BREEDING BIRD HABITAT USE IN RIPARIAN WET MEADOWS: DOES REED CANARY GRASS AFFECT 
BIRD TERRITORY PLACEMENT?

Eileen M. Kirsch, Wayne E. Thogmartin, Tim Fox, and Brian R. Gray 
USGS, Upper Midwest Environmental Sciences Center, La Crosse, WI, 54603.

Riparian wet meadows are rare in the upper Midwestern US due to river 
development, conversion of floodplains to agriculture, and urbanization.  Most 
remaining wet meadows are dominated by invasive reed canary grass (Phalaris 
arundinacea), and wet meadows dominated by native vegetation communities are 
very rare.  The prevailing notion among resource managers is that reed canary 
grass is of little value to birds and other wildlife.  In 2001 and 2002, we 
studied breeding bird habitat use and estimated an index of productivity in 
riparian wet meadows across a range of reed canary grass dominance.  We selected 
12 riparian wet meadows in southeastern Minnesota and southwestern Wisconsin for 
study in 2001 and added another in 2002.  We used spot mapping to estimate 
locations and sizes of bird territories, and indexed productivity from observed 
behavior.  Vegetation features were estimated every 50 meters in a grid pattern 
at each study plot in late July.  Because the sampling units were territories, 
we developed an ArcView III extension to randomly sample areas outside of 
territories (e.g. null-territories).  Inverse distance weighted surfaces for 
percent reed canary grass cover, index of reed canary grass dominance, 
vegetation height density, percent cover of forbs, percent cover of shrubs, and 
average litter depth were estimated in ArcMap 8.1.  Ten a priori logistic 
regression models of associations of vegetation variables with Common 
Yellowthroat and Sedge Wren territory presence/absence were averaged to address 
model selection issues.  The averaged models revealed that for Common 
Yellowthroats amount of shrub was positively associated with territory 
placement.  For Sedge Wrens percent cover of reed canary grass was positively 
associated with territory presence.  For Common Yellowthroats, model averaged 
weights indicated that shrub cover, vegetation height density, and average 
litter depth were more important than reed canary grass cover, reed canary grass 
dominance, and percent forb cover.  However, for Sedge Wrens, model averaged 
weights indicated that none of the other variables approached the importance of 
reed canary grass percent cover in the suite of models.  Thus, Common 
Yellowthroats do not seem to be affected by reed canary grass presence and 
dominance, whereas Sedge Wrens actually show a slight affinity for reed canary 
grass in the study area.

Keywords: breeding birds, habitat use, Phalaris arundinacea, reed canary grass, 
wet meadows