Written by Dan Carter. Photos by Dan Carter
September 23, 2024
In the previous issue of The Prairie Promoter,[1] I introduced stability and discussed how fire can be stabilizing or destabilizing depending on how we use it. Here I will discuss grazing and browsing in the context of managing old-growth Midwestern prairie and oak ecosystems or projects seeking to restore old-growth-like characteristics to degraded sites. Keep that scope in mind. There are cultural reasons for repatriating native herbivores and conservation objectives for which grazing and browsing practices might be explored in surrogate grasslands[2] or for initial opening of brush where there is high confidence that collateral damage will not occur to native biota[3], but there are many reasons to caution against using those practices on remnant old-growth or projects attempting to promote old-growth characteristics. Grassland and savanna species declines are not the result of failure to graze prairie and oak ecosystems; those declines arise from loss of grassland and savanna habitat across the landscape and changes in land use. At one time tremendous variation in habitat existed among the prairie and oak ecosystems that were expressed along gradients of moisture, soil texture and chemistry, aspect, slope, etcetera in the Midwest; and our agriculture was more varied and less industrial. To promote the biotic expressions in need of protection, we should restore landscapes and all the unique ecosystem types that comprise them—not force exquisite and unique remnants to do work that simple plantings or conservation agricultural practices can do.
Large Ungulate Grazing is not Necessary to Maintain or Encourage Old-Growth Characteristics
I became aware of The Prairie Enthusiasts during my graduate study at Kansas State. I was drafting a manuscript, and most publications from the Konza Prairie Biological Station referenced a paper by Knapp et al. (1999)[4]. Knapp et al. argued that bison (Bison bison) played a keystone role promoting “biotic integrity” and healthy plant communities in tallgrass prairie. Having spent a few years working on Konza, I felt uneasy passing along the breadth of their conclusions. While I was in awe of the grassland landscape there, I was underwhelmed by the condition of Konza’s experimental watersheds. Unburned watersheds were becoming forests of eastern redcedar (Juniperus virginiana), frequently burned watersheds (burned after spring green-up) were over-dominated by big bluestem (Andropogon gerardii) and Indiangrass (Sorghastrum nutans). Bison-grazed watersheds were experiencing woody encroachment, and while forbs were conspicuously more abundant, they were predominantly species like plains ragweed (Ambrosia psilostachya), gumweed (Grindelia squarrosa), and goldenrods (Solidago spp.). Except for a few low statured species (e.g., ground plum, Astragalus canadensis), greater concentrations of old-growth-associated species (conservative species)[5] occurred mostly on the steep, rocky slopes. There was little of Konza that I saw as high quality, remnant prairie. As I followed citations of Knapp et al., I stumbled across the letter that Mark Leach, Rich Henderson, and Thomas Givnish[6] published in response, and I found their concerns entirely consistent with my experience. Research into the authors led me to The Prairie Enthusiasts, and I was relieved that there were people that understood what high quality, old-growth prairie was!
How did Knapp et al.3, and later Manning et al. (2017)[7] and Ratajczak et al. (2022)[8] conclude that disturbance from confined bison grazing was beneficial to prairie plant communities?[9] One problem is a reliance on primary productivity, species richness, and other metrics of which richness is part (i.e., diversity and floristic quality index). I explained the spring issue1 how primary productivity and richness can increase while plant communities degrade, as generalist or opportunistic species colonize space opened by disturbance, and competition becomes less driven by low nutrient availability. Manning et al. reported higher floristic quality index in bison-grazed watersheds at Konza, but that effect was driven by richness; mean coefficient of conservatism (affinity of species for remnants—old-growth) was lower with bison grazing. Long-term disturbance history is another important factor at Konza. Frequent early growing season burning (vs. dormant season burning) practices aimed at maximizing grass production for livestock have held sway for a century in the Flint Hills. Those practices favor big bluestem and Indiangrass and had already changed prairie composition by the second half of the twentieth century[10],[11]. Before the 1970s Konza’s prairies were pastured ranchland[12], and there is no reason to suspect they would have been managed differently. Post green-up spring burning between March and May was carried forward in Konza’s most-replicated burn treatments. Spring phenology there is several weeks ahead of southern Wisconsin[13]. This has sustained big bluestem and Indiangrass over-dominance in the frequently burned, ungrazed watersheds that are used to make comparisons with bison-grazed watersheds. In that setting, bison grazing breaks the over-dominance of big bluestem and Indiangrass and allows more opportunistic or generalist plant species to grow, and that leads to higher species richness. Neither condition is like that of less degraded, remnant tallgrass prairies or savannas[14],[15].
Bison roamed the prairies, especially westward, so how could their presence be anything but salubrious? I don’t blame bison so much as settler culture and fences. I suspect there are differences between the effects of bison under confinement—even on thousands of acres—and unfenced bison. Bison were migratory, their behavior was subject to influence by hunting and predation, and their use of the landscape was and still is heterogeneous. We don’t have data to compare the effects bison had historically under those circumstances to those of contemporary confined bison, but the few historical descriptions of tallgrass prairie plant composition we have suggest historical bison numbers and behaviors, whatever they were, led to different expressions of prairie than result from contemporary bison grazing or any fenced scheme involving domesticated herbivores[16][17]. Estimates of past bison populations vary widely and mostly either extrapolate historical observations to large landscapes (e.g., Seton 1929[18]) or estimate carrying capacities based on different levels of consumption of plant biomass that cannot be adequately disentangled from that consumed by other vertebrate consumers, invertebrate consumers, and fire (e.g., Weber 2001).[19] These are only snapshots of the dynamic situation in the historical period. It’s possible that effects of bison as we see them now are more intense in space or time than they were for the several thousand years the tallgrass prairie was developing—even westward where bison numbers were greater.
Bison occurred eastward, but their populations east of the Mississippi River increased vastly following the plagues of disease inflicted by Europeans on Indigenous people. Henderson (2001)[20] summarized evidence, or the lack-thereof, for significant presence of modern bison in Wisconsin before the historical period. Similarly, McMillan (2006)[21] concluded that bison were present in “small scattered herds” in Illinois during that time. Any presence of bison prior to the historical period may have been culturally significant, or ecologically significant in terms of seed dispersal, but bison grazing probably was not a significant factor shaping community structure and composition on most of the landscape for most of the past several thousand years, at least away from localized areas bison may have concentrated.
The restoration or maintenance of high small-scale species richness and conservative vegetation in the absence of prescribed grazing (with frequent dormant season fire) provides additional evidence that bison are not essential to the maintenance of community structure or composition in tallgrass prairie and savanna ecosystems. Even prairie remnants farther west where bison were historically abundant look remarkably like remnant eastern prairies when they share similar European land use histories and contemporary management. These are often remnants along railroad rights of way or areas that were hayed rather than grazed or cultivated. The life history strategies of many conservative species make it unlikely that their assemblage in these areas has been a recent response to cessation of grazing and suggests that prior effects of free bison were less intense in the tallgrass region than those of confined bison. I suspect that as fire’s effects vary with intensity and seasonality, so do bison’s effects by their access to an intact broader landscape replete with elements that affect bison behavior.
I’ve focused on bison because their perceived role is used to justify practices involving other herbivores. More generally, how can fenced grazers and browsers destabilize prairie and savanna ecosystems? As with growing season fire, nitrogen may be part of the story. Herbivores consume the living tissues of herbaceous plants and deposit nutrients that had been held within those tissues during times when opportunistic plant species can capitalize. Frequent burning in the absence of fenced grazing volatilizes nitrogen from fuels and results in less plant-available nitrogen in the soil, but the nitrogen made available in animal waste counters that, and the proportion of plant production consumed by fire is concurrently reduced. Continuous fenced bison grazing leads to as much as six-fold increases in nitrogen mineralization and nitrification[22],[23], which result in increased plant-available nitrogen. Cattle have similar effects[24]. This can promote opportunistic versus old-growth-associated vegetation by shifting the community away from one shaped by competition for nitrogen towards one shaped by tolerance to or avoidance of herbivory or trampling, or competition for light upon cessation of grazing. Plant-available nitrogen may also stimulate germination of opportunistic species, including woody species. Briggs et al. 2005[25] report that introduction of bison to Konza was associated with four- to forty-fold increases in woody encroachment within four years. They speculate about mechanisms without considering the above, but the second to last line below stands out (emphasis added, grammatical error notwithstanding):
“…woody vegetation was restricted to riparian areas or protected escarpments…. Thus, even when bison removed fine fuel by grazing, there was a low probability that seeds of woody plants would reach these patches. In the current fragmented landscape, which is characterized by a higher cover of woody plants…, there is a greater chance that the seeds of woody plants will reach a grazed area with low fire intensity. In addition, bison were migratory grazers; thus, grazing in any given area were probably not chronic. Fuels could accumulate in years when there were no bison grazing in the area and therefore could support intense fires.”
They recognize bison’s past effects may have differed. Indeed, large portions of the bison grazing treatments receiving ignitions at Konza do not burn because of sparse fuels. More than once while serving on Konza burn crews, I was sent deep into bison watersheds searching for patches of adequate fuels to carry ignitions.
Fenced grazing and browsing can also compact soil (e.g., Manning et al. 20066), though there are exceptions, at least with short-term grazing (e.g., Harrington and Kathol 2009[26]). When compaction does occur, it can reduce water infiltration and retention, cause erosion and sedimentation, and physically change the soil growing environment for microbes and plants. Nonetheless, effects of compaction on prairie and oak community structure and composition are poorly understood, in part because changes arising from effects on soil physical properties are difficult to separate from those arising from soil chemical changes and direct effects of herbivores on plants. Also, many studies (e.g., Teague et al. 2011[27]) only compare soil and vegetation effects of different grazing practices in the absence ungrazed controls.
Grazing or browsing with goats or cattle are sometimes used to specifically target problematic woody species. These are seen by some as alternatives to herbicides or fire. Goats can also reach brush on steep terrain. However, goats and cattle also consume a wide range of conservative native plants, sometimes preferentially. On a visit to an open savanna restoration in Iowa that I’ve known since the 1990s, a recent change in management involved introduction of goats in lieu of burning. Inside the goat fence and in areas where goats had recently been, I could find browsed prairie forbs and defoliated oak grubs, but the goats had not yet worked over the smooth sumac (Rhus glabra) and Amur honeysuckle (Lonicera maackii)! Likewise, on a visit to a bluff prairie in southeastern Minnesota, I saw something similar where goats had browsed—heavy impacts to prairie forbs, but unwanted woody species were clearly going to recover. The prairie’s biological crust, important for stabilizing soil and itself capable of , was battered by hooves. In contrast to other bluff prairies in the Driftless, there was no midsummer bloom. I wonder if lepidopterans or oligolectic bees emerged to find the hillside bereft of resources. I am not aware of any study showing a benefit of prescribed goat browsing to old-growth-associated vegetation in the Midwest, nor have I seen one. I have seen photographs showing immediate after-effects—typically reductions of woody vegetation taken shortly after browsing treatments. Longer-term outcomes and effects on conservative species are conspicuously absent.
There is more study of cattle grazing in restoration or conservation contexts. In a Minnesota experiment on effects of canopy thinning, burning, and light cattle grazing treatments in a savanna restoration setting, Yantes et al. (2023)[28] found that cattle grazing did not yield benefits above combined thinning and burning, but that study did not consider effects on mean coefficient of conservatism—just richness, diversity, density, and cover. Harrington and Kothol19 showed reductions in unwanted shrubs in degraded oak savanna and prairie in response to treatment with highland cattle, but fire alone caused greater or similar reductions in savanna and prairie, respectively. Briggler’s (2017)[29] study of short-term effects of patch-burn grazing (one cycle) in Missouri seams to show positive effects on floristic quality index, but the authors don’t report mean coefficient of conservatism. In contrast, Buckles and Harmon-Threatt (2019)[30] show reductions in floristic quality and habitat quality for ground-nesting bees at patch-burn-grazed sites, and Thomas[31] showed longer-term collapse of mean coefficient of conservatism in response to patch-burn-grazing at Taberville Prairie in Missouri. It is worth noting that effects of grazing and burning in Missouri patch-burn grazed sites could also result from a recent tendency to burn sites later in spring versus the dormant season. Regardless, the declines of Taberville and other premier Missouri prairies (e.g., Niawathe Prairie, see Paul Nelson’s presentation[32]) offer cautionary tales and gut-wrenching imagery of well-intended practices profoundly damaging previously well-maintained (with dormant season fire, without prescribed herbivory) old-growth sites. Nor should this be a surprise. Hayden17, Curtis[33], and Weaver16 wrote about how quickly and profoundly fenced livestock can ruin prairie based on research and observations from the late 1800s through the mid-1900s. From Weaver:
“Climax grassland, when grazed lightly, retains essentially its natural composition. It is only when grazing animals are circumscribed in their range by fences and when too large a number are thus confined, that grazing and trampling become so excessive that normal plant cover can not be maintained. Numerous changes in the vegetation then occur…. Where prairie is grazed intensively, in only a few years almost unbelievable changes occur.”
Sometimes the hope is to set the stage for restoration through the temporary use of goats or cattle to remove brush. That may be worth a try where no old-growth-associated biota remain that may be lost. However, unless grazing and browsing treatments are repeated or continuous over long periods, other means of treatment of resprouting woody vegetation are necessary.
Resilience of old-growth prairie and savanna is a myth based on research conducted predominantly on sites already degraded by past land use[34]. The more remnants are disturbed, the more effort is required to promote recovery. Still, remnants are highly resistant to change if the stabilizing ecological processes that promote low nutrient availability, remove smothering thatch and leaf litter, and otherwise maintain healthy and genetically diverse populations[35] are maintained, chief among them is frequent dormant season fire. That’s the common thread among rare sites maintaining or improving condition despite fragmentation, climate change, and nitrogen deposition due to agricultural and industrial practices. Fenced grazing and browsing practices can impair that resistance, destabilizing plant community composition, structure, and everything connected to those in response. While greater site-scale diversity often results, that occurs at the expense of unique and tightly interwoven old-growth, because opportunistic species increase that are much more common on the broader, degraded landscape. Regionally, these site-level increases in diversity represent homogenization—as old-growth loses unique characteristics and becomes more similar to old fields, roadsides, and other disturbed habitats.
[1] Carter, D. (2024). Stability part one: Why I recommend frequent dormant season burning. Prairie Promoter, Spring: 14-19.
[2] Surrogate grasslands are grassland types associated with or arising from European land use practices. Examples of surrogate grasslands include agricultural hayfields, pastures, fallow fields, and certain lower diversity native or non-native conservation grassland plantings. Most grassland habitat consists of surrogate grasslands (vs. old-growth or old-growth-like), so they are important to the conservation of grassland Species of Greatest Conservation Need. In these settings certain grazing practices may be used towards specific habitat goals with lower risk to the ecological integrity of old-growth or old-growth-like prairie and oak ecosystems.
[3] Even at very degraded sites populations of a few to many highly desirable old-growth-associated species often persist as exceedingly small vegetative plants dying slow deaths from shade and excessive litter. These are particularly vulnerable to extirpation from disturbance associated with restoration practices, and it takes skill and luck to detect them.
[4] Knapp, A., et al. (1999). 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. BioScience 49: 39-50.
[5] Conservative plant species are those that are relatively intolerant of degradation. Species are assigned coefficients ranging from zero to ten; those with higher coefficients are more indicative of a relatively stable and undisturbed conditions where they occur. The limited dispersal ability of these species, and the abundance and richness of these species on sites least subject to excessive thatch build-up, European grazing practices, soil disturbance, and regular growing season burning; as well as early historical accounts of tallgrass prairie composition suggest that these are associated with old growth and high quality/integrity sites. There are many native plant species, but these are the ones that define remnant sites.
[6] Leach, M., et al. (1999). A caution against grazing. Bioscience 49: 599-600.
[7] Manning, G., et al. (2017). Effects of grazing and fire frequency on floristic quality and its relationship to indicators of soil quality in tallgrass prairie. Environmental Management, 60: 1062-1075.
[8] Ratajczak, Z., et al. (2022). Reintroducing bison results in long-running and resilient increases in grassland diversity.” Proceedings of the National Academy of Sciences, 119.36: e2210433119.
[9] It is worth noting that in addition to my graduate advisor, two of my dissertation committee members and a former office mate are authors/co-authors on these papers. They are all brilliant and well-meaning people.
[10] Towne, G., & C. Owensby. (1984). Long-term effects of annual burning at different dates in ungrazed Kansas tallgrass prairie. Journal of Range Management 37: 392-397.
[11] Hensel, R. (1923). Recent studies of the effect of burning on grassland vegetation. Ecology 4: 183-188.
[12] Given, C. (2004). History of the Dewey Ranch: https://kpbs.konza.k-state.edu/history/land.html
[13] This means Konza spring burn treatments are phenologically equivalent to burning between mid-April and mid-June in southern Wisconsin. In southern Wisconsin depending on the year, consistently burning after some time roughly between the spring equinox and the second week of April risks destabilizing remnant prairie.
[14] Carter, D. (2023). Change and persistence among prairie grasses. Prairie Promoter, Summer, 10-12.
[15] Many would argue that the degraded condition was the ancestral condition, but the earliest data we have in that region and elsewhere do not support that, fragmentary though it may be. High small-scale richness and high relative abundance of conservative species most likely did not self-assemble into the small fragments we still have today, because the individual life history strategies and complex ecological interactions of those species would make that implausible. They need stability, time, and intact means of dispersal.
[16] Weaver, J. E. (1954). North American prairie. University of Nebraska—Lincoln, URL: https://digitalcommons.unl.edu/agronweaver/15/
[17] Hayden, A. (1918). Notes on the floristic features of a prairie province in central Iowa. In Proceedings of the Iowa Academy of Science, Vol. 25: 369-389.
[18] Seton, E. (1929). The lives of game animals. Doubleday, Doran and Co., Garden City, N.J.
[19] Weber, K. (2001). Historic bison populations: A GIS-based estimate. Proceedings of the 2001 Intermountain GIS Users’ Conference: 45-51.
[20] Henderson, R. (2001). Where did the elk and the buffalo roam? Prairie Promoter. Fall: 15-16.
[21] McMillan, R. (2006). Perspectives on the biogeography and archaeology of Bison in Illinois in Records of early bison in Illinois. Illinois State Museum Scientific Papers, 31: 67-146
[22] Nitrogen mineralization is the conversion of organically bound nitrogen into inorganic forms (e.g., nitrate and ammonium), and nitrification is a microbial process by which primarily ammonia is oxidized to nitrate and nitrate. The important product of these reactions is plant-available nitrogen in the forms of nitrate and ammonium.
[23] Konza bison fact sheet: http://lter.konza.ksu.edu/sites/default/files/BisonFact17.pdf
[24] Johnson, L., & Matchett, J. (2001). Fire and grazing regulate belowground processes in tallgrass prairie. Ecology, 82: 3377-3389.
[25] Briggs, J. et al. (2005). An ecosystem in transition: causes and consequences of the conversion of mesic grassland to shrubland. BioScience, 55: 243-254.
[26] Harrington, J., & Kathol, E. (2009). Responses of shrub midstory and herbaceous layers to managed grazing and fire in a North American savanna (oak woodland) and prairie landscape. Restoration Ecology, 17: 234-244. Note: Grazing treatment duration was two days on several rotations over two years. Also, there were some differences in the woody species more reduced by grazing versus fire.
[27] Teague, W. et al. (2011). Grazing management impacts on vegetation, soil biota and soil chemical, physical and hydrological properties in tall grass prairie. Agriculture, ecosystems & environment, 141: 310-322.
[28] Yantes, A, et al. (2023). Oak savanna vegetation response to layered restoration approaches: Thinning, burning, and grazing. Forest Ecology and Management, 537, 120931.
[29] Briggler, M., et al. (2017). Effects of patchburn grazing on vegetative composition of tallgrass prairie remnants in Missouri. Natural Areas Journal, 37(3), 322-331.
[30] Buckles, B. J., & Harmon‐Threatt, A. (2019). Bee diversity in tallgrass prairies affected by management and its effects on above‐and below‐ground resources. Journal of Applied Ecology, 56(11), 2443-2453.
[31] Thomas, J. (unpublished). The effects of patch burn grazing on a high quality prairie. NatureCite, URL: https://drive.google.com/file/d/0BwTps-dBuxNCenpYbTJWNEVIOWc/view
[32] Nelson, P. (2012). Patch-burn grazing: Is it right for Missouri’s remaining high quality prairies: https://www.youtube.com/watch?v=cqwu5jG96tc
[33] Curtis, J. T. (1959). The vegetation of Wisconsin: an ordination of plant communities. University of Wisconsin Press.
[34] As discussed earlier in this article, most of the western tallgrass prairie has long been utilized as prairie pasture and been dramatically altered in composition and structure for many decades to more than a century prior to more recent study. Research and observation from such areas is not directly applicable to the conservation management of old-growth, true prairie or oak ecosystems.
[35] Don’t mistake this to mean maximizing populations of species. I mean maintaining viable populations within the context of an ecosystem with ecological integrity versus maximizing populations of specific flora or fauna.
Figure 2: A vegetation sampling quadrat in a bison-grazed, biennially burned watershed at Konza in August. Tall goldenrod (Solidago altissima, C=1), stiff goldenrod (Solidago rigida, C=3), plains ragweed (Ambrosia psilostachya, C=3), hoary vervain (Verbena stricta C=1), and Baldwin’s ironweed (Vernonia baldwinii C=2) are the most prominent forbs in the vicinity, none of which are conservative (Coefficients of conservatism are for Kansas).
Figures 3 and 4: Views of minimally grazed/ungrazed (at least by European livestock), mesic prairie in western Iowa (left) and southeast Wisconsin (right). Note the similarity. It is only possible to differentiate them by the presence of nodding onion (Allium cernuum, left of center) in the photo on the right. All other species visible in both images are at least present at both sites at the transect (vs. plot) scale.
Figure 5: This is a reference, old-growth prairie (The Prairie Enthusiasts’ Black Earth Rettenmund Prairie)—not the prairie pastures so often used for research on grazing and fire effects.
Figure 6: Another reference, old-growth prairie, this one on private land in western Iowa.
The Prairie Enthusiasts is an accredited land trust that seeks to ensure the perpetuation and recovery of prairie, oak savanna, and other fire-dependent ecosystems of the Upper Midwest through protection, management, restoration, and education. In doing so, they strive to work openly and cooperatively with private landowners and other private and public conservation groups. Their management and stewardship centers on high-quality remnants, which contain nearly all the components of endangered prairie communities.
