392 ANNUAL REPORTS OF DEPARTMENT OF AGRICULTURE. 



lof]jf!;c(l if possilile within two years aftor the seedlings start. There 

 are certain areas on the Forest on whicli ])hin( injj^ or broadcast sowing 

 will probably be necessary. Under this class ar& included lands 

 which have alread}'" been too heavil)'' cut and certain lands on which 

 satisfactory natural reproduction can not bo obtained within a rea- 

 sonable time even inider the most careful maniigement. Scattering 

 bnish after logging promises to aid natural reproduction materially 

 by reducing evaporation and moderating the effect of frosts. 



rie])r()duction suffers severely from sheep grazing. At present the 

 only practical preventiA'o measure for such damage seems to lie in the 

 total exclusion of sheep from areas to be restocked, until the seedlings 

 have grown out of danger. On an average, this will probably be about 

 twenty j^ears. Exclusion on areas to be logged should begin before 

 the cutting. 



Fires have done serious damage to reproduction in the past. 

 Grazing decreases the danger from this cause by reducing the amount 

 of inflammable material on the ground. In the future damage from 

 fires will be largely eliminated. 



As a means of learning the effect of different methods of cutting 

 upon natural reproduction, cut-over areas in which there is already 

 partial regeneration w^ere studied. From such areas immediate 

 results can be obtained; but because the old cuttings seldom fol- 

 lowed distinct cutting systems, such as it is hoped to adopt in the 

 future, it is necessary also to gather more perfect data by applying on 

 small areas an ideal system of cutting, lliis was done in connection 

 with a timber sale on the Cheyenne National Forest. Definite areas 

 were laid out in a lodgepole pine forest, trees marked for cutting, and 

 after their removal the brush on certain plots was lopped and evenly 

 scattered, and on others piled and burned. On one ])lot approxi- 

 mately 50 per cent of the trees were cut, the selection being as nearly 

 perfect as possible. All overmature trees and all suppressed trees 

 were cut. In contrast with this method, on an adjacent area where 

 the conditions were similar, openings of 30 to 40 feet in diameter were 

 made by removing groups of overmature trees. These openings 

 covered approximately 40 per cent of the area. Throughout the 

 remainder of the stand only suppressed and defective trees were 

 removed. Differences in reproduction in each of these two experi- 

 ments are to be expected. The advantage of brush burning in 

 inducing early reproduction as against the slower and less dense, but 

 perhaps more thrifty reproduction of the unburned area, will be care- 

 fully noted. Under the first system a considerable amount of repro- 

 duction should gradually appear before the next cutting. Under the 

 group system the openings should reproduce rapidly, especially where 

 brush has been burned. Similar experiments have been started in 

 Engelmann-balsam stands and also in yellow pine forests. 



STREAMFLOW STUDIES. 



To determine exactly what effects the forest cover may have upon 

 the disposition of rain and snow water, its run-off, storage, and later 

 appearance in springs, and the possibility of erosion and silting up of 

 streams, an experiment station has been started at Wagon Wheel Gap, 

 on the Rio Grande National Forest. Two similar watersheds w^ere 

 found, of approximately 200 acres each, empt}dng into the valley of 



