Rates of Soil Erosion

The study by Trimble would have been more informative if he had reported what proportion of the 360-km² area he studied was in agriculture from 1930 to 1993; what types of crops were grown during this period; how the crop culture technology changed; and how much of the region was reforested, especially along the creek margin--all factors that would influence the amounts of erosion and sediments being deposited in Coon Creek.

Trimble states that "General and substantial increases of soil erosion in the United States are not borne out by measurements of sedimentation in Coon Creek." Contrary to Trimble's suggestion in this statement that U.S. soil erosion outside of Coon Creek has increased, erosion rates in the United States generally have declined from an average of 17 megagrams per hectare per year (Mg ha^-1 year^-1) (3) to about 13 Mg ha^-1 year^-1 (2, 4) (a megagram is equal to a metric ton). For the 1994 USDA (4) study, investigators sampled 800,000 sites in the United States. A further decline in erosion rates to slightly less than 12 Mg ha^-1 year^-1 was recently reported by the USDA (5). However, this erosion rate is a factor of 12 higher than soil sustainability, on the basis of the average rate of soil formation (6). Uri and Lewis (5) also reported that the social costs of erosion remain high and are estimated to be $29.7 billion annually.

David Pimentel
College of Agriculture and Life Sciences,
Cornell University,
Ithaca, NY 14853, USA

Edward L. Skidmore
USDA-Agricultural Research Service Wind Erosion Research Unit,
Throckmorton Hall,
Kansas State University,
Manhattan, KS 66506, USA

1. The Second RCA Appraisal: Soil, Water, and Related Resources on Nonfederal Land in the United States: Analysis of Conditions and Trends (USDA, Washington, DC, 1989).
2. D. Pimentel et al., Science 269, 461 (1995).
3. D. Pimentel et al., Science 267, 1117 (1995).
4. Summary Report: 1992 National Resource Inventory (USDA, Soil Conservation Service, Washington, DC, 1994).
5. N. D. Uri and J. A. Lewis, J. Sustainable Agric. 14, 63 (1999).
6. F. R. Troeh, J. A. Hobbs, R. L. Donahue, Soil and Water Conservation (Prentice Hall, Upper Saddle, NJ, 1999).

Response
As Pimentel and Skidmore state, land use and soil conservation measures are primary independent variables. For my report it was sufficient to state only that the land use of Coon Creek is generally representative of the region. For details, however, readers were directed to (1), wherein land use and conservation practices have been reconstructed from 1860 to the present. The analysis in that study showed a hysteretic relationship caused by a lag in the response of erosion and sedimentation rates to changes of land use.

With regard to erosion rates in the United States, the values Pimentel and Skidmore refer to of 17 and 13 Mg ha^-1 year^-1 are not measurements, but are estimates from models, and they do not predict movement of sediment to streams. If U.S. soils have indeed been eroding at such rates over the last two or so decades, where are the detritus and efflux?

In regard to soil-loss tolerance (sustainability), Pimentel and Skidmore state that the current average rate of erosion of 12 Mg ha^-1 year^-1 is "a factor of 12 higher than soil sustainability," citing the study by Troeh et al. (2), which would suggest that the tolerance is only 1 Mg ha^-1 year^-1. Yet, Troeh et al., on the basis of USDA information, state that the soil-loss tolerances for U.S. soils range from 2.2 to 11.0 Mg ha^-1 year^-1 (2, p. 115). U.S. agriculture is mostly on soils with a soil-loss tolerance of 11 Mg ha^-1 year^-1 or more (3, p. 678). Hence, there appears to be little disparity between soil-loss tolerance and what Pimentel and Skidmore say is the rate of erosion. Even according to the USDA study cited by Pimentel and Skidmore (4), only one-third of U.S. agricultural land is eroding faster than the sustainable rate--a statement that remains to be proven. Although erosion rates may be periodically high in some regions, U.S. soil erosion remains a problem but does not seem to be a crisis.

Pimentel and Skidmore also mention a USDA study for which 800,000 sites were sampled, seeming to imply a high degree of accuracy. However, these are not physical measures of soil erosion, but are data-gathering sites for models. Moreover, according to Uri and Lewis (5), who they cite, there were only 300,000 such sites. And the annual "social costs" of $29.7 billion in (5) are only asserted; it is not clear what evidence was used to arrive at that figure.

Stanley W. Trimble
Department of Geography and Institute of the Environment,
University of California,
Los Angeles, CA, 90095-1524, USA

1. S. W. Trimble and S. W. Lund, U.S. Geol. Surv. Prof. Pap. 1234 (1982).
2. F. R. Troeh et al., Soil and Water Conservation (Prentice Hall, Upper Saddle River, NJ, 1999).
3. N. C. Brady and R. Weil, The Nature and Property of Soils (Simon & Schuster, Upper Saddle River, NJ, 1999).
4. Summary Report: 1992 National Resources Inventory (USDA, Soil Conservation Service, Washington, DC, 1994).
5. N. D. Uri and J. A. Lewis, J. Sustainable Agric. 14, 63 (1999).