74 

 tissue, and weighing the remaining plant material to the nearest 0.05 kg 

 in order to determine post treatment biomass. Results of phase one were 

 analyzed for the mean percent change in fresh weight and mean percent 

 change in number of plants from initial pretreatment levels. The means 

 of the percent change by treatment were analyzed for the presence of 

 significant differences utilizing Waller-Duncan procedure for the com- 

 parison of multiple means. The Waller-Duncan procedure was employed 

 because it has the advantage over other such methods in that the 

 observed F value is used in the calculation of the LSD (least signifi- 

 cant difference). The use of the observed F value provides a method of 

 accounting for both the comparisonwise and experimentwise Type I error 

 rates (Chew, 1977). 



Phase two efficacy evaluations consisted of visual assessment of the 

 presence of viable meristematic tissue in the treated plots (Seale and 

 Allison, 1946). At the conclusion of 22 days post- treatment, plants 

 treated with 2.24 kg/ha 2,4-D appeared to exhibit near 100 percent 

 control and the decision was made to complete the efficacy evaluation on 

 day 24 post- treatment. Thirty random plants on three transects through 

 the treated plots were examined and the proportion of dead plants per 

 plot calculated. Proportions of dead plants per treatment plot were 

 analyzed for significant differences utilizing a method described by 

 Walpole and Myers (1978). 



An economic evaluation of the use of 2,4-D was performed by calcu- 

 lating the costs of converting the waterhyaci nth control operation con- 

 ducted by the U.S. Army Corps of Engineers on the St. Johns River, 

 Florida, to a control program utilizing various combination of GAg and 



