An experimental study of wetland invasibility by purple loosestrife (Lythrum salicaria)

Abstract

We examined the effects of vegetation and herbivores on the seedling establishment, survival, and inflorescence mass of purple loosestrife (Lythrum salicaria L.), a Eurasian plant that invades North American wetlands. The study was conducted in a stand of reed canary grass (Phalaris arundinacea L.). All grass was removed experimentally from 1.2 × 1.2 m plots to simulate a disturbance. The establishment of sown seeds was compared in disturbed and undisturbed plots. Fifty-three percent of seeds became established in disturbed plots, while no seedlings became established in undisturbed plots, suggesting that disturbance is required for wetland invasion by L. salicaria. In plots without grass, the effect of herbivores on Lythrum's survival and inflorescence mass (i.e., sexual reproduction) was assessed by excluding herbivores from some plots but not from others. After two growing seasons, Lythrum's survival did not differ significantly between plots with herbivores excluded (100% plant survival) and plots with herbivores not excluded (60-80% plant survival). However, only plants protected from herbivores produced inflorescences. Mammals such as deer and rodents ate the shoot tips of unprotected plants, which prevented terminal inflorescences from being produced. These results only partly support the claim that North American herbivores are ineffective predators of L. salicaria. One mechanism underlying wetland invasion by L. salicaria is likely a combination of disturbance permitting seedling establishment and insufficient herbivory in disturbed areas to prevent newly established plants from surviving.