ECOLOGY. 351 



teristic. Similar results were obtained for the eastern desert scrub in 

 New Mexico and its ecotone with the grassland was traced through a 

 new region. 



The short-grass plains, subclimax prairie, true prairie, and mixed 

 prairie were given much detailed attention in eastern New Alexico, 

 Oklahoma, Kansas, Nebraska, Colorado, South Dakota, and Wyoming. 

 The most important results, however, relate to the effect of grazing 

 and are considered later. The occurrence of true prairie on the 

 "breaks" of the Canadian River and of "shinry," a mixed community 

 of dwarf oak and Andropogon, was significant of the associational 

 relations in the short-grass region. Nearly the whole length of the 

 contact between mixed prairie and sagebrush in Wyoming was traced 

 in some detail. While this broad ecotone often takes its color and 

 appearance from the sagebrush, the latter rarel}'" occupies half the 

 ground, except in valleys or on new slopes. 



Changes in Vegetation, by F. E. Clements and E. S. Clements. 



During the past eight years the conviction has steadily gro\\Ti that 

 nearly all grassland has been greatly and some grassland profoundly 

 modified by animals. The effect of gi*azing has often been markedly 

 supplemented by the dry phase of the climatic cycle. The universal 

 importance of grazing as a primary factor in the structm-e of grassland 

 has now been confirmed beyond any question. The evidence is too 

 lengthy to be reviewed here and will be treated in a special paper. 

 Perhaps the most striking evidence has been obtained from Sheep 

 Mountain, an isolated butte with precipitous walls and a gi'assy top 40 

 to 50 square miles in extent. The larger third of this is regularly grazed 

 by cattle, with the result that the short-grasses are dominant and the 

 tall-grasses secondary, having been greatly reduced in the past five years. 

 The middle area is connected with the preceding by a narrow land- 

 bridge which has permitted horses but not cattle to pass. This area 

 seems to have been grazed only by horses a number of years ago. 

 The tall-grasses, with much Stipa, are dominant, while the short-grasses 

 form a carpet over about a third of the surface. The third area is 

 completely cut off from the second by a deep bad-land gorge, and all 

 the evidence indicates that it has never been grazed. Here the tall- 

 grasses are in full control, with Stipa dominant, and only occasional 

 culms of Bouteloua gracilis or tufts of Carex filifolia. 



The most completely modified of all the grassland associations is the 

 bunch-grass prairie of California. In fact, until the present investi- 

 gations were begun, the grass community that preceded and yielded to 

 the exotic Avena, Hordeum, Bromiis, and Festuca was not at all known. 

 The studies of the past four j^-ears have pieced together the original 

 Stipa prairie, until its general outlines can be traced throughout the 

 State by means of innumerable relict communities. Some of these 



