Erosion-Productivity Relationships

Roger Veseth

Chapter l -Erosion Impacts, No.6, Spring 1987

Loss of topsoil from cropland alters nutrient supply and distribution, water storage potential, physical properties of the seedbed and root environment. Growers are well aware of the reduced crop yields where all of the topsoil has been eroded, exposing less productive clayey or calcareous subsoils. Yields from these eroded areas are often less than one-half or one-third of the yields where most of the original topsoil still remains. The “clay knobs” of the Palouse region and the’ ‘white soils” of southern Idaho or eastern Oregon and Washington are stark reminders of lost soil productivity due to erosion.

The gradual year-to-year impact of erosion on yield potential often goes unnoticed because of the annual variations in weather patterns, crop diseases and other factors. The yield impact of the loss of 1 inch of topsoil is also relatively small early in the erosion process, but the impact increases with decreasing topsoil depth. After 30 years or more of farming, growers are faced with major reductions in yield potential from topsoil loss, often measured in feet instead of inches.

Research to Quantify Yield Losses

To help quantify the impact of topsoil loss on wheat yield, Maynard Fosberg, University of Idaho soil scientist and STEEP researcher, initiated an intensive field study in 1982. The research was accomplished with assistance from research associates Mary Bramble-Brodahl and Anita Falen, and research technician Dave Mital.

Field studies were conducted in the northern Idaho Palouse region and in southeastern Idaho. Study sites were located in fields of cooperating farmers who used similar, conventional management practices, Several different soil series were sampled on each farm. Within a series, 10 to 15 sites were sampled where topsoil depth evenly spanned the range of depths expected for each series, ranging from severely eroded to uneroded. Soil profiles were described and sampled for laboratory analysis. Wheat yields were determined from three hand-harvested plots each 1 foot square at each site. From the soil and yield data, mathematical models were developed to estimate yield loss with declining topsoil depth.

Palouse Results

The Palouse study sites were located in Latah County, south and east of Moscow, in a 22-inch annual precipitation area. The fields had been in a crop rotation of winter wheat-spring peas.

The researchers found that the presence of a subsoil horizon, that had physical or chemical properties which sharply contrasted with and were less productive than the overlying surface soils, had a significant influence on topsoil depth-yield relationships. This was true both for soils in the Palouse as well as in southeastern Idaho. Research results from Naff and Palouse soil series in northern Idaho help show the importance of subsoil properties. The two soils were extensively sampled for 4 years, from 1982 through 1985.

The Naff soil typically has a relatively shallow, silt loam topsoil over a heavier, silty clay loam subsoil (argillic horizon). Compared to the topsoil, the clayey subsoil reduces water infiltration, slows root growth, and generally creates a less desirable seedbed. In contrast, the Palouse soil does not have a significant textural change between the topsoil and subsoil. The subsoil of the Palouse series has acceptable physical properties for seedbed preparation and root growth.

The average topsoil depth beyond which yields did not increase was called the “influential soil depth, ” From 1982 through 1985, it averaged 27 inches for the Naff soil, but only 18 inches for the Palouse. Calculated average change in yield per inch of topsoil in this’ ‘influential soil depth” was 2 bushels/acre for the Naff soil and 1.6 bushels/acre for the Palouse soil. Average total yield change over the “influential soil depth” was 46 and 27 bushels/acre for the Naff and Palouse soils, respectively, This indicates that, with similar levels of topsoil erosion, a higher percent yield reduction would occur on Naff soils than on Palouse soils.

Soil properties most strongly correlated with yield reductions and declining topsoil depth were organic matter content and bulk density at the 12-inch depth. Soil water content was also an important factor in the yield-topsoil depth relationship in the dry 1985 crop year. Eroded soils typically had a lower water content than soils with thicker topsoil depths.

Southeastern Idaho

Research on wheat yield-topsoil depth relationships in southeastern Idaho began in 1985. The study was conducted in a 14 inch annual precipitation area in southern Bannock County about 40 miles southeast of Pocatello near Downey. Winter wheat-fallow is the typical crop rotation in this area. Three soil series were studied: Rexburg, Ririe and Watercanyon. All three silt loam soils have calcic horizons (layer of calcium carbonate accumulation/ cementation) in the subsoil and they differ primarily in topsoil thickness. In 1985, average yields typically increased with increasing topsoil depth (Table 1).

The yield-topsoil depth model estimated a 1.1 bushels/acre average yield loss for each inch loss of topsoil when Watercanyon soils were excluded. No soil property consistently correlated with wheat yields. However, the level of organic matter, and concentration of phosphorus and several micronutrients typically decreased with decreasing topsoil depth. In contrast, soil pH and calcium carbonate (free lime) content increased with decreasing topsoil depth.

Table 1. Mean and range values for topsoil depth and yield of hard red winter wheat In 1985 at Downey, ID (Foaberg, Ul).

Current and Future Research

Wheat yield and soil data were collected in 1986 at 146 sites on 10 farms in the Palouse area of northern Idaho. The data collected are currently being analyzed. Sampling was conducted on the Naff, Palouse, Thatuna, Tilma, Santa, Taney, Driscoll and Larkin soil series. Five years of data have been collected on Naff and Palouse series, and 4 years of data on the Thatuna series. The latter five soil series now have 1 to 3 years of data in the study. In southern Idaho in 1986, wheat yield and soil data were collected on 28 sites of the Rexburg and Ririe soil series in the second year of research there.

In 1987, additional data will be collected to refine and verify the wheat yield-topsoil depth relationships for all the previously sampled soil series. The next step will be to analyze and compare the cost of erosion between the soil series.