72 MISC. PUBLICATION 257, U. S. DEPT. OF AGRICULTURE 



the surface. While it is true that a root system is modified somewhat 

 by the water conditions, it is also true that the size and form of root 

 systems are probably determined more by soil conditions and the 

 inherent nature of the plant than by the water relations, which are 

 inconstant and vary markedly from place to place even within narrow 

 geographical limits. 



STORING AND DISPLACING OF WATER 



The storage of water to avoid drought is characteristic of succulents 

 (drought-resistant plants) rather than of trees, although, as has been 

 pointed out, the water in tree trunks, as well as that in the larger 

 roots, serves to some extent as a daily reservoir. Living wood may 

 contain 70 percent and more of water, and losses during dry periods, 

 as well as those caused by daily transpiration excesses, may be re- 

 placed from the amount stored in the stem and roots. That tree 

 stems behave as water reservoirs and change in size accordingly has 

 been shown by the work of MacDougal (142, 143), who measured the 

 daily variations in circumference of trees in the southwestern United 

 States by means of special apparatus (dendrograph) ; in practically 

 all of the species measured, circumferences were found to be greater 

 at night and less in the daytime. 



When the soil-moisture content reaches a critical minimum, plants 

 are able to withdraw water from one cell or organ to supply another. 

 Thus the mucilaginous cells in the leaves of Quercus pedunculated,, 

 birches, and willows may serve as a supply of water to the adjacent 

 cells. When water is plentiful the cells absorb water and retain it, 

 and when a shortage of water sets in these cells sink together as the 

 mesophyll takes water from them. According to the calculations of 

 Haberlandt (81) the quantity of water liberated by the collapse of 

 these cells to one-half of their original volume can supply the water 

 transpired by a horsechestnut leaf for more than 2 hours and by pear 

 trees for 39 minutes. 



That the demands of the leaves seem to have precedence in the 

 water supply of the tree is indicated by the work of Bartholomew (9), 

 who showed that detached lemon twigs with fruits attached wilted 

 much less in a unit of time than did those without fruits. At the same 

 time, the attached fruits decreased in diameter two to three times as 

 much as did detached fruits. Although the leaves on the tree do not 

 wilt until the wilting coefficient is reached, the fruits suffer long before 

 the soil moisture has reached this critical minimum. Similar results 

 were reported by MacDougal {142) with walnuts in Arizona. 



It is also interesting to note that when some of the leaves are re- 

 moved from a rapidly transpiring tree, the remaining leaves may 

 increase their transpiration rate, as reported by Hartig (91) in a 

 5-year-old potted spruce plant. When all of the needles were present 

 on the tree the leaves transpired at the rate of 270 g per square meter 

 of leaf surface. When 40, 70, and 90 percent of the needles had been 

 removed, those remaining transpired at the rate of 272, 460, and 607 g 

 per square meter, respectively. 



Plants are not only able to transfer water from one organ to another, 

 but they are also able to transfer it from moist soil to dry soil, accord- 

 ing to Breazeale and Crider (20). Working with catclaw, date palm, 

 the orange tree, tesota, palo verde, mesquite, and other plants, these 

 workers found that the roots could penetrate soils that were below the 



