84 mvisTox OF water resources 



obtained by averaging the monthly values for each calendar month and 

 then totalino- these twelve monthly averages. 



. This value, 78.45 acre-inches per acre per year, represents the 

 average annual loss of water dnring this period from tule swamps of 

 the area with the water table at the ground surface or above. No reduc- 

 tion factor should be applied to this value, since the tank was set in a 

 SAvamp and completely surrounded by a growth similar to that in the 

 tank. The rim of the tank Avas completely hidden from the rays of 

 the sun so there conld be no rim effect, and the tank was bedded nearly 

 3 feet deep so that there conld be no abnormal temperature variation 

 or restriction of root activity. 



^Measured losses from the moist area on Temescal Creek compare 

 favorably' with the losses from the 6-foot tank at Victorville. The 

 loss for\he 30-day period from April 28 to May 27, 1929, for the 

 swamp growth on Temescal Creek was 12.9 acre-inches per acre.* 

 For the month of May, 1931, tule Tank No. 3 at Victorville nsed water 

 at the rate of 11.62 acre-inches per acre per month, and the maximum 

 rate measured in July, 1931, was 14.65 acre-inches per acre per month. 

 The cross-hatched area on Plate XVIII shows the relatively large 

 loss of water from Tank No. 1 as compared with Tanks Nos. 2 and 3, 

 where natural conditions were replicated. In this connection, it will be 

 noted in Plate XV that the tule growth in the tanks in the swamp was 

 6 feet high, whereas the growth in Tank No. 1 on the bank (Plate XVI) 

 averaged approximately 3 feet in height. These pictures were taken 

 October 31, 1931. This indicates that the number of pounds of dry 

 matter produced in the different tanks bears no rational relationship 

 to the amount of water consumed when tules are grown in exposed 

 tanks, such as tule Tank No. 1. That is, the loss of water from a 

 natural swamp can not be computed from the relation of pounds of 

 water used per pound of dry matter produced as determined from 

 tules grown in isolated tanlcs. During July and August, 1931, Tank 

 No. 1 used between three and four times as much water as either Tank 

 No. 2 or Tank No. 3, yet the ultimate size of the plants produced in 

 Tank No. 1 was only one-half the size of the plants in the swamp tanks. 

 The results indicate that the controlling factor in the consumption of 

 water is the exposure of the tanks and demonstrates, quite forcibly, 

 that natural conditioixs must be replicated before data are of value in 

 estimating field losses. 



The diameter of a tank is a factor of importance when evaporation 

 from a free water surface is being measured, as demonstrated by R. B. 

 Sleight.** For the period from March 5 to November 13, 1916, Sleight 

 found the evaporation from a 2-foot tank to be 117 per cent of that 

 from a tank 6 feet in diameter. 



However, with plants growing in a tank set in a similar growth of 

 sufficient density so that radiation from the sun does not strike the 

 edges of the tanks, the size of the tank should not materially affect 

 the rate of loss per unit area from like densities of plant growth. 

 Before May 15, 1931, the surrounding swamp growth had not shielded 



* Bulletin No. .3 3, "Rainfall Penetration and Consumptive Use of Water In the 

 Santa Ana Hiver Valley and Coastal Plain," Division of Water Resources, California 

 State Department of Public Works (Page 68). 



** "p:vaporation from the Surface of Water and River-bed Material," (Journal 

 of Agricultural Research, Vol. X, No. 5, July, 1917.) 



