THE RELATION OF FORESTS TO STREAM FLOW. 



By James W. Toumey, 

 Collaborator, Bureau of Forestry. 



INTKODUCTION. 



For the purposes of the present discussion ' ' forest " must be under- 

 stood to mean a growth of trees sufficiently dense to form a fairly 

 unbroken canopy of tops, not a scattered growth of low, round-headed 

 trees with bushes and herbage constituting the dominant tj-pes of 

 veg'etation. 



Forests of this kind do not occur in the United States where the 

 mean annual precipitation falls below 18 to 20 inches, except on 

 restricted areas where unusual conditions prevail. The line of separa- 

 tion between the great eastern forest area and the plains approximately 

 coincides with a north and south line marking a mean annual rainfall 

 of 20 inches. The streams which rise in the Rocky Mountains and 

 flow eastward are bordered by forests for long distances into the 

 plains, where the annual rainfall is much less than 20 inches. These 

 forests, however, are not so much a result of the rainfall in the regions 

 where they occur as of surface and seepage flow from adjacent regions. 

 The mesquite forests of the desert regions of southern Arizona, where 

 the mean annual rainfall is but 8 to 12 inches, are made possible by the 

 seepage and surface waters from the adjacent mountains. 



The question of the exact relation which exists between forests and 

 stream flow has long been under discussion. The broad fact that a 

 relation exists is indeed indisputable. Forest destruction always 

 produces a change in the character of the run-off. But the scientific 

 determination of all the causes which produce this effect, and of their 

 relative importance is a difficult and complicated matter. In spite of 

 the fact that for many years European forest experiment stations 

 have been carrying- on observations, measurements, and experiments 

 designed to clear up this subject, final conclusions covering the whole 

 field have not yet been established. In this country almost nothing 

 has ever been done to secure accurate data for the investigation of 

 this problem as a whole. Some light, however, has been thrown on 

 the subject by means of a series of observations which have been 

 going on for several years in the San Bernardino Mountains in 

 southern California. It is the purpose of the present article to make 



279 



