WATICR LOSSES FROi\[ WET AREAS 41 



diurnal changes of temperature by setting all tanks in the ground 

 with tlie tops level with the ground surface. The growth in any tank 

 should not exceed the number of plants of the same crop ordinarilj^ 

 grown in an equal area under normal field conditions. Not only should 

 the area per plant be maintained but also the volume of soil available 

 for plant roots is important. In tanks having high water tables the 

 soil volume is sometimes limited, curtailing growth of the root system 

 and likewise affecting the aerial growth. jMoreover, a small volume of 

 available soil will soon lose its fertility if heavily cropped. This is 

 especially true of anj^ investigation extending over, more than one 

 crop year when the soil in the tank is unchanged. A sufficient lack of 

 soil fertility may result in a higher water requirement per unit of 

 drv matter produced and be the cause of a considerable source of error. 



A further error also exists when the spread of area of foliage grown 

 in a tank exceeds the tank area. Cases of this kind are found when 

 the groAvth droops or spreads beyond the tank limits. In such a case 

 computation of consumptive use per unit of tank area gives an amount 

 in excess of the true consumptive use as the crop area is in excess of the 

 tank area. 



Experimental records of consumptive use of water by plant growth 

 in tanks include evaporation from the soil as well as water transpired 

 through the stomata of the plant leaf. To the agriculturist and 

 others interested in determining the amount of irrigation water which 

 should be applied to soil in order to produce a normal crop, a separation 

 of the water losses into evaporation and transpiration is not important 

 and the water requirement of a crop as determined by experiment 

 generally includes both soil evaporation and plant transpiration. In 

 making such experiments, evaporation and transpiration can not readily 

 be separated from each other except through the use of methods that 

 are inapplicable to field conditions. Evaporation from soil tanks having 

 the same climatic exposure varies with the degree of soil satura- 

 tion, soil texture, and crop shading. In the tanks used in the experi- 

 ments described herein both the degree of crop shading and the soil 

 moisture at the tank surface varied greatly. Both are greatest for 

 those tanks having the highest water tables. It is evident, therefore, 

 that it is not proper to subtract soil evaporation from consumptive 

 use of water by the crop to arrive at the transpiration alone. 



Lack of natural euAdronment is also an important source of error 

 in conducting consumptive use of water studies by the tank method. 

 As transpiration and evaporation are closely related to climatic condi- 

 tions, tank experiments with crops nuist be conducted where the experi- 

 mental growths can be maintained in their natural environment. 

 Experiments with field crops should be carried on in fields of the same 

 crop variety, those with grasses should be in meadows where the same 

 kind of grass has a natural growth and experiments with swamp 

 growth must be conducted in a swamp area where humidity is high, 

 to obtain results that are at all comjiarable with actual swamp con- 

 sumptive use. 



Santa Ana Station 



Because of errors which might occur in the use of water by evapora- 

 tion or consumptive use of water by crops in different soil tanks, it was 

 thought best to operate all tanks in sets of three, the depth to water 



