Physiological integration in an introduced, invasive plant increases its spread into experimental communities and modifies their structure.

What determines the invasiveness of introduced plants is still poorly known. Many of the most invasive plant species are clonal, and physiological integration between connected individuals (ramets) of clonal plants may contribute to their ability to spread into communities and reduce performance of existing species. This contribution of integration to the invasiveness of clonal plants may be greater in denser communities. A greenhouse study was conducted to test these two hypotheses. High- and low-density communities were created by sowing seeds of eight grassland species. Each community was planted with three ramets of the stoloniferous, introduced plant Alternanthera philoxeroides that were disconnected from or left connected to ramets growing on bare soil. Connection increased the spread of Alternanthera within a community, but did not reduce community biomass. Alternanthera grew less in high-density communities, but connection did not improve its growth more than in low-density communities. Low-density communities had higher evenness when Alternanthera was connected than when it was disconnected because shoot mass was lower in the more abundant species in the community and higher in the less abundant ones. These results partly supported the first hypothesis, but not the second. The effect of integration on community structure could be due to higher resource import by the ramets of Alternanthera closer to the dominant species. Integration therefore can increase the initial spread of new clonal plant species into communities and modify the effects of this spread on community structure.