60 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol 123, No. 1, March 2011 
habitats (e.g., Swanson et al. 1999, Warren and 
Anderson 2005, Zuckerberg and Vickery 2006), 
but few studies have evaluated the effects of 
mowing on wintering bird communities. We 
studied the response of wintering birds to fall 
mowing of herbaceous buffers. We hypothesized 
that wintering bird abundances, species richness, 
and total avian conservation value would be less 
in mowed than in unmowed buffers. We focus on 
the response of grassland and scrub-shrub birds 
because they are experiencing substantial popula¬ 
tion declines (Sauer et al. 2008) and are of high 
conservation concern (Hunter et al. 2001, Askins 
et al. 2007). 
METHODS 
Study Area— The Eastern Shore of Maryland 
(east of Chesapeake Bay) has —46% of land- 
cover in farms (USDA 2009a) and 77% of the 
CREP buffers in the state (USDA 2007). Filter 
strips (USDA Practice CP21) are the most 
common type of herbaceous buffers in Maryland 
(USDA 2010). We conducted an experiment in 13 
filter strips (hereafter, buffers) among two coun¬ 
ties (Queen Anne’s and Talbot) on Maryland’s 
Eastern Shore. 
All buffers selected were installed between 
1997 and 2004, and were >3 years of age at the 
time of the study. Each buffer was between a 
rowcrop field and a forested wetland, which is a 
common location of buffers in Maryland. The 
adjacent rowcrops had been planted to either com 
or soybeans in the previous growing season, and 
most were planted to winter wheat after fall 
harvest. 
Nine buffers were planted with cool-season 
grasses and four were planted with warm-season 
grasses. Common warm-season grasses were big 
bluestem ( Andropogon gerardii ), little bluestem 
(Schizachyrium scoparium ), indiangrass ( Sor - 
g hast rum nutans ), and broomsedge bluestem (A. 
virginicus). The most common cool-season grass 
in buffers was orchardgrass (Dactylis glomerata), 
but other cool-season grasses including red 
(Festuca rubra ) and sheep (F. ovina) fescue were 
also planted. 
We established two treatments in each buffer: 
(1) a section (experimental treatment) mowed in 
late summer or fall (Aug-Dec) to 10-15 cm in 
height, and (2) an unmowed section. Mowed and 
unmowed treatments were randomly located 
along the length of the buffer and spanned the 
entire width of the buffer. We established one 
study site in each treatment. Each study site 
spanned the width of the buffer, was >50 m from 
the ends of the buffer and from the interface with 
the other treatment, and > 100 m from the other 
study site in the same buffer. Mowed and 
unmowed study sites among all buffers were 
similar (x ± SD) in length (mowed: 176.0 ± 
50.0 m; unmowed: 176.6 ± 50.3 m). 
We defined the width of each buffer as the 
distance from the crop edge to the wooded edge 
and calculated width by averaging measurements 
taken every 50 m over the length of the buffer. 
Buffers ranged in width from 11 to 91 m, and 
average buffer width was 40.9 ± 35.7 m. We 
measured the length of each study site in a 
Geographic Information System (GIS) and calcu¬ 
lated the area of each site by multiplying site 
width by site length. 
Vegetation Surveys. —We conducted vegetation 
surveys once at each study site in winter 2007. We 
established one transect line through the center of 
the site in buffers <45 m wide, and two transect 
lines spaced evenly across the width of the site in 
buffers >45 m wide. We measured vegetation 
structure characteristics within 1-m 2 sampling 
plots at random distances perpendicular to five 
points spaced evenly apart along each transect 
line. Thus, we surveyed vegetation at five plots in 
buffers <45 m wide and 10 plots in buffers >45 m 
wide. We visually estimated the percent cover 
(non-overlapping) of grasses, forbs, trees, bare 
ground, and litter in each plot. We also measured 
vertical vegetation density (Robel et al. 1970), 
litter depth, and maximum vegetation height. 
Bird Surveys. —We conducted three bird sur¬ 
veys at each study site between 19 January and 10 
March 2007. All surveys were between 1 hr after 
sunrise and 1 hr before sunset. We did not conduct 
surveys in precipitation, fog, or wind >16 km/hr. 
Bird surveys in the two study sites in the same 
buffer were subsequent to one another and in 
random order. Individual birds observed in one 
study site were not observed to move to any other 
study sites, and study sites were considered 
independent. 
We surveyed birds across the entire area of 
each study site. All surveys were conducted 
simultaneously by P. J. Blank and J. R. Parks. 
We walked parallel to the wooded edge of the 
buffer <20 m apart. The distance between us 
varied depending on width of the buffer. Nine 
buffers were <40 m wide and required only one 
pass. Four buffers were >80 m wide and required 
