Soil Science Society of America Journal 65:391-402 (2001)
© 2001 Soil Science Society of America
DIVISION S-5-PEDOLOGY
Assessing the Impact of Land Conversion to Urban Use on Soils with Different Productivity Levels in the USA
Egide L. Nizeyimanaa,
G.W. Petersena,
M.L. Imhoffb,
H.R. Sinclair, Jr.c,
S.W. Waltmanc,
D.S. Reed-Margetand,
E.R. Levineb and
J.M. Russoe
a Dep. of Agronomy, College of Agricultural Sciences, and Environmental Resources Research Institute, 101 Land and Water Research Building, The Pennsylvania State Univ., University Park, PA 16802
b NASA Goddard Space Flight Center, Code 923/Biospheric Sciences Branch, Greenbelt, MD 20772
c National Soil Survey Center, USDA-NRCS, 100 Centennial Mall North, Lincoln, NE 68508
d Statistical Lab., Iowa State Univ., 212 Snedecor Hall, Ames, IA 50011
e ZedX Inc., 369 Rolling Ridge Drive, Bellafonte, PA 16823
Corresponding author (egideN{at}psu.edu)
 |
ABSTRACT
|
|---|
There has been increased public concern in the USA over the long-term impact of urbanization on the available land used to produce food, feed, and fiber. Concern that urban use of highly productive soils may threaten our food security and sustainability has been debated for nearly three decades. This study was primarily initiated to compare different soil productivity classes in terms of areas and proportion of land converted to urban uses in the USA. The methodology consisted of analyzing data resulting from a geographic information system (GIS) overlay of urban land use maps derived from the Defense Meteorological Satellite Program's Operational Linescan System (DMSP/OLS) nighttime imagery and layers of potential soil productivity. Soil productivity distributions were obtained using the Soil Rating for Plant Growth (SRPG) model based primarily on soil and landscape parameters contained in the State Soil Geographic (STATSGO) database. Currently, the urban land use covers 
3% of the conterminous USA and is primarily on areas that were originally of low and moderate soil productivity. Only 6% of the total land under urbanization had consisted of highly productive soils. However, land with highly productive soils, roughly 3% of the total U.S. area, has a higher level of urbanization (5%) than that of any other soil productivity category. States differ in the areas and proportion of land converted to urban uses in each soil productivity class. These results are a first step in determining the current status of soil resources in relation to urbanization and should be interpreted according to the scale and resolution of data sources and assumptions made in the soil productivity modeling.
Abbreviations: DMSP/OLS, Defense Meteorological Satellite Program's Operational Linescan System • GIS, geographic information system • MLRA, Major Land Resource Area • NOAA/NGDC, National Oceanic and Atmospheric Administration's National Geophysical Data Center • NRI, National Resources Inventory • SRPG, Soil Rating for Plant Growth • STATSGO, State Soil Geographic Database • UN/FAO, United Nation's Food and Agriculture Organization
|
INTRODUCTION
|
|---|
CONCERNS ABOUT THE IMPACT of human settlements on the available land suitable for agriculture have been rising rapidly during recent years. At the global scale, the discussion has focused primarily on the issues of Earth's carrying capacity in the face of population increase and its consequence on food security and environmental sustainability (Daily and Ehrlich, 1990). In this case, the land use conversion is driven by increasing populations in search of new habitats and is particularly intense in already highly populated regions of the world, especially in Africa and Asia (Brown, 1995). The world's population is estimated to reach 8 billion by the year 2025, a 38% increase from its current population. Yet, the expansion of cropland base has not been growing at the same pace (Richards, 1990). Most of the projected population increase (88%) is in Africa and Asia, where land development has been increasing faster than anywhere else in the world and where food shortages are common (UN/FAO, 1996; Cornelius and Cover, 1997). Cropland acreage is also decreasing rapidly in China, Thailand, India, Indonesia, and Vietnam as a result of urbanization (Engelman and LeRoy, 1995; Gardner, 1996). According to recent estimates, the global demand for food is expected to rise in the future, whereas the world's capacity to supply it is to decrease (Nelson, 1996).
In the USA, public concern over the expansion of urban development into agricultural landscapes has been expressed in terms of the impact of suburban sprawl on farmland extent and local agricultural communities. Suburban sprawl consists of an encroachment of residential developments over open farmland at the urban–rural fringe or along highways and has been for many years the dominant form of growth of U.S. metropolitan areas (Stoel, 1999). This has often resulted in a decrease in the number of farms, farmland extent, and connectivity. A possible consequence has been for some farmers to relocate on less productive soils or be forced out of agricultural business (Berry, 1978; Sampson, 1981). Nonetheless, state and local land use policies (e.g., agricultural zoning ordinances, agricultural districts, transfer or purchase of development rights) adopted by counties and cities in the late 1970s and early 1980s to alleviate conflicts between agricultural activities and urbanization have recently intensified. These policies are commonly part of "comprehensive plans," forms of guidelines to land use planning designed to maintain a balance between natural ecosystems, cropland, and urban development while preserving the local economy base (Vesterby and Heimlich, 1991).
While the decline of agricultural activities as a result of urbanization pressures concerns local communities, the national perspective focuses attention on the impact of highly productive farmland's loss on the future capacity of land to produce food in the USA. Whether current rates of rural land conversion to urban uses is or is not threatening U.S. food security has been debated for nearly three decades. Much of the alarm was prompted very early by assessment results of the National Agricultural Lands Study (1981) led by the USDA-SCS (now NRCS) and the Council on Environmental Quality. This study estimated that the annual conversion of rural nonfederal land to urban uses increased from 445000 ha between 1958 and 1967 to 850000 for the period of 1967 to 1975. A series of discussion papers followed a year later (Brown et al., 1982; Fischel, 1982; Fisher, 1982; Raup, 1982; Simon and Sudman, 1982) in which authors disagreed and questioned the methodology used to determine the extent of urbanization. Later analyses based on the National Resources Inventory (NRI) reported only a moderate annual increase, 570000 ha of rural land conversion to urban uses between 1982 and 1992 (USDA-NRCS, 1997). The NRI has been conducted every 5 yr and consists of a database of land use and land cover information and soil characteristics that have been recorded at georeferenced sampling points on nonfederal lands throughout the USA (Goebel, 1992). Although only 28% were from cropland, the concern has continued to exist through the years until today. While results of all these studies are areal estimates of rural land or farmland converted to urban uses, the adequacy of the nation's land to produce food, feed, and fiber can only be measured in terms of potential soil productivity of the land converted. It is obvious that all rural lands, even those presently used as cropland, are not equally productive.
Previous studies dealing with land quality–urbanization relationships used prime farmland as a measure of soil productivity (Vining et al., 1977; Plaut, 1980; Sorensen et al., 1997). Prime farmland is land that has the best combination of physical and chemical characteristics for producing food, feed, forage, fiber, and oilseed crops. Results of these studies coupled with historical experience and observations of past and current land use distributions in the USA indicate that prime farmland is likely to be developed for urban uses (Dillman and Cousins, 1982). There is also a general agreement that to hold construction costs to minimum, urban developers prefer land with deep, well-drained, and nearly level soils, areas that happen to be best suited for agricultural production. The only source of the extent of prime farmland land and prime farmland converted to urban uses at the national level is the NRI data set (USDA-NRCS, 1999). Prime farmland, however, is a generalized and qualitative indication of soil limitations as related to soil management and does not consider the entire range of soil, landform, and climate characteristics of the land.
The extent of urbanization has been traditionally identified and mapped from census data and aerial photography. These approaches are time-consuming, expensive, and inappropriate for regional and conterminous scale analyses. The integration of satellite remote sensing and GIS technology appears to be a viable alternative for mapping and analysis of urban land use for the conterminous USA in a timely and efficient manner. The DMSP/OLS nighttime imagery has been evaluated by the National Oceanic and Atmospheric Administration's National Geophysical Data Center (NOAA/NGDC) and appears to be an appropriate source for locating human settlements (Elvidge et al., 1997a) at the continental scale. The objectives of this work were to determine the areal extent of land converted to urban uses and distribution of soil productivity in the USA and to compare soil productivity classes in terms of areas and proportion of land converted using remote sensing and soil productivity modeling.
|
MATERIALS AND METHODS
|
|---|
Data Sources and Methodology
The extent of urbanization was determined by generating an urban land use GIS layer from the U.S. Air Force DMSP/OLS nighttime imagery. Soil productivity ratings were computed using a soil productivity model, the SRPG, based on climate, landscape, and soil parameters contained in the STATSGO database (Soil Survey Staff, 1994). The level of soil productivity for STATSGO mapping units presently under urbanization was obtained by analysis of data resulting from overlays of urban land use maps and soil productivity layers. All GIS computations and coverage overlays were performed using ARC/INFO software (Environmental Systems Research Institute, 1998).
Mapping Urban Land Use
The DMSP program is part of the Department of Defense and is operated by the Air Force Space and Missile Systems Center in Los Angeles, CA. Its satellites are in a near-polar, sun-synchronous orbit at 830 km above the Earth and cross any point on the Earth twice a day. The OLS, one of the six DMSP satellite sensors, was designed to monitor the distribution of the sun- and moon-reflected off clouds during daytime and nighttime conditions in the visible and infrared ranges for aircraft missions. However, its nighttime capability made possible by a very sensitive photometer makes it ideal for acquiring images of light sources on the Earth's surface (NOAA, 1998). The imagery is acquired at a nominal spatial resolution of 2.7 km at nadir.
The data set used in this study is a digital imagery of stable lights for the conterminous USA previously created using DMSP/OLS composite images from 231 orbital swaths that were gathered in 1994 and 1995 (Elvidge et al., 1997b). The NOAA/NGDC created this layer of consistently lit areas by screening images for clouds and eliminating ephemeral light sources such as fires and lightning (see Elvidge et al., 1997b, for a detailed description of the methodology). At this point, this stable-light imagery product still contained ephemeral light sources and the reflection of light onto water when adjacent to cities, a state that overestimated the spatial extent of urban areas. A final urban land use layer was derived using a thresholding technique described by Imhoff et al. (1997b). The procedure applied a threshold value of 89% to lit areas in the stable light image composite and left the dense urban core or areas illuminated 89 to 100% intact. The digital map of urban land use thus produced contained not only urban areas as defined by the U.S. Census Bureau, but also suburban developments, commercial and industrial facilities, and well-lit transportation corridors. Imhoff et al. (1997a) determined that the lit area remaining in the thresholded image represented areas having an average population and housing density of 1033 persons km-2 and 427 housing units km-2. The lit area defining urbanized areas compared very well with the urbanized area estimates determined by the 1990 U.S. Census (U.S. Census Bureau, 1991).
Assessing Potential Soil Productivity Distributions
Potential soil productivity ratings for the conterminous USA were determined using the SRPG model based primarily on soil and landscape parameters contained in the STATSGO database. The model was developed by the USDA-NRCS National Soil Survey Center in collaboration with the Iowa State University's Statistical Laboratory. It consists of a computerized procedure that classifies soils for their inherent ability to produce fiber, vegetative growth, and grains for nonirrigated commodity crops (Sinclair, 1996; Soil Survey Staff, 2000). It uses a wide range of soil morphological, physical, and chemical properties in addition to climatic variables (soil moisture and temperature regimes) and landscape features. Each parameter in the SRPG model has been given ranges of values with corresponding ratings that show its suitability to nonirrigated crop production. For each soil component in a given STATSGO mapping unit, the model assigns a percentage rating at each of its parameters, multiplies all parameter ratings, and then normalizes the resulting values to a percentage rating. The overall rating has values that vary from 100 to 0. Unlike other soil productivity schemes commonly used in the USA, the SRPG takes into account variations in crop types grown in the county (or soil survey area) and Major Land Resource Areas (MLRA), which are geographic regions that were identified by the USDA-NRCS and are characterized by common patterns of soil, climatic, water resources, and land uses (USDA-SCS, 1981).
The STATSGO database was developed by the USDA-NRCS at 1:250000 scale for use in regional and statewide resource assessments (Bliss and Reybold, 1989). Each mapping unit is linked to relational databases including tables for soil components, layers for each component, and associated soil properties. First, STATSGO soil components were rated from 100 (most productive soils) to 0 (least productive soils) using the SRPG model. Soil component rating values were then weighted by percentage composition to determine a single landscape rating value for each STATSGO mapping unit. The ratings were finally grouped into the following four categories: high (100–76), moderately high (75–51), moderate (50–26), and low (25–0). This soil productivity grouping is similar to other generalized parameter-based land suitability schemes commonly used in other countries for regional and national land evaluation (e.g., Dumanski and Stewart, 1983; Brklacich and MacDonald, 1992; Ogunkunle, 1993).
Urbanization preceded the soil survey for some areas in the USA. In this case, STATSGO soil components of areas under urbanization did not have soil property values and were therefore listed as "unsuited" for agricultural production (SRPG rating of zero). This resulted in underestimation of the preurbanization landscape SRPG rating for STATSGO mapping units dominated by urbanization. In this case, components were assigned the average value (weighted by component percentage) of ratings of the nonurban soil components contained in the same STATSGO mapping units.
Assessing the Areal Extent of Soils Converted to Urban Uses in Each Productivity Class
The areas of STATSGO soil mapping units converted to urban uses in each soil productivity category were determined by overlaying urban thematic maps with GIS soil productivity layers. Results are areal estimates and proportion of areas of STATSGO mapping units converted to urban land use that fall in each soil productivity category for the conterminous USA. Using the same data set, further analyses were performed to determine the proportion of land in each soil productivity class that has been converted to urban uses. In addition to continental level assessments, the above analyses were also performed for each of the 48 states by intersecting GIS state boundaries with STATSGO-derived soil productivity layers. The rationale for performing state analyses is the fact that states are presently playing an important role in land use planning policies. Results from these analyses could therefore be used as a basis for developing sound statewide farmland preservation programs. Finally, a general discussion at the distribution of potential soil productivity and urban land use is provided for different USDA-NRCS regions indicated in Fig. 1
. Region and state boundaries were derived from 1:2000000-scale U.S. Geological Survey GIS state delineations (Negri, 1994).
|
RESULTS AND DISCUSSION
|
|---|
Figure 2
shows the distribution of soil productivity of land across the USA. Table 1 lists the percentage of land area in each soil productivity category for the entire USA and for individual states. In general, the most productive soils, that is those in the high soil productivity class are concentrated in the Midwest states (Illinois, Indiana, Iowa, Minnesota, Missouri, Ohio, and Wisconsin) and in two of the Northern Plains states (Nebraska, Kansas). These areas have deep, nutrient-rich soils developed from glacial drift and loess over glacial drift. Although not shown, Midwestern states accounted for 66% (or 16.2 million ha) of the land that is classified as having highly productive soils, and most of it is distributed in Illinois, Iowa, and Indiana. Illinois has the highest amount of land in the high soil productivity class (6.4 million ha). Land with soils in the high productivity class is also in delta and river valleys (e.g., Mississippi Valley and Delta), river and coastal alluvial plains (e.g., Lower Rio Grande Plain, Western Gulf Coast Plain), and valleys along mountains (Northern Piedmont) in other states. The total U.S. land is distributed as 3, 26, 38, and 33% in high, moderately high, moderate, and low soil productivity classes, respectively. About 29% of the total land is classified in the two top categories, high or moderately high soil productivity. Soils classified as highly productive occupy a substantial portion of the land area in Illinois (44%), Iowa (34%), Indiana (22%), and Nebraska (12%) (Table 1). Land with low soil productivity is primarily in states located in the West Region and in the western part of the Northern Plains (Montana, Wyoming, and Colorado).
The urban land use represents 3% of the total U.S. land area (Table 2). The distribution of urban land use in the USA is also presented in Fig. 3
. The urbanization is concentrated in the Midwest, South Central, Eastern, and Southeast Regions. More than 80% of the total urban land use is in these regions. California, Texas, and Florida have the most urban land area (>1 million ha each) in the country. Most states have <10% of their land under urbanization, except Connecticut (12%), Delaware (10%), Massachusetts (13%), Florida (11%), Maryland (13%), New Jersey (22%), and Rhode Island (14%). It appears that most of the land currently under urbanization (83%) in the USA had consisted of moderately high or moderate soil productivity (Table 2). Only 6% of the total USA urban land, 1 million ha, is on land that was originally of high soil productivity. In several states, such as Alabama, California, Colorado, and North Dakota, the urbanization occurs mostly on land that had either moderately high or moderate soil productivity prior to development. Nebraska is the greatest exception; the urban land in this state is primarily on areas that had highly productive soils (51%). In Illinois, Indiana, Iowa, Kentucky, and Nebraska, a quarter or more of the urban land occurs on land that was originally of highly productive soils. These states also have little or no development on the land with low soil productivity.
Figure 4
shows the level of preurbanization soil productivity for land presently under development. The area and proportion of land in each soil productivity class that is under urbanization for the USA and individual states are presented in Table 3. The percentage of land area in this table indicates the level of urbanization for land in a given soil productivity class. Therefore, within each state the higher the number, the greater the urban pressure on land in this soil productivity class. For example, of the total land area classified in the high soil productivity class in the Commonwealth of Pennsylvania, 79000 ha are urbanized. This represents 15% of the total land area in this soil productivity class. In comparison, urbanization occupies 6, 2, and 1% of the land that was of moderately high, moderate, and low soil productivity, respectively, in this state. Therefore, land in the high soil productivity class in Pennsylvania experiences more urban pressure than land in any other class.
The degree of urbanization of USA land increases with increasing soil productivity. It is 1, 2, 4, and 5% for land in the low, moderate, moderately high, and high soil productivity classes, respectively. Illinois has more land in the high soil productivity class under urbanization (277 000 ha) than any other state. Of the total land area with highly productive soils under urbanization, a little more than 50% of it is distributed in four states: Illinois, Indiana, Iowa, and Texas. In almost one-half of the states, the degree of urbanization increases from land in the low to that in the high soil productivity category, thus indicating that land with the most productive soils experiences higher urbanization pressure. Alabama, California, Utah, and Vermont have a higher portion of land with highly productive soils (
25%) under urbanization than other states. Land in the high soil productivity class for these states along with Kansas, Kentucky, Louisiana, Maryland, Missouri, Pennsylvania, Tennessee, Texas, and West Virginia is also the most urbanized as compared with that in other soil productivity classes. In contrast, Florida, Mississippi, and South Carolina have the land with the least productive soils being the most urbanized. Moreover, the level of urbanization increases from the land having the most potentially productive soils to that with least productive ones in Florida.
Data Interpretation and Limitations
The DMSP/OLS-derived maps provided continental and statewide views of urban land use. It should be understood that small towns and villages, isolated housing developments and industrial sites, and intensively wooded urbanized areas may not appear on this map due to the relatively low resolution of the satellite sensor (2.7 km), atmospheric interference, and the thresholding technique used in this study.
The SRPG model rates soils based on the assumption that crop yields are a function of soil characteristics available in soil survey manuals and USDA-NRCS databases, landform, and climate. The ratings are produced using parameter ranges determined from existing knowledge of plant growth principles and therefore reflect long-term average crop performances rather than actual soil productivity. The latter would take into account soil and crop management (e.g., irrigation, organic and inorganic fertilization, and pest control) and productivity losses, such as those due to erosion or nutrient depletion. Moreover, the SRPG criteria are related to commodity crops that appear in county soil survey reports and are not necessary equally efficient for rating all crops. Different crops have different soil requirements. The actual soil productivity of various mapping units at a given time can be better estimated by computing yield quantities of a specific plant per unit land area and comparing the results to an optimum yield value, a process almost impossible to accomplish at a continental scale. In addition to inherent soil properties, variations in yields are the result of differences and/or interactions among crop varieties, tillage methods, rotation history, level of fertilization, and other factors (Gersmehl and Brown, 1990).
As indicated earlier, SRPG ratings were further grouped into four convenient classes: high (rank 100–76), moderately high (rank 75–51), moderate (rank 50–26), and low (rank 25–0). It should be understood that the lowest-ranked mapping unit of a class and the highest-ranked mapping unit of the next consecutive class might not be significantly different in soil productivity. For example, soil mapping units ranked 76 and 75 or 51 and 50 may not be significantly different in soil productivity. Furthermore, the degree of uncertainty and error propagation associated with GIS analysis operations, urban land use locations, STATSGO mapping units, and soil productivity ratings are presently unknown. However, having established GIS procedures and spatial data sets developed from remote sensing and soil productivity modeling, it would be relatively easy to reassess these interpretations as more detailed information and reliable data become available.
Results of this study should be considered a first step in determining the current status of soil resources in the USA as affected by urbanization. Maps and tabular data presented above are subject to the scale and resolution of original data sources used in the analysis and therefore should be interpreted only as guidelines in state, regional, and national land use planning efforts. County level and site-specific assessments will need more detailed data sets and may yield different results depending on the level of soil productivity and development pressures in the area.
|
SUMMARY AND CONCLUSIONS
|
|---|
This study assessed the distribution of urban land use and soil productivity in the USA. Soil productivity classes were also compared in terms of the areas and level of urbanization. The location and extent of urbanization was obtained by generating an urban land use map from the DMSP/OLS satellite nighttime imagery. Soil productivity distribution was determined using the SRPG model based on the STATSGO soil database and landscape attributes and climate. The level of urbanization for land falling into each soil productivity class was obtained by analysis of data resulting from the GIS overlay of the urban land use map and the soil productivity layer. Results consisted of a set of digital maps and tabular data including summary statistics of soil productivity levels for land under urbanization in the USA and individual states. As expected, the Midwest, Northern Plains states, and river valleys and deltas in other states have highly productive soils. Urbanization occupies about 3% of the U.S. land area and makes up 6, 48, 35, and 11%, respectively, of the land in the high, moderately high, moderate, and low soil productivity categories. The land with highly productive soils in the USA, that is that in the high soil productivity class, represents 3% of the total land area. Unfortunately, this land also has the highest level of urbanization (5%). In general, the level of urbanization increases with increasing soil productivity for the USA and for several states. Although the land with the most productive soils represents a small fraction of the total land area in several states, it also experiences the highest level of urbanization.
Results of this study should be interpreted according to the scale and resolution of data sources. Nevertheless, they provide a basis for developing sound management and/or land use plans for state and regional agencies involved in land use planning. Finally, this study elucidates some of the effects of human appropriations of terrestrial ecosystems, and results can be used in other applications varying from Global Circulation Models to regional assessments of land use and biodiversity.
|
ACKNOWLEDGMENTS
|
|---|
This work was supported in part by the National Aeronautics and Space Administration/Office of Earth Science (NASA/OES) under Grant NAG 5-3856.
Received for publication November 9, 1999.
|
REFERENCES
|
|---|
- Berry, D. 1978. Effect of urbanization on agricultural activities. Growth Change 9:2–8.
- Bliss, N.B., and W.U. Reybold. 1989. Small-scale digital soil maps for interpreting natural resources. J. Soil Water Conserv. 44:30–34.
- Brklacich, M., and K.B. MacDonald. 1992. Prototype agricultural land evaluation systems for Canada: II. Selected applications and prospects. Soil Use Manage. 8:8–15.
- Brown, L.R. 1995. Facing food scarcity. World Watch 8:10–20.
- Brown, D.L., M.F. Brewer, R.F. Boxley, and C.L. Beale. 1982. Assessing prospects for the adequacy of agricultural land in the United States. Int. Regional Sci. Rev. 7:273–284.
- Cornelius, D., and J. Cover. 1997. Population and environment dynamics. Data sheet. Population Reference Bureau, Washington, DC.
- Daily, G.C., and P.R. Ehrlich. 1990. Population, sustainability, and earth carrying capacity. BioSci. 42:761–771.
- Dillman, B.L., and C.F. Cousins. 1982. Urban encroachment on prime agricultural land: A case study. Int. Regional Sci. Rev. 7:285–292.
- Dumanski, J., and R.B. Stewart. 1983. Crop production potential for land evaluation in Canada. Agriculture Canada Research Branch, Ottawa, ON, Canada.
- Elvidge, C.D., K.E. Baugh, V.R. Hobson, E.A. Kihn, H.W. Kroehl, E.R. Davis, and D. Ceceros. 1997a. Satellite inventory of human settlements using nocturnal radiation emissions: A contribution for the global toolchest. Global Change Biol. 3:387–395.
- Elvidge, C.D., K.E. Baugh, E.A. Kihn, H.W. Kroehl, and E.R. Davis. 1997b. Mapping city lights with nighttime data from the DMSP Operational Linescan System. Photogramm. Eng. Remote Sens. 63:727–734.
- Engelman, R., and P. LeRoy. 1995. Conserving land: Population and sustainable food production. Population and Environment Program, Population Action International, Washington, DC.
- Environmental Systems Research Institute. 1998. ARC/INFO. User's manual. Vers. 7.2.1. Environ. Systems Res. Inst., Redlands, CA.
- Fischel, W.A. 1982. The urbanization of agricultural land: A review of the National Agricultural Lands Study. Land Econ. 58:236–259.
- Fisher, P.S. 1982. Introduction: Public policy and the urbanization of farmland. Int. Regional Sci. Rev. 7:249–256.
- Gardner, G. 1996. Asia is losing ground. World Watch 9:18–27.
- Gersmehl, P.J., and D.A. Brown. 1990. Geographic differences in the validity of a linear scale of innate soil productivity. J. Soil Water Conserv. 45:379–382.
- Goebel, J.J. 1992. Description of the national resources inventory. p. A1–A8. In R.L. Kellogg et al. (ed.) Agricultural chemical use and ground water quality: Where are the potential problem areas. USDA-NRCS. U.S. Gov. Print. Office, Washington, DC.
- Imhoff, M.L., W.T. Lawrence, C.D. Elvidge, T. Paul, E. Levine, M.V. Privalsky, and V. Brown. 1997a. Using nighttime DMSP/OLS images of city lights for estimating the impact of urban land use on soil resources in the United States. Remote Sens. Environ. 59:105–117.
- Imhoff, M.L., W.T. Lawrence, D.C. Stutzer, and C. Elvidge. 1997b. A technique for using composite DMSP/OLS "city light" satellite data to map urban area. Remote Sens. Environ. 61:361–370.
- National Agricultural Lands Study (NALS). 1981. Final report. USDA-Council on Environmental Quality. U.S. Gov. Print. Office, Washington, DC.
- National Oceanic and Atmospheric Administration. 1998. DMSP satellite program description. NOAA/NGDC. Available at http://www.ngdc.noaa.gov/dmsp/dmsp.html (verified 2 Oct. 2000).
- Negri, M. 1994. Digital map of state boundaries of the conterminous United States. U.S. Geological Survey, Reston, VA.
- Nelson, T. 1996. Closing the nutrient loop. World Watch 9:10–17.
- Ogunkunle, A.O. 1993. Soil and land suitability evaluation: An example with oil palm in Nigeria. Soil Use Manage. 9:35–40.
- Plaut, T.R. 1980. Urban expansion and the loss of farmland in the United States: Implications for the future. Am. J. Agric. Econ. 62:537–542.
- Raup, P.M. 1982. An agricultural critique of the National Agricultural Lands Study. Land Econ. 58:260–274.
- Richards, J.F. 1990. Land transformation. p. 163–178. In B.L. Turner II (ed.) The earth as transformed by human action. Cambridge Univ. Press, Cambridge, UK.
- Sampson, R.N. 1981. Farmland or wasteland. A time to choose. Rodale Press, Emmaus, PA.
- Simon, J.L., and S. Sudman. 1982. How much farmland is being converted to urban use. Int. Regional Sci. Rev. 7:257–272.
- Sinclair, H.R. 1996. Soil rating for plant growth. p. 279. In 1996 Agronomy abstracts. ASA, Madison, WI.
- Soil Survey Staff. 1994. National collection-State Soil Geographic (STATSGO) Database. Digital general soil maps and associated attribute tables. [CD-ROM.] USDA-NRCS, National Soil Survey Center, Lincoln, NE.
- Soil Survey Staff. 2000. Soil rating for plant growth, a system for arraying soils according to their inherent productivity and suitability for crops. C.S. Holzhey and H.R. Sinclair, Jr. (ed.) USDA-NRCS, Natl. Soil Survey Ctr., Lincoln, NE.
- Sorensen, A.A., R.P. Green, and K. Russ. 1997. Farming on the edge. American Farmland Trust, Ctr. for Agric. in the Environment, DeKalb, IL.
- Stoel, T.B., Jr. 1999. Reining in urban sprawl. Environment 41:7–33.
- United Nations, Food and Agriculture Organization. 1996. World Food Summit. Rome. 28–30 Oct. 1996. FAO, Rome.
- U.S. Census Bureau. 1991. Summary tape file 1B, 1990 census of population and housing. U.S. Dept. of Commerce, Economic and Statistics Administration. U.S. Gov. Print. Office, Washington, DC.
- USDA-Soil Conservation Service. 1981. Land resource regions and major land resource areas of the United States. USDA-NRCS Agric. Handb. 296. Rev. 1984. U.S. Gov. Print. Office, Washington, DC.
- USDA-NRCS. 1997. The 1992 National Resources Inventory. State rankings by acreage and rate of non-federal land developed. Available at http://www.nhq.nrcs.usda.gov/CCS/devtable.html (verified 29 Nov. 2000).
- USDA-NRCS. 1999. Index of maps, facts, and figures. Prime farmland. Available at http://www.nhq.nrcs.usda.gov/land/index/prime.html (verified 2 Oct. 2000).
- Vesterby, M., and R.E. Heimlich. 1991. Land use and demographic change: Results from fast growing counties. Land Econ. 67:279–291.
- Vining, D.R., T. Plaut, and K. Bieri. 1977. Urban encroachment on prime agricultural land in the United States. Int. Regional Sci. Rev. 2:143–156.
This article has been cited by other articles:
|
|
|
|
 
E. G. Irwin and N. E. Bockstael
From the Cover: Land Change Science Special Feature: The evolution of urban sprawl: Evidence of spatial heterogeneity and increasing land fragmentation
PNAS,
December 26, 2007;
104(52):
20672 - 20677.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
T. Chirenje, L. Q. Ma, M. Szulczewski, R. Littell, K. M. Portier, and E. Zillioux
Arsenic Distribution in Florida Urban Soils: Comparison between Gainesville and Miami
J. Environ. Qual.,
January 1, 2003;
32(1):
109 - 119.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
