ABSTRACT
Many chemical and non-chemical strategies have been applied to control weeds in agricultural and industrial areas. Knowledge regarding the effects of these methods on roadside vegetation is still poor. A 2-year field experiment was performed along a road located near Livorno (Tuscany, central Italy). Eight plots/strips were identified, of which four were subjected to periodical mechanical mowing and the remaining four were treated with a chemical herbicide based on glyphosate (the producer's recommended rates were used for the selective control of broad-leaved weeds). Our results clearly showed that roadside soil and vegetation are a significant reservoir of anthropogenic activities which have a strong negative effect on several phytosociological, pedochemical and biological parameters. Compared with conventional mechanical mowing, chemical treatment induced (i) a significant increase in organic matter in the upper plot layers (+18%), and (ii) a marked reduction in weed height throughout the entire period of the experiment. Irrespectively of the kind of treatment, no significance differences were detected in terms of (i) biological quality of soil (the abundance and diversity of arthropod communities did not change), and (ii) plant elemental content (bulk concentrations of analysed trace elements had a good fit within ranges of occurrence in the "reference plant"). The glyphosate partially controlled broad-leaved weeds and this moderate efficacy is dependent upon the season/time of application. In conclusion, the rational and sustainable use of chemical herbicides may be a useful tool for the management of roadside vegetation.