702 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 123. No. 4. December 2011 
TABLE 1. BBS routes used in aerial photograph and NLCD analyses. Aerial photographs were compared between the 
early (1967/1971) and middle (1982) time periods, and between the middle (1982/1985) and late (2000/2003) time periods. 
BBS Route 
State 
Name 
Aerial photograph analysis 
NLCD analysis 
1992-2001 
Early/Middle 
Middle/Late 
84039001 
KY 
Cumberland Gap 
X 
84039011 
KY 
Monticello 
X 
84066043 
OH 
Meadowbrook 
X 
84066049 
OH 
Dresden 
X 
X 
X 
84066072 
OH 
Rav 
X 
84066073 
OH 
New Marshfield 
X 
84066074 
OH 
Omega 
X 
84066087 
OH 
Scioto Run 
X 
84066181 
OH 
South Olive 
X 
84066901 
OH 
Wavne National Forest 
X 
84066903 
OH 
Zaleski State Forest 
X 
84066904 
OH 
Tar Hollow 
X 
84090005 
WV 
Bramwell 
x 
84090014 
WV 
Spencer 
X 
84090016 
WV 
Nicut 
x 
84090022 
WV 
Canaan 
x 
84090024 
WV 
Three Forks 
x 
x 
84090026 
WV 
Cedarville 
x 
x 
84090028 
WV 
Strange Creek 
x 
84090037 
WV 
Moundsville 
x 
84090038 
WV 
Monongah 
X 
x 
X 
84090039 
WV 
McDonald 
84090041 
WV 
84090044 
WV 
x 
84090051 
WV 
Martinsburg 
84090052 
WV 
Clinton 
84090053 
WV 
RuthBelle 
84090147 
WV 
Bismark 
X 
developed and agriculture into a non-forest lane 
cover class. Finer distinctions between specific 
land cover types are probably unnecessary and 
may not be valid (Thogmartin et al. ^004 
Wickham et al. 2007) at a broad scale in assessing 
amount ot potential Cerulean Warbler habitat. 
Land cover was quantified within a 300-m 
buffer around each stop from aerial photographs 
and the NLCD. Cerulean Warblers rarely are 
detected beyond 100 m on point counts (Bosworth 
2003) and their territories average 0.3-I.0 ha in 
size (Oliamvk and Robertson 1996. Barg et al. 
2005, Perkins 2006); thus, the 300-m (28.3 ha) 
buffer is sufficiently large to accommodate 
territories ot several Cerulean Warblers. We 
calculated area of each land cover polygon, and 
summed area lor each land cover type per BBS 
stop. The NLCD was reclassified to the three land 
cover types and (he area was tabulated for each 
each c eaCh S '° P - We “ sed the ventage of 
each land cover lype for analyses because the 
raster grid cells of the NLCD do not create a 
perfect circle around each of the route stops and. 
as a result, the total area around each buffered 
stop was -28.3 ha. 
We also calculated change over time in several 
forest fragmentation metrics for the aerial photo¬ 
graph and NLCD data sets. Cerulean Warblers 
prefer large tracts of unfragmented deciduous 
forests (Hamel 2000a. Weakland and Wood 
2005), and we included maximum size of 
deciduous/mixed forest patch (ha), core area of 
deciduous/mixed forest (%), and edge density (m/ 
ha). Maximum size forest patch was the size of 
the largest deciduous/mixed forest polygon within 
the 300-m buffer around the stop. Core forest area 
was the amount of deciduous/mixed forest >60 m 
from an edge converted to percentage of core 
lorest in each buffer. Edge effects are known to 
occur within 50 m of a forest edge (Paton 1994) 
and other studies have used 50 m in analyses 
(Hazier et al. 2006, Wood et al. 2006). Edge 
