Studies in Prairie and Woodland 1 1 



of the first foot in July because of local showers. By comparing 

 this graph with figure 22 of Weaver and Thiel, it will be seen 

 that the sumac station was characterized by considerably less 

 available water in the upper layers of soil during the season of 

 1917 than in 1916. However, there was an abundance of mois- 

 ture at the lower depths in the sumac thicket throughout the sum- 

 mer, ranging from a minimum at 4-5 feet of 21.0 per cent, to a 

 maximum of 27.0 per cent, for the same level. This is a sig- 

 nificant fact in connection with the possibility of colonization of 

 such stations by the deeper rooted shrubs and later, by trees. The 

 ecological position and possibilities of the sumac thicket as com- 

 pared with adjacent prairie habitats in so far as soil moisture is 

 concerned is strikingly illustrated by the fact that the water con- 

 tent of the 4-5 ft. level in the prairie was barely one half that of 

 the same depth in the sumac. And, furthermore, it was found 

 that this contrast in the two sites continued throughout the vege- 

 tative season. These facts are of utmost importance when con- 

 sidering the possibility of the invasion of the prairie by any but 

 the most deeply rooted species and consequently also when con- 

 sidering the successional relations involved. 



Figure 5 portrays the march of soil moisture in the prairie 

 station at depths of 0-6 in., 6-12 in., and 12— 24 in. from April to 

 September, 191 7. The horizontal lines represent the wilting co- 

 efficient, the upper being that for the 12-24 in. layer and the lower 

 that for the 0-12 in. layer. 



The rapid depletion of soil moisture from about June 15 until 

 late in August is the striking feature of the conditions which 

 this graph represents. Available soil moisture was completely ex- 

 hausted by late June or early July and the depression of the mois- 

 ture content continued rapidly until late July. The wilting co- 

 efficient was not again exceeded until August 7 for the 0-6 in. 

 level and August 28 for the 6-12 in. level. The moisture content 

 in the second foot of soil was still below the wilting coefficient at 

 the close of August. Thus during practically the entire latter 

 half of the vegetative period very little or no soil moisture was 

 available for those plants whose root systems are confined to the 

 upper two feet of the soil. Determinations which are not re- 



17 



