2. MATERIALS AND METHODS 



a. Estimating Growth Responses 



Most of the experiments initiated from 1961 to 1963 utilized stands 

 of American beachgrass, Ammophila breviligulata , and saltmeadow cordgrass, 

 Spartina patens, already established; some natural, but most of them planted 

 by the National Park Service from I to 3 years prior to 1961 on Ocracoke, 

 Bodie and Hatteras Islands. These stands tended to be much more variable 

 than would be desirable for field plot purposes, resulting in quite variable 

 data. 



Since sand trapping was considered a major function of the cover pro- 

 duced, it was not deemed desirable to harvest the growth to measure treatment 

 effects. Consequently, we were faced immediately with the problem of devising 

 a suitable technique that would measure field responses under these limita- 

 tions. A considerable amount of time was devoted to counting stems; esti- 

 mating heights of stems, perimeter of clones, and various combinations of 

 these. It soon became apparent that American beachgrass, as it normally 

 occurs both on the dunes and in the nursery, is a mixture of types having 

 a wide range in stem size and length. Also, individual clones, due to 

 variations in age and environment, produce stems that vary considerably in 

 their dimensions. Of greater concern was the fact that a single plant, under 

 suitable conditions, can produce 50 to 100 stems in one season, making stem 

 counting a very laborious task. Cordgrass is even worse in this respect. 

 This experience led to the conclusion that data based on stem counts was 

 not too reliable and was much too expensive. 



Fortunately, in the summer of 1962, a device was developed which 

 estimated bulk (dry weight) of grass in rows by measuring resistance to 

 compression (10). This procedure was found to be quite rapid and accurate 

 for both American beachgrass and saltmeadow cordgrass as long as the plants 

 remained in rows. However, under favorable treatment, new plants soon 

 appear between rows and by the summer of 1964, a method which would handle 

 broadcast stands became necessary. This was done by modifying a method, 

 based on resistance to vertical compression, developed for use on forage 

 plants (I). The modified method consists of dropping a reinforced sheet of 

 aluminum (2x2 feet) over the grass, and measuring the average height the 

 square remains above the ground. These data are then converted to pounds 

 per acre by the formula: 



(h 2 ) (c) = lb/acre dry weight 



where h = average height and c is an appropriate constant developed for 

 each species. The method ■ i s rapid and reasonably accurate, yielding cor- 

 relation coefficients of 0.90 or better on stands of American beachgrass, 

 saltmeadow cordgrass and dune panic grass. It is not satisfactory on sea 

 oats after the large seed stalks on this species have fully emerged. 



b. Field Plot Technique 



Essentially all of the experiments have been in randomized blocks 

 with at least three replications. Plot size and shape varied considerably 



