Habitat characteristics of Spiraea virginiana Britton, an imperiled riparian shrub

Background/Question/Methods

Spiraea virginiana Britton is a rare, clonal riparian shrub, considered endangered in the state of North Carolina and federally listed as threatened.  Anthropogenic alteration of habitat has been cited as one of the primary causes of the species’ decline.  As such, identification of habitat characteristics is a crucial step in the protection of the species.  In the summer of 2011, we expanded previous research done at the Cheoah River corridor in Graham County, NC to include all known occurrences of S. virginiana in western NC.  The study included populations on eight rivers.  A plot was established at each occurrence (67 total), along with a corresponding control plot.  Abiotic characteristics measured were canopy openness, substrate type, slope, and aspect.   Biotic characteristics examined were herbaceous cover, presence of vines, shrub density, and tree influence index.  Wilcoxon Two-Sample Tests were used to compare characters between S. virginiana and control plots.  The Cheoah River habitat, which had different fluvial geomorphology than the other rivers, was compared to all other rivers combined using a two-sided Wilcoxon Two-Sample Test.  Habitat characteristics were then analyzed separately for the Cheoah population and all other populations pooled using Wilcoxon Two-Sample Tests.

Results/Conclusions

When all data were analyzed together, we found no abiotic factors were correlated with the shrub’s presence.  Several biotic factors including, total grass cover (N = 134, p = 0.0139), total forb cover (N= 134, p = 0.0048), Parthenocissus quinquefolia cover (N=134, p = 0.0232), and Ipomoea spp. cover (N=134, p = 0.0280) were all significantly higher in plots without S. virginiana.  Most characteristics were significantly different between the Cheoah and all other rivers pooled.  When these groups were examined separately, canopy openness (N = 62, p = 0.0134) and the extent of boulders in the substrate (N = 64, p = 0.0306) were significantly higher in S. virginiana plots compared to the controls at the Cheoah River.  Biotic factors were significantly different between S. virginiana  and control plots at all other rivers pooled, but not at the Cheoah.  These differences are likely related to differing hydrology, which at the Cheoah might be more representative of S. virginiana’s natural habitat due to high-flow dam releases.  Failure to find significant abiotic habitat characteristics at additional rivers may also reflect flaws in our sampling design at these rivers.  Future studies should examine the hydrology and geomorphology of all rivers and use consistent methods for establishing reference plots at each river.