350 CARNEGIE INSTITUTION OF WASHINGTON. 



The reconstruction of the original grasslands of the Great Valley and the 

 coastal valleys and hills of California has reached the point where it is possible 

 to recognize the major dominants and their associates, and the role of each, 

 to determine their serai relations, and to evaluate the relationship to the 

 similar bunch-grass prairie of Oregon and Washington and to the mixed 

 prairie of northern Arizona. Of 20 dominants, 16 occur in the northern 

 bunch-grass and 12 in the prairies to the southeast. In addition, the southern 

 portion contains 7 relict species that occur abundantly in southern Arizona, 

 5 of which belong to Bouteloua. It seems evident that the grassland of 

 California was originally in contact with the mixed prairie, as the bunch-grass 

 of the Northwest is to-day, and that the climatic change which produced the 

 Mohave and Colorado deserts led to its sharp differentiation. A similar 

 effect is now taking place in the north, where the isolation of the California 

 and Oregon portions of the bunch-grass climax by mountain ranges has 

 already produced an important difference in the major dominants. 



The Original Grassland of Mohave and Colorado Deserts, by F. E. Clements. 



The comparative study of grasslands in California and Arizona indicated 

 that these must once have been continuous across the Mohave and Colorado 

 Deserts. A special search was consequently made across both areas for 

 relict grasses, and this not only resulted in proof of the assumption, but it 

 also furnished evidence for the probable sequence of climatic and vegetational 

 changes since the Pleistocene. Probably the most significant discovery was 

 that of a well-developed community of Sporobulus cryptandrus flexuosus, 

 Muhlenbergia porteri, and Artistida purpurea in sandy plains near Goffs and 

 Yucca. These have been able to persist in a rainfall of about 6 inches, since 

 the efficiency of the latter was increased several inches by the sand mulch. 

 The floor of the Mohave with a rainfall of 2 inches is unable to support even 

 the most intense xerophyte among the grasses, the shrubby Hilaria rigida, 

 which is found there only in washes and sand. Eriocoma cuspidata is nearly 

 as resistant, and this is followed closely by Stipa speciosa. At Cabezon, on the 

 western edge of the Colorado Desert, Stipa speciosa, coronata, setigera and 

 eminens, Koeleria cristata, and Poa scabrella suggest the grassland that 

 must have covered much of the two deserts when they had a rainfall of 

 about 10 inches. This is essentially in agreement with the presence of 

 Hilaria jamesi, Stipa comata, and Poa scabrella in the Argus Mountains 

 at about the same rainfall. The grassland at Kingman, above the eastern 

 edge of the Mohave Desert, with a rainfall of 12 inches, consists of Bouteloua, 

 eriopoda, Aristida divaricata, Sporobolus strictus, S. c. flexuosus, and Hilaria 

 rigida, while further east at Seligman, with a rainfall of 15 inches, Bouteloua 

 gracilis, B. racemosus, Andropogon scoparius, and Stipa pennata have appeared 

 also. The significance of this is emphasized by the occurrence of Bouteloua 

 gracilis, hirsuta, rothricki, and bromoides as relicts at Jamacha, near San 

 Diego, in a similar rainfall. Since grasses are the most important indicators 

 of climatic and vegetational changes in semi-arid and arid regions, it becomes 

 possible to estabhsh a sequence of communities from the absence of grasses in 

 the midst of the Mohave and Colorado Deserts in a rainfall of 2 to 3 inches 

 to mixed prairie under 15 inches at Seligman. This sequence is regarded as 

 indicating the changes by which these regions have passed from mixed- 



