planting site at Drum Inlet. There were no noticeable differences in 

 the performance of the selections of bitter panicum from Hatteras and 

 Ocracoke. The low survival of Hatteras beachgrass cannot be explained; 

 it is possible that the transplants were initially of low vigor. The 

 low survival of sea oats is consistent with results from other 

 experiments and those of Dahl, et al. (1975). 



b. 1974. Plant cover was estimated to be about 70 percent in June 

 1974. Of the three dune grasses planted in April 1973, only bitter 

 panicum contributed an appreciable amount of this estimate (Fig. 8). 

 The remainder of the cover was composed of native plants that invaded the 

 planting (Table 3); a few seedlings of American beachgrass and sea oats 

 were observed. Cenchrus tvihuliodes L. (sand spur), D-tod-ia teres Walt, 

 (buttonweed) , Oenothera humifusa Nutt. (evening primrose), and purple 

 sandgrass were the most widely distributed invading species (Table 4) . 

 Only 0. himifusa is a perennial. 



By September 1974, plant cover was estimated to be about 75 percent 

 (Fig. 9). Again, only bitter panicum contributed an appreciable amount 

 of this estimate, with the remainder composed of the invading nati-"'e 

 plants (Table 3). Purple sandgrass, one of the four abundant inv ders, 

 was the most prominent. Although the shoots of bitter panicum were well 

 developed, the rhizomes had not proliferated rapidly over the area. 

 This growth response was possibly due to the lack of sand movement. 

 Bitter panicum is most healthy and vigorous in areas accumulating 

 moderate amounts of sand. 



Plant frequency, the percentage of sample plots in which a plant 

 species occurs, is a relative measure of distribution. In the October 

 sampling, values of 60 percent and above were obtained for bitter 

 panicum in all plots of original planting (Table 4). However, for both 

 American beachgrass and sea oats, frequency values were 60 percent or 

 below in the experimental plots; in several plots these species were not 

 recorded in the sample. 



Except for plot 18, cover values were 70 percent and above (Table 

 5) . The relatively high cover values were largely due to the invading 

 species, particularly the annual, purple sandgrass. 



Bitter panicum contributed about 27 percent of the aboveground dry 

 weight, and averaged about 30 grams per square meter in the plots of 

 original planting, 1, 2, 3, 5, and 6, and where purple sandgrass biomass 

 was also determined (Table 6) . Purple sandgrass was sampled for biomass 

 in only six plots, but since it was uniformly distributed it probably 

 contributed at least a minimum biomass of 37 grams per square meter 

 which would be more than that of bitter panicum (Table 6) . 



2. Drum Inlet Planting . 



a. 1973 . Survival of transplants in June 1973 was estimated at 

 75 percent, except for the two rows adjacent to the 0.61 -meter sand 



20 



