Revegetation with Native Plants: a Test of Best Practices

Abstract

Restoration practitioners are tasked with recreating ecosystems using appropriate plant material that will provide ecological goods and services. However, best-practices for this type of intervention are not well developed for the southern Ontario landscape. Therefore, we evaluated approaches from four different aspects of seed-based restoration. First, we quantified the impact of seeding rate and application method on the success of grassland recreation. We also measured the impact of this restoration on the local bee community. Next, we compared a suite of native and nearly native wetland plants for their potential to prevent the establishment of invasive Phragmites australis. We measured the effect of competition on Phragmites across soil moisture and salinity gradients. Finally, we sought evidence for local specialization in a grassland forb, Monarda fistulosa, that would warrant policies to prevent the transfer of grassland seed for revegetation. In re-creating grasslands from seed, we found an interaction between seeding rate and application method. At a high rate, both methods had the same outcome, but at a low rate, a two-phase application method produced better results than a single-phase method. However, we also found that a single-phase method produced target plant cover with a higher floristic quality index after three years. In one study region, restored sites supported a greater bee abundance than un-restored sites, but bee abundance did not change after restoration in all regions. Of all the native species tested, Phragmites was supressed most by Bidens frondosa, a fast growing annual. We also found evidence that Phragmites may be less competitive at low soil moisture, and more competitive at high soil salinity. Finally, we found no evidence of local adaptation in M. fistulosa at the watershed scale; instead, we see independent effects of site and seed origin. This implies that current site conditions may not be favorable to the offspring of relic populations, and that local genotypes may not always be the best choice for restoration.