104 BULLETIN 1105, U. S. DEPARTMENT OF AGRICULTURE. 



for study under a. great variety of conditions. At the same time an 

 average drought through June and the first half of July, 1920, 

 resulted in heavy thinning, thus bringing about the desired differ- 

 entiation between various sites. 



The results of the 1920 seedling counts under different methods 

 of brush disposal are given in Table 25 and Figure 14. With the 

 exception of groups 6, 7, and 8. seedlings are strikingly less abun- 

 dant in brush than in open places. These three groups of plots 



Depth 6 inches 



3^8 



_ >• 



~ i. 



OX)' 





(0 *- 



> £ 



Between branches — — 



Under branches 



• Interpolated for'no bru.sh' 



2£* 



Precipifarion (inches) 







* 



; 



. 





/ 



\ 



• 



• 

 



* '•»•- 



• 





• 

 • 





MAY 



JUNE 



JULV 



AUGUST 



SEPT 



OCT 



Fig. 14. — Available soil moisture, brush experiment, Fort Valley, 1920. 



are on closely grazed or overgrazed areas. Under prevailing range 

 conditions, the herbaceous vegetation, especially bunch grass, is 

 taller and more luxuriant among tree tops (Pis. XVI, Fig. 2, XVIII, 

 Fig. 2, and XIX, Fig. 1), in the edges of brush piles, and in scattered 

 brush than in open places. Exceptions to this rule occur where the 

 ground cover was largely destroyed by grazing or other causes dur- 

 ing or prior to logging, as in Groups 6, 7, and 8, Table 25 (PL XVI, 

 Fig. 1), or where the branches fell in such deep masses as com- 

 pletely to kill out the original vegetation. Even in the latter case, 

 however, the grasses often develop very strongly in the edges of 

 the brush cover. 



