Applied Ecologist Ecostudies Institute Olympia, Washington, United States
Abstract: In Washington State, coastal meadows comprise one of the last habitats for the Federally Endangered Taylor's checkerspot butterfly (Euphydryas editha taylori). A major recovery goal is to amend sites with the butterfly's preferred host plant species. One of the butterfly's most important host and larval food species is the hemi-parasitic plant harsh paintbrush (Castilleja hispida), itself locally rare in these same meadows. The butterfly's dependence on a parasitic plant poses a unique conundrum since the persistence of the butterfly could depend on third-order effects related to the paintbrush's host species. To address this problem, we set up several field experiments using several thousand out-planted paintbrush individuals to disentangle the myriad factors affecting harsh paintbrush establishment and persistence. We also wanted to determine which amendment methods were the most cost-effective. We used hierarchical modeling to evaluate the importance of the reintroduction method (out-planting nursery-grown plugs or direct seeding), the spatial scale of site selection, and microsite factors (microtopography, aspect, soil temperature, vegetation community composition) on four demographic parameters for harsh paintbrush (establishment rate, total seed production, spatial expansion beyond initial planting footprint, and the number of host plant patches). There are several key results. First, the reintroduction method drives the importance of microsite factors for paintbrush establishment. For nursery-grown plugs, nursery factors – such as tray quality and nursery lot, which relate to initial plant quality, are the primary drivers of out-planting success or failure. After accounting for initial plant quality, the presence and cover of Oregon sunshine (Eriophyllum lanatum), the primary host species for harsh paintbrush, was the most important factor for success. For directly sown seed, successful populations were in deeper soil microsites with plant communities with high phylogenetic diversity, with at least 7 to 9 perennial forbs that also included Oregon sunshine. Second, direct seeding resulted in more host plant patches (n =25) than out-planting (n = 13). Seeded microsites also had 54% more individuals (n = 17) than out-planted microsites (n = 11), and direct seeding resulted in more total individuals (n = 1,028) than out-planting (n = 535). Third, the cost of producing suitable checkerspot host plant patches is significantly cheaper with direct-seeding (~ $4/patch) than out-planting plugs (~$1,233/patch). This effort demonstrated a successful framework for applied research that addressed basic questions in community ecology and applied questions in conservation, management, and project planning that practitioners can use for other butterfly species.
