308 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1953 



affect adequate bark absorption. In addition, the person doing the 

 spraying must be able to recognize plant species, for unnecessary 

 spraying of permissible shrubs not only needlessly increases presejut 

 costs but may increase future costs by allowing new trees to i-einvade 

 those spots. Knowledge of such factors as concentration of chemical 

 used, type of oil solute, kind and combination of esters, season of treat- 

 ment, soil moisture conditions, and products of different manufacturers 

 are also important. 



This discussion of herbicides for brush control has until now pur- 

 posely been restricted to the effects on certain plant species, mainly 

 trees. One must not forget, however, that the vegetation of rights- 

 of-way is not composed only of unwanted trees and wanted grass: 

 it is an infinitely complex assemblage of plant communities, each com- 

 posed of various grasses, forbs (herbaceous plants, not grasses) , ferns, 

 shrubs, and trees. These communities vary regionally, according to 

 floristic area, climate, soil, fauna, and human history. The problem 

 is far more than one of "brush" and "grass." It is a problem for the 

 plant ecologist who understands the ramifications of these plant com- 

 munities, rather than for the contractor or maintenance engineer whose 

 objective is simply to destroy the brush in order to get grass to beautify 

 the right-of-way. The rest of this paper will deal with certain phases 

 of vegetation science that apply to the problem at hand. 



SOME PRINCIPLES OF VEGETATION DEVELOPMENT 



In forested regions of the globe, all new or disturbed areas tend to 

 progress, in a predictable or unpredictable manner and more or less 

 quickly, to some forest type. In academic lingo, this is "plant suc- 

 cession" — a very unsatisfactory term, for the word "plant" gives no 

 indication that reference is to a community, not a species or an indi- 

 vidual, and the word "succession" implies a series of discrete steps, 

 which usually do not occur. The theory of plant succession was 

 developed from early studies of quiescent sand dunes and of floating 

 bogs, types of habitat that really do show a succession of vegetative 

 stages. Actually, the theory of plant succession has done much to 

 retard the development of a rational vegetation management for 

 rights-of-way, since it presupposes that every shrub stage is relatively 

 quickly followed by a tree stage, a situation that may occur but is 

 relatively rare, for reasons discussed below. 



Rights-of-way and roadsides are nothing more than nonforested 

 lands which are tending to develop into forests. They may start 

 off as bare raw road cuts or fills, or abandoned agricultural lands, or 

 lumbered areas. In each instance the normal trend is back toward 

 forest. Forest is the one vegetation type that here is not wanted, 

 and thus management of these lands involves (a) the destruction of 

 the incipient forest and (b) the prevention, insofar as possible, of 



