Dennis Schlicht, Iowa Lepidoptera Project, 1108 First Ave. Center Point, IA 52213
First presented in a public lecture in Iowa City, Iowa, on October 3, 1993.
Index Descriptors: Lepidoptera, butterflies, prairie management, biodiversity, prairie insects, prescribed fire.
In 1962 Thomas Kuhn proposed that science describes the world and how it functions through ideas he called paradigms. A paradigm may do several things for its followers. It tells us how we perceive the world, what it is all about, so that it makes sense. It makes order out of chaos. It describes how and what science practice is legitimate; for instance, which theories and methods should be used to substantiate current thought. In this way it puts forward boundaries that keep out ideas that challenge current thought.
He described normal science as scientists going about the business of doing experiments and studies that prove the existing paradigms. When ideas arise that don’t fit in the currently held paradigm they are ignored, either as being irrelevant to the issue or as bad science by someone who doesn’t perceive the real issues as defined by the conventional paradigm.
Kuhn stated that new paradigms will arise when enough compelling anomalies raise the concerns of enough people that research is stimulated that searches for a different view of the issue. If the results of that research are compelling, there is the possibility that the paradigms may change, but Kuhn suggests that the original paradigm supporters, who have spent their careers looking for data that supports their paradigms will not relent.
It has become apparent that many people around the world, especially those who study prairie obligate insects, view prairies with a different set of paradigms than do most prairie managers. Those studying prairie insects view prairies in light of a new conceptual framework based on biodiversity and ecological interaction. These ideas have long been noted but have not been widely regarded in the current self-feeding fervor of prairie management. Current prairie management here defined as a very young technology based on botanical paradigms where large areas are repeatedly treated with fire.
These concerns come from data on prairie invertebrates, including insects and butterflies in particular, even though most investigators believe butterflies are not among the most ecologically sensitive insects. Prairie obligate invertebrates often use only portions of a prairie site, so the treatment of that particular portion is key to a species’ survival. Prairie species, even whole genera and families, are missing from some managed prairies. In fact by 2004 five obligate species of prairie butterflies and skipper were completely missing from the largest preserves in northern Iowa.
Andrew Williams (1996) work on the invertebrates that use the prairie plant Onosmodium molle is pivotal to our understanding of the magnitude of this biodiversity tragedy. He discovered that this plant species is used or visited by a least 112 species of invertebrates in Wisconsin. At least 5 of these appear to be specific to this plant and several where previously undescribed. Multiply some large part of this number (100) by the 200 plus species of plants (some say 400) on a native prairie and the result is staggering (20,000 interactions with a mathematic potential of 399,980,000 species interactions). There are over 12,656 possible interactions for just the species on O. molle alone. To extrapolate further many undescribed species may be threatened. A most conservative estimate of the magnitude of this issue would be 5 obligate species or 30 interaction including the plant and 1000 invertebrates plus 200 plant species for a total of 1200 species on a prairie. That is 1,438,800 interactions possible between these species. Obviously, we know almost nothing of the effects of extensive use of fire on prairie organisms.
The following points stand as anomalies to current prairie paradigms:
1. On unmanaged prairies, obligate butterflies often show more diversity than on managed sites.
2. Plots that are being grazed are often more diverse in prairie obligate butterflies than fire managed prairies.
3. Currently a number of railroad prairies hold better diversity than large fire managed prairies.
4. Many prairie obligate insects are threatened throughout their range. The Regal Fritillary (Speyeria idalia) for example was found from New Brunswick to North Carolina and west to Colorado. It is now all but gone from New England and east of the Appalachians except for a few places. Ohio lost their Regals by 1991 and Indiana by 1992. By 1995 Wisconsin had 6 sites in 4 counties, which was down from 16 counties in the 1980’s.
5. Prairie skipper species are declining and /or absent from many fire managed prairies, Hesperia dacotae, Hesperia ottoe, Oarisma poweshiek, Ceononympha inornata, Lycaeides melissa and Atrytone arogos are now missing (2008) or of special concern in Iowa as well as nearly all of the tall-grass prairie biome..
6. Many fire-managed prairies in the mid‑west are not getting botanically better but in fact the problems that were to be cured with fire have been made worse.
The paradigms that we entomologists use to understand prairie insects are the following:
1. As the interactions of insects with plants in a community are of primary ecological importance, the endangerment of the insect community is a grave concern.
2. Plot size is less important for survival than management regime.
3. There are magnitudes more invertebrate species than plant species on prairies. Some authors (Clements and Shelford) believe that there are as many as 10 million individual insects per acre of prairie.
4. Insect populations naturally display dramatic population fluctuations on Iowa prairies which combined with excessive disturbance could extirpate local populations.
5. Invertebrates that reside in any life stage above ground do not survive dormant season fire.
6. Most prairie insect do not use the whole plot evenly but are scattered in patches which may be contained in a burn unit. Therefore the whole population is threatened.
7. Prairie insects recolonize slowly.
8. Recolonization must be complete before management repeats.
9. Prairie insects cannot recolonize from oblivion (a new form of creationism).
10. Local extirpation occurs one patch at a time until the fragmented populations are all lost.
Population biologists (Slobodkin, 1986) tell us that for prairie obligate insects to survive natural pressures long‑term they must fulfill the following criteria:
‑Individual populations must be numerically large, they are not.
‑They need to have highly resistant life stages, few do.
‑The whole population must not be in any one life stage at a time, they are.
‑They must have a continual or a long breeding season, they are short.
‑They must have adults that live through more than one breeding season, they do not.
‑They must have migratory rates between populations that will insure recolonization. Most do not in the current insular island situation.
‑They must not be subject to excess interspecific competition. We don’t know about the effects of alien species on these natives.
In Iowa, fires are currently conducted by private citizens and agencies, such as the DNR, County Conservation Boards, roadside managers, U.S. Fish and Wildlife, nature centers, The Nature Conservancy and the Iowa Prairie Network. Most Universities, Colleges and Technical Schools offer course work on fire management. The public is invited to get involved and all with hardly any knowledge of or consideration for invertebrates survival.
We have concerns over today’s prairie paradigms based on their expression in management, whether or not they are written somewhere in a management handbook or on a Nature Conservancy preserve signpost. Let’s examine some of the apparent paradigms about prairie and fire along with some challenges to those ideas. First let me state here that we recognize the role of fire in the prairie biome but that the extent and manner in which it is used and the results achieved are points of contention.
We also recognize that butterflies are one of the smallest groups of invertebrates on the prairie but effects on their life cycles represent an indication of what is happening with other invertebrates.
The paradigms upon which prairie management is currently based follow:
Paradigm 1: Prehistorically most of the prairie region burned often.
Practice: To burn prairies as a whole every 1, 2, or 3 years.
Response: Prehistorically the prairie peninsula had a lot of water on the landscape. Because of this, fire would have burned randomly, burning some large areas but often only consuming strips and patches between the rivers, marshes and sloughs. The effect of large grazers on the way a prairie burned also led to increased patchiness. This left refuges (survivia) for invertebrates that served as sources for repopulating the burned areas, unlike todays complete fires in undisturbed prairie. Even if we agreed on complete burns over huge areas the situation today is different. Today the prairie is fragmented, but often with the original diversity intact, so that we have a situation that is more like island biogeography. If the disturbance from fire is too extensive species are lost with no source for recolonization available. Disturbance return intervals of 5 to 10 years may be more natural and have been supported by several studies. Other studies are often quoted that show fire return intervals of up to 35 years. Despite that, Southern Iowa savannas are being burned annually. Local extirpation occurs one patch at a time until the fragmented populations are all lost.
Historical anecdotal evidence from pioneer accounts (mostly from the second wave settlers) are helpful, but only from the point of view of a people in a strange land who wanted the savage natives out so that they could take their land. These people lived in flammable houses in a flammable landscape and of course took note of the occurrence of fires. But would they have stayed if the state burned every one, two, or three years? It seems unlikely they would have been able to get a foothold. Smoke on the horizon surely got their attention just as it does ours, and I think that in a time of slow communication it would have been reported with more fervor than its effects warranted. If a fire burned even 10% of an average Iowa county (that’s about 126 square miles) it surely made the papers, but it wasn’t the whole county, only 10%. Historical accounts also tell that these fires were mostly in the fall with some in the summer.
Paradigm 2: Lightning started fires on the prairie enough to burn it frequently.
Response: The Iowa Fire Marshal’s office reports only one fire in cropland in 1991 from natural causes and no one recalls a lightning started crop fire in 1992. I witnessed a fire in June of 1993 which was apparently caused from lightning hitting an electric power pole. There surely are some in most years but it seems to be a rare event. In the early 70’s, I worked for the U. S. Forest Service for 3 summers in the Medicine Bow N.F. in Wyoming as a fire lookout. In that time, lightning started fires on only one day on the several million acres I watched over. Western forests are nothing like tallgrass prairie when lightning is considered as an ignition factor.
During the spring and fall when the fuels are combustible the dew points are high. You could walk through grass even at mid‑day and get wet feet. How could these fires start or carry in these conditions? Man‑made prairie fires are in a narrow window for date, time, and dew point. Add to that the chance of a lightning strike and the odds are indeed small of having a natural fire that would spread. Most lightning in the tall-grass prairie area happens when the fuels are at least partly green. We believe that lightning‑started prairie fires were very rare.
If this paradigm was true, natural fires would occur on large prairie tracts like the Loess Hills prairies in Iowa or the Konza Prairie in Kansas often enough to burn the majority of the area in every three year period. It does not! Nor do the large prairies in Iowa.
I recently had a conversation with a rancher from Thomas, in western Oklahoma, about natural fire on the bluestem rangelands in his area. He related that in 40‑50 years of memory he considered natural range fires to be very rare and he couldn’t recall any particular occurrence.
Suppression of these very rare fires is said to be the cause for current problems with woody growth.
Paradigm 3: All of the prairie must be burned to achieve management goals. In practice unburned patches which are naturally left in a burn are sought out and burned.
Response: Even today we hear horror stories of whole prairies being burned. But even in such cases patches are naturally left unburned. These natural survivia in a burn are one of the opportunities for biodiversity to be supported. In a prehistoric prairie situation when there were more wetlands and sloughs on the landscape, fast‑moving fires surely left large and small patches scattered throughout the burn that were survivia for invertebrates.
To intentionally go in after a burn to seek out the unburned spots demonstrates an attitude that fire management is the goal and not the tool, as a doctor might treat all diseases with a single favorite prescription.
Paradigm 4: Fire removes woody growth.
Response: In many, if not most cases we have observed in Iowa, fire has not eliminated woody growth. Sumac and aspen are stimulated to produce more stems when they are burned. Sumac stands in the loess hills often look better a few years after burning as old and dead plants are removed to the benefit of young shoots. In some cases a miniature three year forest cycle has been achieved but the woody plants are not gone. It could be argued that in fact the use of fire alone stimulates woody growth by selecting for fire tolerant variations in woody populations. Williams Prairie in Johnson County, Iowa, which has a woody growth problem that has only been controlled by hand work, yet the grazed field to the south which is nearly brushless, offers an interesting contrast in woody growth. We believe that the presence or absence of grazers has a lot to do with control of woody growth. Fire disturbance can even encourage the growth of some undesirable species such as leafy spurge, sweet clover, sumac, aspen and poison ivy.
During the spring of 1996 a large fire burned nearly 1100 acres of Bluestem Prairie near Glyndon, Minnesota. A walk-through in July revealed that nearly all woody plants were NOT killed but were resprouted. Fire did not kill willows, aspen, honeysuckle, cottonwood, bog birch or poison ivy.
Unburned prairies can be seen that have not been overgrown with woody plants. Grazing is usually key to that situation. Consider the effects of bison and elk life cycles on the herbaceous and woody plants on a prairie. They also produce many of the same desirable effects as fire. Rodents are often considered to eat more plant matter that even the bison did. A new paradigm might see the role of fire as it interacted with other forms of disturbance rather than being the only form of disturbance that is important.
Paradigm 5: Prairie fires invigorate new growth making the prairie look like a fresh flower garden.
Response: Fire opens the soil surface to the sun to warm the soil causing a flourish of plant growth that makes the prairie look like a beautiful, even‑standing field. Is this look one of the objectives of management? I believe that we have been sold the idea that a prairie should look like a naturalized flower garden. Look at recent books or calendars on tall‑grass prairies. You will find that most of the photographs were taken the year of a burn. In fact, this has become the paradigm most accepted and I believe most dangerous to invertebrate diversity. With large grazers and infrequent fire on the prairies, is this what they really looked like? The action of grazers made the prairie more heterogeneous not less so as fire does.
Some plant species do not respond favorably to fire and it may change bloom patterns (purple prairie clover) or even inhibits flowering (phlox and the western prairie fringed‑orchid). In the case of prairie annuals repeated burning eliminates the seed supply.
Much research has be done to prove that there is increased biomass after burning and that the species composition can be altered. But what does this have to do with biodiversity? Or is tons of grass per acre the objective? Several recent investigations show that the most productive grasslands are not the most diverse.
Paradigm 6: Prairies should be divided into 3 parts with one part to be burned each year to simulate a natural occurrence.
Response; Prairies are indeed patchy but 3 parts do not allow for a continuation of the natural biodiversity. These units are usually set up without regard for the natural patchiness of the site. For example, if there were only one or two patches of a particular species of plant, those patches should not be in the same burn unit. Each plant species may host obligate prairie insects that could be wiped out if the whole patch of plants is burned. Division in three parts has the effect of reducing the numerous patches of diversity to only three, each maintaining only those species that are in a life cycle stage that can resist fire in the spring. Management patches need to be smaller than the natural patchiness or made to overlap portions of multiple natural patches.
Paradigm 7: There is a period of time in March or April when prairies can be burned without negatively impacting invertebrates.
Response: Insects stagger their life cycles in most habitats to partition the resources. At any one date there are insects in all life stages on the prairie. Some will not be effected but many will. Continued burning in the same place on the same date alters the species composition, probably permanently, in this prairie island situation.
Paradigm 8: All prairie invertebrates have been naturally selected over millennia to withstand fire.
Response: While it is true that these animals have survived on a landscape basis, individuals cannot be burned and survive. They survived because the whole landscape was not burned and/or the natural burns were patchy, leaving areas from which recolonization could occur. Patchy management is the only way to keep them in this non‑prairie landscape situation.
Paradigm 9: Small sites cannot hold rare biodiversity for long periods.
Response: Most of the prairies we have in Iowa have come down to us through 100 years of haying, grazing or by being ignored. These small unmanaged sites hold rare plants and insects today. Railroad prairies are as good as managed prairies: Hayden Prairie with its lack of some prairie species as opposed to the unmanaged Florenceville Prairie with those species; Gitchie Manitou State Preserve where the Prairie Ringlet and the Regal Fritillary were lost in the 70’s versus the Granite site where they remain in a pasture. An isolated small unmanaged prairie in Linn County, Iowa has maintained a population of the threatened Byssus Skipper, probably since settlement, while the nearby fire managed Rock Island Preserve prairie has not maintained the species. The challenge is to manage as well as that first 100 years of “neglect”..
Small fens in Iowa make up .01% of the landscape but 44% of the 320 +‑ fens in the state hold 19% of our rare plants, 50% of our rare butterflies and 30% of our rare snails. The :management of our fen biota is even more delicate than prairies.
Paradigm 10: Fire removes thatch.
Response: Grazing also helps to remove thatch. And thatch is not all bad. The greatest number of prairie invertebrates are involved in the decomposition of the organic matter in the thatch. A lack of thatch means that the soils are open to the seeds of invading species as well as native prairie plants. The Prairie Moonwort is an example of a rare prairie fern species that is thatch dependant. Henslow=s Sparrow, the prairie bird declining most precipitously, requires heavy thatch.
Paradigm 11: By managing to preserve the prairie on the landscape level we will also preserve the rare prairie elements.
Response: Can we focus on saving the prairie landscape and in that way save the prairie’s biodiversity? Studies in England have shown that the insects have declined despite management to preserve plant communities (42% loss of butterflies in Suffolk).
Remember that species go extinct one site at a time until they are gone from the landscape. This paradigm is exactly the opposite of what it will take to save prairie biodiversity.
Paradigm 12: Prairies are static systems where woody species are detrimental to the long‑term maintenance of prairie species.
Disturbance and the interaction of prairie species with disturbance loving species is responsible for much of the diversity on the prairie. Many species use the edges of woody patches that cannot find a foothold in pure prairie and many insects that pollinate prairie plants carry out parts of their life cycles on woody plants. The absence of buffalo wallows, the species that used them and the species that then reclaimed them is one example of this lost biodiversity.
It has become apparent that the pollinators of some herbaceous prairie plants use woody species as their larval foodplant. The larval foodlant of the pollenator of the Western Prairie Fringed Orchid is a woody plant that grows in patches. For this relationship to persist both plant types must be present in the landscape.
From our experience on tall-grass prairies the following are less dangerous than is fire for prairie butterflies.
‑association with active highways and railroads
Note: What about transplanting lost species back to the prairie? Some experiments have been conducted in Illinois on transplanting prairie butterflies, but Illinois has few good species left on their prairies to transplant, whereas Iowa still has about 10 prairie obligate butterfly species. Save them in place first, when they are safe in there survivia consider re-establishment to damaged areas.
1. Survey first: what is there, where is it, when is it active or growing?
a.non-vascular and vascular plants.
b. invertebrate animals, including soil and duff species, snails and insects.
2. Determine goals and objectives based on your survey. Is it tall plants, blooming flowers, or biodiversity?
3. Divide your prairie into as many management plots as possible so as not to include all of one habitat patch or plant type in any one plot.
Long narrow units that randomly cross the site (depending on wind direction in the case of fire) would afford a high degree of heterogeneity.
4. Water down or otherwise protect small areas of critical habitat within a larger management plot.
5. Plan so that plants will be available in all stages of their life cycle at any one time. Invertebrates need all stages from new sprouts to dead standing stems.
6. Don’t burn all of one watershed, but part of each.
7. Except in emergency, burn in 5‑10 year cycles. Remember you have several other effective management options available and that not doing anything is one of those options.
8. Don’t burn without a compelling reason: on small plots any action may cause an irreversible change. Excess disturbance is not a normal condition in low energy ecosystems (as opposed to beaches and dunes).
9. Randomize your management as to type (mowing, grazing, and burning), time (season), and placement.
10. Never burn, mow, or graze the whole thing: that effectively results in one patch with no or little biodiversity.
11. Don’t burn or mow adjacent units in consecutive years. For one unit to provide individuals to re‑establish populations in another unit the populations need more than one year to recover.
12. Determine the effectiveness of your management based on your goals at 1, 3, and 5 years.
13. A document should be kept for each prairie. It will include its history, original condition, goals and objectives for management, management used, by whom, when and to what extent. What are the outcomes of management?
14. Try a strip burn arrangement. Burn 6‑8 narrow strips across the prairie each year from different sides, depending on wind direction. This will maximize perimeter and patchiness.
15. Managers should be evaluated on preserving diversity and therefore be held responsible for losses, not on how many acres they burn.
We believe that a massive experiment is being conducted on our prairie biodiversity, an experiment It is however without controls: no parts are willfully left untested. We risk of losing organisms that we know almost nothing about. This is not good science, it is irresponsible behavior without anyone being responsible. We rarely get the opportunity to survey a prairie “before” it is burned in our time, so we don’t know what the pieces of the puzzle are!
Future research is being jeopardized by fire sorting of the fauna. Studies of genetic isolation, climatic effects, successional effects, preserve size, recolonization and biodiversity richness and comparisons to other sites are all compromised.
We don’t expect a change to take place very fast. First, most of your prairie paradigms don’t allow you to be comfortable with this information, and second, there isn’t a mechanism in place to change what is now happening. Time, money, and labor are always in short supply. While consternation and excuses can be heard at prairie meetings and workshops, the tendency is to claim it’s all too complicated and give up, exclaiming that we just can’t save everything.
If the reason or at least one of the reasons for preserving prairies is biodiversity, we need to reconsider our methods and consider a new look at prairies with management options to achieve that objective.
Clements, Frederic E., and Victor E. Shelford, Bio-Ecology. New York:John Wiley, 1939.
Hamilton, Ray., Native Prairie Management Guide, Iowa Prairie Network, Maquoketa, Iowa. 1994.
Kirby, Peter. Habitat Management for Invertebrates:.., Joint Nature Conservation Committee, Bedfordshire, 1992.
Kuhn, Thomas S., The Structure of Scientific Revolutions, Chicago: University of Chicago Press. 1962
New, T. R., Butterfly Conservation, Oxford University Press, Oxford, 1991.
Schlicht, Dennis, John Downey and Jeff Nekola, 2007. Butterflies of Iowa, U. of Iowa. Press, Iowa City.
Shimek, Bohumil., The Plant Geography of Iowa, University of Iowa Studies in Natural History, XVIII, 4, Iowa City, 1948.
Slobodkin, Lawrence B., “On the Susceptibility of Different Species to Extinction: Elementary Instructions for Owners of a World, in The Preservation of Species, The Value of Biological Diversity, Bryan G. Norton, Ed., Princeton: Princeton University Press, 1986.
Swengel, Ann B. and Scott Swengel, Research on the Community of Tallgrass Prairie Butterflies, Baraboo WI. 1993. (Unpublished)
Swengel, Ann B., 1996, Effects of fire and hay management on abundance of prairie butterflies, Biological Conservation 76 (1996) 73-85.
Williams, Andrew H., 1996. Conservation of the Plant Onosmodium molle and the Beetle Longitarsus subrufus in Wisconsin, Masters Thesis, University of Wisconsin, Madison, 424pp .
Updated April 2010.