422 



SCIENCE 



[N. S. Vol. XXXVII. No. 950 



An extension of observations on tree dis- 

 tribution, as related to the depth of perennial 

 water, to regions outside of southern Arizona, 

 gives interesting, if not entirely conclusive, 

 results. A comparison of the depth to ground 

 water of the Coastal Plain of Texas, as given 

 by Taylor,^ with the tree distribution, as given 

 by Bray,° for example, offers important sug- 

 gestions in the present connection. In gen- 

 eral, it may be said, that the stream bottoms 

 of the Coastal Plain support a hardwood for- 

 est, which also extends over such upland as 

 has a fairly shallowly placed wate* table. 

 Such of the deciduous trees as are marked 

 xerophytes, for example, the post oak, occur 

 on dry ridges where pines of various sorts are 

 also to be found, and where the depth to per- 

 manent water is considerable. Of these trees, 

 the root habit of the long-leaf pine is known. 

 This species has a long tap root which pene- 

 trates to a great depth and which renders the 

 species in a measure independent of surface 

 conditions of soil and moisture. In the more 

 arid southern portions of the Coastal Plain, 

 where the water table lies below 50 feet, cha- 

 parral is characteristic of the upland, and, 

 along the streams, where the water table is 

 less deep, forests occur. 



Northward from Texas, as well as westward 

 from the Coastal Plain of the state, are to be 

 found conditions analogous to much already 

 noted for southern Arizona and the Coastal 

 Plain. That is, other things being equal, 

 trees and forests, especially deciduous forests, 

 are limited to areas where the depth to the 

 water table is not great. Thus, in Kansas and 

 Nebraska, the deciduous forests are mostly 

 confined to the flood plains of streams, while 

 the adjacent upland is treeless. 



As one examines other regions (reference is 

 made more in particular to those that are 

 semi-arid) he finds forests confined to such 

 areas as are underlain by ground water not 

 beyond the attainment by the roots of trees. 



^ U. S. Geological Survey, Water-Supply and 

 Irrigation Paper No. 190, 1907. 



»TJ. S. Dept. Agric, Bureau of Forestry, Bull. 

 No. 47, 1904. 



Although it is not practicable at present to 

 give in detail the relation of tree roots to the 

 water table in the more humid regions, enough 

 is known to justify the belief that often there 

 is a very intimate relation between the two, 

 according to Bowman.* For example, the level 

 of the ground water is said always to be lower 

 in a forested tract. The same writer states 

 that the greater supply of moisture for trees 

 is derived from deeper lying sources, i. e., than 

 which supplies shallowly rooted plants. The 

 roots, also, which supply the moisture, descend 

 to a point a little above the surface of the 

 ground water. If the level of the water table 

 changes greatly, the trees suffer either from 

 lack of moisture, or from poor aeration, ac- 

 cording as it is lowered or raised. The varia- 

 tion in depth to the ground water, however, 

 does not affect trees having superficial roots, 

 or at least roots which do not attain it, and 

 such species are well adapted for growth where 

 the water table is high, or the upper soil is 

 shallow. The ecological importance of this is 

 apparent, and may be illustrated by a single 

 example. Rossmassler' mentions trees which 

 are characteristic of two habitats, of which 

 one is rough and stony, and the other is 

 underlaid by an impervious clay. Oaks and 

 pines form a mixed forest in the first habitat, 

 and of these the oaks at least have deeply 

 penetrating root-systems. In the second hab- 

 itat there is only Picea, since the soil depth 

 prevents such root development as is charac- 

 teristic of the other species. 



An important phase of the study of the 

 relation of roots to the water table lies in 

 observing the range of variation under nat- 

 ural conditions. Specialized roots, such, for 

 example, as were mentioned at the beginning 

 of this note, are, generally speaking, not ca- 

 pable of great variation. Hence, plants with 

 this character of a root-system, and for this 

 reason only, may have sharp bounds placed 

 on their distribution. On the other hand, 

 generalized root-systems are often variable to 

 a high degree, and, corresponding to this fact, 



'"Forest Physiography," p. 42, etc., 1911. 

 '"Der Walde," p. 31, 1881. 



