22 EXPERIMENT STATION BULLETIN. 



BARRIER OR SHELTER METHOD. 



Sand blows often occur which do not warrant extensive outlay of time 

 or money. The private owner is willing to do something, but he wants 

 to do it gradually, a little at a time as his means and time warrant. To 

 such conditions the Barrier Method of control is advisable. The vul- 

 nerable point of attack must first be determined. This is indicated by 

 the general travel direction of the sand. On shore sand dune work there 

 is only one point of beginning and this is at the water edge. With in- 

 land blows however this point may be more difficult to determine 

 especially if one cannot watch I he blow throughout the year. Often 

 there are seasonal blows from diU'erent directions, in which case dif- 

 ferent sides of the blow should necessarih come under control simul- 

 taneously. 



, . 





Fig 13. Birds eye view of tin- -ovcriiiiirnt planting of permanent clumps on the \\vst 

 slope of Creeping Joe. The haphazard location of the groups of trees furnish little or no 

 protection to the entire cover planting on the upper slope of the dune. A few years more 

 would have witnessed the complete burial of the temporary grass and poplar cover at the 

 crown of the dune. 



The tirst operation consists in establishing a living cover of sand- 

 binding material and low woody plants with a permanent setting of 

 small trees for both temporary and permanent shelter. For the pur- 

 pose of sand-binding work, any herbaceous plant that will grow on the 

 sand in the, locality should be used. Such plants as the valuable and 

 common beach grass ( Ammophylla arenaria), the beach or sand pea (La- 

 thyrus maritimus), Rosa virginiana, and other marsh or upland grasses 

 and wild vetches are suitable. These are planted with 2 by 2 ft. spacing 

 over the entire belt and later protected by the dead cover. Willow 

 cuttings are next made from the abundant marsh willows which are 

 common all along the Lake Michigan shore. The variety does not seem 

 to make much difference so long as good fresh new wood from rapid 

 growing trees is used. The cult ings should be twelve to eighteen inches 

 long and not over three-fourths inches in diameter at the butt end. 

 These are stuck into the sand to a depth of ten to fourteen inches so as 

 to insure being constantly in moist sand. A row of willow cuttings is 



