The Keystone Role of Bison in North American Tallgrass Prairie
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
The Keystone Role of Bison in
North American Tallgrass Prairie
Bison increase habitat heterogeneity and alter a broad array of
plant, community, and ecosystem processes
Alan K. Knapp, John M. Blair, John M. Briggs, Scott L. Collins, David C. Hartnett,
Loretta C. Johnson, and E. Gene Towne
T
hroughout the history of Great range (Samson and Knopf 1994).
Plains grasslands, North The near-simultaneous reduction in
American bison (Bos bison, Ungulate grazing herbivore abundance and grassland
also known as Bison bison; Jones et extent left little opportunity to assess
al. 1992) and other large herbivores
activities and fire are bison–tallgrass prairie interactions.
were abundant and conspicuous com- key to conserving and Today, thanks to conservation
ponents of the biota (Wedel 1961, efforts (Berger and Cunningham
Stebbins 1981). Many of the earliest restoring the biotic 1994), bison numbers in the Great
herbivores, particularly those that Plains have rebounded (to approxi-
were primarily browsers, are now integrity of the remaining mately 150,000), and significant
extinct, but their consumption of public and private herds are main-
woody vegetation is thought to have tracts of tallgrass prairie tained in several mixed- and short-
played a critical role in the post- grass prairie preserves and ranches.
Pleistocene rise of the grassland steppe, some have speculated that a It is from these semi-arid grasslands,
biome and the subsequent increase greater density of bison could be many of which escaped cultivation,
in bison populations (Axelrod 1985, supported in eastern tallgrass prai- that the most extensive knowledge
Hartnett et al. 1997). Indeed, the ries than elsewhere in the plains of bison–grassland interactions has
large herds of bison encountered by (McHugh 1972). It is unfortunate, been generated (Peden et al. 1974,
early Europeans on the Great Plains then, that despite the historic abun- Coppock et al. 1983). By contrast,
were likely the result of the rapid dance of bison in tallgrass prairies, the current understanding of tallgrass
early-Holocene proliferation of this their ecological effects in these mesic prairie structure and function has
ungulate into a relatively young and grasslands are poorly understood. been developed almost exclusively
expanding “treeless” gras sland Knowledge of the bison’s role in from studies of ungrazed tracts or
biome (Stebbins 1981, Axelrod tallgrass prairies is lacking because from sites grazed by domestic cattle
1985). In the most productive re- the extent of this grassland and the (Risser et al. 1981, Collins 1987, Howe
gions of the Great Plains, the eastern abundance of its dominant ungulate 1994, Leach and Givnish 1996). Only
tallgrass prairies, abundant bison have declined dramatically and in recently have bison been reintro-
herds were noted by early explorers tandem over the last 150 years. Al- duced to tallgrass prairie sites that
(Shaw and Lee 1997). Although herds though there is debate over the num- are large enough to assess both their
were larger in the western shortgrass bers of bison inhabiting the Great influence on other biota and ecosys-
Plains before the 1800s (estimates tem processes, as well as their inter-
Alan K. Knapp, John M. Blair, John M. range from 30 million to 60 million; actions with other important fea-
Briggs, Scott L. Collins, David C. Hartnett, McHugh 1972, Flores 1991), it is tures of these grasslands, particularly
and Loretta C. Johnson are professors, well documented that from 1830 to fire (Collins et al. 1998, Coppedge
and E. Gene Towne is a research associate, 1880 the slaughter of bison in the and Shaw 1998, Knapp et al. 1998b).
in the Division of Biology, Kansas State Great Plains reduced their numbers The Konza Prairie Research Natu-
University, Manhattan, Kansas 66506. to an estimated low of a few thou- ral Area in the Flint Hills of north-
Collins is also an adjunct professor with
the Department of Zoology, University of
sand individuals. Widespread culti- eastern Kansas is the largest tract of
Maryland, College Park, MD 20742, and vation of the plains, which accompa- unplowed tallgrass prairie (3500 ha)
a program director in the Division of En- nied the near extirpation of the bison, in North America dedicated to re-
vironmental Biology, National Science reduced the once-vast tallgrass prairie search (Knapp et al. 1998b). Konza
Foundation, Arlington, VA 22230. © 1999 (approximately 68 million hectares) Prairie was one of the original sites
American Institute of Biological Sciences. to less than 5% of its presettlement selected in 1981 for inclusion in the
January 1999 39
Since that time, the herd has been
maintained at approximately 200
individuals, who have had unre-
stricted access to a 1012 ha portion
of the landscape. Within this area
are 10 watersheds that are subjected
to different frequencies of late-spring
prescribed fire. The target animal
density of 200 animals was selected
so that approximately 25% of
aboveground primary production is
consumed annually. This consump-
tion rate is approximately half that
of tallgrass prairie managed for do-
mestic cattle. The bison herd at Konza
Prairie is not provided supplemental
feed in winter, nor is it actively man-
aged. Thus, this herd provides an
opportunity to document the impact
of bison on a native tallgrass prairie
ecosystem.
Although research on bison–
tallgrass prairie interactions began
soon after bison were reintroduced
to Konza Prairie, comprehensive
studies spanning scales from the leaf
to the landscape level began in the
early 1990s. In this article, we pro-
vide the first synthesis of these re-
search efforts, with the goal of high-
lighting the keystone role (sensu
Power et al. 1996) that bison played
in the tallgrass prairies of the past.
Within this overview, we address
two general questions: What are the
direct and indirect effects of bison
on patterns and processes in tallgrass
prairie? What factors influence the
spatial and temporal patterns of graz-
ing activities by bison? Perhaps as
important as addressing these ques-
tions, the data that we present indi-
cate that large-ungulate herbivory
Figure 1. Grazing by bison increases spatial variability in tallgrass prairie. Immediately can, and should, play a key role in
after fire (April) in grazed watersheds on the Konza Prairie, a native tallgrass prairie in the management and conservation
northeastern Kansas, the uneven consumption of fuel by fire in grazed areas (upper of the remaining tracts of this once-
photo) is caused by the patchiness of bison grazing the previous year. Dark areas widespread grassland.
indicate sites where bison did not graze or where grazing was light, and fuel loads were
therefore sufficient to carry a fire through the patch. Lighter areas are patches that were
grazed repeatedly by bison, such that fuel loads were reduced and the impact of fire was Bison grazing activities
minimal. Later in the season, bison graze preferentially in recently burned watersheds and plant responses
(lower photo). In this late spring (May) view of a burned watershed, unburned
watersheds can be seen in the background. Bison at Konza Prairie have access to 10 Like all large herbivores, bison do
watersheds that differ in fire regime. not graze indiscriminately across the
landscape or even within a local area
(Senft et al. 1987, Wallace et al.
National Science Foundation’s Long variability (Knapp and Seastedt 1995). Rather, they graze in two
Term Ecological Research (LTER) 1986, Knapp et al. 1998b), but a patterns in tallgrass prairie, creating
program, and ongoing experimental majority of these data were collected both distinct grazing patches (typi-
treatments on some parts of the site in the absence of bison grazing. cally 20–50 m2 in Flint Hills tallgrass
date to 1972. Past syntheses of re- In 1987, 30 bison were reintro- prairie; Catchpole 1996) and more
search at Konza Prairie have focused duced to Konza Prairie, and the herd extensive grazing lawns (larger than
on the effects of fire and climatic was allowed to increase until 1992. 400 m2; McNaughton 1984). In both
40 BioScience Vol. 49 No. 1
cases, bison revisit grazed sites Figure 2. Response of tallgrass
throughout the season, such that re- prairie vegetation to bison her-
peated defoliation of grazed plants is bivory at three temporal scales.
the norm and relatively sharp bound- (a) Instantaneous leaf-level pho-
aries between grazed and ungrazed tosynthetic responses of Andro-
pogon gerardii to grazing (Erin
vegetation become evident (Figure
Questad and Alan K. Knapp, un-
1). Bison are primarily graminoid published data). (b) Relative
feeders and consume higher propor- growth rates of tillers of A.
tions of the dominant grasses than gerardii in sites that differ in graz-
would be predicted based on grass ing history (data from Vinton
availability in the landscape (Peden and Hartnett 1992). Both un-
et al. 1974, Van Vuren and Bray grazed tillers and grazed tillers
1983, Steuter et al. 1995). Bison tend with a history of grazing had
to avoid forbs and woody species, similar growth rates, whereas
which usually constitute less than grazed tillers without a history of
grazing had significantly greater
10% of their diet. Thus, within a growth rates. (c) End-of-season
bison grazing area, forbs are often above ground biomass in
conspicuously left ungrazed and are ungrazed sites that differ in graz-
surrounded by grazed grasses (Fahne- ing history (data from Knapp et
stock and Knapp 1993, Damhoureyeh al. 1998a). Asterisks in panels (a)
and Hartnett 1997). and (b) indicate significantly dif-
Preferential grazing of the domi- ferent values (P < 0.05) for com-
nant grasses by bison sets the stage parisons at specific dates; in panel
for significant alterations in com- (c) they indicate comparisons
petitive interactions among the C4 within growth forms.
grasses and the C3 forbs. Such shifts
are important for plant community averaged 53% higher in
structure because in ungrazed and grazed tillers (individual
frequently burned prairie, a small grass stems) than in ungrazed
group of grass species (Andropogon plants, with a maximum
gerardii, big bluestem; Sorghastrum stimulation of 150% (Figure
nutans, Indian grass; Pani cum 2a). Mechanisms for this en-
virgatum, switchgrass; and Andro- hancement of photosynthe-
pogon scoparium, little bluestem) sis include increased light
account for most biomass, density, availability and reduced wa-
leaf area, and resource consumption ter stress for all species in
(Knapp 1985, Briggs and Knapp grazed patches (Fahnestock
1995). However, it is the species- and Knapp 1993) and greater
rich forb component (more than 350 tissue nitrogen concentration
species are recorded on Konza Prai- in A. gerardii leaves as nitro-
rie; Freeman 1998) that is critical for gen is reallocated from roots. grazing history on productivity in
the maintenance of high levels of These potential compensatory in- tal lgrass prai rie by measuring
biotic diversity in tallgrass prairie creases in photosynthesis after graz- aboveground primary production in
(Gibson and Hulbert 1987, Glenn ing may be augmented by the trans- a number of sites with different graz-
and Collins 1990, Turner et al. 1995). location of carbon reserves from ing histories. They found that com-
Thus, by grazing on grasses and al- belowground to aboveground tissues. pensatory regrowth of biomass oc-
lowing forbs to flourish, bison have Vinton and Hartnett (1992) found curred in sites with little history of
the potential to significantly influ- that in the first year of grazing, grazing but not in sites that had been
ence biodiversity in these grasslands growth and biomass of grazed A. grazed heavily in previous years.
(Collins et al. 1998). gerardii tillers had completely com- Some researchers have argued that
The short-term effects of bison pensated for the loss to grazing by the inability of tallgrass prairie
herbivory on the most abundant prai- season’s end. But after several years grasses to compensate for the bio-
rie grass, A. gerardii, are different of grazing, the ability of tillers to mass lost in frequently grazed areas
from the long-term effects. At the compensate for lost tissue was re- is evidence that prolonged, intensive
leaf level, short-term responses to duced (Figure 2b). Vinton and bison grazing did not occur in these
leaf removal are typical of many Hartnett (1992) attributed these dif- grasslands (Shaw and Lee 1997). If
graminoids in grazing systems ferences in short- and long-term re- this hypothesis were true, then grass
(McNaughton 1983). Wallace (1990) sponses to reductions in below- abundance would decline continu-
reported a postgrazing enhancement ground carbon allocation and stored ally following repeated grazing.
of photosynthesis in A. gerardii in carbohydrate reserves after several However, the tallgrass prairie’s loss
Oklahoma. Similarly, on Konza Prai- years of grazing. Turner et al. (1993) of production potential due to graz-
rie, midseason photosynthetic rates also demonstrated the importance of ing is short lived. At Konza Prairie,
January 1999 41
10 watersheds that are subject to
different fire frequencies. Since 1991,
twice-weekly observations of the dis-
tribution of bison within this area
have been made to assess patterns of
bison grazing (Nellis et al. 1992).
The results confirm that bison do
graze preferentially within burned
watersheds from April (burning takes
place in late March–early April)
through June and July, and, in some
years, through August (Figure 3;
Vinton et al. 1993, Nellis and Briggs
1997). In addition to grazing prefer-
entially in burned sites, bison in-
crease their selective consumption of
some grass species in burned sites
relative to unburned sites (Pfeiffer
and Hartnett 1995). Late in the sum-
mer, lowland topographic positions
with deeper soils (and therefore
greater soil moisture and plant pro-
ductivity; Knapp et al. 1993) in
burned watersheds become preferred
grazing locations as the uplands dry.
This preference for burned areas in
Figure 3. Spatial distribution of bison in 10 watersheds in the central portion of Konza tallgrass prairie is consistent with
Prairie after the completion of spring burning. During the months of April, May, and postfire responses in mixed grass prai-
June 1992, bison locations were determined twice per week and recorded on large-scale, rie (Coppock and Detling 1986), as
spatially rectified aerial photographs overlaid with a 30-meter grid. Bison were well as with large ungulates’ winter
observed most frequently in the recently burned watersheds. B, burned watersheds; U, preference for burned sites in the north-
unburned watersheds. ern mixed grasslands of Yellowstone
National Park (Pearson et al. 1995).
permanent fenced exclosures have can be enhanced by the selective con- Within a watershed or at a spe-
excluded bison from experimental sumption of grasses by bison (Fahn- cific topographic position in tallgrass
plots within grazed watersheds for estock and Knapp 1993, Hartnett et prairie, several factors may influ-
several years. When adjacent grazed al. 1996, Damhoureyeh and Hartnett ence initial patch selection and
sites were also protected from graz- 1997). Given the importance of past reselection (see also Wallace et al.
ing with temporary exclosures, grazing pressures to the direct re- 1995). In addition, patch selection
aboveground production in the first sponses of grasses to herbivory and has several long-term consequences.
year in these newly protected sites the indirect responses of forbs, iden- When bison were first reintroduced
was reduced relative to that in the tifying those factors that influence to Konza Prairie, they encountered a
adjoining long-term ungrazed areas the selection and reselection of graz- mosaic of burned and unburned wa-
(Figure 2c); however, these sites re- ing patches by bison in tallgrass prai- tersheds, with significant differences
covered their production potential rie is critical for understanding the among watersheds in the spatial het-
by the second year. Thus, produc- long-term consequences of bison erogeneity of plant community com-
tion potential can recover if bison grazing patterns. position. For example, frequently
grazing is sufficiently dynamic, ei- burned but ungrazed watersheds are
ther spatially or temporally, such Factors influencing bison dominated by C4 grasses and have
that sites are grazed intermittently. selection of grazing sites low species richness and diversity,
Removal of grass leaf area by bi- whereas less frequently burned sites
son, and reductions in the capability Historical information regarding bi- have higher species richness and forb
of the dominant grasses to compen- son grazing patterns in the Great cover (Gibson and Hulbert 1987,
sate for tissue lost after multiple Plains is replete with anecdotal ac- Collins 1992). Initial studies on
years of grazing, suggest that the co- counts of herds attracted to recently Konza Prairie indicated clearly that
occurring subdominant forbs may burned grasslands (Figu re 1; bison established grazing patches in
benefit from bison grazing. Indeed, McHugh 1972, Pyne 1982), but areas strongly dominated by C4
comparisons of forbs inside grazing quantitative evidence of this prefer- grasses and that these patches were
patches with those in adjacent ence in tallgrass prairie has been reselected at a high rate (Vinton et
ungrazed prairie have shown that lacking until recently (Coppedge and al. 1993). Six years later, a survey of
gas exchange and aboveground bio- Shaw 1998). As noted earlier, bison floristic composition indicated that
mass production, density, and cover at Konza Prairie have free access to established bison grazing patches had
42 BioScience Vol. 49 No. 1
Figure 4. Conceptual model of the spatial and temporal dynamics of bison grazing activities and the responses of tallgrass prairie to the reintroduction of bison. Before the reintroduction of bison, watershed attributes at Konza Prairie differed strongly depending on the fire regime imposed, with frequently burned sites notable for their high productivity but low plant species diversity. Selective bison grazing at the watershed level (preference for burned sites, particularly in the spring) and the patch level (selection of patches dominated by C4 grasses) has led to increased similarity in watershed attributes despite differences in fire history. Moreover, significant increases in plant species diversity and spatial heterogeneity have been noted in all watersheds. Superscripts denote studies on which this synthesis is based: 1, Nellis and Briggs (1997); 2, Vinton et al. (1993); 3, Briggs and Knapp (1995); 4, Collins (1992), Collins et al. (1995); 5, Vinton and Hartnett (1992), Vinton et al. (1993); 6, Hartnett et al. (1997); 7, Fahnestock and Knapp (1993, 1994); 8, Catchpole (1996); 9, Collins and Steinauer (1998); 10, Hartnett et al. (1996). a higher abundance of forbs and a nitrogen content of plants is highly Day and Detling (1990) have shown lower cover of grasses than adjacent variable, both spatially and tempo- that grasses growing on urine patches ungrazed patches (Catchpole 1996). rally. Bison contribute to this patchi- in mixed-grass prairie have higher Similar small-scale patterns of forb ness through deposition of nitrogen- leaf nitrogen content, and are there- and grass abundance were observed rich urine. Steinauer (1994) applied fore more nutritious per bite, than by comparing the floristic composi- synthetic bovine urine at randomly grasses growing on patches without tion inside and outside grazing selected locations along eight urine. Not only was grass cover lower exclosures in grazed watersheds at transects at Konza Prairie (four each on urine patches at Konza Prairie, Konza Prairie (Hartnett et al. 1996). in grazed and ungrazed areas). In but the total area of grazed patches These observations suggest that bi- ungrazed areas, grass cover was sig- on the urine-treated transects was son alter plant community composi- nificantly higher in plots that were significantly larger than the area of tion at the patch scale by selecting fertilized with urine than in plots grazed patches on transects that were species-poor, grass-dominated sites without urine. But in the grazed area, not treated with urine (Steinauer and converting them to sites of lo- by contrast, grass cover was signifi- 1994). Thus, the enhanced produc- cally higher diversity (Figure 4). cantly lower on the urine-treated tivity of grasses growing on urine A second factor that may influ- plots than in plots without urine patches represents a potential stimu- ence patch selection and reselection because bison preferentially grazed lus for the initiation and reselection by bison is plant quality. The foliar the grasses on the urine-treated plots. of grazing patches by bison. January 1999 43
Given that bison prefer grazing Other impacts of bison synthetic bison urine increased con- patches that are initially dominated in tallgrass prairie centrations of ammonium and ni- by C4 grasses, but that their selective trate in Konza Prairie soils over 130- foraging and reselection habits con- Effects of ungulates, especially graz- fold and 30-fold, respectively, 8 days vert these patches to sites with a ing, on plant community composi- after application (J. R. Matchett and greater abundance of nonforage spe- tion and structure have been studied Loretta C. Johnson, unpublished data). cies, it is likely that patch locations in many grasslands worldwide Bison grazing can decrease the are spatially dynamic across the land- (McNaughton 1984, Milchunas et export of nitrogen from tallgrass scape. Patches can “move” by two al. 1988, Frank and McNaughton prairie by altering the magnitudes of mechanisms: patch abandonment, 1992, Milchunas and Lauenroth two major pathways of nitrogen followed by selection of a new patch; 1993). Ungulate activities, however, loss—combustion and ammonia or patch migration, wherein portions affect many other aspects of grass- volatilization. Fire is the major path- of a patch are abandoned and the land structure and function, including way of nitrogen loss from ungrazed patch expands into adjacent areas. the physical structure of the environ- tallgrass prairie (Dodds et al. 1996, Over a 3-year period, Catchpole ment and the rates of a number of Blair 1997); nitrogen loss from burn- (1996) found the rate of patch aban- ecosystem-level processes (McNaugh- ing averages 1–4 g·m–2·yr–1 (Blair et donment to be approximately 6–7% ton 1993, Frank and Evans 1997, al. 1998). Grazing lowers combus- per year in both burned and un- McNaughton et al. 1997). There are tion losses of nitrogen in tallgrass burned watersheds; thus, at least several other important mechanisms prairie by reducing the aboveground portions of previously established by which bison alter ecosystem-level plant detritus and increasing the grazing patches were reselected at a processes and physical habitat struc- patchiness of a prairie fire (Figure 2; high rate (Figure 4). However, when ture in tallgrass prairie. Hobbs et al. 1991). Although vola- grazing patches were mapped and tilization of ammonia can be in- compared across years, the extent of Nutrient redistribution and cycling. creased by grazing in other types of spatial reselection—although highly Bison can substantially alter nutri- grassland (Detling 1988), Hobbs et variable due to differences in total ent cycling processes and patterns of al. (1991) suggested that any increase burned area available to bison in any nutrient availability in tallgrass prai- in ammonia volatilization in tallgrass one year—averaged approximately rie. Their effects on nitrogen cycling prairie will be more than compen- 50% per year. Thus, grazing patches are critical because nitrogen avail- sated for by a reduction in combus- in both burned and unburned water- ability often limits plant productiv- tion losses of nitrogen. sheds appear to migrate significantly ity in these grasslands (Seastedt et al. Finally, bison grazing affects the from year to year. This local migra- 1991, Blair 1997, Turner et al. 1997) amount and quality of plant litter re- tion permits periodic release of por- and influences plant species compo- turned to soils. Grazing increases plant tions of the grassland from grazing sition (Gibson et al. 1993, Wedin uptake of nutrients (Ruess 1984) and pressures (Figure 4) and provides a and Tilman 1993). Simulation mod- shoot nitrogen content in many grass- mechanism for recovery of below- els of tallgrass prairie responses to lands (Holland and Detling 1990, ground carbohydrate storage reserves grazing (Risser and Parton 1982) Milchunas et al. 1995), including and production potential. and studies of grazers in other grass- tallgrass prairie (Turner et al. 1993). At the watershed and landscape lands (Frank and Evans 1997, However, the effects of grazers on scales, the long-term consequences McNaughton et al. 1997) have dem- root growth and chemistry vary of bison activities include a reduction onstrated a disproportionate influ- among grasslands (Milchunas and in cover, dominance, and productivity ence of ungulates, including bison, Lauenroth 1993). On Konza Prairie, of grasses; the competitive release of on the regulation of nitrogen cycling root productivity and root biomass many subdominant species, resulting processes. Preliminary data from were 30% and 20% lower, respec- in an increase in the abundance of Konza Prairie suggest that bison are tively, in bison grazing lawns than in forbs; an overall increase in plant similarly important in controlling ungrazed exclosures. In addition, the species richness and diversity; and nitrogen cycling in tallgrass prairie. nitrogen concentration of new root increased spatial heterogeneity (Fig- Bison influence nitrogen cycling, growth in bison grazing lawns at ure 5; Hartnett et al. 1996). Although conservation, and availability in Konza Prairie increased significantly, alterations in plant community com- tallgrass prairie ecosystems by alter- from 0.6% to 0.9%, and the C:N ratio position can be attributed, in large ing several soil and plant processes. of roots decreased. A lower C:N ratio part, to the direct effects of grazing Ungulates in grasslands consume reduces microbial immobilization and by bison, increased plant species rich- relatively recalcitrant plant biomass enhances nitrogen availability within ness is also likely to be a product of and return labile forms of nitrogen grazed areas. Indeed, recent studies increased microsite diversity gener- (i.e., urine) to soils (Ruess and on Konza Prairie indicated that net ated by nongrazing activities, such McNaughton 1988), thus bypassing nitrogen mineralization in bison graz- as dung and urine deposition, tram- the otherwise slow mineralization of ing lawns was 153% greater, and net pling, and wallowing. These and nitrogen in plant litter. Nitrogen in nitrification 126% greater, than in other bison activities contribute sig- bison urine is largely urea, which can ungrazed prairie (Figure 6). Further- nificantly to the increase in spatial be hydrolyzed to ammonium in a more, net nitrogen mineralization rates heterogeneity that is characteristic matter of days (Ruess and McNaugh- were proportional to the intensity of of grazed tallgrass prairie (Figure 5). ton 1988). Indeed, application of bison use of a given area. Thus, the 44 BioScience Vol. 49 No. 1
net effect of bison grazing appears to tial scales, grazed prairie that con- of bison mortality. Given the enor-
be increased rates of nitrogen cycling, tains bison wallows has higher plant mous size of the bison population
coupled with a significant increase in species diversity than grazed prairie before their widespread slaughter in
spatial heterogeneity in nitrogen avail- without wallows (Collins and Bar- the 1800s, annual mortality was
ability; together, these effects can alter ber 1985). Thus, bison can physi- probably high. High death rates
patterns of plant productivity and spe- cally alter grasslands in ways that would have been especially common
cies composition in tallgrass prairie increase environmental heterogeneity during droughts, when there would
(Figure 5; Steinauer and Collins 1995). and enhance both local and regional be the potential for large numbers of
biodiversity (Hartnett et al. 1997). carcasses to occur across the land-
Wallowing. One aspect of bison be- scape. Even though predators and
havior that differs from that of cattle, Bison carcasses. Bison not only af- scavengers may have consumed and
and is primarily a physical activity, fect vegetation patterns and soil pro- relocated many of these carcasses,
is wallowing. Wallows in Flint Hills cesses through their grazing activi- decomposition of the remaining and
tallgrass prairie, which are estab- ties but also have profound and partial carcasses would still have re-
lished primarily in level upland or lasting localized effects after they sulted in patches of locally high nu-
lowland sites, dramatically alter the die. Although legal requirements and trient concentration. Thus, although
patch structure of this prairie. Bison management practices dictate the it was variable, bison mortality would
wallows develop as the animals paw removal of carcasses of domestic have led to a continual cycle of dis-
the ground and roll in the exposed herbivores from public and private turbance and recovery of these
soil. Continued use of wallows by grasslands, native herbivores rou- patches in presettlement grasslands.
bulls, cows, and calves creates a soil tinely die of natural causes and their
depression of 3–5 m in diameter (and bodies remain in situ. As part of the Are bison keystone species?
10–30 cm in depth) that is devoid of minimal management strategy at
vegetation. These denuded patches Konza Prairie, bison that die on site The net effects of selective bison graz-
either gradually revegetate or remain are not removed. As a result, these ing activities at the landscape, patch,
as bare soil, depending on the fre- carcasses create unique local distur- and individual plant level include
quency of revisitation by bison. With bances (Figure 7) that are the focus shifts in plant species composition,
the vast numbers of bison that once of studies to assess their effects on alterations of the physical and chemi-
occupied the Great Plains, these soil soil nutrients and vegetation re- cal environment, and increased spa-
depressions were probably abundant sponses in tallgrass prairie. tial and temporal heterogeneity in
and widespread features of the land- When an individual bison dies, vegetation structure, soil resource
scape (England and DeVos 1969). copious quantities of fluids (with availability, and a variety of ecosys-
For example, a number of relic wal- high nitrogen concentration) are re- tem processes (Figures 4, 5, and 6).
lows had to be filled to level the play- leased during decomposition. Adult Before bison reintroduction at Konza
ing field for the first University of bison can weigh more than 800 kg, Prairie, the long-term burning ex-
Oklahoma home football game in 1895 and these carcasses typically kill un- periments produced clear patterns of
(University of Oklahoma Athletic De- derlying and adjacent plants, creat- response in the vegetation. As fire
partment 1986). Relic wallows still ing a denuded zone of 4–6 m2 (Figure frequency increased, the dominance
exist in many areas where bison have 7). Although the fluids that are ini- of C4 grasses increased, and the cover
not occurred in the past 125 years. tially released are toxic to vegeta- of C3 grasses, forbs, and woody spe-
Environmental conditions in relic tion, these sites eventually become cies decreased (Figure 4; Gibson and
and newly established wallows zones of high fertility. For example, Hulbert 1987). Overall, plant spe-
strongly influence prairie patch dy- soil cores extracted from the center cies diversity declined as fire fre-
namics (Polley and Collins 1984, of carcass sites on Konza Prairie 3 quency increased in ungrazed tall-
Polley and Wallace 1986). Because years after death had inorganic ni- grass prairie (Collins et al. 1995).
of soil compaction, wallows often trogen concentrations that were two These patterns in community struc-
retain rainwater in the spring, creat- to three times higher than the sur- ture, which had developed over 20
ing localized habitats that are suit- rounding prairie. This nutrient enrich- years of burning treatments at Konza
able for ephemeral wetland species, ment may extend up to 2.5 m away Prairie, are being rapidly and dra-
similar to vernal pools in California from the original carcass site and matically altered by the grazing ac-
(Holland and Jain 1981, Uno 1989). results in patches dominated initially tivity of the reintroduced bison. In
In the summer, however, the same by early successional species. The particular, grazing by bison has low-
wallows support only plants that can aboveground primary production in ered the abundance of the dominant C4
tolerate severe drought. Vegetation these patches is two to three times grasses, increased the abundance of
composition and structure in wal- higher than in undisturbed prairie. the subdominant C3 grasses and forbs,
lows is different from that in the Although disturbances created by and markedly increased plant species
surrounding prairie (Polley and bison carcasses are sporadic and lo- diversity (by 23%), richness (by 38%),
Collins 1984), and these differences calized on Konza Prairie, they pro- and community heterogeneity (by
are enhanced by fire (Collins and vide nutrient pulses that exceed all 13%) relative to ungrazed sites, even
Uno 1983), which may not spread other natural processes, even urine under annual burning conditions
through wallows because of low fuel and fecal deposits. Overall, we can (Hartnett et al. 1996, Collins and
loads. Consequently, at larger spa- only speculate about historical rates Steinauer 1998, Collins et al. 1998).
January 1999 45
Because of the multiple and dra- Plains grassl ands is undisputed wooded and grassland habitats op-
matic effects of bison on this land- (McHugh 1972), and we have em- portunistically (Hartnett et al. 1997).
scape, we believe that bison are key- phasized the keystone role that bison Studies that have focused exclu-
stone species in the tallgrass prairie. played in determining the structure sively on cattle generally concur that
Other authors have noted the poten- and function of tallgrass prairies at their grazing activities increase spa-
tial of large grazing mammals to act multiple spatial and temporal scales. tial heterogeneity and enhance plant
as “keystone herbivores” capable of With the replacement of native bison species diversity, so long as stocking
maintaining open grassland vegeta- by domesticated cattle in the remain- density is not too high (Collins 1987,
tion that would otherwise undergo ing grasslands, an obvious issue is Hartnett et al. 1996). Because bison
succession to shrubland or wood- the degree of similarity between these grazing in tallgrass prairie has a simi-
land (Owen-Smith 1987). Indeed, the two ungulates with respect to their lar effect, one could conclude that
disappearance of a grazing mega- effects on tallgrass prairie. In other either herbivore can alter resource
fauna at the end of the Pleistocene words, can bison and cattle be con- availability and heterogeneity and
may have played a major role in the sidered ecological equivalents? reduce the cover of the dominant
widespread transition from steppe to There have been several previous grasses sufficiently to enhance the
tundra at that time (Zimov et al. 1995). attempts to answer this question, success of the subdominant species.
However, the concept of keystone but the results have been equivocal Perhaps of greater importance than
species has been controversial since its at best (Plumb and Dodd 1993, differences in foraging patterns be-
inception (Power et al. 1996). One of Hartnett et al. 1997). The primary tween bison and cattle, however, are
the problems with this concept has barrier to resolving this issue rests the number of nongrazing activities,
been the variable interpretation of cri- with a lack of comparative studies in such as wallowing and horning (i.e.,
teria by which species are determined which management is held constant rubbing on trees) that are associated
to be keystone. Power et al. (1996) and the type of grazer is varied. Such exclusively with bison (Coppedge
consider a keystone species to be “one studies have recently been initiated and Shaw 1997, Hartnett et al. 1997).
whose impact on its community or at Konza Prairie. Results after 3 years These activities, when combined with
ecosystem is large, and disproportion- indicate that the abundance and rich- the spatial redistribution of nutri-
ately large relative to its abundance.” ness of annual forbs, and the spatial ents and selective consumption of
To make this definition operational, heterogeneity of biomass and cover, the dominant grasses, may further
these authors proposed a measure of are higher in sites with bison than in increase plant species richness and
community importance (CI) to be used sites with cattle. No dramatic differ- resource heterogeneity, particularly
as an index of the strength of the ences have been detected, however, at the landscape scale.
impact of a given species: between cattle- and bison-grazed sites Nevertheless, it is likely that be-
in cover of the dominant C4 grass, A. cause bison and cattle are function-
CI = [(tN – tD)/tN][1/pi] gerardii, or the dominant forb, Am- ally similar as large grass-feeding
brosia psilostachya; total plant spe- herbivores, management strategies
where tN is a quantitative measure of cies richness is also not dramatically (stocking intensity and duration) will
a trait (e.g., diversity) in an intact different (E. Gene Towne and David have a greater influence on the de-
community, t D is the measure of the C. Hartnett, unpublished data). gree of ecological equival ency
trait when species i has been deleted, Results at Konza Prairie are con- achieved than inherent differences in
and pi is the proportional abundance sistent with previous assessments these ungulates (Hartnett et al.
(biomass) of species i before it was (e.g., Schwartz and Ellis [1981], Van 1997). Clearly, the degree of over-
deleted. CI values “much greater than Vuren and Bray [1983], and Plumb lap in diet and foraging patterns is
1” indicate that a species is key- and Dodd [1993] in mixed and short- greater between bison and cattle than
stone. We estimate bison biomass at grass prairie), which noted that both between cattle and other historically
Konza Prairie to be approximately bison and cattle are generalist herbi- important native herbivores (Hart-
11–12 g·m2 (Collins and Steinauer vores that graze preferentially on nett et al. 1997), such as antelope
1998), which is approximately 1% graminoids. Nevertheless, some dif- (Antilocapra americana), deer (Odo-
of the total vegetative biomass. On ferences in the foraging patterns of coileus virginianus), and elk (Cervus
Konza Prairie, diversity is signifi- bison and cattle have been docu- canadensis). Indeed, the loss of ante-
cantly higher (10–33%) on grazed mented that may have long-term im- lope and elk from the tallgrass prai-
sites than ungrazed sites (Hartnett et plications for grasslands. For example, rie, coupled with dramatic increases
al. 1996), and these values yield a bison have a higher proportion of in deer populations, presents addi-
range of CIs from 6 to 25. For this graminoids in their diet than do tional challenges for managing these
reason and others, we consider bison cattle; consequently, forb and browse ecosystems.
to be keystone species in tallgrass species are more common in cattle
prairie ecosystems. diets (Van Vuren and Bray 1983, Conservation implications
Hartnett et al. 1997). Also, bison
Are bison and cattle functional spend less time grazing than cattle Conserving small and moderate-sized
equivalents in tallgrass prairie? and more time in nonfeeding activi- tracts of once-vast biomes, such as
ties (Plumb and Dodd 1993), and the tallgrass prairie, presents a unique
The historical presence of immense bison strongly prefer open grassland set of problems that are distinct from
herds of large ungulates in Great areas for grazing, whereas cattle use those associated with spatially re-
46 BioScience Vol. 49 No. 1stricted ecosystems because many of Figure 5. Landscape-
the defining forces that historically level changes in spatial
were important in structuring these heterogeneity in tall-
systems occurred at spatial scales grass prairie induced by
that no longer exist. For example, in bison grazing activities.
pre-1900s grasslands, fires were not False-color composite
of Thematic Mapper
plot-level or even watershed-level (TM) data from an area
events but operated at spatial scales on Konza Prairie grazed
encompassing thousands of hectares. by bison (upper left)
This large spatial scale resulted in po- and from a nearby area
tentially high fire frequencies protected from grazing
throughout the tallgrass prairie be- (upper right). Both sites
cause any point of ignition in this were burned and have
“inland sea of grass” could affect grass- similar soil types, as-
lands hundreds of kilometers distant. pect, and slopes. Red
Today, the fragmentation of Great colors represent areas
of high productivity,
Plains grasslands is recognized as a and blue colors corre-
key factor in reducing the frequency spond to areas of low
of fire, which in turn contributes to productivity or bare
species loss (Leach and Givnish ground. (lower panel)
1996). Indeed, the primary manage- Percentage difference in
ment strategy for small prairie pre- spatial heterogeneity of
serves, which are most prone to in- biomass (estimated
vasion by woody vegetation and from remotely sensed
exotic species, is to burn them as spectral reflectance
frequently as possible to suppress data) between areas
grazed by bison and
invasion by undesirable plants (Leach adjacent ungrazed ar-
and Givnish 1996). Unfortunately, eas on Konza Prairie.
frequent (annual or biannual) spring Before bison reintro-
fire maintains dominance by C4 duction, watersheds
grasses but reduces plant species di- scheduled to remain
versity relative to grasslands that are ungrazed appeared to have greater spatial heterogeneity than those scheduled to be
burned infrequently. One alterna- grazed (negative values in 1987). After the reintroduction of bison in 1988, spatial
tive is to conduct burns at different heterogeneity in the grazed watersheds increased substantially. Spatial heterogeneity
times of the year (Howe 1994), but was assessed using the TEXTURE algorithm, which involves passing a moving 3 ´ 3
summer fires, for example, may not pixel window through the images and determining the differences between the mini-
mum and maximum values for each subset of pixels (Briggs and Nellis 1991). Pixels
prevent invasion or reduce the abun- represent derived Normalized Difference Vegetation Index (NDVI) values from TM
dance of woody vegetation (Adams data (30 m2 spatial resolution) from 1988 to 1991 and from 1993 (TM data were not
et al. 1982). In addition, burning in available for 1992). Previous studies have confirmed that the use of the TEXTURE
late summer may be difficult because algorithm with NDVI data is useful for estimating patterns of spatial heterogeneity in
of other considerations, including tallgrass prairie (Nellis and Briggs 1989, Briggs and Nellis 1991).
reduced ability to control the fire
under dry, windy conditions. Not
only are prairies threatened by frag- duced to grasslands through pre- Alternatively, mowing can be used
mentation and invasion by undesir- scribed fire, the key elements of bi- to reduce the dominance of the tall
able species, but grasslands through- son grazing activities can and should grasses and to enhance species rich-
out the Great Plains are now affected be incorporated into conservation ness (Gibson et al. 1993, Collins and
by increased atmospheric nitrogen and restoration strategies for rem- Steinauer 1998). Results from a long-
deposition (Wedin and Tilman 1996). nant prairies (Steuter 1997). One term experiment at Konza Prairie
Thus, remnant grasslands are sub- approach to accomplish this goal is incorporating annual fire, nitrogen
jected to a variety of anthropogenic the substitution of cattle for bison. addition, and mowing (Collins et al.
factors that can reduce the diversity Plumb and Dodd (1993) argued that 1998) indicated that on annually
of native prairie species. the choice of whether to use cattle or burned and fertilized treatment plots,
The spatial and temporal impacts bison as a management tool in grass- productivity of the grasses was
of bison grazing activities caused by lands is scale and context dependent. higher, and plant species diversity
the historically large and nomadic Clearly, reintroducing bison may not lower, than in control plots. How-
herds are also best characterized as be appropriate for small prairie rem- ever, on burned, fertilized plots that
landscape-level forces. These too are nants with public access and low were mowed (with removal of the
difficult to replicate in today’s frag- economic resources. But cattle, man- foliage; a rough substitute for graz-
mented grassland remnants. Yet just aged for their ecological rather than ing), plant species diversity was re-
as some of the ecological character- their economic value, may be suit- stored to levels similar to control
istics of natural fires can be reintro- able in such cases. plots (Collins and Steinauer 1998).
January 1999 47Figure 6. Net nitrogen gulate herbivory to this grassland is
mineralization and net evident. Indeed, it is the interaction
nitrification rates in of ungulate grazing activities and
grazed and ungrazed fire, operating in a shifting mosaic
areas of Konza Prairie.
Measurements were across the landscape, that is key to
made in situ for a 1- conserving and restoring the biotic
month period in June integrity of the remaining tracts of
1996 and May 1997 tallgrass prairie.
using a modified bur- Before bison were reintroduced to
ied soil core technique Konza Prairie, Knapp and Seastedt
(Raison et al. 1987). (1986) speculated that bison grazing
Replicate study plots and fire could act in similar ways by
were located on upland reducing the accumulation of detri-
areas of adjacent
tus in this system. It is primarily the
grazed and ungrazed
annually burned wa- blanketing effect of the accumula-
tersheds. Values are means ± 1 SE and are significantly different (P < 0.05, n = 120 tion of dead plant material above
per grazing treatment). ground that limits productivity in
undisturbed tallgrass prairie. Like
fire, bison grazing reduces above-
A combination of frequent spring nificant and sustainable biotic diver- ground standing dead biomass. But
fire to maintain populations of the sity in tallgrass prairie is a goal. it is now clear that the unique spatial
desirable C4 prairie grasses, decrease and temporal complexities of bison
nitrogen availability (Blair 1997), grazing activities (Figure 5) are criti-
Conclusions cal to the successful maintenance of
and suppress growth of weedy annu-
als and woody vegetation, coupled Despite less than a decade of re- biotic diversity in this grassland. This
with mowing portions of the site to search at Konza Prairie on bison– grazing-induced heterogeneity con-
reduce the competitive dominance tallgrass prairie interactions, the key- trasts sharply with the spatial homo-
of C4 grasses, can enhance the abun- stone role that bison must have geneity induced by fire in an ungrazed
dance of forbs and maintain high historically played in this grassland landscape (Figure 6).
plant species diversity in small rem- is clear. Moreover, much as fire is Tallgrass prairie, by virtue of its
nant prairies. Ultimately, manage- now recognized as an essential com- inherently variable climatic, grazing,
ment designed to increase the spatial ponent of tallgrass prairie manage- and fire regimes, is an ecosystem that
heterogeneity of resources in a man- ment (because without fire this grass- requires long-term study to docu-
ner analogous to that imposed by land disappears), the need for ment patterns and quantify processes
ungulate activities is essential if sig- reintroducing the forces of large un- (Knapp et al. 1998b). Through the
partnership of The Nature Conser-
vancy, the National Science Foun-
dation’s LTER program, and Kansas
State University, ongoing studies at
this site will continue to explore the
ecological interactions of fire and
grazing in the tallgrass prairie land-
scape. Such research is timely be-
cause conservation and management
issues have intensified in the remain-
ing tracts of this once-vast biome,
particularly in response to predicted
alterations in global climate and land-
use changes. Interdisciplinary eco-
logical research, such as that ongo-
ing at Konza Prairie, will provide the
basic information necessary for de-
signing optimal conservation, resto-
ration, and management strategies
in this and other grasslands.
Acknowledgments
Figure 7. Bison carcass approximately 9 months after death in burned tallgrass prairie. Research summarized here was sup-
In this spring (May) photo, vegetation is completely lacking within the area in which the ported by the National Science Foun-
animal died. A zone of high fertility exists around the edge of this disturbance, and dation’s Long-Term Studies, Ecol-
within 1–2 years the site will be dominated by early successional (annual) species. ogy, and Ecosystems Programs; the
48 BioScience Vol. 49 No. 1US Department of Agriculture’s Na- versity by grazing and mowing in native prairie. Vegetatio 72: 175–185.
tional Research Initiative Program; tallgrass prairie. Science 280: 745–747. Gibson DJ, Seastedt TR, Briggs JM. 1993.
Coppedge BR, Shaw JH. 1997. Effects of Management practices in tallgrass prai-
the National Aeronautics and Space horning and rubbing by bison (Bison bi- rie: Large- and small-scale experimental
Administration; The Nature Conser- son) on woody vegetation in a tallgrass effects on species composition. Journal of
vancy; and the Kansas State Univer- prairie landscape. American Midland Applied Ecology 30: 247–255.
sity Agricultural Experiment Station Naturalist 138: 189–196. Glenn SM, Collins SL. 1990. Patch structure
______. 1998. Bison grazing patterns on sea- in tallgrass prairies: Dynamics of satellite
(99-156-J). sonally burned tallgrass prairie. Journal species. Oikos 57: 229–236.
of Range Management 51: 258–264. Hartnett DC, Hickman KR, Fischer-Walter
References cited Coppock DL, Detling JK. 1986. Alteration of
bison and black-tailed prairie dog grazing
LE. 1996. Effects of bison grazing, fire
and topography on floristic diversity in
Adams DE, Anderson RC, Collins SL. 1982. interaction by prescribed burning. Jour- tallgrass prairie. Journal of Range Man-
Differential response of woody and her- nal of Wildlife Management 50: 452–455. agement 49: 413–420.
baceous species to summer and winter Coppock DL, Detling JK, Ellis JE, Dyer MI. Hartnett DC, Steuter AA, Hickman KR. 1997.
burning in an Oklahoma grassland. South- 1983. Plant–herbivore interactions in a Comparative ecology of native versus in-
western Naturalist 27: 55–61. North American mixed-grass prairie. I. troduced ungulates. Pages 72–101 in
Axelrod DI. 1985. Rise of the grassland biome, Effects of black-tailed prairie dogs on Knopf F, Samson F, eds. Ecology and
central North America. The Botanical intraseasonal aboveground plant biomass Conservation of Great Plains Vertebrates.
Review 51: 163–201. and nutrient dynamics and plant species New York: Springer-Verlag.
Berger J, Cunningham C. 1994. Bison: Mat- diversity. Oecologia 56: 1–9 Hobbs NT, Schimel DS, Owensby CE, Ojima
ing and Conservation in Small Populations. Damhoureyeh SA, Hartnett DC. 1997. Ef- DJ. 1991. Fire and grazing in tallgrass
New York: Columbia University Press. fects of bison and cattle on growth, repro- prairie: Contingent effects on nitrogen
Blair JM. 1997. Fire, N availability and plant duction, and abundances of five tallgrass budgets. Ecology 72: 1374–1382.
response in grasslands: A test of the tran- prairie forbs. American Journal of Botany Holland EA, Detling JK. 1990. Plant responses
sient maxima hypothesis. Ecology 78: 84: 1719–1728. to herbivory and belowground nitrogen
2359–2368. Day TA, Detling JK. 1990. Grassland patch cycling. Ecology 71: 1040–1049.
Blair JM, Seastedt TR, Rice CW, Ramundo dynamics and herbivore grazing prefer- Holland RF, Jain SK. 1981. Insular biogeog-
RA. 1998. Terrestrial nutrient cycling in ence following urine deposition. Ecology raphy of vernal pools in the Central Val-
tallgrass prairie. Pages 222–243 in Knapp 63:180–188. ley of California. American Naturalist 117:
AK, Briggs JM, Hartnett DC, Collins SC, Detling JK. 1988. Grasslands and Savannas: 24–37.
eds. Grassland Dynamics: Long-Term Eco- Regulation of energy flow and nutrient Howe HF. 1994. Response of early- and late-
logical Research in Tallgrass Prairie. New cycling by herbivores. Pages 131–148 in flowering plants of fire season in experi-
York: Oxford University Press. Pomeroy LR, Alberts JJ, eds. Concepts of mental prairies. Ecological Applications
Briggs JM, Knapp AK. 1995. Interannual Ecosystems Ecology. Ecological Studies 4: 121–133.
variabili ty in primary production in 67. New York: Springer-Verlag. Jones JK, Hoffmann RS, Rice DW, Jones C,
tallgrass prairie: Climate, soil moisture, Dodds WK, Blair JM, Henebry GM, Koelliker Baker, RJ, Engstrom MD. 1992. Revised
topographic position and fire as determi- JK, Ramundo R, Tate CM. 1996. Nitro- checklist of North American mammals
nants of aboveground biomass. American gen transport from tallgrass prairie by north of Mexico, 1991. The Museum,
Journal of Botany 82: 1024–1030. streams. Journal of Environmental Qual- Texas Tech University, Occasional Papers
Briggs JM, Nellis MD. 1991. Seasonal varia- ity 25: 973–987. 146: 1–23.
tion of heterogeneity in the tallgrass prai- England RE, DeVos A. 1969. Influence of Knapp AK. 1985. Effect of fire and drought
rie: A quantitative measure using remote animals on pristine conditions on the Ca- on the ecophysiology of Andropogon
sensing. Photogrammetric Engineering and nadian grasslands. Journal of Range Man- gerardii and Panicum virgatum in a
Remote Sensing 57: 407–411. agement 22: 87–94. tallgrass prairie. Ecology 66: 1309–1320.
Catchpole FB. 1996. The dynamics of bison Fahnestock JT, Knapp AK. 1993. Water rela- Knapp AK, Seastedt TR. 1986. Detritus lim-
(Bos bison) grazing patches in tallgrass tions and growth of tallgrass prairie forbs its productivity in tallgras s prairie.
prairie. Master’s thesis. Kansas State Uni- in response to selective herbivory by bi- BioScience 26: 662–668.
versity, Manhattan, KS. son. International Journal of Plant Sci- Knapp AK, Fahnestock JT, Hamburg SP,
Collins SL. 1987. Interaction of disturbances ence 154: 432–440. Statland LJ, Seastedt TR, Schimel DS.
in tallgrass prairie: A field experiment. ______. 1994. Responses of forbs and grasses 1993. Landscape patterns in soil–water
Ecology 68: 1243–1250. to selective grazing by bison: Interactions relations and primary production in
______. 1992. Fire frequency and community between herbivory and water stress. tallgrass prairie. Ecology 74: 549–560.
heterogeneity in tallgrass prairie vegeta- Vegetatio 115: 123–131. Knapp AK, Briggs JM, Blair JM, Turner CL.
tion. Ecology 73: 2001–2006. Flores D. 1991. Bison ecology and bison 1998a. Patterns and controls of above-
Collins SL, Barber SC. 1985. Effects of dis- diplomacy: The southern plains from 1800 ground net primary production in tallgrass
turbance on diversity in mixed-grass prai- to 1850. Journal of American History 78: prairie. Pages 193–221 in Knapp AK,
rie. Vegetatio 64: 87–94. 465–485. Briggs JM, Hartnett DC, Collins SC, eds.
Collins SL, Steinauer EM. 1998. Disturbance, Frank DA, Evans RD. 1997. Effects of native Grassland Dynamics: Long-Term Ecologi-
diversity and species interacti ons in grazer s on gras slan d N cycl ing in cal Research in Tallgrass Prairie. New
tallgrass prairie. Pages 140–156 in Knapp Yellowstone National Park. Ecology 78: York: Oxford University Press.
AK, Briggs JM, Hartnett DC, Collins SC, 2238–2248. Knapp AK, Briggs JM, Hartnett DC, Collins
eds. Grassland Dynamics: Long-Term Eco- Frank DA, McNaughton SJ. 1992. The ecol- SC. 1998b. Grassland Dynamics: Long-
logical Research in Tallgrass Prairie. New ogy of plants, large mammalian herbi- Term Ecological Research in Tallgrass Prai-
York: Oxford University Press. vores, and drought in Yellowstone Na- rie. New York: Oxford University Press.
Collins SL, Uno GE. 1983. The effect of early tional Park. Ecology 73: 2043–2058. Leach MK, Givnish TJ. 1996. Ecological de-
spring burning on vegetation in buffalo Freeman CC. 1998. The flora of Konza Prai- terminants of species loss in remnant prai-
wallows. Bulletin of the Torrey Botanical rie: A historical review and contemporary ries. Science 273: 1555–1561.
Club 110: 474–481. patterns. Pages 69–80 in Knapp AK, Briggs McHugh T. 1972. The Time of the Buffalo.
Collins SL, Glenn SM, Gibson DJ. 1995. JM, Hartnett DC, Collins SC, eds. Grass- New York: Knopf.
Experimental analysis of intermediate dis- land Dynamics: Long-Term Ecological McNaughton SJ. 1983. Compensatory plant
turbance and initial floristic composition: Research in Tallgrass Prairie. New York: growth as a response to herbivory. Oikos
Decoupling cause and effect. Ecology 76: Oxford University Press. 40: 329–336.
486–492. Gibson DJ, Hulbert LC. 1987. Effects of fire, ______. 1984. Grazing lawns: Animals in
Collins SL, Knapp AK, Briggs JM, Blair JM, topography and year-to-year climate varia- herds, plant form and coevolution. Ameri-
Steinauer EM. 1998. Modulation of di- tion on species composition in tallgrass can Naturalist 124: 863–868.
January 1999 49You can also read
