Restoring Abandoned Agricultural Lands in Puget Lowland Prairies: A New Approach

Puget Lowland Prairies are one of the most endangered ecosystems in the United States. Restoration has been occurring for more than thirty years and has focused on extant sites with some native species present. Generally, actions first began with invasive species control, followed by reintroduction of fire, and then seeding and planting. Significant improvements have been made, however existing habitat continues to be converted and additional strategies to restore the ecosystem need to be explored. Developing techniques and methods for restoring abandoned agricultural lands, and other lands with no native species present, to native prairie is a critical information need, as it will likely be an important future strategy. This dissertation explored techniques to restore native prairie habitat on a suite of abandoned agricultural lands over a large spatially and temporally replicated experiment, the Prairie Habitat Restoration Project (PHRP). I first explored techniques and factors affecting the seedling establishment and survival of a rare species, a hemiparasitic plant endemic to the Puget Lowland Prairies, Castilleja levisecta. I seeded C. levisecta in combination with different site preparation and seeding treatments across 4 sites over 2 years. Next, I tested the effect of host plant identity and container size on survival and reproduction of C. levisecta in a replicated outplanting study with two potential host species. Finally, I tested a novel approach we call “Staged-Scale Restoration” (SSR) that rigorously explored multiple habitat treatments within an adaptive management framework. I implemented SSR by identifying several promising restoration strategies and testing them in small, replicated experimental plots. Based on the results of these small-scale tests, the most successful treatments were applied to increasingly larger scaled-up areas. My first study demonstrated that direct seeding is an effective strategy for establishing large populations of C. levisecta. My second study found that the presence of a host species, and the identity of that host, significantly affected the survival and performance of C. levisecta. My final study demonstrated that SSR enables adaptive management to be implemented in restoration projects while minimizing risks, improving scientific rigor, and providing a cost effective approach.