Society for Range Management
Bison Use of Fire-Managed Remnant Prairies Author(s): Mario E. Biondini, Allen A. Steuter, Robert G. Hamilton Source: Journal of Range Management, Vol. 52, No. 5 (Sep., 1999), pp. 454-461 Published by: Allen Press and Society for Range Management Stable URL: http://www.jstor.org/stable/4003772 Accessed: 22/03/2010 00:25
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=acg.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
Allen Press and Society for Range Management are collaborating with JSTOR to digitize, preserve and extend access to Journal of Range Management.
http://www.jstor.org J. Range Manage. 52:454-461 September 1999 Bison use of fire-managed remnant prairies
MARIO E. BIONDINI, ALLEN A. STEUTER, AND ROBERT G. HAMILTON
Authors are professor, Department of Animal and Range Sciences, North Dakota State University, Fargo, N.D. 58105; director of science and stew- ardship, Nebraska, The Nature Conservancy, 431 North Maple Street, Ainsworth, Neb. 69210; and director of science and stewardship, Okla., The Nature Conservancv, P.O. Box 458, Pawhuska, Okla. 74056.
Abstract Resumen
This study was designed to: 1) compare the landscape dis- El objetivo de este estudio fue: 1) comparar el pastoreo de tribution patterns of bison on fire-managedprairie remnants bisontes en pastizales nativos de Oklahoma y Nebraska in the tallgrass (Oklahoma), and mixed prairie (Nebraska); manejadoscon fuego; y 2) determinarsi el fuego influencia el uso de diversos sites" site es una combinacion and 2) identify the extent to which fire and range site [topo- "range (range nespecificade topografiay suelos). edaphic classifilcationof the landscape] affect bison distribu- El experimento se condujo en dos lugares: la Niobrara tion patterns. Valley Preserve (1990-1996), y la Tall Grass Prairie Preserve This research was conducted at 2 sites: the Niobrara Valley (1993-1996). En ambos lugares, durante la epoca the crec- Preserve (1990-1996), and the Tallgrass Prairie Preserve imiento, los bisontes pastorearon de 1 a 3 afios seguidos sec- (1993-1996). At both preserves, bison selected burned areas ciones recientementequemadas y evitaron secciones no que- during the growing season for 1-3 years and mostly avoided madas o quemadas mas the 3 afios atras. La intensidad de old burns and unburned areas. There was an interaction uso fue influenciada por una interaccion entre el fuego y los between fire and range site in selection patterns. In the range sites. Los bisontes pastorearon livianamente los absence of fire, bison mostly avoided both the Choppy Sand Choppy Sand range sites en la Niobrara Valley Preserve y los range sites at the Niobrara Valley Preserve and Loamy Loamy Prairie range sites en la Tallgrass Prairie Preserve. Prairie range sites at the Tallgrass Prairie Preserve. When Los mismos range sites, sin embargo, fueron intensamente they were burned, however, these sites were highly selected. pastoreados despues de la quema. Durante el otofio y el The main difference in bison selection patterns between the invierno los bisontes en la Tallgrass Prairie Preserve pas- Tallgrass Prairie Preserve and the Niobrara Valley Preserve toreando intensament por 2 aniosseguidos las secciones que- was observed during the dormant season. In the Tallgrass madas, mientras que en la Niobrara Valley Preserve no se Prairie Preserve, burned sites continue to be preferredduring observaron diferencias en el uso de secciones quemadas y no the dormant period for an average of 2 years while in the quemadas.Esta disparidadse debio a: (a) la continuaciondel Niobrara Valley Preserve selections were random. These dif- crecimiento de los pastos en el otofio y parte del invierno en ferences can be explained by 2 mechanisms:the fall and win- la Tallgrass Prairie Preserve; y (b) la alteracion que el fuego ter re-growth of forage at the more southern latitude and the causo a la filsionomiade las praderas de la Tallgrass Prairie significant physiognomic changes that fire can cause in tall Preserve. Demostramos una interaccion entre el pastoreo de grass prairies. Our study documents a continuing interaction los bisontes y los procesos ecologicos generados por distintos regimenes de between the ecological processes of the fire regime and bison fuego y proveemos detalles tecnicos para el uso de fuego en estas praderas. distribution and abundance within 2 of the major prairie landscapes of the Great Plains, and provides critical details for understandingthis relationship.
DeVos 1969, Malin 1953), thus, are generally unknown. Key Words: Fire, grazing patterns, tallgrasss prairie, mixed Following the near extinction of bison, the Great Plains were prairie, range site. developed for intensive agriculture,transportation corridors, and cities. Bison herds are once again expanding, although this time on prairiefragments within the developed landscape. Plains bison (Bison bison bison) were extirpatedas a free- These remnant prairies are critical to conserving biological rangingspecies in the last 2 decades of the nineteenthcentury. diversity within the Great Plains (Samson and Knopf 1996). The factors that influenced pre-historic bison distribution The interruptionof landscape-scaleprocesses, however, con- (Hanson 1984, McHugh 1972), and the distributionof their tinues to be an overlooked mechanism reducing biodiversity impacts on the biotic and abiotic environment (England and (Leach and Givnish 1996). In commerciallivestock operations neither large grazers nor fire are managed, for the most part, Research was funded by grants from The Nature Conservancy Ecosystem to preserve or restore ecological processes. Instead, they are Research Program, USDA NRICGP 93-005 1, and the National Science managed to achieve individual species, or management unit Foundation DEB-9627928. Manuscript accepted 20 Oct. 1998. objectives.
454 Journal of Range Management 52(5), September 1999 Abiotic factors such as weather, Prairie preserves and native grassland (1990-1996), and the Tallgrass Prairie topography, soils, and distance from pasturesare relatively small considering Preserve (1993-1996). Both preserves water are suggested as the primary the mobility of bison. Most of the abiotic are owned and managed by The Nature determinantsof the large-scale distribu- factors associatedwith these prairiefrag- Conservancy. The Niobrara Valley tion patterns of large herbivores (see ments (e.g. climate, weather, topogra- Preserveis locatedin Brown,Cherry, and review by Bailey et al. 1996). At smaller phy, soils) remained scaled similar to Keya Paha counties, Nebr. on the north- scales, the animal's spatial memory their original condition. The fire regime, ern edge of the Sandhills(42? 46'N, 1000 regardingfactors such as forage compo- however, can be re-scaled through the 00' W). The Sandhills are an extensive sition, quantity and quality serve to use of prescribedburning. Localized ver- area of stabilized dunes dominated by improve foraging efficiency and define sions of the regional fire regime driven bluestem (Schizachyrium scoparium grazing patterns (Bailey et al. 1996). by a real-time fuel-forage interaction Michx., Andropogon gerardii Vitman) Biotic factors, therefore, should play a (Steuteret al. 1990) are implementedon and prairiesandreed (Calamovilfa longi- more significant role in defining bison several Great Plains preserves managed folia Hook.) (Shiflet 1994). The Niobrara distribution than abiotic ones within by The NatureConservancy. Bison have riverand valley are significantfeatures of fragmentedprairie remnantsthan on an been introducedas year-long, unsupple- the study site. The climate is continental unboundedprairie landscape. The modi- mentedresidents of these prairies. with an averageprecipitation of 559 mm fications of many ecosystems processes The native rangelands of the mixed- year', and mean daily temperaturesthat (net primary production, nutrient grass and tallgrass prairie have been rangefrom -5.7?C in Januaryto 240 C in cycling, etc.) by large grazersare widely highly fragmented by farmland. July. The experimentwas conductedin a recognized (Hobbs 1996). At moderate Although the ecology of cattle grazing 3,000 ha pasture established in 1988, stocking rates, large grazers do not on prairie fragments has been studied with a bison stocking rate of 1.0 AUM appear to influence, in the long term, intensively for 75 years [e.g Weaver and ha-'year'. An AUM is the amountof for- plant species assemblages within plant Clements 1938, Dyksterhuis 1949], age that a 454 kg bison consumes in 1 communities, or alter the trajectory of there have been no opportunities, until month (360 kg dry-wt., Heitschmidtand plant succession at the landscape level recently, to study the ecology of bison Taylor 1991). The bison pasture has 4 (Jefferies et al. 1994, Biondini et al. grazing on relatively large mixed and dominantrange sites (Voightlanderet al. 1998). Large grazers, however, have tallgrass prairies managed with fire. 1992) (Table 1). considerableshort-term effects on vege- During the last 20 years private bison The TallgrassPrairie Preserve is locat- tation at small scales. Examples include ranches have developed throughoutthe ed in the northernhalf of the Osage Hills, changes in bunchgrass size and tiller central and eastern Great Plains. Many Okla. (360 50' N, 96? 25' W). The Osage dynamics (Briske and Anderson 1990, of these operations are based on native Hills, an extension of the Flint Hills, are Pfeiffer and Hartnett 1995), plant prairie remnants of high conservation characterized by shallow to deep soils species diversity (Collins 1987, Hartnett value due to their rarity. Since bison over limestone and limy shale (often et al. 1997), and plant above-ground offer previously unavailable manage- exposed) and a well establisheddrainage biomass and live:dead ratio (Coppock et ment alternativesfor these native prairie system. The study site is dominatedby a al. 1983, McNaughton 1984). remnants, we have initiated studies to bluestem prairie (Shiflet 1994) with Great Plains land conversions and explore the contemporary response of savannatype vegetationoccurring in the managementhave probablycaused larg- these rangelands within the context of stream courses and in exposed bedrock er changes in landscape-scaleprocesses the evolutionaryfire regime. areas. The climate is continentalwith an than the replacement of bison grazing The objectives of this study were: 1) average precipitationof 877 mm year-', with cattle (Bos taurus)grazing (Hartnett to compare the landscape distribution and average monthly temperaturesthat et al. 1997). Yet, significant differences patterns of bison on fire-managed rangefrom a low of -3.3 C in Januaryto exist between the physiology, morpholo- prairie remnants in the Tallgrass a high of 33.30 C in July . The study site gy, social behavior, and foraging ecolo- (Oklahoma), and Mixed prairie was established in 1993 in a 1,970 ha gy of bison and cattle which may make (Nebraska);and 2) to identify the extent pasturewhich was subsequentlyexpand- bison more adaptedto some GreatPlains to which fire and range site [topo-edaph- ed to 2,575 ha in the Fall of 1995, and grasslands (Christophersonet al. 1978, ic units of the landscape] affect bison 2,875 ha in the Springof 1996. The bison Peters and Slen 1964, Plumb and Dodd distribution patterns. Results from this stockingrate, however, was kept constant 1993, Towne et al. 1988). Bison select study should provide part of the frame- at 1.2 AUM ha-'year-'. The bison pasture for the dominantperennial grasses form- work for placing the aboriginal land- has 6 dominantrange sites (Bourlieret al. ing the prairie matrix, show a strong scape-scale process of bison grazing 1979) (Table2). preference for areas with young within the context of contemporary regrowth following fire and grazing, prairieremnants. Fire Management avoid wooded or brokentopography dur- The study sites are managedwith pre- ing the breeding season, and spend rela- Materials and Methods scribed burning designed to mimic the tively little time around water sources seasonal distributionand returninterval (Coppock and Detling 1986, Plumb and of fires for the pre-Europeansettlement Dodd 1993, Shaw and Carter 1990, Site Description period (Steuter et al. 1990). Approxi- Steuter et al. 1995, Vinton and Hartnett The researchwas conductedat 2 sites: mately 13% of the Niobrara Valley 1992, Vinton et al. 1993, Coppedge et the Niobrara Valley Preserve Preserve and 20%oof the Tallgrass al. 1998).
Journal of Range Management 52(5), September 1999 455 Table 1. Niobrara Valley Preserve range site description, dominant vegetation, and estimated production.
Range Site Area Description DominantVegetation EstimatedProduction (%) (kg ha-') Sand 35 Deep, well drained,coarse Andropogonhallii, Panicum virgatum, 2600 texturedsoils on level to Schizachyriumscoparium, Calamovilfa gently rolling uplands longifolia, Eragrostis trichodes,and Bouteloua hirsuta.
Choppy Sands 40 Deep well drainedcoarse Andropogonhallii, Panicumvirgatum, 2400 texturedsoils on hilly Calamovilfalongifolia, Schizachyrium uplandswith stablized scoparium,Stipa comata, Eragrostis trichodes, dunes Bouteloua hirsuta, and Muhlengbergia pungens.
Shallow to 20 Moderatelyfine to coarse Boutelouagracilis, Bouteloua hirsuta, 1600 Gravel texturedsoils underlainwith Calamovilfalongifolia, Stipa comata, gravel and coarse sand at Andropogonhallii, and Schizachyrium depths 20cm to 50cm scoparium
Sandy Savanna 5 Level to very steep areas on Grasslandsinterspersed with maturetrees Variable breaksor riparianareas and brushes characterizedby grasslands interspersedby maturetrees and brush
Prairie Preserve pastures are burned sources, fuel conditions, fire intensity, mum threshold for fine fuel capable of each year to simulate the average fire- and rates of fire spread (Higgins 1986, carrying a fire is set at 1,000 kg ha-' as return interval of 7.5 and 5 years, Cheney et al. 1993). Areas to be burned estimated from Forest Service models respectively, estimated for the study are selected from landscape units that using local data (Burger and Rothermel sites in question (Wright and Bailey meet a minimum size (>15 ha) using a 1984, Andrews 1986). 1982). On both study sites, approxi- weighted random procedure. The mately 80% of the burns occur during weighting factor is fine fuel accumula- Bison Location the dormant season with the remainder tion immediately prior to burn date (the We recorded the landscape distribu- taking place during the growing season higher the fuel accumulationthe higher tion of "mixed" bison groups which to reflect seasonal differences in ignition the probability of selection). A mini- consist of adult females, juvenile
Table 2.. Tall Grass Prairie Preserve range site description, dominant vegetation, and estimated production.
Range Site Area Description DominantVegetation EstimatedProduction (%) (kg ha-') Claypan 6 Level to gently sloping soils, fine to Schizachyriumscoparium, Anidropogon 3000 medium-texturedtopsoils, impervious gerardii, and Panicum virgatum subsoils within 50cm of the surface.
Loamy Bottomland 5 Alluvial soils that are subjectto Panicum virgatum,Sorghastrum nutans, 6100 flooding (riparianzones and overflow and Andropogongerardii with mature areas) trees and shrubscompromise 20% of the vegetation).
Loamy Prairie 65 Upland soils more than 50 cm deep, Andropogongerardii, Panicum virgatum, 5200 with a loamy textureand permeable Sorghastrumnutans, and Schizachyrium subsoils. scoparium
Shallow Savanna 5 Loamy uplandsoils that are less than Maturetrees and shrubscompromise 20% 2100 50 cm deep with exposed ledge rock and of the vegetation with Schizachyrium very shallow soils intermixedwith scopariumand gerardii as the dominant deeper soils. herbaceousspecies.
Shallow prairie 16 Rocky and rough uplandsoils, less than SchiZachyriumscoparium, Bouteloua 2300 50 cm deep, slopes of less than 20% curtipendula,Andropogon gerardii, Sorghastrumnutans, and Panicum virgatum.
Sandy Savanna 3 Loamy or sandy texture uplandsoils SchiZachyriumscoparium, Andropogon 3300 more than 50 cm depth with gently to gerardii, and Sorghastrumnutans. steep slopes.
456 Journal of Range Management 52(5), September 1999 females and males, and bulls generally where Zoc/(2NR)is the Bonferroni adjust- under a null hypothesis that incorporates younger than 4 years old. During the rut, ed z value for the standardnormal distri- the knowledge that J number of bison older bulls were also partof these mixed bution; wi= ratio between the proportion have alreadyselected resourcej (expect- groups. The location of bison groups of bison using resource i and the one ed number of u+-J bison selecting each within the pasture was determined expected under the null hypothesis. of the remaining NR-1 resources). This throughoutthe year by regular (once or Deviations of selection ratios from the analysis was used to determine: (a) more per week) visual census conducted null hypothesis (wi?l with P X(cLxNR-ld_i (1) :5 50 a. where NR= number of categories 40 9 (burned sites by year, or range sites); 0 ui= the number of animals using resource i; u+= total numberof animals; 30 ji= proportionof resource i (proportion E of burnedor range site area i in the pas- ~ 0 ture). Interactionsbetween burned area 20 and range site selection were tested in a similar fashion with Chi square contin- 0 10 gency tables (Mosteller and Rourke 1973). If either test was significant (null 0 hypothesis rejected) we proceeded with 90 91 92 93 94 95 96 the following analysis. The 100(1-oc)%multiple confidence intervalfor selection ratios was calculat- ed as follows (Manly 1993): Fire Fire Fire Fire Fig. 1. Niobrara Valley Preserve. Aggregate bison selection of burned areas (sum of all CI(wi ) =?Z/,(2NR) ; gwhere = U= (2) burned areas) during the growing season. The bars represent significant departure (P<0.01) from null model predictions(bison use burned areas randomlyand in proportion to their area) Journal of Range Management 52(5), September 1999 457 80 200 1996 results are a clear example of fire 5 0 2~~~~~~~~~~ dominatingrange site effects. As shown 90 91 92 93 94 95 96 9100 in Fig. 3, bison in 1996 substantially 50 -~ un selected Choppy Sands, but when the 0 fire effect is taken out of their selection e(20) e 250 199 Burns *Bur19nnd196Bredaeas ((40) 1991 Burns ((0 patterns, there is no selection for 1(60) (100) Fire Choppy Sands (Fig. 4). 1 (80;(8) (150) 90 91 92 93 94 95 96 90 91 92 93 94 95 96 Non-GrowingSeason 250 25( Bison selection patterns during the .200 Fr 01994 Burns c2200- non-growing season were markedlydif- ;a 150 5 1991 and 1996 Burnedareas ~~~~~~~~~~~~~~~thatOverlap ferent from the ones observedduring the ;150 0 0 ~~~~~~~~~~~~~~~~~~~~Firegrowing season. There was no signifi- 550 2 100 cant effect of fire upon selection patterns for the entire 7 year period. Choppy 00 50 Sands were avoided only 2 out of the 7 0 years, while Sands were selected above 90 91 92 93 94 95 96 90 91 92 93 94 95 96 (100; null model predictionsonly once during Fig. 2. Niobrara Valley Preserve. Bison selection of individual burn units during the growing the same periodof time (Fig. 3). season. The bars represent signiricant departure (P<0.01) from null model predictions (bison use burned areas randomly and in proportion to their area). (A) 1991 burns; (B) 1993 burns; (C) 1994 burns; (D) 1991 and 1996 burned areas that overlap. Tallgrass Prairie Preserve GrowingSeason aggregate selection of burned areas gent upon the selection of burned areas Contrary to the selection patterns returnedto null model predictions. The within it. Burned areas within the observed at the Niobrara Valley 1996 changes in selection patterns Choppy Sands were consistently select- Preserve, there was not an aggregate occurred because, for the first time, ed by bison overridingthe overall avoid- selection of burned areas (sum of all bison began to select new burns and ance of the site when unburned(Fig. 3). burned areas) early in the study period. avoid old ones (Fig. 2). When the effect of burning was taken This result is probably due to the fact The 1991 burnswere selected by bison out of bison selection patterns,Choppy that while forage production in the for 3 consecutive years (Fig. 2), but Sands were either consistently avoided TallgrassPrairie Preserve was twice that avoided in 1996. The area burned in or followed null model predictions.The of the Niobrara Valley Preserve, the 1993 had been avoided by bison in 1992 but was highly selected when bumed in 1993 (Fig. 2). Unlike the 1991 burn, 150 60 however, the 1993 fire effect only lasted *Choppy Sands ]Sands *Choppy Sands ]Sands for 1 year primarilybecause of a switch in selection to the 1994 burn (Fig. 2). Selection of areas burned in 1994 had 100 40 0 followed null model predictions from 0 1991 to 1993, but were substantially selected in 1994 and 1995. As with the 1991 burn, however, bison avoided this D 20) la a50) ~~~Fire 10Fire areain 1996. The fire of 1996 overlapped 0 9 by 60% with the areabumed in 1991 and was substantially (> 200% above null z z model predictions)selected by bison. E l E The Choppy Sands range site was a. ~ ~ ~ ~ (0. 2 0 9 2 mostly avoided by bison by (an average FieFire Fire of 35% below null model predictions) from 1990 to 1993 but was highly Fig.3 sFire Fire (20) selected in 1996 (Fig. 3). Sand range in(50) T sites were selected above null model Fire A predictions only twice during the 7 years of the study. (4)89/90 91/92 93/94 95/96 The 1996 selection of the Choppy (0)90 91 92 93 94 95 96 90/91 92/93 94/95 Sands reflects the capacity of fire to override range site effects since 93% of Fig. 3. Niobrara Valley Preserve. Bison selection of Sands and Choppy Sand range sites for the 1996 burns occurredover this range both the growing and non-growing season. The bars represent significant departure site (Fig. 4). Fig. 4 shows the bison (P<0.01) from null model predictions (bison use range sites randomlyand in proportionto their area). (A) Growing season; (b) Non-growingseason. selection ratios of Choppy Sands contin- 458 Journal of Range Management 52(5), September 1999 300 were mostly avoided (40% reduction from null model predictions)(Fig. 5d). ! 250 Burned Non-Burned ~ ~ ~ ~ ~ ~ HArea WiArea t 200 Non GrowingSeason Non-growing season response to fire .L 150 was similar to that observed in the grow- 2 100 - ing season, but lasted for 2 years rather than 1 and departures from null model 2 50 E predictions were on average 60% lower (Fig. 5a-d). These patternswere substan- '.t 0 c 0 tially different from those observed at .!(50)- the Niobrara Valley Preserve where there was no response to fire during the (100) 93 94 96 non growing season. A similar pattern was observed vis a vis range site. Again Fig. 4. Niobrara Valley Preserve. Bison selection of Choppy Sands range sites as affected by the only significant interaction was fire. The bars represent significant departure (P Joumal of Range Management 52(5), September 1999 459 of forage quality and productivity with Preserve do not regrow sufficiently to ing by bison within tallgrass prairieand the grazing behavior of large herbivores attractbison during the winter (Pfeiffer mixed prairie, in the Great Plains. The by hypothesizing a direct link between and Steuter 1994). On the Niobrara powerful attraction of bison to burned herbivore aggregation and a substantial Valley Preserve spring burns have been patches within large heterogeneous spatial variability in forage quality or grazed more intensively during the pre- landscapes may be used to achieve sev- production. Our experiments replicated vious summer, or have a mature forage eral management objectives. Since the the model assumptionsof Fryxell (1991) quality more similarto unburnedareas. bison-fire managementregime results in by creating a complex spatial and tem- Collins (1987,1992) showed a sub- between-year rotations of grazing and poral mosaic of unburned and burned stantial decline in plant community spa- rest, ratherthan within-season rotations patches within a large landscape. tial heterogeneity coupled with an of grazing deferment, there are larger, Results at both sites support Fryxell's increase in the grass component with longer and possibly more useful con- (1991) predictionsby showing a consis- fire in tallgrass prairies. This physiog- trasts in standingcrop on different parts tent spatial aggregation of bison in nomic change should attractbison, since of the landscape for other wildlife recently burnedareas. they prefer to graze large, un-fragment- species to use. For example, species pre- The temporal differences that we ed, and relatively uniformplant commu- ferring high residual cover for nesting observed between the Niobrara Valley nities that are dominated by grasses [waterfowl and Henslow's Sparrow Preserve and the Tallgrass Prairie (Vinton and Harnett 1992, Plumb and (Ammodramus henslowii)] would be Preserve in the selection of burnedareas Dodd 1993, Vinton et al. 1993, Steuter undisturbed in areas unburned and highlight the significance of the forage et al. 1995). The subsequentbison graz- avoided by bison for several years. production- bison stocking rate - season ing forms a unique landscape patch in Species preferring short stature grass- of burn interactionon landscape hetero- which rhizomatus grasses are favored land for nesting, feeding, or lekking geneity. Althoughthe stockingrate on the over bunchgrasses (Pfeiffer and Steuter [Upland Sandpiper (Bartramia long- Tallgrass Prairie Preserve is 0.2 AUM 1994, Pfeifer and Hartnett 1995), and icuada), Chestnut-collared Longspur higher than on the Niobrara Valley forb composition and density changes as (Calcarius ornatus), Greater Prairie Preserve (1.2 vs. 1.0 AUM ha-' year-'), a function of season of burn. Chicken (Tympanuchuscupido)] would forage productionis on average 2 times Pre-fire results from the Niobrara be able to select recently burned and higher (4,420 kg ha-' vs. 2,200 kg ha-'). Valley Preserve suggested that range intensively grazed areas. In addition, Also, the average fire-returninterval on site would have a significant effect on forb-rich "food plots" are created by Tallgrass Prairie Preserve is 5 years, bison selection of burned areas (Steuter intensive bison grazing and wallowing while on the NiobraraValley Preserveit et al. 1995). It now appearsthat in both of late summer burn patches, which is 7.5 years. This results in a 7% larger the tallgrass and mixed-grass prairies, experience a short-termincrease in low acreage, with twice the post-fire fire can override range site effects. This succession plants such as Ambrosiaspp. regrowth, available to bison on the allows for a more even, or more In general, the almost exclusive grazing Tallgrass Prairie Preserve. At the bison clumped, distribution of bison use of graminoidsthat has been documented stockingrate used on the TallgrassPrairie across the landscape depending on the for bison should lead to an increase in Preserve, forage regrowthfollowing fire specifics of the fire regime. forb diversity and abundance. is more slowly and less completely Our results suggest that the details of The new growth from prescribed removed by grazing and it maturesmore the fire regime, and bison distribution burns could be used to draw bison away rapidly.This leads to a steeperdecline in and abundance within the tallgrass and from grazing sensitive resources, or forage quality on newly bumed areas of mixed prairie have a substantial effect attract them to accessible viewing, the TallgrassPrairie Preserve when com- on the spatial and temporal distribution round up, or harvest areas [an ages-old pared to the Niobrara Valley Preserve of habitat patches. These details were idea]. Areas burned to control woody (Tieszenet al. 1998). largely controlled by a sequence of plants, and selected by bison during the The attraction of bison to recently Plains Indian cultures during the spring shed of their winter coats, would burned sites during the dormant season Holocene (Hanson 1984, Higgins 1986). have additional small tree kill due to on the TallgrassPrairie Preserve, but not For example, although Shaw and Lee rubbingand horning. on the NiobraraValley Preserve, is like- (1997) report little bison use of the Furthermore,the early growth of these ly explained by 2 mechanisms: the fall southern Tallgrass prairie from 1806 to burnedareas should improve bison qual- and winter regrowth of forage at the 1857, summer burning by earlier ity for lactation and subsequentconcep- more southern latitude and the signifi- Tallgrass prairie inhabitants may have tion rates following winter nutrient cant physiognomic changes that fire can greatly increased winter bison use of stress. Fall and winter forage quality cause in tallgrass prairies. Although this prairie type. Our results re-enforce could be improved by late summer warm season [C41grasses dominateboth the notion that human management of burns in the southern tallgrass prairie, sites, temperaturesfollowing late sum- the fire regime and large grazers was a thus improving bison winter condition mer and fall burns are more conducive basic, but not necessarily static, compo- without supplementalfeeding. With the to forage regrowth on the Tallgrass nent of GreatPlains grasslands. proper stocking rate and fire regime, PrairiePreserve. A significant portionof planned grazing might be distributed this regrowth consists of cool season with much lower investment in cross- [C3] grasses and sedges (Coppedge Management Implications fencing. Less cross-fencing may result 1996, Coppedge et al. 1998). Late sum- in a more dynamic spatial and temporal mer burns on the Niobrara Valley Our research provides detail on how expression of plant communities. fire determines the distributionof graz- 460 Journal of Range Management 52(5), September 1999 LiteratureCited Canadian grasslands. J. Range Manage. Pfeiffer, K.E. and A.A. Steuter. 1994. 22:87-94. Preliminaryresponse of SandhillsPrairie to Fryxell, J.M. 1991. Foragequality and aggre- fire and bison grazing. J. Range Manage. Andrews, P.L. 1986. Behave: fire behavior pre- gation by large herbivores. Amer. Nat. 47:395-397. diction and fuel modeling system/bum subsys- 138:478-498. Plumb, G.E. and J.L. Dodd. 1993. Foraging tem, Part I. USDA-FS GTR INT 194. Hanson, J.R. 1984. Bison ecology in the ecology of bison and cattle on a mixed Bailey, D.W., J.E. Gross, E.A. Laca, L.R. NorthernPlains and a reconstructionof bison prairie: Implicationsfor naturalarea man- Rittenhouse, M.B. Coughenour, D.M. patternsfor the NorthDakota Region. Plains agement.Ecol. Appl. 3:631-643. Swift, and P.L. Sims 1996. Mechanismsthat Anthropologist29:93-113. Samson, F.B. and F.L. Knopf (eds.). 1996. result in large herbivore grazing distribution Hartnett, D.C., A.A. Steuter, and K.R. Prairie conservation: preserving North patterns. J. Range Manage. 49:386-400. Hickman. 1997. Comparativeecology of America's most endangered ecosystem. Biondini, M.E., B.D. Patton, and P.E. Nyren. native and introducedungulates, p. 72-101. IslandPress, Covela, Calif. 1998. Grazing intensity and ecosystem In: F.L. Knopf and F.B. Samson (ed.), Seastedt,T.R., J.M. Briggs, and D.J. Gibson. processes in a northern mixed-grass prairie, Ecologyand conservation of GreatPlains ver- 1991. Controlof nitrogenlimitation in tall- USA. Ecol. Appl. 8:469-479. tebrates. Ecological Studies Vol. 125. grassprairie. Oecologia 87:72-79. Briske, D.D. and V.J. Anderson. 1990. Tiller Springer-Verlag,Inc. New York,N.Y. Shaw, J.H. and T.S. Carter. 1990. Bison dispersion in populations of the bunchgrass Heitschmidt, R.K. and C.A. Taylor. 1991. movementsin relationto fire andseasonality. Schiz-achvriurnscoparium: implications for Livestock Production,p. 161-177. In: R.K. Wildl. Soc. Bull. 18:426-430. herbivory tolerance. Oikos 59:50-56. Heitschmidtand J.W. Stuth (eds.), Grazing Shaw, J.H. and M. Lee. 1997. Relativeabun- Britton, C.M. and A.A. Steuter. 1983. Management: An ecological perspective. dance of bison, elk, and pronghornon the Production and nutritional attributes of TimberPress Inc., Portland,Ore. southern Plains, 1806-1857. Plains tobosagrass following burning. Southwest Higgins, K.F. 1986. Interpretationand com- Anthropologist42:163-172. Naturalist 28:33-41. pendium of historical fire accounts in the Shiflet, T.N. 1994. Rangelandcover types of Burger, R.E. and R.C. Rothermel. 1984. northernGreat Plains. USDI Fish and Wildl. the United States. Soc. Range Mange. Behave: fire behavior prediction and fuel ServiceRes. Publ. 161. Denver,Colo. modeling system/fuel subsystem. USDA-FS Hobbs, N.T. 1996. Modificationof ecosystems Steuter, A.A., C.E. Grygiel, and M.E. GTR INT-167. by ungulates.J. Wildl.Manage. 60:695-713. Biondini. 1990. A synthesis approach to Bourlier, B.G., J.D. Nichols, W.J. Ringwald, Jefferies, R.L., D.R. Klein, and G.R. Shaver. researchand managementplanning: the con- P.J. Workman, and S. Clemmons. 1979. 1994. Vertebrateherbivores and northern ceptual development and implementation. Soil survey of Osage county, Oklahoma. plantcommunities: reciprocal influences and NaturalAreas J. 10:61-68. USDA-SCS. responses.Oikos 71:193-206. Steuter, A.A., E.M. Steinauer,G.L. Hill, P.A. Cheney, N.P., J.S. Gould, and W.R. Leach, M.K., and T.J. Givnish. 1996. Bowers, and L.L. Tieszen. 1995. Catchpole. 1993. The influence of fuel, Ecological determinantsof species loss in Distribution and diet of bison and pocket weather and fire shape variables on fire- remnantprairies. Science 273:1555-1558. gophers in a sandhills prairie. Ecol. Appl. spread in grasslands. Int. J. Wildland Fire Malin, J.C. 1953. Soil, animal,and plantrela- 5:756-766. 3:31-44. tions of the grasslands,historically reconsid- Tieszen, L.L., L. Stretch, and J. Vander Christopherson,R.J., R.J. Hudson, and R.J. ered.Sci. Monthly76:207-220. Kooi. 1998. Stableisotopic determination of Richmond. 1978. Comparative winter bioen- Manly, B.F. 1993. Resourceselection by ani- seasonaldietary patterns in bison at fourpre- ergetics of American bison, yak, Scottish mals: statisticaldesign and analysisfor field serves across the GreatPlains, p. 130-140. highlander and Hereford calves. Acta studies.Chapman & Hall,New York. In: L.R. Irby and J.E. Knight (ed.). Theriologica 23:49-54. McHugh, T. 1972. The time of the buffalo. InternationalSymposium on Bison Ecology Collins, S.L. 1987. Interaction of disturbances Universityof NebraskaPress. Lincoln, Nebr. andManagement in NorthAmerica. Montana in tallgrass prairie: a field experiment. Ecol. McNaughton, S.J. 1984. Grazinglawns: ani- StateUniversity, Bozeman, Mont. 68:1243-1250. mals in herds, plant form, and coevolution. Towne, G., T.G. Nagaraja, R.C. Cochran, Collins, S.L. 1992. Fire frequency and commu- Amer.Nat. 124:863-886. D.L. Harmon, C.E. Owensby, and D.W. nity heterogeneity in tallgrass prairie vegeta- Milchunas, D.G., A.S. Varnamkhasti, and Kaufman. 1988. Comparisonof ruminalfer- tion. Ecol. 73:2001-2006. W.K. Lauenroth. 1995. Forage quality in mentationcharacteristics and microbialpop- Coppedge, B.R. 1996. Range ecology of bison relationship to long-term grazing history, ulation in bison and cattle. Appl. Env. on tallgrass prairie in Oklahoma. Ph.D. Diss., current-yeardefoliation, and waterresources. Microb.10:2510-2514. Oklahoma State University, Stillwater, Okla. Oecologia101:366-374. Vinton, M.A. and D.C. Hartnett. 1992. 98 p. Mosteller, F. and R.E.K. Rourke. 1973. Effectsof bison grazingon Andropogonger- Coppedge, B.R., D.M. Leslie, and J.H. Shaw. Sturdy statistics: non parametricand order ardii and Panicum virgatumnin burned and 1998. Botanical composition of bison diets statistics. Addison-Wesley Publishing unburned tallgrass prairie. Oecologia on tallgrass prairie in Oklahoma. J. Range Company,Reading, Mass. 90:374-382. Manage. 51:379-382. Ohr, K.M. and T.B. Bragg. 1985. Effect of Vinton, M.A., D.C. Hartnett, E.J. Finck, and Coppock, D.L. and J.K. Detling. 1986. fire on nutrientand energy concentrationof J.M. Briggs. 1993. Interactiveeffects of fire, Alteration of bison and black-tailed prairie five prairiegrass species. PrairieNaturalist bison (Bison bison) grazing,and plant com- dog grazing interaction by prescribed burn- 17:113-126. munity composition in tallgrass prairie. ing. J. Wildl. Manage. 50:452-455. Ojima, D.S., D.S. Schimel, W.J. Parton, and Amer.Midland Nat. 129:10-18. Coppock, D.L., J.E. Ellis, J.K. Detling, and C.E. Owensby. 1994. Long- and short-term Voightlander, A.L., R.R. Hammer, R.R. M.I. Dyer. 1983. Plant-herbivore interac- effects of fire on nitrogencycling in tallgrass Zink, M.K. Babcock, C. Evens, M.D. tions in a North American mixed-grass prairie.Biogeochemistry 26:67-84. Patterson, and H. Schultz. 1992. Soil suvey II. prairie. Responses of bison to modifica- Peters, H.F. and S.B. Slen. 1964. Hair coat of Brown county, Nebraska. USDA-SCS. tions of vegetation by prairie dogs. characteristicsof bison, domestic x bison Weaver, J.E. and F.E. Clements. 1938. Plant Oecologia 56:10-15. hybrids,cattalo, and certaindomestic breeds ecology [2nd edition]. McGraw-HillBook Dyksterhuis, E.J. 1949. Condition and man- of cattle.Canadian J. Anim. Sci. 44:48-57. Company,Inc., NY USA. agement of range land based on quantitative Pfeiffer, K.E. and D.C. Hartnett. 1995. Bison Wright, H.A. and A.W. Bailey. 1982. Fire ecology. J. Range Manage. 2: l0zl1- 15. selectivity and grazing response of little Ecology: United States and Canada. John England, R.E. and A. DeVos. 1969. Influence bluestem in tallgrass prairie. J. Range Wiley andSons, New York. of animals on pristine conditions on the Manage.48:26-31. Journal of Range Management 52(5), September 1999 461