242 THE FUTURE OF ARID LANDS 
flows are heavily laden with mud. The aim of management in such 
low-lying areas is to reduce sediment movement. Most of the large 
water reservoirs lie within this zone and catch the debris dumped 
into the the tributary streams. Although the scanty rainfall sup- 
plies plants, they are not competing with man for their supply 
but rather are racing evaporative forces to take water from the 
soil before it returns to vapor. These plants provide soi! protec- 
tion, forage, and other products from soil moisture that would 
otherwise evaporate. Phreatophytes are an exception. These 
water-loving plants are not dependent on local rainfall but grow 
along streams and rivers where they steal water that could be put 
to other uses. Our justified fight against these water thieves must 
not be carried to the upland vegetation (4). 
The permanent rivers, on which the Southwest depends for its 
water supply, tap the higher elevations where precipitation is 
more abundant and water to feed streams is in surplus, at least 
at certain seasons. Just as on lower lands, the protective effect of 
vegetation is important. Where plant cover is reduced, the capac- 
ity of the soil to absorb and store water is diminished. Flash floods 
and mud flows are all too common after excessive grazing use, 
careless logging, or wild fire. Where the protective cover remains, 
streams run clear and contribute little sediment to aggravate 
downstream troubles. The good quality of mountain waters is 
appreciated by irrigators as well as by domestic and industrial 
users. 
In some situations, mountain lands can be managed to increase 
water production. Enough research has been done to show the 
possibilities of increasing streamflow, but not enough has been 
done to make general recommendations. Where winter snow is 
heavy, dense stands of evergreen trees hold snow on their foliage 
and let it evaporate without reaching the ground. In snow coun- 
try, thinnings in young forest, selection cutting, and patch clear 
cuttings to open up areas of dense forest have been shown to in- 
crease the amount of snow finally contributing to streamflow. 
These measures are good forest management practices and can 
be carried out so that both wood and water production are bene- 
fited. Another possibility is the replacement of deep-rooted species 
of trees with shallow-rooted vegetation. This would reduce the 
