Long-Term Effects of Initial Site Treatment on Fescue in a Novel Prairie Ecosystem (Washington)

In prairie restoration, site preparation is a key step in altering environmental conditions to favor the establishment of desirable plants over invasive species. Initial site preparation is especially important in restoration scenarios where resources may not exist for follow-up treatments and consistent maintenance. In the Pacific Northwest, for example, prairie restoration commonly involves practices such as prescribed fire, herbicide application, and manual vegetation removal before native species are planted or seeded (Sinclair et al. 2006). Despite the common use of these methods (e.g., Barnes 2004), how effectively single treatment events perpetuate desired vegetation over long timeframes has not been adequately evaluated.

In 2008, we revisited prairie plots restored in 1994 to examine the long-term impacts of initial site treatments on the survival and expansion of planted fescues (Festuca spp.). We focused our study on the perennial bunchgrass, Idaho fescue (Festuca idahoensis), because it is a fire-tolerant keystone prairie species and stress tolerator that can survive in nutrient poor soils with low moisture. Idaho fescue is common throughout Puget Sound lowland prairies and has also responded favorably to reintroduction at prairie restorations near our research site and within the Union Bay Natural Area (UBNA) in Seattle, Washington.

As part of the University of Washington Botanic Gardens, UBNA is presently managed as a public green space of grasslands, forests and wetlands that provide valuable research and demonstration sites. It is a former landfill that was retired in 1966 and in 1971 was graded, capped with a dense aggregate of nutrient- and organic matter-poor clay and gravel, and seeded with a mixture of predominantly rhizomatous pasture grasses (Kentucky bluegrass [Poa pratensis], Canada bluegrass [Poa compressa], tall fescue [Schedonorus arundinaceus], and sweet vernalgrass [Anthoxanthum odoratum], among others) to reduce erosion. Some forbs have been planted or colonized since retirement (Ewing 2002b), but they currently make up a small component (<3% in 2008) of vegetation. Soil properties are similar to native prairies: nitrate levels are low (2–11 ppm in 1994; Ewing 2002a, 2002b), while soil moisture declines throughout the growing season as spring rains dissipate (Ewing 2002a). The location, soil properties, and non-native-dominated grass community make UBNA an ideal site to examine the effectiveness of techniques to restore bunchgrass communities similar to local Puget Sound prairies. Due to limited institutional resources, restored ecosystems at UBNA receive minimal maintenance, often limited to Himalayan blackberry (Rubus discolor) removal and noxious weed control with herbicide; within our plots, only blackberry removal occurred. A full site description is provided by Ewing (2002a, 2002b).

Table 1.

Summary of initial site treatments applied in the Union Bay Natural Area in 1994 (Ewing 2002a) and mean (±SE) Idaho fescue (Festuca idahoensis) survival and weed biomass; NA = not applicable. For all treatments, n = 10, except the control (n = 6).

Five site preparation techniques that altered soil resources were implemented in a pasture grass dominated portion of UBNA in March and April of 1994 (Table 1): herbicide and till (HT), burn (B), remove vegetation (V), remove vegetation and add compost (VC), and remove vegetation and add fertilizer (VF). It was thought that HT, B, and V would reduce soil resources available to aggressive non-native grasses and create growing space for Idaho fescue transplants, while the VC and VF treatments would increase the initial physiological performance of transplanted Idaho fescue by improving soil resources. Treatments were randomly allocated to 50 experimental plots (2 m × 2 m), ten per treatment. In May of 1994, 16 Idaho fescue plugs were planted in each plot, for a total of 800 plants. Six untreated, unplanted control (C) plots were interspersed among the experimental plots and were not analyzed by Ewing (2002a) with treated plots. Plots in which vegetation was removed (V, VC, VF, HT) all experienced similar, minimal levels of soil disturbance; soil was relatively undisturbed in B and C plots. Although Idaho fescue was the focus of the original restoration, some of the plants appear to be red fescue...