KEPRODUCTION OF WESTEEH YELLOW PINE. 101 



INVESTIGATIONS FROM 1919 TO 1921. 



SOIL MOISTUEE. 



Since all available information pointed to soil moisture as the 

 key to the situation, this phase of the problem was again attacked 

 in 1919, and the study continued through 1920. Soil-moisture de- 

 terminations were made at intervals of from two weeks to a month 

 through the growing season. The samples were taken near the 

 Fort Valley Experiment Station on plots on which brush had 

 been placed with a special view toward simulating conditions in 

 typical brush scattering. Having in mind the discrepancies in 

 earlier experiments, where samples were taken more or less at 

 random, samplings were confined to localized areas, with the object 

 of eliminating as far as possible inequalities of soil, ground cover, 

 and action of tree roots. Each condition, such as " under branches," 

 " between branches," and " no brush," or control, was represented by 

 two sampling points. Since natural conditions on the 1919 plots 

 were more or less disturbed by moving branches, digging, and 

 trampling, new plots were established in 1920. On the whole, the 

 scattering was somewhat lighter on the 1920 plots, and they were 

 closer to growing trees, the "distance being about 60 feet for the 

 1919 and 50 feet for the 1920 plots. In both instances the herbace- 

 ous vegetation was closely cropped by cattle. In this respect con- 

 ditions were very different from the 1909 and 1912 series, which 

 were on ground protected against all grazing. Another difference 

 is in the soil. The 1919 and 1920 series were on the loamy clay soil 

 characteristic of the greater portion of the yellow pine type on 

 the Coconino and Tusayan Forests. The 1909 and 1912 series were 

 on a soil which differs from the above in that it contains consider- 

 able quantities of cinders in mixture or sometimes forming a thin 

 surface layer. 



As in the 1909 and 1912 series, the data point to a saving of 

 soil moisture by the brush cover. (See Table 24 and Figs. 13 and 14.) 

 In 1919 the continued heavy rainfall vitiated the results, so far as 

 critical points are concerned; nevertheless, there was more moisture 

 both under and between branches than in the open in June and 

 September, the only times at which there was even a suggestion 

 of drought. In 1920, the' month of June and the first half of July 

 provided a real test by giving values below wilting coefficient at a 

 depth of 6 inches. The points during this period, June 19, July 6, 

 and July 13, show some irregularity, and on the whole there seems to 

 be no more or even less moisture under the brush 'than elsewhere. 

 This disagreement may be partly explained by the fact that the 

 cover in 1920 was lighter than in any of the previous years, and 

 therefore less effective. The large irregularities in the 12-inch series 



