56 



SCIENCE 



[N. S. Vol. XXXIV. No. 863 



quadrat methods, and he is especially fortu- 

 nate in his application of the principles of the 

 development and structure of vegetation to a 

 complex vegetational problem. The close and 

 careful analysis of vcater factors and the inti- 

 mate correlation of the natural vegetation 

 with them will cause the present study to 

 long remain a model for work in similar 

 fields. The v&lue of the natural crop as an 

 index of agricultural possibilities is so clearly 

 worked out that it must henceforth be taken 

 fully into account in the survey of a new 

 region. The present bulletin merits further 

 praise for the happy way in which the newest 

 ecological facts are combined with a knowl- 

 edge of crop production in such a fashion as 

 to yield results usable by both the scientists 

 and the layman. It gives further evidence of 

 the fact that the best scientific work is the 

 most practical, and that practise can be 

 permanent or successful only in so far as it 

 is scientific. 



In laying down principles for the use of the 

 natural vegetation as an indicator, correla- 

 tion with the physical or chemical nature of 

 the soil, with rainfall or with temperature is 

 held to be impossible for the region of the 

 great plains. As the ecologist would expect, 

 the water content of the soil furnishes the 

 most reliable correlation, as the most impor- 

 tant and controlling of all direct factors in an 

 arid region. The author is probably correct 

 also in insisting that the entire vegetation is 

 a better indicator of conditions than any 

 single species of it. Since the structure of a 

 plant group varies considerably, however, it 

 is not improbable that further study wiU re- 

 veal a few species which are the essence of 

 the group, and hence the clue to it. Indeed, 

 this is not far from the method used, as 

 shown by the terms short-grass land, bunch- 

 grass land, etc. The second step in the prob- 

 lem is to correlate the native vegetation, 

 which should always be regarded as a crop 

 grown by nature, with the culture vegetation, 

 i. e., the crop production. This correlation, 

 of which the ecologist requires no proof, is 

 practically possible only in so far as actual 

 experiments in cropping have been made. In 



an arid region, all cropping is essentially ex- 

 perimental, and the necessary evidence, 

 quantitative in a large degree, is at hand. 



The first essential in correlating vegetation 

 and conditions is a careful analysis of the 

 former into its formations, associations and 

 societies. Over a vast grass-land area, such as 

 the great plains, this is peculiarly difficult, 

 not only because of the disturbing effect of 

 succession, but also on account of the ease of 

 migration in all directions. Two typical 

 grass-land formations are recognized, the 

 prairie grass formation of the prairies proper, 

 which extends westward into the shortgrass 

 formation characteristic of the plains. The 

 prairie grass formation in its plains portion 

 falls into three groups, here called the bunch- 

 grass, sandhills mixed and blowout associa- 

 tions. The shortgrass formation comprises 

 three associations, grama-buffalo grass, wire- 

 grass and Gutierrezia-Ariemisia. These as- 

 sociations are not all stable groups of the 

 final association, but some, notably the blow- 

 out and wiregrass associations, are initial or 

 intermediate stages of a succession. Indeed, 

 it is the skillful working out of the time se- 

 quence of the various associations which has 

 made possible the correlations suggested. 



The typical short-grass association is made 

 up chiefly of two species, grama grass 

 (Bouieloua oligostachya) and buffalo grass 

 (Buchloe dactyloides) . It is characteristic of 

 loam or clay soils, the so-called " hard lands," 

 upon which the author has worked out in 

 convincing fashion the essential water rela- 

 tions. A thorough study of rainfall, run-off 

 and penetration, rate of water loss, non-avail- 

 able water, and root systems brings out clearly 

 the fact that it is the slight penetration of 

 the rainfall on hard land which controls the 

 root development, and consequently the es- 

 tablishment of the plant. The roots are from 

 12-18 inches deep, corresponding to a pene- 

 tration of little more than 20 inches after the 

 heaviest rains, and a consequent water con- 

 tent which repeatedly falls to the non-avail- 

 able during the summer. In accordance with 

 this, the shortgrass association is an indi- 

 cator of a small amount of available water, of 



