A NATIONAL PLAN FOR AMERICAN FORESTRY 373 



Not only did the timbered lands produce a negligible run-off but 

 there was no erosion from such areas. 



As shown in these investigations, the effectiveness of forest stands 

 of different densities in holding back run-off on these uplands varies 

 materially with density of stand and intensity of grazing. 



BEHAVIOR OF SMALL STREAMS 



Although reliable quantitative measurements on intermittent small 

 streams are unavailable, field observations indicate that those origi- 

 nating on field and pasture areas may yield 50 percent or more run-off 

 from hard rains. Thus a watershed of 100 acres may readily develop 

 a stream the peak flow of which exceeds 100 second-feet. It would 

 be virtually impossible for such a flow to be developed by an area 

 having a complete forest cover. 



Of greater interest is the behavior of permanent small streams 

 draining watersheds so small that the entire area is likely to be 

 affected by a single local storm. A watershed area of 5,000 acres, for 

 example, may deliver normally a stream of about 5 second-feet. 

 Under extreme conditions such a stream may be swollen to 800 or 

 more times its normal volume, developing a peak flow of 0.8 second- 

 foot per acre of watershed. On larger watersheds, although under 

 similar conditions the total flood volume may be proportionately the 

 same, because of the greater length of stream and greater diversity of 

 watershed conditions, there is a tendency for the flood to be relatively 

 more prolonged and for the peak flow to be relatively lower. 



The flashy run-off of the small streams is strikingly illustrated by 

 Gilmore Creek, near Winona, Minn., having a watershed area of 

 only about 15 square miles, of which about 40 percent has been 

 cleared and having, under normal conditions, a discharge of 10 second- 

 feet. During August 1932 the discharge of this stream rose in 2 

 hours to nearly 5,000 second-feet, sweeping everything in its path, 

 including finally the concrete weir at which the discharge was meas- 

 ured. Such phenomena have not been infrequent in the history of 

 this stream since the advent of agriculture. 



SILT LOADS OF STREAMS 



According to conservative estimates by the Lake States Forest 

 Experiment Station based on sampling, the Wisconsin River carries 

 2 million cubic yards of silt, despite the fact that some of the load 

 from the upper three fourths of its drainage now settles above power 

 dams. A proportionate silt contribution from the Black River, on 

 which conditions are similar, would probably be about 1 million cubic 

 yards annually. 



The smaller streams that rise within the unglaciated area are like- 

 wise heavily loaded with silt. Their contributions of water may 

 never be large enough to affect appreciably the flow of the Mississippi, 

 but they bring silt loads entirely disproportionate to their water 

 discharges. Temporary streams that empty directly into the Mis- 

 sissippi often carry 5 to 10 percent of solid matter. Larger and 

 more permanent streams such as the Buffalo River sometimes carry 

 5 percent of silt to their outlets. 



