116 CLIMAX FORMATIONS OF WESTERN NORTH AMERICA. 



geographical contact. Still more eloquent is the fact that overgrazing favors 

 BouteUnia and Butbilis at the expense of Stipa and Agropyrum, and thus 

 frequently converts the base association of Slipa-Bouteloua into a pure short- 

 grass cover. Concrete evidence of this has been obtained in widely separated 

 areas and has led to the working hypothesis that a pure short-grass cover is 

 partly if not largely a response to graaing animals. The evidence for this is 

 discussed in Chapter VI. 



Correlation with climate. The apparent objection to the view of the grass- 

 land advanced here is that the climates of Saskatchewan, Nebraska, Arizona, 

 and CaUfornia, for example, are vastly different, and hence the same climax 

 can not exist in all of them. This objection is partly met by the fact that it is 

 impossible to speak of the climate of Arizona or California in particular, since 

 even from the human viewpoint each shows several climates. The conclusive 

 answer, however, is that the objection is based upon a definition expressed in 

 human terms or in physical measures. The everyday conception of cUmate 

 emphasizes temperature, especially the extremes, and rainfall. It ignores 

 water relations very largely and in particular the compensating r6le of water- 

 content. Humanly, the Palouse region of Washington and the prairies of 

 Kansas possess distinct climates, but in terms of wheat production and grass- 

 land vegetation they are very similar. Likewise, the winter in Saskatchewan 

 is long and the summer short, while in Texas just the reverse is true. But the 

 short growth period of Bovleloua gracilis fits into the short summer of Saskat- 

 chewan as readily as it does into the early summer of Texas, with the result 

 that this dominant covers large areas in both. 



Examples of this sort can be multiplied almost indefinitely to prove that in 

 the study of vegetation the plant must be taken as the best if not the only 

 judge of climate. However sympathetic one may be with the use of physical 

 factor instruments, he can not afford to minimize the unique importance of 

 the plant for the analysis of cUmates. To do otherwise is to substitute human 

 judgment for plant judgment in the plant's own field. Hence, in the correla- 

 tion of vegetation and climate, it has seemed imperative to determine at the 

 outset and at first hand just where each formation or association is found. 

 The next step is to accept the judgment of the formation as final, and to 

 regard the climatic region as identical with the area of the formation. This 

 done, it at once becomes possible to correlate cUmate and vegetation by means 

 of phytometers and permanent quadrats, and to check the correlations in 

 some degree by means of physical instruments. 



Use of weather records. The tendency to approach the problem by the 

 use of weather records and floristic reports is almost irresistible, especially 

 in view of the time and effort involved in obtaining an adequate first-hand 

 knowledge of climaxes. However, the latter not only seems indispensable, 

 from the vantage ground of a continuous study of the problem, but its para- 

 mount importance seems to be shown also by the endeavors to correlate an 

 unknown vegetation with imperfect records of climate. The most interesting 

 attempts have been those of Merriam (1898) and Transeau (1905), partly 

 because they have endeavored to determine the limits of vegetational zones 

 by means of climatic Unes. In so far as Merriam 's life-zones dealt with natural 

 vegetation, they are necessarily unsatisfactory, since temperature is far less 

 critical than water for native species. 



