83 



to give much color to the association, Avhile the more abundant in- 

 terstitials have for the most part veiy small flowers. There is but 

 one season when the association has a well-marked floral aspect. 

 That is during the aestival period when the reddish spikelets of Pani- 

 cuni pseudopiibesccns give a general red tone to the whole. Local 

 displays of color, caused by single plants or groups of Oenothera 

 rhoinbipctala, Monarda punctata, or other species, are conspicuous, 

 but not distributed generally over a whole station. 



The duration of the association is usually not great. Since both 

 its beg'inning and end are caused by wind action, its ag'e depends 

 somewhat upon the rate at which sand is removed. If the destruc- 

 tion of vegetation by the wind is aided by heavy pasturing, its dura- 

 tion is still further shortened, and one station in the Havana area 

 contained at the same time relics of the bunch-grass. Car ex Muhlen- 

 bergii and Leptoloma cognatuni, and pioneers of the blowsand asso- 

 ciation, Paspaluni setaceuni and Sporobolus cryptandnts. Under 

 other circumstances the blowing may cease, and the association then 

 gradually reverts to the original bunch-grass association. 



REVERSION TO THE BUNCH-GRASS ASSOCIATION 



The dominant or climax nature of the bunch-grass association 

 has already been mentioned. Whenever those physical conditions 

 which are concerned in producing the Panicum pseudopubescens asso- 

 ciation become inoperative or ineffective, a reversion to the original 

 vegetation begins. This may take place with considerable rapidity, 

 because of the usual proximity of the two associations and the con- 

 sequent readiness with which migration may take place. Reversion 

 begins not only near the margin of the association, but in the center 

 as well, if that part has ceased blowing. This 'succession has been 

 observed only in the Hanover area, but undoubtedly occurs at any 

 other place where both associations are present and the environmental 

 conditions are suitable. 



The pioneer invader in the Hanover area is Sclagiiiclla riipcstris. 

 Its habit of growth in circular patches allows a comparative esti- 

 mate of the age of dift'erent stages in the succession. Some stations 

 of the Panicum pseudopubescens association were observed which 

 were apparently normal except for a few small, regularly circular 

 mats of Sclaglnella near the margin. The number, size, and regu- 

 larity of the mats all indicate an early stage in the reversion. Later 

 it becomes so abundant that it may form a solid mat on the ground, 

 in which the ring's are of large size and overlap each other. Ac- 



-5 



