Pre-settlement Vegetation of Greene, Jersey and Macoupin counties along the prairie/forest border in Illinois.
The Public Land Survey (PLS) was a survey conducted by the U.S. government (Government Land Office) on all lands west of the original thirteen states. Its purpose was to survey land prior to sale and subsequent settlement. The survey created a grid of square "townships" which were 6 miles (9.7Km) on a side. Each township contains 36 square "sections" a mile on a side. The sections are further subdivided into four 1/4 square mile "quarter sections". In the PLS "bearing trees" (sensu Grimm 1984) were blazed with an axe at each section and quarter section corner throughout the survey area to mark the land for later sale to pioneer farmers. The center section corner was not surveyed. Surveyors also identified and measured so called "line trees" that they found on section lines. The bearing trees were identified, measured, and located with respect to the quarter section corner. Line trees were not located with respect to quarter section corner, but with respect to the section line. The survey can be used today to reconstruct the vegetation present at the time of the survey--just prior to settlement. The PLS was completed in SW Illinois between 1818 and 1820. It provides a database that can be used to examine the relationships among vegetation, topography, drainage and soils. It also provides an environmental baseline for conditions at the time of the survey.
Over approximately the past eighty years, analyses of the original survey notes have been carried out in many states providing valuable basic ecological information that in turn has been used to reconstruct original vegetation patterns (Kilburn 1959; Grimm 1983; Brugam and Patterson 1996).
Over time in Illinois more and better uses of the PLS records have been made as researchers have been able to go well beyond the simple enumeration of the various tree species. Savannas (barrens) have been mapped (Kilburn et al. 2009). King and Johnson (1977) compared vegetation with topography in the Sangamon River watershed. Development of computer systems has enabled a useful analysis of these data (Grimm 1983; Schaetzl and Brown 1996; Schulte and Mladenoff 2005). The application of digital methods has allowed investigators to ask and answer questions about tree distributions at the time of the survey.
Analyses of these data have become increasingly sophisticated because of the application of Geographical Information Systems (GIS). In the Sylvania Wilderness Area, an undisturbed wilderness reserve in the Upper Peninsula of Michigan, investigators confirmed the accuracy of the PLS using GIS (Manies and Mladenoff 2000). Vadeboncoer et al. (2012) and Thompson et al. (2013) compared the PLS with the USDA Forest Service Forest Inventory Analysis data in the Northeast to determine how forests have changed since the PLS. Hanberry et al. (2014) completed a similar study in Northern Missouri. Thomas-Van Gundy et al. (2015) categorized the fire adaptations of the PLS survey trees in the Northeast to reconstruct the fire frequency in various landscapes. In Illinois 48 of 103 counties have been either completely or partially analyzed (Kilburn and Brugam 2014). However, most studies are too old for GIS analysis. In Illinois, the original surveyor plat maps have been georeferenced for GIS and published in CD format (Illinois Natural History Survey 2005). Data analysis in Illinois has not yet been as sophisticated as in Missouri and the Northeast because, as yet, there is no state-wide database available for analysis.
This paper analyzes the pre-settlement forest-prairie transition east of the Illinois River ranging from the heavily forested western portions of Greene and Jersey Counties to the upland prairies of Macoupin County in the east (Fig. 1). This paper is based on the original survey notes and plat maps recorded by the Deputy Surveyors of the Public Land Survey (PLS). Our goal is to use the PLS in Green, Jersey and Macoupin counties to understand the causes of the complex pattern of prairie and forest in these southwest Illinois counties. We will compare vegetation, topography and soils to reconstruct the patterns of vegetation distribution. We will extend the work of King and Johnson (1977) testing the significance of topography and soils in controlling tree distribution. We will expand the study area and apply modern data analysis techniques to the dataset.
Location. The study area covers the counties of Jersey, Greene and Macoupin in southwestern Illinois (Fig. 1). The area ranges approximately 42 miles (66 km) north and south and 52 miles (83 km) east and west. This tri-county area lies east of and adjacent to the southern portion of the Illinois River and a few miles north of the Mississippi River which here is flowing mainly west to east. The Mississippi forms the southern border of Jersey County.
The study area has a continental climate typical of SW Illinois (Natural Resources Conservation Service 2007). The average monthly range of high temperature is from 40[degrees]F (4.4[degrees]C) in January to 90[degrees]F (32.2[degrees]C) in July. The lower average temperature ranges from 21[degrees] F (-6.1[degrees]C) in January to 66[degrees]F (18.9[degrees]C) in July. Total average precipitation is 34.3 inches (87.1 cm) with the lowest monthly average 1.6 inches (4.1 cm) in February and highest average of 4.4 inches (11.2 cm) in June. The growing season is approximately 180 days.
Soils and Topography. Floodplains lie adjacent to the Illinois River extending for one to eight km east These are covered by alluvial soils (Natural Resources Conservation Service 2007, 1974). Bottomland soils extend east along major streams sometimes through the Greene and Jersey counties and into Macoupin County (Natural Resources Conservation Service 2004). This is particularly true of Macoupin Creek (Fig. 1). There the floodplain is often more than 1.6 km wide through Greene County and into Macoupin County. Apple Creek which runs through Greene County turning north and having its headwaters in Scott and Morgan Counties to the north. Many other creeks run into these rivers but are much shorter, have steeper gradients and little or no alluvial soils.
To the east of the Illinois River floodplain limestone bluffs form escarpments rising up to 75m. These are capped with 15-20m of loess. These bluffs represent a boundary between the Illinois River floodplain and the uplands. They are penetrated by stream valleys. The loess cap thins sharply to the east until it is no more than an average of two meters in depth in Macoupin County at the eastern edge of the study area. This average depth varies and in some areas is completely absent exposing Illinoisan glacial age tills. East of the bluffs is a 6 to 9 km wide belt of very irregular topography with deep ravines.
Prairie and forest are mixed in the eastern portions of the county. The upland loess-capped prairie soils are easily distinguished from the forest soils by the black color imparted by the decay of prairie roots and above ground growth (mollic epipedon). These prairie soils (mollisols) are extensive on level to gently rolling lands. They generally have slopes less than four percent and are some of the most productive soils in the world. Where cut by streams, valleys and ravines the vegetation is dominated by forest and forest soils (alfisols).
Data were obtained from microfilm copies of the Public Land Survey Field Notes housed in Lovejoy Library of Southern Illinois University Edwardsville. The mapping program, National Geographic TOPO (version 4.5.0), was used to locate all outside quarter section corners and all line trees. TOPO is a digitized version of the U.S. Geological Survey 7.51 topographic map series that allows the user to mark locations and to enter notes at those locations. We transcribed data from the Field Notes into a file maintained in TOPO. The TOPO file recorded the coordinates of each quarter section corner or line tree. Trees were listed according to the species identified by the surveyors (Table 1).
The taxonomic abilities of the surveyors is unclear (Table 1). Most commonly names used by the surveyors have clear modern botanical meanings. "Spanish oak" is an exception. Edgin (1997) and Edgin and Ebinger (1997) in Illinois and Bragg (2002) in Arkansas interpreted this species as Quercus falcata. Stolynoff and Hess (2002) did a survey of the genus Quercus in Illinois and found the taxonomic confusion of Quercus falcata found in southern Illinois and Quercus ellipsoidalis found in northern Illinois. The PLS was conducted in winter when it was probably difficult to differentiate between the species. Here we follow the convention of Edgin (1997), Edgin and Ebinger (1997) and Bragg (2002) in calling the surveyor's "Spanish oak" Quercus falcata.
Where no tree was indicated, the quarter section corner was identified as prairie. In addition to coordinates and tree species, the diameter of each tree was noted in inches by the surveyors. Distance from the quarter section corner was measured in links and copied from the field notes. Slope was measured on TOPO and tabulated as a percent. Topography was determined for each section corner based on TOPO slope measurements.
The latitude/longitude-indexed spreadsheet that resulted from digitization of the vegetation data using TOPO contained 9,520 data points. At many quarter section corners surveyors listed 2 trees. We entered each tree as a separate data-point. Prairie points were entered as single data points. This data-set was exported from TOPO and converted into an excel spreadsheet. The spreadsheet was uploaded into MAPINFO 6.5, a geographic information system (GIS) to create maps of tree locations (Figs. 2 through 5).
Georeferenced digital maps of soil series for all three counties were obtained from the Natural Resources Conservation Service. These maps showed the locations of all soil series. The excel vegetation data-set was overlain on the soil maps using MAPINFO and the files were combined. The result was that the official soil series name was determined for each data point.
Non-metric multidimensional scaling (NMDS) ordination was performed on tree species and soils using PC-Ord version 6.01 (Peck 2010). Minchin (1987) showed that this non-parametric method is relatively robust to variations in community models. It is an appropriate choice for most ecological applications. In our analysis we used 41 soil series with > 10 quarter section corners and line trees as "plots" (Table 2). We also used 16 tree species present in [greater than or equal to] 25 soil series (Table 1). This number is smaller than the total number of trees in the dataset. The reduced dataset was used for NMDS analysis because many species from the total dataset were too rare for inclusion in the statistical analysis (Peck 2010). This use of ordination asks how tree species are organized among soils.
The total data-set includes 29 tree species that have more than 10 recorded individuals (Table 1). Not all species were included in the ordination. There were 2,234 prairie quarter section corners (Fig. 2). The 5 most abundant tree species (Table 1) were Quercus alba (1,759), Quercus velutina (1,657), Carya spp. (1,152), Quercus stellata (598) and Ulmus sp. (429).
Geographic Distribution of Prairie and Tree Species. We can locate trees on a map to show their geographic distribution. Quercus alba is abundant along the river bluffs and up river valleys (Fig. 2). A narrow band along the Illinois River does not have prairie or Quercus alba but does have floodplain species (Fig. 5). Our results show that prairie covers the uplands in the eastern 3/4 of the study area. Our prairie section corners coincide with the locations of prairies indicated in PLS summary maps. The prairie-forest borders on Figures 2 through 5 were traced from the Illinois Natural History georectified survey maps. (Illinois Natural History Survey 2005). These boundaries were copied by the INHS from survey maps prepared shortly after the completion of the original survey.
Carya spp. and Quercus velutina are less abundant in the east, which had large areas of prairie and forests of Quercus stellata and Quercus palustris. Quercus alba and Carya spp. were more common in the western part of the study site along the Illinois River bluffs (Fig. 3). The trees extend up stream valleys on more sloping ground.
Quercus stellata, Quercus marilandica, and Quercus palustris are abundant on the edges of prairie (Fig. 4). All three species are found in narrow areas of forest surrounded by prairie. They do not appear in the 5 to 13 km of land between the river bluffs and the westernmost extension of prairie. This belt is populated with Quercus alba, Quercus velutina and Carya spp. (Figs. 2 and 3). Quercus stellata, Quercus palustris and Quercus marilandica segregate themselves geographically. Quercus stellata is most abundant in the forested areas on the southern border of the study site and in the upstream parts of Macoupin Creek and its tributaries. Quercus palustris is abundant along the downstream parts of Macoupin Creek in areas of the floodplain surrounded by prairie. Quercus marilandica is widely scattered in the same areas as Quercus stellata and Quercus palustris.
Salix sp. and Populus deltoides are all located near the Illinois River. Ulmus sp. are scattered throughout the study area (Fig. 5).
Size, Distance and Slope. Some investigators (Fralish et al. 1991; Anderson et al. 2006) have used the distances of trees from quarter section corners to estimate the density of trees in the pre-settlement forests using various plotless sampling techniques. We have not followed that approach because our large study area makes it difficult to relate a tree density to a county-wide area. However, distance from the section corner does indicate distance between forest trees of particular species and, thus is a rough indicator of fire frequency. It is likely to reveal responses of species to fire. Woodlands burned frequently are more open than ones that have less frequent fire (Fralish et al. 1991; Anderson et al. 2006). The 5 species located, on average, farthest from the quarter section corner were Quercus marilandica (28.6[+ or -]2.2m), Quercus stellata (24[+ or -]2m), Quercus palustris (18[+ or -]2m), Quercus velutina (16[+ or -].6m) and Carya spp. (16[+ or -].6m). Thomas-Van Gundy et al. (2014) classified these species as "pyrophyllic". They are all highly fire tolerant species. The 5 species located nearest to the quarter section corner were Cercis sp. (4[+ or -].6m), Morus sp. (5[+ or -].8m), Acer saccharum (6[+ or -].6m), Juglans cinerea (6.5[+ or -]1m) and Acer sp. (7[+ or -].6m). All are "pyrophobic" using Thomas-Van Gundy's et al.'s (2014) scheme.
The data-set also contains information on tree diameter. The five trees with largest diameter are Quercus macrocarpa (62[+ or -]16cm), Quercus falcata (56[+ or -]9cm) Quercus lyrata (52[+ or -]6cm), Populus deltoides (48[+ or -]5cm) and Platanus occidentalis (48[+ or -]6cm). The 5 species with the smallest diameter are Cornus sp. (19[+ or -]2cm), Cercis sp. (20[+ or -]5cm), Aesculus sp. (24[+ or -]5cm), Quercus marilandica (28[+ or -]2cm) and Morus sp. (28[+ or -]6cm).
A plot of tree diameter against distance from the quarter section corner indicates a wide range of variation in our data-set (Fig. 6). Quercus marilandica individuals are farthest from the section corners but each tree species has a relatively small diameter. Quercus falcata, Quercus macrocarpa and Quercus lyrata are some of the largest diameter trees in the data set, but they are not particularly far from the quarter section corner. Quercus alba and Quercus velutina are very common in the data-set, but neither is particularly large or widely spaced. Quercus stellata is the second most widely spaced tree in the data-set. The smallest trees nearest to the quarter section corner are Cornus sp., Cercis sp. and Aesculus sp. An exception is Quercus marilandica whose diameter is among the smallest in the data-set, but whose distance from the quarter section corner is the largest.
Slope, Species and Soils. King and Johnson (1977) working in the Sangamon River Valley argued that slope is a proxy for fire frequency because areas of steep slopes represent fire breaks. For this reason, we recorded the slopes of trees and prairie points in our study area (Fig. 7). The five trees located on the highest slopes were Cornus sp. (10.3[+ or -]4.0), Sassafrass sp. (6.9[+ or -]2.6), Quercus alba (6.0[+ or -]2.6%), Quercus falcata (5.2[+ or -]2.6%) and Acer saccharum (4.8[+ or -]1.8%). Plants on the shallowest slopes were Acer negundo (.2[+ or -].3%), Acer saccharinum (.2[+ or -].3%), Salix sp. (.2[+ or -].3%), Populus deltoides (.8[+ or -].6%), and prairie grasses (1.0[+ or -].8%).
In addition to relating species to slope, it is important to relate slope to soils. The Soil Surveys for all three counties (NRCS 1974, 2004, 2007) provide a range of slopes for each Official Soil Series Description. However, we noted the slope and soils for each of our quarter section corners and line trees so that we could directly compare soil, slope and trees. The five steepest slope soils are: Sylvan (13.0[+ or -].6%), Fayette (7.4[+ or -].2%), Memfro (7.2[+ or -].5%), Elsah (5.6[+ or -].9%), and Hickory (5.6[+ or -].2%). These soils vary from 88 to 100% forested. The soils with least slope are: Petrolia (.12[+ or -].06%), Beau coup (.16[+ or -].05), Darwin (.17[+ or -].1%), Titus (.22[+ or -].08%) and Sable (.44[+ or -].1%). These soils are very level. Petrolia, Beaucoup and Darwin soils are classified as fluvaquentic in the soil surveys (Table 2). All are located adjacent to streams and are heavily forested with floodplain species. Soils that supported prairie are nearly level but often more sloping than the floodplains. These soils are Virden (.44[+ or -].04% slope, 92% prairie), Cowden (.48[+ or -].06% slope, 99% prairie), Herrick (.53[+ or -].03% slope, 93% prairie) and Ipava (.61[+ or -].06% slope, 83% prairie).
Pre-settlement Vegetation and Soils. Because georeferenced soil surveys are available for the three counties in this study, it is possible to relate vegetation with official USDA-NRCS soil descriptions. We can determine whether species found in the PLS are associated with particular soil series. Figure 8 shows an NMDS ordination of soil series versus the vegetation found on that soil. The data points are marked to differentiate among mollisols, alfisols and wetland soils. The species plot of the NMDS (Fig. 8) shows that prairies are plotted at high values on Axis 1 among mollisols. An exception is the Cowden soil which is classified as an alfisol, but had 99% prairie. Axis 1 is positively correlated with percent prairie points associated with a particular soil (Fig. 8). Mollisols lower on Axis 1 support Populus deltoides, Fraxinus spp., Acer negundo, Ulmus sp. and Acer saccharum. It is also clear from Figure 8 that mollisols have high Axis 2 values and alfisols have low values. Oak species are all located at negative values of Axis 2 among alfisols suggesting an association between woodlands and alfisols.
Our hypothesis is that the distribution of vegetation in the study area is controlled by fire frequency which will vary in different geographical areas as a result of topography. King and Johnson (1977) hypothesized that variations in slope will result in firebreaks. Furthermore, the effectiveness of firebreaks will vary with topography. Differences in topography will limit the distribution of fire-sensitive species to particular low-fire locations.
Our maps of the three counties support this hypothesis. Prairie is very fire tolerant requiring fires at frequent and sometimes nearly annual intervals (Grimm 1986, Kilburn et al. 2009, Kilburn and Brugam 2010). Our maps show a series of bands of vegetation along the Illinois River and extending eastward across the counties (Figs. 2 to 5). These bands (Fig. 9) are based on the distribution of vegetation that we found. They reflect differences in topography. They are very similar to the physiographic regions of the state outlined by Schwegman (1974).
Nearest to the Illinois River is a band about 8 km wide that supported flood plain species including a group of species that are highly intolerant of fire. In this location the topography is very level (Floodplain mean slope .86[+ or -]17%, Fig. 10), but fire is suppressed by broad expanses of moist floodplain lakes, and marshes. This region has large areas of frequently flooded soils (Natural Resources Conservation Service 1974, 2004, 2007). There was usually a strong spring flood along the Illinois River (Junk et al. 1989).
East of the Illinois River bluffs, on hilly land, was a 13 km band of moderately fire tolerant oak and hickory species--Quercus alba, Quercus velutina, Carya spp. (Fig. 3). This region (Fig. 9) has the highest mean slope of the dataset (6.3[+ or -].14%, Fig. 10). The steep slopes allow only occasional fires by providing firebreaks. King and Johnson (1977) and Brugam and Patterson (1996) emphasize the role of steep slopes in preventing the spread of fires. On flat land all areas will dry at a similar rate with large areas being capable of supporting fire at the same time. However, in broken topography different patches will dry at different rates reaching combustibility at different times. This difference in patches will impede the spread of fires.
Still further to the east highly fire tolerant oak species (Quercuspalustris, Quercusstellata and Quercus marilandica) are added to Quercus alba, Quercus velutina and Carya spp. (Fig. 9) These are intermixed with prairie on flat uplands and with forests primarily confined to the slopes. Forests are located on intermediate slopes (2.2[+ or -].05% slope, Fig. 10). Prairies are located on level uplands with low slope (.99[+ or -].03% slope, Fig. 10). These level uplands had few firebreaks and supported frequent fires. It is likely that these fires burned into adjacent forests, establishing forests of relatively high fire frequency. King and Johnson (1977) sought to relate fire frequency to slope. Our dataset supports their hypothesis, but because it is larger and more topographically complex it reveals more about the ecology of the southwestern Illinois landscape. King and Johnson (1977) argued that because of the lack of firebreaks, prairie developed on flat uplands. This was also the case with our dataset.
The floodplain along the western border of the study site is even more level than the prairies. This region was covered by flood-tolerant species like Acer sp., Salix spp. and Populus deltoides. Their presence suggests that these sites were frequently flooded. There were many section corners without trees on the floodplain but it is unclear whether these locations were true prairies or marshes. True prairie is found on slightly higher average slope.
Trees on the steepest slopes were fire intolerant species (Cornus sp, Sassafras sp. and Acer saccharum). Quercus alba and Q. falcata are the only oaks inhabiting steeper slopes. Our results support King and Johnson showing prairie on level ground without flooding. However, our results support the basic understanding that the floodplain environment is very different from uplands because of periodic flooding and poorly drained soils and resulting in strikingly different vegetation. Flooding can be added to the other abiotic controls on forest composition.
Quercus stellata is a species which is frequently found in a peculiar Illinois vegetation type called the "Illinois Post Oak Flatwoods" (Coates et al. 1992, Taft et al. 1995, Edgin et al. 2003, Taft 2005). This vegetation type grows on relatively flat topography where soils are underlain by a "hard pan" that inhibits drainage through the soil column.
Distance, Size and Slope. In our dataset, known fire-tolerant genera (Quercus and Carya, Abrams 1992, Abrams 2003) are generally far from quarter section corners indicating wide spacing in the forest. In contrast, closely spaced species are fire intolerant supporting the argument that the forests receiving frequent fires have more broadly spaced individuals. In addition the largest trees are mostly fire-tolerant Quercus. The smallest are mostly fire intolerant successional species. An exception is Quercus marilandica. In our dataset it is both far from the quarter section corners and small in diameter. Carey (1992) reports that Quercus marilandica is even more fire-tolerant than Quercus stellata and Quercus velutina. This species has a peculiar ecology because it root sprouts vigorously after fire. Kilburn et al, (2009) reported that it is a major component of Illinois barrens. These are areas of woody vegetation that support fires. However, fires in barrens are probably not as frequent as in prairies allowing some woody vegetation to survive. Figure 6 shows how anomalous Quercus marilandica is. On the basis of distance from quarter section corners, it seems to be adapted to a higher fire frequency than other tree species in the dataset.
Species, Slopes and Soils. Another question about vegetation in the study site is the relationship among vegetation slope, fire frequency and soils. How does soil develop in the Southwestern Illinois mosaic of prairie, forest and wetland? Brady and Weil (1996) identified 5 soil forming factors, 1) parent material, 2) climate, 3) topography, 4) organisms, 5) time. In our location parent material has limited effect because the study area is covered by at least two meters of loess nearly everywhere. To the east on slopes some Illinoisan till is exposed. Climate is similar across the counties and today's soil types developed in this overlying material. The time interval since loess deposition ceased is similar in all study locations. Topography and vegetation (organisms) vary across the landscape. It is clear that soil development will depend on the complex of factors described here and that soils, topography and vegetation in our study counties will not be independent variables. All are be interrelated and variations in any of these three factors will influence the others.
We can use NMDS ordination techniques to discover the relationships among Brady and Weil's (I996) soil forming processes in Greene, Jersey and Macoupin Counties. Mollisols are darkly colored soils (mollic epipedon) with high amounts of organic matter (Foth 1984). These include prairie and wetland soils. Alfisols are light colored soils (ochric epipedon) that develop under forests (Foth 1984). NMDS ordination can be used to relate vegetation type to soils answering the question; are particular vegetation types associated with particular soil series? Each soil series is used as a sample "plot" for vegetation (Fig. 8).
The ordination associates soils with vegetation. The result of the ordination is that prairie vegetation clusters at high Axis 1 values. Forests have low Axis 1 values. Oak species of mesic upland forests are located on alfisols and have low Axis 2 values. Trees of wetland forests have high Axis 2 values and are located on mollisols.
The average percentage of prairie section corners is strongly correlated with Axis 1 of the ordination (Fig. 8). High prairie percentages are associated with high axis 1 values. These are located on prairie soils. There is a weaker correlation between average percent slope and Axis 1 (Fig. 8). The heavily forested soils (Sylvan, Fayette and Memfro) have the steepest average slope. The soils with highest percent prairie (Virden, Cowden, Herrick, Ipava) are also extremely level. It is clear that soil series are strongly associated with slope and vegetation type. It is likely that topography controls fire frequency which, in turn, controls vegetation type and soils.
Pre-Columbian Human Activity and SW Illinois Vegetation. Human activities are another potential influence on the vegetation mosaic in the study area. Attention has been focused on the extensive Indian settlements in and around Cahokia, on the Mississippi floodplain only about 30 km to the southeast (Emerson and Lewis 1991). This was one of the largest Indian settlements in North America at the time (1000 CE), with extensive cultivated fields growing four varieties of maize. Around 1500 CE the population declined, and their consequent influence on surrounding vegetation greatly diminished.
In contrast, the Illinois River Valley has been densely populated for at least 8700 years (Brown and Viera 1996). However, it is still unclear what part Native Americans have played in landscape management. Many investigators have argued that they used fire as a tool to drive game and to develop edible wild plant resources (review in Abrams and Nowacki 2015). The vegetation of SW Illinois could have been heavily modified by Native Americans.
There has been a long-running controversy among ecologists whether the prairies of Illinois are a result of climate or of human activities (Transeau 1935, Abrams and Nowacki 2015). Wright (1968) demonstrated that the climate of Illinois is more drought-prone than regions to the north east and south. The climate is more like regions to the west. Illinois has been termed the "Prairie Peninsula" (Transeau 1935) as a consequence of the climatic conditions.
It may be that the PLS data is a reflection of Native American use of the landscape. Unfortunately, there is little direct experimental evidence for or against Indian landscape modification, however, there is lots of anecdotal evidence for Native American use of fire. They might well have provided the ignition source for the highly fire tolerant vegetation that we have described. Grimm (1984) emphasized that Indians were a major ignition source for Minnesota forests. He emphasized that lightning strikes, though present, were rare and that Indian fires were likely to have been more frequent. Hotchkiss et al. (2007) working in the large Northwestern Wisconsin sand plain found that changes in the pollen record occurred at the same time as changes in Indian cultural practices. Munoz and Gajewski (2010) working in southern Ontario using pollen and charcoal analysis in lake cores showed increases in human indicators in pollen and increases in charcoal occurring together. Thomas Van-Gundy et al. (2014) associated pyrophyllic vegetation in the Northeast with likely villages and trails in the PLS record. This result shows increased fire frequency near Indian settlements.
The climate of Illinois likely permits fire as Wright suggested (1968), but the large population of people in presettlement times probably increased the frequency of ignition as it did in Minnesota (Grimm 1984). The SW Illinois vegetation was clearly heavily influenced by fire. However, we do not yet understand the role of active fire management in Illinois by Native Americans.
Our examination of the PLS record for Greene, Jersey and Macoupin Counties show a clear variation in vegetation from west to east that is probably dependent on variations in fire frequency. The western part of the study area, adjacent to the Illinois River had floodplain vegetation. The bluffs to the east of the River had forests of Quercus alba, Quercus velutina and Carya spp. These forests had an intermediate tolerance for fire. To the east of these woodlands was a mosaic of prairie and patches including very fire-tolerant trees like Quercus stellata, Quercus palustris and Quercus marilandica.
The variations in vegetation corresponds strongly with topography. Both the floodplain and the prairie are nearly level. The bluff lands have most variable topography. Upland forest have intermediate slopes. The differences in slope result in differences in fire-breaks and fire frequency.
Variations in topography result in variations in soil series. Varying soils support different vegetation. Thus, the complex variations in topography, vegetation and soils result in a complex mosaic of vegetation and fire-frequency.
Abrams, M. D. 1992. Fire and the development of oak forest. Bioscience 42:346-353.
Abrams, M.D. 2003. Where have all the white oak gone? Bioscience 54: 927-939
Abrams, M.D., and G. J. Nowacki. 2015. Exploring the early Anthropocene burning hypothesis and climate-fire anomalies for the Eastern U.S. J. Sustainable For 34:30-48.
Anderson, R.C. and M.R. Anderson 1975. The presettlement vegetation of Williamson County, Illinois. Castanea 40:345 363.
Anderson, R.C., S.L. Jones, and R. Swigart. 2006. Modifying distance methods to improve estimates of historical tree density from General Land Office survey records. J. Torrey Bot. Soc. 133:449-459
Brady, N.C. and Weil, R.R. 2007. The Nature and Properties of Soils, 14th Edition Pearson Education Limited, Essex, England.
Bragg, D.C. 2002. Checklist of major plant species in Ashley County, Arkansas, noted by General Land Office surveyors. J Ark Acad Sci 56: 32-41.
Brown, J. A. and R. L. Vierra. 1983. What happened in the Middle Archaic? An introduction to and ecological approach to Koster site archeology. J.L. Phillips and J. A. Brown. eds., Archaic Hunters and Gathers in the American Midwest. Academic Press, New York.
Brugam, R.B. and M.J. Patterson. 1996. Application of a geographic information system to mapping presettlement vegetation in southwestern Illinois. Trans 11l. Acad. Sci. 89:125141.
Carey, Jennifer H. 1992. Quercus marilandica. In: Fire Effects Information System, [Online]. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory (Producer). Available: http://www.fs.fed.us/database/feis/ [2015, May 4].
Coates D. T., K.J. Lyman, and J. E. Ebinger. 1992. Woody vegetation structure of a post oak flatwoods in Illinois Castanea 57:196-201.
Edgin B. 1996. Barrens of presettlement Lawrence County, Illinois. Proceedings of the 15th North American Prairie Conference pp. 59-65.
Edgin, B.R. and J.E.Ebinger. 1997. Barrens and the presettlement prairie/forest interface in Crawford County, Illinois. Castanea 62:260267.
Edgin, B., W.E. McClain, R. Gillespie, and J.E. Ebinger 2003, Vegetation composition and structure of Eversgerd Post Oak Flatwoods, Clinton County, Illinois. Northeast Naturalist 10:111 -p 118.
Emerson, T.D., and R. B. Lewis. 2000. Cahokia and the Hinterlands: Middle Mississippian Cultures in the Midwest. University of Illinois Press. 347pp.
Foth, H.D 1984. Fundamentals of Soil Science, Seventh Edition. John Wiley and Sons, New York 435 pp.
Grimm, E.C. 1984. Fire and other factors controlling the Big Woods vegetation of Minnesota in the mid nineteenth century. Ecol Monog 54:291 311.
Fralish, J.S., F.B. Crooks, J.L. Chambers, and F.M. Harty. 1991. Comparison of presettlement, second growth and old growth forest on six site types in the Illinois Shawnee Hills. Am Midl Nat 125:294 309 .
Hanberry, B. B., J.M. Kabrick, and H.S. Hu. 2014. Changing tree composition by life history strategy in a grassland-forest landscape. Ecosphere 5:34-44.
Hotchkiss, S.C., R. Calcotte, E.A. Lynch. 2007. Response of vegetation and fire to Little Ice Age climate change: regional continuity and landscape heterogeneity. Lands Ecol 22:25-41
Illinois Natural History Survey. 2005. Federal Township Plats of Illinois (1804-1891). geo-rectified. CD-ROM.
Junk, W., P.B. Bayley, and R.E. Sparks. 1989. The flood pulse concept in river floodplain systems. Can J Fish Aquat Sci Special Publications 106:110 127.
Kilburn, P. D. 1959. The forest-prairie ecotone in northeastern Illinois. Amer. Midl. Nat. 62: 206-217.
Kilburn, Paul, and R. B. Brugam. 2010. How natural is nature? The Confluence.Spring/Summer: pp 42-55.
Kilburn, P and R. Brugam 2014 Inventory of Vegetation Studies in Illinois Based on the Public Land Survey Records. . Trans. Ill. State Acad. Sci. 107: 13-17
Kilburn, P, B. Tutterow, and R. B. Brugam. 2009. The tree species composition and history of barrens identified by government land surveyors in southwestern Illinois. J. Torrey Bot. Soc. 136:272-283.
King, F. B. and J. B. Johnson. 1977. Presettlement forest composition of the Central Sangamon River Basin, Illinois. Trans. Ill. State Acad. Sci. 70; 153-163.
Manies, K.L. and D. J. Mladenoff 2000. Testing methods to produce landscape-scale presentment vegetation maps from the U.S. public land survey records. Landsc Ecol 15:741-754.
Minchin, PR. 1987. Simulation of multidimensional community patterns: towards a comprehensive model. Vegatio 71:145-156.
Natural Resources Conservation Service 2007. Soil Survey of Jersey County, Illinois. United States Department of Agriculture 616 pp.
Natural Resources Conservation Service 2004. Soil Survey of Macoupin County, Illinois. United States Department of Agriculture 341 pp.
Natural Resources Conservation Service 1974. Soil Survey of Greene County, Illinois. United States Department of Agriculture 85 pp.
Peck, J.E. 2010. Multivariate Analysis for Community Ecologists: Step by Step UsingPC-ORD. MjM Software Design, Gleneden Beach, Oregon. 162 pp.
Rodgers, C. and R.C. Anderson 1979. Presettlement vegetation of two Prairie Peninsula counties. Bot. Gaz. 140:232-240.
Schaetzl, R.J., and Brown, D.G. 1996. Forest association and soil drainage classes in presettlement Baraga County, Michigan. The Great Lakes Geographer 3:57-74.
Schulte L.A, and D.J. Mladenoff 2005. Severe wind and fire regimes in Northern forests: Historical variability at the regional scale. Eco 86: 431-445.
Schwegman, J. E. 1974. Comprehensive Plan for the Illinois Nature Preserves System, Part 2: The Natural Divisions of Illinois. Ill. Nature Preserves Commission. 32 pp.
Taft, J.B., M. Schwartz, L.R. Philippe 1995. Vegetation ecology of flatwoods on the Illinoisan till plain. J Veg Sci 6: 647-666.
Taft, J. B. 2005. Fire effects on structure, composition and diversity in a South-Central Illinois Flatwoods Remnant. 2005, Castanea 70: 298-313
Thomas-Van Gundy, M.A, G.J. Nowacki, C. V Cogbill. 2015. Mapping pyrophilic percentages across the Northeastern United States using witness trees, with focus on four National Forests. USDA Forest Service Northern Research Station General Technical Report NRS-145. 30pp.
Thompson J.R., D.N. Carpenter, C.V Cogbill, and D. R. Foster. 2013. Four centuries of change in Northeastern United States Forests. PLOS ONE e72540.
Transeau, E. N. 1935. The Prairie Penninsula. Ecol 16: 99 112
Vadeboncoer, M.A., and S.P Hamburg, C. V Cogbill, and WY. Sugimura. 2012. A comparison presettlement and modern forest composition along elevation gradient in central New Hampshire. Canadian Journal of Forest Research 42:190-202.
Wright, H.E. 1968. History of the Prairie Peninsula. in R.E. Bergstrom, ed. The Quaternary of Illinois. Special Publication 14, College of Agriculture, University of Illinois.
(1) * Richard B. Brugam, (2) Paul Kilburn, and (1) Laura Luecking
(1) Department of Biological Sciences, Southern Illinois University, Edwardsville IL 62025
(2) Jefferson County Nature Association, 6695 Terry Court, Arvada CO 80007
* corresponding author
Table 1. Tree species represented at more than 10 quarter section corners or section lines. Surveyor Scientific Name Number of Tree designation Section Abbreviation corners and Line Trees Prairie 2234 PR White Oak Quercus alba L. 1759 QA Black Oak Quercus velutina Lam. 1652 QV Hickory Carya spp. 1151 C Post Oak Quercus stellata Wangh. 597 QS Elm Ulmus spp. 429 U Pin Oak Quercus palustris 225 QP Black Jack Oak Quercus marilandica 225 QM Muenchh. Ash Fraxinus spp. 151 A Hackberry Celtis sp. 107 CE Black Walnut Juglans nigra L. 106 JN Red Oak Quercus rubra L. 92 QR Overcup Oak Quercus lyrata Walt. 85 QL Cottonwood Populus deltoides Marsh. 76 PD Dogwood Cornus sp. 69 C Sycamore Platanus occidentalis L. 61 P Maple Acer spp. 58 AI Sugar Maple Acer saccharum Marsh. 46 AU Spanish Oak Quercus falcata Michx. 38 QC Willow Salix spp. 37 S Sassafras Sassafras albidum Nutt, 32 SA Lyn Tilia americana L. 32 TA Mulberry Morus rubra L. 22 M Buckeye Aesculus sp.. 22 A Butternut Juglans cineria L. 20 JC Birch Betula nigra Ehrh. 19 B Burr Oak Quercus macrocarpa Michx. 16 QM Black Locust Robinia pseudoacacia L. 14 RP Honey Locust Gleditsia triacanthos L. 13 GT Box Elder Acer negundo L. 13 AN Red Bud Cercis canadensis 13 CE Table 2. Soils containing more than 10 quarter section corners or line trees. Number of Section Names Corner and Line Suborders Soil Series Trees Soil Fayette 966 hapludalf Hickory 661 hapludalf Herick 498 argiudoll Virden 406 argiaquoll Rozetta 358 hapludalf Keomah 329 endoaqualf Clinton 298 hapludalf Lawson 287 hapludoll Homen 235 hapludalf Muscatine 212 argiudoll Clarksdale 172 endoqualf Elco 172 hapludalf Wakeland 171 fluvaquent Ipava 158 argiudoll Sylvan 148 hapludalf Marine 127 albaqualf Downs 116 hapludalf Bunkum 110 hapludalf Titus 89 endoaquoll Cowden 86 albaqualf Oconee 85 endoqualf Menfro 81 hapludalf Beaucoup 73 endoaquall Winfield 62 hapludalf Sable 50 endoaquoll Petrolia 48 endoaquepts Blyton 45 udifluvent Coffeen 45 hapludoll Haymond 44 fluvaquent Greenbush 36 hapludalf Atterberry 35 endoqualf Fishhhook 32 hapludalf Worthen 32 hapludoll Rushville 31 albaqualf Elsah 30 udifluvent Quiver 30 fluvaquent Darwin 29 endoaquoll Keller 29 argiudoll Assumption 28 argiudoll Emery 27 endoqualf Tama 25 argiudloll
Please note: Some tables or figures were omitted from this article.
|Printer friendly Cite/link Email Feedback|
|Author:||Brugam, Richard B.; Kilburn, Paul; Luecking, Laura|
|Publication:||Transactions of the Illinois State Academy of Science|
|Date:||Jan 1, 2016|
|Previous Article:||Effects of lowhead dams on freshwater mussels in the Vermilion River Basin, Illinois, with comments on a natural dam removal.|
|Next Article:||Herpetological diversity of Stemler Cave Nature Preserve, St. Clair County, Illinois.|