Figure 15. — Soil erosion 

 on the Trinity 

 Mountains study area 

 in relation to 

 percent cover of 

 plant and litter at 

 2y 6, 10 percent 

 organic matter in 

 the surface inch of 

 soil. 



20 



30 



40 50 60 70 80 90 



PLANT AND LITTER COVER {%) 



graphically in figure 15. When more than 50 percent of the soil surface is protected 

 by vegetation and litter, organic matter favors resistance to erosion. At less than 

 50 percent cover, there is a reversal of this influence and a tendency for the more 

 highly organic soils to be more erodible. This is an important consideration because 

 the average plant and litter cover on the plots in this study area was only 52 percent 

 and organic matter content averaged 6 percent, a high organic matter content for soils 

 averaging 72 percent sand. 



Since this equation accounts for only 57 percent of the variance, there are obvi- 

 ously other factors affecting erosion on this site. Further study is needed, particu- 

 larly on inorganic bonding between sand particles and on the nature of the adverse 

 effects of organic matter on stability of sandy soils. 



APPLICATION 



The regression equations presented in this paper may be used to estimate the amount 

 of sheet erosion expected under design rainstorm conditions. These estimates are rela- 

 tive values to be used for comparisons among sites. The equations are derived from 

 small plots with a fixed amount of simulated rain and cannot be expected to yield abso- 

 lute estimates of erosion because variations in rainfall characteristics and plot size 

 will influence the actual amount of erosion. 



Since direct solution of the equations is tedious unless a computer is available, 

 tables and nomograms are provided in the Appendix to facilitate calculations. Cool- 

 water Ridge is not included because its equation is based on too few observations to be 

 useful for estimation. 



Estimates may be made for areas other than those studied only if there is reason- 

 able assurance that the area in question closely resembles one of the study areas and 



16 



