Soil organic matter content appears in five of the equations. The favorable ef- 

 fects of organic matter in promoting aggregation of clay are well documented. However, 

 the equations for the Diamond Mountain, Basalt, and Trinity study areas imply definitely 

 adverse effects of organic matter on the stability of sandy soils; and it is exp.ected 

 that more intensive sampling would have revealed similar effects on Coolwater Ridge. 

 It appears that while organic matter binds clay and silt particles into aggregates that 

 resist erosion, it has an adverse effect on aggregation of sand particles. It is 

 hypothesized that this adverse effect results from the hydrophobic character of organic 

 coatings on sand particles, which causes the particles to resist wetting and, possibly, 

 to possess mutual electrostatic repulsion, thus making the sand particles more easily 

 detached and transported. No report of this phenomenon has been found in the literature 

 but its occurrence in widely separated areas, as found in this study, indicates that it 

 is not a mere coincidence, but an actual effect that should be recognized and 

 investigated further. 



The effects of cover, slope, organic matter content, and other site factors are 

 discussed in detail for each study area in the following sections. Results on two of 

 the study areas appear in other papers (Meeuwig 1969, 1970), but they are also pre- 

 sented here in a revised form to serve the purposes of this paper. 



Great Basin Experimental Area. --In this area of calcareous fine-textured soils, 

 bulk density was found to be the most important secondary factor affecting soil erosion. 

 The proportion of soil surface protected from direct raindrop impact explains 52 percent 

 of the variance of the log of soil eroded. Bulk density of the surface 4 inches of soil 

 in combination with cover explains 62 percent of the variance. Plot slope gradient 

 accounts for an additional 4 percent of the variance. 



The regression equation for sheet erosion on this study area is: 

 y = -3.12 - 0.618B - 2.5052 + 5.92ff - 2.53/72 + 1.445/7 + 0.02216" 



in which and G are: protective cover (plant, litter, and stone); bulk density; and 



slope, as defined previously. This equation is based on 162 plots and has a coefficient 

 of determination (R^) of 0.66. Its standard error of estimate is 0.58. Since the de- 

 pendent variable is a logarithm, the standard error of estimate is also a logarithm and 

 not easily interpreted. To overcome this difficulty, erosion as estimated by this equa- 

 tion is plotted logarithmically in figure 2 against erosion as actually measured. 



The relation of erosion to protective cover and bulk density, as defined by this 

 equation, is shown graphically in figure 3. Slope gradient was held constant at its 

 average of 18 percent for the calculation of curves presented in figure 3. l\Tiile cover 

 percentage exerts the major controlling influence on the weight of soil eroded, soil 

 bulk density has an important influence. At any fixed cover percentage the amount of 

 soil eroded is about twice as great at a bulk density of 1.1 g./cc. as it is at 0.9 

 g./cc. Bulk density influences erosion because aggregation and porosity are inversely 

 related to bulk density. Wei 1 -aggregated soils tend to have low bulk densities and they 

 also tend to resist erosion. Soils of high porosity have good infiltration character- 

 istics and, consequently, produce less overland flow and erosion. 



Correction factors for deviations of slope gradients from an average of 18 percent 

 are tabulated in table 1. Weights of soil eroded in figure 3 should be multiplied by 

 the appropriate factor in table 1 to correct for slope effects. At any given cover 

 percentage and bulk density, the amount of erosion is about 3 times greater on 40 per- 

 cent slopes than on 18 percent slopes. 



6 



