COS 48-6 - Investigating the role of host plants in recovering golden paintbrush (Castilleja levisecta)

Tuesday, August 7, 2012: 9:50 AM
C120, Oregon Convention Center
Eric G. Delvin1, Jonathan D. Bakker2 and Peter W. Dunwiddie2, (1)The Nature Conservancy, Olympia, WA, (2)School of Environmental and Forest Sciences, University of Washington, Seattle, WA
Background/Question/Methods

The establishment of parasitic plants may be limited both by their requirements and by their host plants.  These issues are magnified when the parasitic species is rare.  Castilleja levisecta (golden paintbrush) is a federally threatened hemiparasite associated with one of the most endangered ecosystems in the United States, the prairies of western Washington.  Recovery of this species requires understanding how sensitive it is to presence and identity of host plants, and whether this sensitivity depends on plant size.  We tested the effects of host plant presence and identity (none, Festuca roemeri, or Eriophyllum lanatum) and initial C. levisecta size (grown in 49 or 393 ml pots) on the performance of 560 C. levisecta ouplanted in a tilled field.  Performance was characterized using four metrics: survival, vigor, number of flowering stems, and seed production one growing season after outplanting.

Results/Conclusions

C. levisecta survival was high (81%) when planted with no host and enhanced by the presence of a host plant.  Host plant presence also enhanced the other C. levisecta performance metrics (all p < 0.05).  Host plant identity did not affect vigor or flowering stem production, but strongly affected seed production: each surviving C. levisecta plant produced ~ 500 seeds if grown with no host compared to ~1,400 if grown with F. roemeri and ~3,000 if grown with E. lanatum.  Host plant identity also appeared to affect the flowering phenology of C. levisecta.  C. levisecta plants grown in smaller pots had higher survival but poorer vigor than those grown in large pots.  Significant host plant × pot size interactions were detected for some metrics, but were generally much smaller than main effects.  These results improve our understanding of interactions between parasitic plants and their hosts.  Further research should test whether C. levisecta can use other host plants and the mechanisms by which host plant identity might affect seed production, including differences in rooting morphology, haustorial connections, and ability to attract pollinators that also visit C. levisecta.  Recovery of C. levisecta would be enhanced by planting E. lanatum as its host.