Oak savannas were common throughout the edge of the prairie-deciduous woodland border area of Minnesota before European settlement. The habitat is typified by scattered oak trees, usually bur oak (Quercus macrocara) and northern pin oak (Q. ellipsoidalis), and large open areas of prairies (Tester 1995). Multiple factors probably affected the distribution of oak savannas in the area, including climate, fire, soil type, soil drainage, and topography (Grimm 1984). Fire, however, was probably the most important factor determining the distribution of this habitat, and the fire regime, in turn was effected by multiple factors.

Prairies covered much of Minnesota during the post-glacial maximum of warmth and dryness around 7200 years ago, except in a few well fire protected sites, which harbored forests. Around 5000 to 2400 years ago, woodland expansion began with Quercus as the primary invader, and was dominant woodland taxon until the "little ice age," 300 years ago. Other tree species expanded drastically at this time due to the climate becoming cooler and wetter. The woodland species, however, could not easily invade the prairie because of the prairie's susceptibility to fire (Grimm 1984, Clark 1988). Thus a prairie woodland border evolved in this area. Oak savannas were established along this border, especially in areas without definite firebreaks, a barrier between the two habitats, as oaks and the other associated vegetation are less susceptible to fire than the species of the deciduous forests (Grimm 1984).

The role of fire in maintaining oak savannas can be observed through their recent disappearance in Minnesota since fire suppression began. Very few areas are left today that fit the description of an oak savanna. Before European settlement fires swept through oak savannas (and the associated prairies) annually to biennially (Tilman et al 2000, Grimm 1984). Numerous early European settlers and travelers recount wildfires in the southern Minnesota (Grimm 1984). Fires usually occurred in the spring or autumn when vegetation was, for the most part, dry (Clark 1989, Grimm 1984). Fire suppression throughout Minnesota began around 1850 through 1900 as settlement increased and a need arose to protect crops, buildings, and livestock from being engulfed by flames (Tester 1989). With the absence of fire, the understory has grown in and non-fire resistant species have taken hold. The former savannas now resemble oak forests (Peterson and Reich 2001, Tilman et al 2000). This is in part due to the fact that the precipitation and soil moisture of many of the areas where oak savannas historically existed is high enough to support forests (Peterson and Reich 2001).

The dominant vegetation of savannas can be classified as resistors and/or endurers in regards to their strategy towards fire. Resistors, such bur oak, are shade intolerant, have thick insulated bark, can survive low-severity fires with little damage, have wood that resists decay after fire scarring, and may have a long life span. Northern pin oak is slower to develop resistance to fire but can be classified as a resistor once they reach sufficient size (Peterson and Reich 2001). Endurers resprout damaged parts after a fire, and include oaks, many shrub species, and grasses (Peterson and Reich 2001). Both strategies allow for the perpetuation of these species after fires and the maintenance of savannas. Species common to the deciduous woodlands, such as maples (Acer sp.), are often avoiders, i.e. they are shade tolerant, fire sensitive species (Peterson and Reich 2001).

Fire greatly effects the make-up of an oak savanna beyond just species composition. In areas exposed to pre-European settlement fire regimes in the study area of Cedar Creek Natural History Area, a decrease in overall tree density and basal area, a decrease or absence of seedlings, sprouts, and saplings; decreased or absence of canopy ingrowth, and a decrease in nitrogen and organic matter content in soils (Peterson and Reich 2001, Tester 1989). Low tree density and basal area, low seedling, sprout, and sapling densities, and little canopy ingrowth are all characteristics of savannas. These characteristics allow for grasses, herbs, and forbs to become a dominant part of the landscape in these areas. Decreased nitrogen and organic matter relates to the frequency of fires and the low total biomass of this habitat type.

The probability of fire in any area is related to a variety of factors, including climate, firebreaks, topography, soil, vegetation type and patterns, and human interactions. Climate contributes to fire regime trends over the short and long term. Centuries of dryness are correlated to centuries with high fire occurrence (Clark 1988). Over the shorter term, fires tend to increase after years with low annual water balance (Clark 1989). Climate may have been especially important in the establishment of oak savannas in the area, especially at the time of major species shifts during the "little ice age." However the general climate of the area during the nineteenth century was relatively constant and did not greatly contribute to the distribution of oak savannas at this time (Grimm 1984).

Firebreaks, greatly contributed to the occurrence of fire in the area, and hence the distribution of oak savannas. Firebreaks usually consist of water bodies and other topographical features that shield areas from fire. Throughout much of southern Minnesota during the nineteenth century, woodlands existent on the protected side of the firebreak, whereas prairies and savannas existed on the other. Savannas also were common along the prairie-deciduous woodland border where distinctive firebreaks did not exist; they served as a buffer between the fire-tolerant prairie and the fire-intolerant deciduous forest (Grimm 1984).

Topography is another important factor contributing the probability of fire. Fire spreads more readily and rapidly along flat topography as compared to rough topography. Hence prairies and savannas occurred more often on flat plains than rugged hills. Uneven topography creates irregular and slower wind movements, and has discontinuous fuel supplies, leading to more protected areas.

Soil also contributes to the probability of fire in an area. Coarse textured soils can hold less water than fine textured soils. Thus vegetation in coarse textured soils dry out faster during times of drought, increasing the chance of fire. However, there tends to be less vegetation in coarse textured soils, hence less fuel. The overall contribution of soils is to fire probability is not as high as firebreaks and other factors (Grimm 1984, Dorney and Dorney 1989).

The vegetation greatly influences the fire probability as well. The above ground plant matter of grasses and herbs, for instance, die late summer to early autumn every year. This creates a large amount of flammable biomass during this time and in the spring before the plants begin to grow, and contributes to the fires on prairies (Grimm 1984). Oak savannas also have a large proportion of grasses and herbs, and hence may also experience this phenomenon. The presence of shrubs however helps to retain moisture and allows fire to be less prevalent or, at least, less intense than prairie fires. The absence of fire however, allows dead biomass to accumulate and increases the probability of an high intensity fire. When an intense fire comes, it removes the fire-intolerant species and helps rejuvenate the oak stand (Grimm 1984). Restoration ecology today is focusing on the proper spacing of burns to maintain oak savannas. Studies show that around three fires per decade seem necessary to prevent the development of a sapling layers and canopy ingrowth. Any more burns and it may lead to an unsustainable oak population, as the trees cannot handle that many fires (Peterson and Reich 2001). In contrast, deciduous forests have a dense shade over the forest floor, which deters the growth of herbs and shrubs, thus limiting fuel accumulation. Biomass turnover occurs quickly, as well, further decreasing the potential fuel supply (Grimm 1984). Thus the vegetation helps perpetuate the habitats that were established due to climate, fire regimes, and other factors.

There is debate over the role of Native Americans in maintaining oak savannas and prairies via prescribed burning, or at least careless fire practices. Both the Sioux and the Ojibwe, the two main Indian groups in Minnesota during the nineteenth century, used fire for hunting. The Sioux also used fire for warfare. There is not documentation for other common uses of fire, such as clearing land (Clark and Royall 1996). Through charcoal analysis, Clark and Royall have shown a correlation between the expansion of oak savannas and the presence of Indians in Ontario (1996). Dorney and Dorney also suspect that Native Americans contributed to the presence of an oak savanna in northern Wisconsin mainly because the habitat type is not found anywhere near the area – northern Wisconsin is dominated by mixed conifer – hardwood forests, not oak forests as in southern Wisconsin (1989). Grimm states that Native Americans caused most tallgrass prairie fires, either accidentally or purposefully (1984). According to Grimm, lightening strikes are very uncommon during the burning season and if they occur almost always occur in heavy rain, hence Native Americans must be a major cause of fires (1984). Thus it seems likely that Native Americans also caused most fires in oak savannas, directly or indirectly.

Fire was a major factor in the distribution of oak savannas before European settlement of Minnesota. Oak savannas were common in areas between the prairie and deciduous forests, and often served as a buffer. They did not burn as frequently and/or intensely as prairies. The fire regime of oak savannas, however, removed non-fire tolerant species, limited canopy ingrowth, and prevented a sapling understory from forming, while allowing for the perpetuation of oak trees and some shrub species. The advent of European settlers and the resulting fire suppression has lead the widespread decline of this habitat type in Minnesota.

Works Cited
Clark, JS and PD Royall. 1996. Local and regional sediment charcoal evidence for fire regimes in presettlement north-eastern North America. Journal of Ecology 84: 365-382.

Clark, JS. 1988. Effect of climate change on fire regimes in northwestern Minnesota. Nature 334:233-235.

Clark, JS. 1989. Effects of long-term water balances on fire regime, north-western Minnesota. Journal of Ecology 77:989-1004.

Dorney, CH and JR Dorney. 1989. An unusual oak savanna in northeastern Wisconsin: the effect of Indian-caused fire. American Midland Naturalist 122:103-113.

Grimm, EC. 1984. Fire and other factors controlling the big woods vegetation of Minnesota in the mid-nineteenth century. Ecological Monographs 54:291-311.

Peterson, DW and PB Reich. 2001. Prescribed fire in oak savanna: fire frequency effects on stand structure and dynamics. Ecological Applications 11:914-927.

Tester, JR. 1989. Effects of fire frequency on oak savanna in east-central Minnesota. 116: 134-144.

Tester, JR. 1995. Minnesota's natural heritage: an ecological perspective. Minneapolis : University of Minnesota Press.

Tilman, D, P Reich, et al. 2000. Fire suppression and ecosystem carbon storage. Ecology 81:2680-2685.