bison from the great plains and the eastern US respectively.These factors, combined with long winters, may have contributed to the development of a large, forwardly placed hump. The evolution of a massive, forwardly placed hump may be more suited to carrying weight and designed for efficient cratering / foraging through the long months of winter, than for agonistic or defensive purposes.
Three general questions were posed to determine whetherconsistent taxonomic differences exist between adult, breeding aged male plains and wood bison.
1. Are there inter- and intra-specific differences in the four angles of the hump?
2. Are there inter- and intra-specific differences in the area of each component of the hump?
3. Are there inter- and intra-specific differences in the relative proportions of the torso and legs of the body?
METHODS
Sampled bison populations.
The following 13 populations were selected primarily because a photographic collection acquired for the research of van Zyll de Jong et al. (1995) was in existence and could be employed for this research. Additional photographs were acquiredopportunistically and were added to the original photographic database.
PLAINS BISON
1. Alaska, Delta Junction (ADJ), Alaska.
The plains bison populations in Alaska are from a 1928 translocation from the National Bison Range, Montana in. At that time 23 bison were released along the Delta River, near the mouth of Jarvis Creek (Dubois and Rogers 2000). Through emigration and translocations, additional herds have been established at Copper River, Farewell and Chitna(Berger and Cunningham 1994, Boyd 2003, Halbert 2003). Only photographs of bison from the Delta River herd were used for morphometric analysis.
2. Badlands National Park (BLNP),Nebraska.
In 1963 three plains bison from Fort Niobrara, and 50 from Theodore RooseveltNational Park were reintroduced to Badlands National Park. Another 20 bison were added to the herd in 1983 from the Colorado National Monument population (Berger and Cunningham 1994, Boyd 2003, Halbert 2003).
3. Custer State Park (CSP), South Dakota.
The Custer State Park bison population was established in 1914 with the acquisition of 36 plains bison fromthe herds of Scotty Philips and Fred Dupree (Boyd 2003). The herd was supplemented during the 1950’s with a movement of about 800 bison from Wind Cave National Park into Custer State Park.
4. Elk Island National Park-Plains (EINP-P), Alberta.
The plains bison in EINP originated from the well documented Pablo – Allard herd (American Bison Society 1913; Bork 1990; Coder 1975; Corner 1958; Douglas 1909; Gates et al 1992; MacEwan 1995; Michiel 1990; Polziehn, Beech, Sheraton and Strobeck 1996; Reynolds and Parkinson 1983; Reynolds 1989; Seton 1926; Shackleton 1968; van Camp 1975; Wilson and Zittlau 2004). Between June and October 1907, 419 plains bison were transported north to the newly established BuffaloNational Park near present day Wainwright Alberta (Brower 2002; Fuller 2002). The new park was not fenced and ready for the bison so they were temporarily off-loaded into ElkIslandNational Park, where they were confined until October 1909 when BuffaloNational Park was ready for them. Attempts were made to capture the entire herd, but approximately 48 plains bison remained in ElkIslandNational Park and formed the foundation of the present population. No additional plains bison have been added to this population since the initial translocation(Boyd 2003). Population size has ranged from a high of about 2500 in 1936 to the present when a base population of about 350 is maintained (Blyth 1987; Olson 2004).
The population has a fairly broad genetic base, despite its several genetic bottlenecks.
5. Fort Niobrara Wildlife Refuge, (FNWR), Nebraska.
This population was established in 1913 with 6 bison from Nebraska, and was supplemented during latter years with bison from Custer State Park and the National Bison Range (Boyd 2003, Halbert 2003).
6. Grand Tetons National Park (GTNP), Wyoming.
The first bison to repopulate the region that is now GTNP came from Yellowstone National Park in 1948. A total of 20 bison were moved to the Jackson Hole Wildlife Park. These were supplemented in 1964 with the addition of another 12 bison from Theodore Roosevelt National Park.(Boyd 2003, Halbert 2003).
7. National Bison Range (NBR) Montana.
The National Bison Range population began with the arrival of 34 bison from the Conrad herd, 3 from the Corbin herd, and 1 bull from the Goodnight herd in 1908. No further additions to the population occurred until 1939 when 2 bulls were added from the 7-Up Ranch, in 1952 with 4 bulls from FNWR, 2 bulls from YNP in 1953 and 4 females in 1984 from the Maxwell State Game Refuge (Halbert 2003). The Elk Island plains bison population and the Delta River Alaska population were both established from the NBR herd, when it only contained animals of 1908 origin.(Boyd 2003, Halbert 2003).
8. Wind Cave National Park (WCNP), South Dakota.
The WCNP population was established in 1913 with 14 bison from the New York Zoological Park. These were supplemented in 1916 with an additional 6 bison from YNP. No further additions to the population have taken place.(Boyd 2003, Halbert 2003).
9. Yellowstone National Park (YNP), Wyoming.
By the end of the 19th century, the Yellowstone population contained the only free-roaming wild plains bison population in North America. By 1902 the population had been reduced to about 30 wild bison. The herd was supplemented in 1902 with 18 cows from the Pablo-Allard herd and with 3 bulls from the Texas Goodnight herd. (Meagher 1973, Boyd 2003, Halbert 2003).
WOOD BISON
10. Elk Island National Park-Woods (EINP-W), Alberta.
EINP wood bison originated from the NeedleLake – NyarlingRiver region of north-western Wood Buffalo National Park (WBNP). In 1958 a remote herd of bison was observed in the Needle Lake region, and due to their isolation by geographic distance and inhospitable terrain, combined with their wood bison phenotypic expression, the group was thought to be a remnant population of pure wood bison (Banfield and Novakowski 1960). Shortly after the discovery of this herd an outbreak of anthrax along the eastern portion of WBNP impelled conservationists to attempt to salvage this herd before it too was decimated by the disease. This led to the live capture and subsequent translocation of 23 wood bison from WBNP to EINP in August of 1965 (Blyth and Hudson 1987).
Despite rigorous testing for the presence of Bovine brucellosis and tuberculosis prior to and following their arrival in EINP, in 1968 the first confirmed case of TB was found. This discovery eventually led to the slaughter of all founder bison and the bottle-raising of their calves in an attempt to eradicated the disease. This was successful and by 1970 the herd was certified as disease free (Blyth, Gates and Reynolds 1991) and has remained so since that time (Olson 2007c).
To have contracted the disease in WBNP, the wood bison must have had physical contact with the diseased hybrids, and thus probably shared genetic material along with the disease. This realization led to a phenotypic survey of the EINP population (Geist and Karsten 1977) who then described seven traits which served to define the physical and pelage differences between plains and wood bison. While they concluded that “to a trained observer, the wood bison bull and cow differ significantly from those of prairie bison” it is generally accepted that the EINP wood bison must have some plains bison genetic introgression.
11. Mackenzie Bison Sanctuary (MBS), Northwest Territories
The MBS population was established in 1963 with the translocation of 18 bison from the Needle Lake region of WBNP, as part of the project described above. At the time of the capture of these bison it was assumed that they had not come into contact with the introduced plains bison or their descendant hybrids. The fact that the population has remained free of the introduced diseases bovine tuberculosis and brucellosis supports this. No further additions have been made to the herd (Gates et al 2001, Gates et al 1991).