Land Use History of North America
This web site serves as an archive to maintain the visibility of the important work done by the LUHNA project team.
|Table of Contents LUHNA Program Project History LINKS Contact Us|
Clues from the Past about our Future
Expanding Agriculture and Population
Night Lights and Urbanization
Patterns in Plant Diversity
Great Lakes Landscape Change
Upper Mississippi River Vegetation
Greater Yellowstone Biodiversity
Southwestern US Paleoecology
Palouse Bioregion Land Use History
Northeastern Forest Dynamics
The LUHNA Book!
Keywords: endangered ecosystem, grassland, wetland, conservation, landscape, human influence, biodiversity.
Keywords: endangered ecosystem, grassland, wetland, conservation, landscape, human influence, biodiversity.
What is the value of history to ecology? More specifically, what can historical, time-series data tell us that is relevant to current land management? We addressed this question by looking at how humans have influenced the Palouse Prairie of west-central Idaho. Our objectives were to (1) develop a history of European-American settlement and biological diversity in the Palouse bioregion, (2) use this history to understand how human activities have altered land cover and ecological integrity of the bioregion, and (3) assess the utility of information of differing scales and time periods. We believe that understanding the biophysical changes that have occurred in this region will provide a useful starting point for establishing research hypotheses, conservation goals, and ecological restoration efforts.
Land use is a function of culture and settlement pattern as well as environmental characteristics (Meinig 1968; Rappaport 1968; Bennett 1976; Robbins et al. 1983). The interactions of social, economic, and ecological factors are described in a large, diverse literature. Measures of social and economic conditions that have been shown to influence or be correlated to land-use patterns include historical land use (Turner and Meyer 1991; Savisky 1993; Flamm and Turner 1994), rural population density (Clawson 1971; Jobes 1991; McKendry and Machlis 1993), economic land value (Odum 1936; Alig 1986), tax status (Odum 1936; Fortmann and Huntsinger 1989; Savisky 1993), access (Turner et al. 1991; Skole et al. 1994), type of owner (Tosta and Green 1988; Fortmann and Huntsinger 1989; Turner and Meyer 1991; Savisky 1993; Skole et al. 1994), and residency of owner (Fortmann and Huntsinger 1989; Rudzitis and Johansen 1989; Fortmann and Kusel 1990; Jobes 1991).
Measures of ecological conditions demonstrated to influence or interact with land use include land cover and successional stage (Odum 1936; Rudel 1984; Forman and Godron 1986; Johnson 1987; Tosta and Green 1988; Boyle 1991; Dale et al. 1993; Flamm and Turner 1994; Koopowitz et al. 1994), soil type (Johnson 1987; Iverson 1988; Aspinall et al. 1993; Dale et al. 1993; Savisky 1993; Flamm and Turner 1994), physiography (Johnson 1987; Iverson 1988; Tosta and Green 1988; Ludeke et al. 1990; Savisky 1993; Schreier et al. 1994; Flamm and Turner 1994; Riebsame et al. 1994), climate (Hendrix et al. 1988; Ludeke et al. 1990; Riebsame et al. 1994), and biodiversity (Boyle 1991; Dale et al. 1993; Flamm and Turner 1994; Koopowitz et al. 1994).
The Palouse bioregion (Bailey 1995) covers 16,000 km2 in west central Idaho, southeastern Washington, and northeastern Oregon between the western edge of the Rocky Mountains and the Columbia River basin (Fig. 10-1). The region is characterized by a moderate climate and loess soils deposited on plateaus dissected by rivers deeply incised through layers of bedded basalt.
The Palouse Prairie, on which we focus, lies at the eastern edge of the Palouse bioregion, north of the Clearwater River. Here, where the loess hills are most developed, soils are often more than 100 cm deep. The depth and fertility of the soils make the region one of the world's most productive grain-growing areas (Williams 1991).
It is well known that the Palouse bioregion has undergone extensive change over the past 125 years. Both biophysical and human changes have been closely associated with advances in agricultural technology. Therefore, we structured our analysis using four time periods that bracket major changes in technology: European-American settlement (1870-1900), horse-powered agriculture (1901-1930), industrial agriculture (1931-1970), and suburbani-zation (1971-1990).
Because we knew both social and ecological changes in the bioregion would occur over and be detectable at different spatial scales, we conducted our analyses at three different levels: bioregion, county, and an 875-ha focal near Viola, Idaho (Fig. 10-1). This multiscale approach also allowed us to take advantage of data collected and reported at different spatial scales and geographic extents. In general, data available for the entire basin were coarser than those obtained for the county; data for the small area were at a yet finer spatial resolution.
Data for the broad scale encompassed the entire Palouse bioregion. We used both historical and current vegetation, mapped at a 1-km2 spatial resolution, from the Interior Columbia River Basin Ecosystem Management Project (Quigley and Arbelbide 1997). The historical vegetation data were derived from a combination of expert opinion, modeling, and historical records. The current vegetation was classified from satellite imagery. Historical and current fire regimes, which describe the frequency and severity of typical fires, were modeled by expert opinion from these layers (Morgan et al. 1996, Quigley and Arbelbide 1997).
At the county scale, we used soil surveys and published literature to predict reoccurrence of historical distribution (Kincaid and Harris 1934; Daubenmire 1942; Tisdale 1961; U.S. Department of Agriculture 1978; Ratti and Scott 1991). The U.S. Department of Commerce (1890, 1900, 1930, 1950, 1959, 1970, 1974, 1990, 1992) provided data on human settlement and social patterns (economic, demographic, and land use changes); these data have been recorded at the county level since the mid-1800's. Where data were available for all counties in the bioregion, we report bioregion-wide information. Where data were less accessible or unavailable for all counties, we concentrated on three counties that typify the region (Lewis and Latah Counties in Idaho and Whitman County in Washington).
Bioregional and county-level measures are often too coarse for tracing change in local biological resources because ecological boundaries are often smaller than those of the county or they do not coincide with political boundaries. To more accurately track land use and landcover changes at a finer scale, we examined an 875-ha area near Viola, in Latah County, Idaho, that contained the four major Palouse Prairie vegetation types (grass, shrub, dry forest, and riparian). This area was selected because historical aerial photographs were available from the Latah County courthouse and the area is typical of the bioregion. Vegetation and land use were interpreted from the aerial photographs (1940, 1965, and 1989, approximately 1:16,000). Vegetation and land-use boundaries were interpreted on mylar overlays, then digitized, georeferenced, and entered into a geographic information system (GIS) database. Information on tax status, market value, parcel size, and type and residency of owners for each parcel of land was obtained from county courthouse records.
To describe past-to-current change in vegetation at the fine scale for a longer but unknown time interval, we used a digital map of soils and potential vegetation for Latah County (Barker 1981) to project past vegetation based upon soil characteristics. Where soils were hydric, as indicated by the color of reduced iron compounds when described in the original field survey, the past vegetation was mapped as wetlands. Soils described as including hydric soils were mapped as potential wetlands to reflect the high probability that the area included small areas of wetlands. Soils that now support forests or had characteristics suggesting development under mostly forested conditions, such as low organic matter and light color, were mapped as forest. All others soils were mapped as grasslands. In addition, we generated and overlaid a map of areas that have a perched water table within 1.5 m of the surface to highlight areas that would have significantly more water available.
Our analyses and interpretations were primarily graphical. We used a GIS extensively to assemble and compare mapped information. For instance, we calculated the differences in extent of historical and current vegetation at the broad scale. Our synthesis of information on human demographics, land use, land cover, and biophysical resources since the turn of the twentieth century was necessarily descriptive. Because information was not available for or reported at consistent spatial and temporal scales, opportunities for quantitative comparisons across either time or space were limited. To strengthen and support our analysis, we relied on multiple lines of evidence at multiple spatial scales.
The Palouse Bioregion Prior to European-American Settlement
The vegetation encountered and documented by the earliest European-Americans cannot be considered to represent "stable" conditions, conditions that had been consistent for several generations of inhabitants and processes. The first inhabitants of the Palouse region, ancestors of the present-day Nez Perce Indians, probably arrived more than 12,000 years ago. For at least the last 2,000 years of that time, they lived in the river canyons, harvesting salmon during spring, summer, and fall and traveling to the prairies to harvest camas bulbs (mostly Camassia quamash) in the spring (Josephy 1965; Meinig 1968). Their economy was based on locally harvested wildlife, including salmon (Oncorhynchus spp.), elk (Cervus elaphus), and mule deer (Odocoileus hemianus), and was supplemented by traded goods from both the west coast and interior areas (Josephy 1965; Chalfant and Ray 1974).
Based on early settlers' descriptions of vegetation and wildlife habitats (e.g., Buechner 1953; Kaiser 1961; Meinig 1968) and more recent botanical assessments of prairie remnants (e.g., Weaver 1917; Daubenmire 1942; Tisdale 1961), we know that prior to settlement by European-Americans, bunchgrasses dominated the Palouse bioregion. Most of the original perennial grass prairie, though, was gone by 1900. One of the hindrances to fully understanding the ecological changes that have occurred on the Palouse Prairie is the lack of early natural history studies in the area. Weaver (1917) noted, "In the great inland province, practically no botanical work except of a taxonomic character has been done."
Prior to 1900, the native grasslands occurred in three zones (Daubenmire 1942; Tisdale 1961). The more mesic zone, on the wetter, eastern edge of the Palouse Prairie, was dominated by two perennial grass species, Idaho fescue (Festuca idahoensis) and bluebunch wheatgrass (Psuedoregneria spicata). Climax shrub communities, particularly bluebunch wheatgrass-snowberry (Symphori-carpos spp.) but also black hawthorn (Crataegus douglasii) and rose (Rosa spp.), grew on the northern sides of many of the loess hills (Lichthardt and Moseley 1996). Throughout this zone, summer moisture was too low to sustain trees except near streams (Lichthardt and Moseley 1996). The western portion of the Palouse Prairie was drier, though also dominated by bluebunch wheatgrass (Tisdale 1961). A third distinctive community occurred in the Snake River and Clearwater River canyons. These areas were far hotter and drier than the prairies and supported a sparser bunchgrass/shrub community (Tisdale 1986). Draws and waterseeps in the canyons supported a rich variety of tree species, including hawthorn and mock orange (Philadelphus lewisii).
True riparian communities were largely limited to the Palouse and Potlatch Rivers and to the broad outwash plains along sections of the Snake and Clearwater Rivers. These riparian zones supported a narrow gallery forest of plains cottonwood (Populus deltoides), quaking aspen (P. tremuloides), mountain maple (Acer glabrum), and red alder (Alnus rubra; Daubenmire 1942; Tisdale 1986). Wetlands were important but scattered. The vegetation was diverse and typically dominated by camas, a mixture of forbs, and many grasses.
Forest communities occupied the higher elevation mountains and ridges. On warmer sites, ponderosa pine (Pinus ponderosa) and Douglas fir (Pseudotsuga menziesii) grew with a rich shrub understory dominated by oceanspray (Holodiscus discolor), ninebark (Physocarpus malvaceus), serviceberry (Amelanchier alnifolia), snowberry, and wild rose. Cooler north- and west-facing canyons supported some western redcedar (Thuja plicata), grand fir (Abies grandis), and western larch (Larix occidentalis; Daubenmire 1942; Tisdale 1986).
European-American Settlement (1870-1900)
Fig. 10-4. Changing proportion of the human population living in rural areas (U.S. Department of Commerce 1900, 1930, 1970, 1990).
The European-American settlement and land-use patterns differed dramatically from Native American practices. Native Americans lived in the river valleys, while European-Americans lived on the prairies. Native Americans were hunter-gatherers or low-impact agriculturists of native species; the European-Americans were high-impact agriculturists of introduced species.
Initially, European-Americans used the Palouse hills as pasture, but farmers soon discovered the soil's fertilit (Prevost 1985). Fruit was an important early, though short-lived, commercial crop in the Snake River canyon and other areas in the Palouse. Apples, peaches, prunes, plums, apricots, and pears thrived (Williams 1991). However, both competition from areas better suited for fruit production and a better return on investment for wheat farming effectively killed the local fruit industry (Williams 1991).
Grain farming initially began on drier meadows and lower side slopes, leaving the steep hills, hilltops, and wet meadows for livestock pasture. Within as few as 10 years after first plowing, however, water tables in low-lying meadows dropped sufficiently to allow tilling (Victor 1935).
Lack of efficient transportation systems posed marketing problems for early farmers. Though steamboats operated up to the junction of the Snake and Clearwater Rivers, the rivers were low during the critical summer months, and the roads to the river ports were deep with dust in the summer and mud during the long spring and fall. Moving grain 30 miles from farms to ports on the Snake River often took 2 days if the wagons mired in the mud or dust (Williams 1991). Grain farming was also labor intensive, relying on human- and horse-power. Though steam-driven threshers appeared on the Palouse Prairie in the late 1800's, each machine required 18 horses and many men to operate (Prevost 1985). Many horses and cattle were used as draft animals on the farms, and all of these animals needed year-round pasture.
The combination of relatively inaccessible markets, considerable untillable lands, large pasturage requirements, and high labor requirements created a landscape that contained considerable refugia for native flora and fauna.
Horse-powered Agriculture (1901-30)
Major changes in land use between 1901 and 1930 resulted from the intensification of agriculture. Development of an extensive railroad network just after the turn of the century opened markets outside of the Palouse area. Farming became commercialized. Wheat and other cereals well adapted to the hillsides and climate of the Palouse region (Williams 1991) emerged as the dominant crops. The human population continued to grow (Fig. 10-4); by 1910, there were 22,000 people in 30 communities across the Palouse Prairie.
Farming remained labor-intensive and still relied heavily on human- and horse-power. An organized harvesting/threshing team in the 1920's required 120 men and 320 mules and horses (Williams 1991). The quest for a less labor-intensive bushel of wheat continued, but combine use lagged behind other farming areas in the United States (Williams 1991). It was only when the Idaho Harvester Company in Moscow began to manufacture a smaller machine that widespread combine harvesting became feasible. By 1930, 90% of all Palouse wheat was harvested using combines (Williams 1991). Such improvements enabled farmers to use lands previously left for grazing and as "waste," but the steepest hills and hilltops were still left as pasture for cattle and horses.
Industrial Agriculture (1931-70)
The era between 1931 and 1970 was one of continued mechanization, and especially industrialization. With the development of each new technology, farming became less labor intensive, allowing fewer people to farm larger areas (Fig. 10-5). Petroleum-based technology replaced horse and most human labor early in the era. By 1970, most farm workers used motorized equipment, which removed the need for pasture lands and provided equipment that could till even the steepest slopes. Fertilizers, introduced after World War II, increased crop production by 200%-400%. Federal agricultural programs encouraged farmers to drain seasonally wet areas, allowing farming in flood plains and seasonally saturated soils. With the advent of industrial agriculture, the last significant refugia for native communities were plowed.
Since 1970 major changes have occurred in the composition of the rural population and land use. Rural population began to rise as more town and city residents sought rural suburban homesites. Ironically, yesterday's farmers and their children who still own farm land have moved to ever more distant towns and cities (unpublished data from Latah County records). The influx outweighs the exodus, however, and human populations in rural areas are growing. Little change has occurred in overall hectares devoted to agriculture; however, some lands with highly erodable soils have been temporarily removed from crop production under the Federal Conservation Reserve Program (CRP). In Latah County alone, this program removed about 14,000 ha from agricultural production (U.S. Department of Agriculture 1996a), most planted with introduced perennial grasses. Because the Palouse Prairie has been declared an endangered ecosystem (Noss et al. 1995), current CRP guidelines encourage restoration of endangered native ecosystems (U.S. Department of Agriculture 1996b). Within Latah County, an additional 1,559 ha have been included in the Idaho Fish and Game Department's Habitat Improvement Program. Ponds and plantings on the latter lands are designed to enhance nesting habitat for upland game birds.
The effects of prairie conversion can be seen in both social organization and ecosystem composition. Instead of living in the river canyons and foraging on the prairies, people now live on the prairies, cultivate the former wild meadows, and recreate in the river canyons. Local economies are based on extraction rather than subsistence. With each advance in agricultural technology, crop production has increased and more native prairie vegetation has been converted to field or pasture. First the draining of wetlands, then equipment that enabled farming of steep slopes, then the introduction of chemicals; each effectively shrank remaining refugia for native flora and fauna. Grazing and farming introduced new species and imposed a different set of disturbance regimes on the landscape.
Suburbanization of the rural landscape represents the second major land-use conversion in the last century. As with the first, this change will have a profound influence on ecological processes and biological diversity.
Change in Vegetation
We used broad-scale information to monitor changes in grass, shrubs, and forest vegetation across the bioregion. Since 1900, 94% percent of the grasslands and 97% of the wetlands in the Palouse bioregion have been converted to crop, hay, or pasture lands (Fig. 10-6). Approximately 63% of the lands in forest cover in 1900 were forested in 1990, 9% were grasslands, and 7% were regenerating forest or shrub vegetation. The remaining 21% of previously forested lands have been converted to agriculture or urban areas.
Fig. 10-6. Historical (circa 1900) and current (circa 1990) vegetation in the Palouse bioregion (Quigley and Arbelbide 1997).
We believe that these results vastly underestimate the past abundance of riparian areas and the small patches of wetlands and shrubs once common on the Palouse Prairie. The fine-scale topography of the Palouse hills would have harbored many wetlands of a size too small to be captured at a 1-km2 scale. In addition, such changes were captured only over the last 90 years, 40 years after European-Americans began to settle in the area.
To identify probable locations and extents of habitats under-represented at the broad scale, we conducted two finer-scale analyses. We found ecologically significant changes in small patches of brush, grass, and riparian vegetation from 1940 to 1965 and 1965 to 1989 in the 875-ha area evaluated with historical aerial photographs (Fig. 10-7). A total of 170 ha (19.4% of the total area) was converted to agriculture between 1940 and 1965, mostly from open shrublands and riparian areas. Most forest lands were logged during this period, creating open forests with shrubs. Although significant conversions of riparian areas to fields and pastures probably occurred between 1880 and 1940, 61% of riparian areas existing in 1940 were gone by 1989. Stringers of riparian vegetation shrunk to thin, broken tendrils, and shrub vegetation virtually disappeared. If this 875-ha area is indeed representative of the entire bioregion, the cumulative effects of such changes are enormous. Alteration in the size, quality, and connectivity of habitats may have important consequences for wildlife species (Forman and Godron 1986; Soule 1986).
Fig. 10-7. Changing vegetation over time as interpreted from aerial photographs of a small (875 ha) area near Viola, Idaho.
We also predicted past vegetation extent based on soil characteristics. This process was particularly useful for mapping areas where wetlands existed long enough to result in hydric characteristics. It revealed much more extensive riparian or wet areas than shown over the last century at the broad scale or over the last 50 years at the fine scale (Fig. 10-8). One plausible reason for this difference is that extensive changes undoubtedly occurred prior to either 1900 (the date of our earliest broad-scale vegetation map) or 1940 (the date of our earliest aerial photograph). Another plausible explanation is that the soil characteristics are a legacy of past climate, such as the Little Ice Age that ended in the early 1800's (Pielou 1991). The soils may have developed during a preceding cooler, wetter period when forest cover was more extensive than at present. Under the slightly warmer and drier present-day climate regime, one would expect a retreat of the forest margins. We hypothesize that the forest extent before European settlement was greater than today, but not as extensive as our "potential vegetation" map suggests, and that wetlands were more extensive in the past than they are today.
Of the once-continuous native prairie dominated by midlength perennial grasses, only little more than 1% remains. It is one of the most endangered ecosystems in the United States (Noss et al. 1995), and all remaining parcels of native prairie are subject to weed invasions and occasional drifts of aerially applied agricultural chemicals (J. Lichthardt, personal communication). Two of the native plant communities, bluebunch wheatgrass-snowberry and bluebunch wheatgrass-rose, are globally rare, and eight local plant species are threatened globally (Lichthardt and Moseley 1996). Many once-intermittent streams are now farmed; many perennial streams with large wet meadows adjacent to them are now intermittent or deeply incised, and the adjacent meadows are seeded to annual crops. Few areas of camas bloom in the spring. Clean farming practices (field burning, herbicide use, and roadbed-to-roadbed farming) leave few fences and fewer fencerows, negatively impacting even those edge species which can flourish in agricultural areas (Ratti and Scott 1991).
With the virtual elimination of native prairies, species dependent on grassland ecosystems have declined or disappeared as well. Formerly abundant sharp-tailed grouse (Tympanchus phasianellus) occur only in highly fragmented, marginal, and disjunct populations (Kaiser 1961; Burleigh 1972; U.S. Department of Agriculure 1978; Ratti and Scott 1991). The white-tailed jack rabbit (Lepus townsendii) and ferruginous hawk (Buteo regalis) have been nearly extirpated as breeding populations.
At the same time, new land uses offer habitats for a different suite of species (Table 10-1). Humans have intentionally introduced the gray partridge (Perdix perdix), ringnecked pheasant (Phasianus colchicus), and chukar (Alectoris chukar), species which generally fare well in agricultural landscapes. Grazing, agriculture, and accidents have introduced a variety of exotic plants, many of which are vigorous enough to earn the title "noxious weed" (Table 10-2).
Conversion of agricultural lands to suburban homesites invites a second new suite of biodiversity onto the Palouse Prairie. A University of Idaho wildlife professor has documented changes in bird community composition over the past 10 years as he converted a wheat field into a suburban wildlife refuge. His 6-ha yard now attracts 86 species of birds, an increase from 18 (Ratti and Scott 1991). While many of the plant species he planted are nonnative, the majority of avian species using the habitat are native (95%).
Suburbanization of agricultural lands does not necessarily favor native species, however. Rural residents in Latah County have constructed some 1,500 ponds. These ponds are facilitating rapid colonization by an exotic bullfrog (Rana catesbeiana) which experts fear may compete with and/or eat native amphibians, including the sensitive spotted frog (Rana pretiosa; Monello and Wright 1997). The brown-headed cowbird (Molothrus ater) and European starling (Sturnus vulgaris) have taken advantage of the new habitats and moved into the area. The black-tailed jack rabbit (Lepus californicus) has largely displaced the white-tailed jack rabbit (Tisdale 1961; Johnson and Cassiday 1997).
Changes in biodiversity in the canyonlands follow a parallel track, though from slightly different causes. Due to steep slopes and infertile soils, the canyonlands have been used for grazing instead of farming (Tisdale 1986). Intense grazing and other disturbances have resulted in irreversible changes, with the native grasses being largely replaced by nonnative annual brome grasses and noxious weeds, particularly star thistles.
Change in Physical Resources
The Palouse region has one of the highest soil erosion rates in the country (U.S. Department of Agriculture 1978). Breaking of the original perennial grass cover left the soil vulnerable to erosion by wind and water. Commercial farming practices exacerbated these problems. Summer fallow leaves the soils with poor surface protection during the winter; burning straw and pea crop residues leave the soil with less organic binding material; and heavier, more powerful farming equipment pulverizes the soil, leaving it more vulnerable to wind and water erosion (Kaiser 1961).
Erosion measurements and control efforts began in the early 1930's. Soil loss by water erosion in the Palouse River basin from 1939 to 1972 was most severe in the heavily farmed areas of Whitman County, Washington, where soil losses of 15-18 tons per acre per year were mapped (U.S. Department of Agriculture 1978; Fig. 10-9). One major study reports that an average of 358 tons of soil was lost from every cropland acre in the basin from 1939 to 1972. This translates into an average of 0.2 tons of soil for every bushel of wheat grown on the Palouse from 1939 to 1972 (U.S. Department of Agriculture 1978).
Intensification of agriculture has affected both water quantity and quality as well. Replacing perennial grasses with annual crops resulted in more overland flow and less infiltration, which translates at a watershed level to higher peak flows that subside more quickly than in the past. The result is more intense erosion and loss of perennial prairie streams. Once-perennial streams are now often dry by mid-summer. As early as the 1930's soil scientists were noting significant downcutting of regional rivers (Victor 1935) due to higher, faster runoff, effectively lowering the water table in adjacent meadows. Nitrates from fertilizers have found their way into the surface aquifers as well. Nitrogen levels in groundwater have increased significantly (Jones and Wagner 1995; Wertz and Kinney 1995; U.S. Department of Agriculture 1995).
Change in Disturbance Regimes
Changes in vegetation and settlement pattern have changed the frequency, size, and pattern of the region's two major disturbances: fires and floods. While there is some debate over how frequently the Palouse Prairie burned historically, there is consensus that fires are generally less frequent today than in the past (Morgan et al. 1996). Historians recount lightning-ignited fires burning in the pine fringes bordering the prairies in late autumn, but the extent to which forest fires spread into the prairie or the converse is not known. In addition, some fire ecologists believe the Nez Perce burned the Palouse Prairie to encourage growth of camas (P. Morgan, University of Idaho, personal communication). Camas grows in seasonally wet areas and was harvested in the spring when it was easy to dig the bulbs from the damp soil; how camas responds to dry-season burning is unknown today.
European-American settlers used fire to clear land for settlement and grazing. Since then, forest fires have become less common because of fire suppression, human settlement, the presence of roads which act as fire breaks, and the conversion of grass and forests to cropland (Morgan et al. 1996). One result of the lower fire frequency has been increasing tree density on forested lands and encroachment of shrubs and trees into previously open areas. Consequently, when fires occur in forests they are more likely to result in mixed severity or stand-replacing events instead of the low severity fires of the past. Fires are still frequent in canyons, though today, fires give exotic annual grasses an edge over native species in burned areas.
Flooding on the major rivers has been curtailed in the region by large hydroelectric projects on the Snake and Clearwater Rivers. In addition to altering stream flow and channel scouring, the dams are major barriers to anadromous fish. Changes in hydrology, such as drainage tiles placed under seasonally wet areas to allow agricultural production, removal of riparian vegetation, channeling of prairie streams, and building in flood plains, contribute to more severe localized flood events during winter and spring.
Conversion of the Palouse bioregion, particularly of the Palouse Prairie, from perennial native grass, shrub, and forest vegetation to agriculture has been so complete that it might seem a moot point to study its change. However, the processes by which the conversion occurred and the interactions between human cultures and environment influenced the extent and spatial pattern of landscape change, and therefore influenced wildlife population dynamics and viability. Linking these underlying processes to the resulting landscape patterns can provide new insight and make significant contributions to science and policy (Black et al. 1998). The conversion of more than 94% of the areas occupied by native landcover types makes the Palouse grasslands one of the most endangered ecosystems in the United States (Noss et al. 1995). Despite this loss of native habitats, no plant species (Lichthardt and Moseley 1996) or animal species (Buechner 1953; Tisdale 1986) have been lost from the Palouse. The earliest collections postdate extensive human use, however, so human impacts on native flora and fauna are probably greater than we were able to document. Several once common animal species, including ferruginous hawk, white-tailed jack rabbit, and sharp-tailed grouse, are rare and survive only as small relict populations in isolated fragments of habitat. Six globally rare plant species are endemic to the Palouse region (Lichthardt and Moseley 1996), and the integrity of remaining habitats for these and other species are low.
We used data from multiple spatial scales to analyze change. At the coarsest scale, native grasslands declined by more than 97% across the entire bioregion. However, certain critical habitat features, including remnant prairie vegetation, wet meadows, and shrubfields, were not adequately captured in the coarse resolution of the broad-scale data. Finer-scale ecological features and social system factors influencing conservation were examined in greater detail within the 875-ha study area. While this area was too small to allow extrapolation of results to the entire bioregion, its small riparian patches and wetlands largely disappeared between 1940 and 1989, further reducing refugia and connectivity for native flora and fauna.
Projections of past vegetation based on soil characteristics documented in this century suggest that historical records may underestimate the total change from human influences. Yet, historical data are useful for evaluating changes in land cover and understanding the drivers of that change (Morgan et al. 1994). If collected over large areas, time-series data such as those compiled here offer an unprecedented opportunity to study species response to human land uses, including (1) the degree of change in habitat type and area, (2) the displacement of one suite of species with another as habitats change, and (3) the mechanisms, vectors, and rates of expansion and contraction of species. Further, understanding landscape changes provides critical information for outlining future research agendas, regulating development (Black et al. 1998), setting conservation goals, and targeting ecological restoration efforts.
The current suburbanization of the Palouse region provides opportunities for restoring native vegetation and enhancing animal species and populations (Ratti and Scott 1991). The loss of more than 98% of the native vegetation communities of the Palouse has not resulted in a single known species extinction or extirpation; however, the populations of six globally rare plant species endemic to the region (Lichthardt and Moseley 1996) have been reduced dramatically and survive only on isolated grassland remnants. These areas provide a logical starting point for efforts to restore native Palouse ecosystems.
Our analyses show that landscape changes are rarely gradual; they are episodic and at times the result of breakthroughs in agriculture technology. In the Palouse bioregion, episodes of change were based on breakthroughs associated with ever more intensive land uses. Railroads enabled agricultural settlement, engines and fertilizers commercialized farming and facilitated intensified land use, and now the information superhighway is enabling former city dwellers to relocate to rural areas (Rudzitis 1989). Each transition has been accompanied by changes in physical and biological conditions, including significant shifts in composition. At some point in time, if soil erosion is not curtailed, the physical capacity of this area to produce food will diminish; then we can expect a dramatic shift in both social and ecological systems. Whether these transitions are deemed favorable or not, time-series studies such as this provide new insights on regional change and offer critical information upon which to base land and resource management.
These figures where created for the original version of this paper, however they were not used in the official publication "Perspectives on the Land Use History of North America: A Context for Understanding Our Changing Environment". They are included here as additional references.
This paper resulted from the cooperation of many individuals and institutions. The Landscape Dynamics Research Lab provided facilities and covered overhead costs. We are particularly grateful to our reviewers from the University of Idaho: Minoru Hironaka (range ecology), Bill Swaggerty (environmental history), and Tim Forseman and Joel Hamilton (agricultural economics). We also thank Ken Huska (soil scientist, USDA Natural Resources Conservation Service), Robin Johnston (historian, USDA Forest Service), and Jay Pengilly (member, Latah County Planning and Zoning Commission) for their review comments. Writing by committee is never easy; many thanks to all who participated!
Regionalization Studies at Harvard Forest LTER