Relationships between allelopathy and invasiveness by Eupatorium adenophorum at different sites.

With an increase in the abundance and cover of Eupatorium adenophorum [E. trapezoideum], there is a concurrent decrease in the abundance and cover of native species; however, this decrease varies at different sites and in different forests. The relative abundance and cover of E. adenophorum in deciduous broadleaved forests and along the roadside were 3–5 times greater than in evergreen broadleaved forests. The composition, relative abundance and cover of native plants in deciduous broadleaved forests and along the roadside decreased by 30, 50 and 70%, respectively, compared with the evergreen broadleaved forest. We investigated the allelopathy of aquatic extract from E. adenophorum that grew under evergreen broad leaved forests, deciduous broadleaved forests and along the roadside. We found that allelopathy of aquatic extract of E. adenophorum was different among sites, i.e. roadside> deciduous broadleaved forests> evergreen broadleaved forests. Correlation analysis showed that allelopathy of E. adenophorum was significantly correlated to the relative abundance of native species at each of the sites. These results showed that differences in invasiveness among sites are due to differences in allelopathy at the different sites. The cabbage biomass showed that the allelopathy of aquatic extract from E. adenophorum shoots was greater among sites than its roots. This suggests that the allelopathy of aquatic extract from E. adenophorum shoots contributes to different levels of invasiveness among sites more than its roots. The growth of E. adenophorum is hearty and the biomass per unit area is great along roadsides and under deciduous broadleaved forests, so the allelopathy of E. adenophorum populations per unit area along roadsides and under deciduous broadleaved forests must be much more than in evergreen broadleaved forests. Hence, the affects of E. adenophorum will be greater on these communities accelerating their decline.