20 University of New Hampshire [Sta. Bull. 326 



The LaCoss pasture had the least lirush, about 14 per cent ; good 

 grass, largely bents, occupied about 16 per cent; poor grasses about 

 22 per cent; and weeds, including moss, covered 42 per cent; with 

 about 6 per cent of the ground bare. 



The Stewart pasture had the heaviest stand of brush, an almost 

 pure stand of gray birch with some pine, poplar, juniper, and hard- 

 hack. This growth was so heavy that only 3 per cent of the ground 

 cover was grass. About 8 per cent of the area was rock ledge. The 

 brush growth was so heavy that when removed more than half the 

 area was without any vegetation except lichens and moss. Yet this 

 pasture had one of the highest pH concentrations, even on plots not 

 treated with lime. 



On the LaCoss pasture, after fertilizing in the spring of 1937. the 

 herbage shifted from 14 per cent to 3 per cent brush ; from 16 per 

 cent good grasses to 65 per cent, including approximately 12 i)er 

 cent of wild white clover ; a slight reduction in poor grasses from 9 

 to 3 per cent ; and moss from 25 per cent to 4 per cent. 



On the Stewart pasture the brush cover was reduced from about 

 85 per cent to 1 1 : while good grass increasecfc, from 5 per cent to 38; 

 weeds increased from 10 per cent to 43 per cent ; while 8 per cent of 

 the area was rock ledge. The lack of good herbage or even weeds 

 made recovery to good pasture herbage much slower than on the 

 LaCoss pasture. The gray birch, however, was practically eliminat- 

 ed bv the spring of 1940. 



Plot Results 



On the basis of method of removal of In-ush. pulling required the 

 most work and was most effective in preventing regrowth of brush. 

 I'ulling required an average of 101 hours per acre. Burning was 

 next most effective and required only about 60 per cent as much 



Table XII. Herbage composition of pasture plots in per cent of ground cover 



