Long-Term Assessment of Alternative Strategies for the Restoration of Floodplain Forest in the Presence of an Invasive Grass, Phalaris arundinacea

In 1998, a 6.1-ha wetland restoration project along the Rock River, IL, USA was designed to test five afforestation methods on former agricultural land, including planting bareroot trees, balled-and-burlapped trees, seedlings, acorns, or allowing natural regeneration. Fifteen years later, we assessed vegetation at the site to determine the effectiveness of alternative strategies and compare treatments to an adjacent floodplain forest. We also compared the cost of treatments to determine whether long-term restoration outcomes justified initial costs. After 15 years, lower cost treatments (acorn plantings and passive restoration) were dominated by dense reed canarygrass (Phalaris arundinacea) with sparse trees, whereas higher cost treatments (bareroot and balled-and-burlapped tree plantings) had developed closed tree canopies, and tended to have greater plant species richness, tree basal area, and density of stems >7.5-cm diameter. For every additional $10,000 per ha spent on restoration, predicted richness increased by 1 species per 250-m² plot, predicted P. arundinacea biomass decreased by 61 g m⁻², and predicted tree basal area increased by 3 m² ha⁻¹. Although some studies have indicated that passive regeneration alone is effective for afforestation, restoration of floodplains in the presence of P. arundinacea will require a more intensive approach.