704 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 123, No. 4. December 20/1 
TABLE 2. Land cover and fragmentation metrics based on aerial photographs for the early (1967/1971) and middle 
(1982) time periods and Cerulean Warbler detections during 5 years surrounding each aerial photograph year for 68 stops 
on two BBS routes in West Virginia and Ohio (df = 66). 
Category 
Early 
X ± SE 
Middle 
X * SE 
F 
p 
Landcover % 
Deciduous/mixed forest 
36.3 ± 3.1 
49.3 
± 3.40 
3.82 
0.06 
Coniferous forest 
1.5 ± 0.6 
2.3 
± 1.15 
0 
0.98 
Developed 
16.8 ± 2.4 
14.7 
± 2.11 
0.32 
0.57 
Agriculture 
45.4 ± 3.1 
33.9 
± 3.16 
5.30 
0.02 
Fragmentation metrics 
Max forest patch (ha) 
6.8 ± 0.6 
8.8 
± 0.6 
1.10 
0.30 
Core forest (%) 
7.8 ± 1.2 
10.6 
± 1.8 
0.68 
0.41 
Edge density (m/ha) 
138.6 ± 5.3 
136.8 
± 4.6 
0 
0.99 
Cerulean Warblers detections/stop 
Average 
0.05 ± 0.02 
0.01 
± 0.007 
5.25 
0.03 
Maximum 
0.16 ± 0.05 
0.03 
± 0.020 
5.48 
0.02 
differed between the early and middle time 
periods. Average and maximum Cerulean Warbler 
detections per stop decreased (Table 2) and 
Cerulean Warblers were detected at fewer stops 
(15% in the early period vs. 3% in the middle 
period). 
Aerial Photograph Analysis for 1982/1985 
vs. 2000/2003. —A mount of developed land cover 
increased and agriculture decreased around 240 
stops on six BBS routes between the middle and 
late time period (Table 3). Core forest increased, 
but other forest metrics (forest land covers, max 
size forest patch, and edge density) did not change 
(Table 3). Average number of Cerulean Warbler 
detections per stop decreased, but the maximum 
number counted did not change between these two 
time periods and percent stops with detections 
was similar (27% stops in the middle period and 
30% in the late period). Cerulean Warblers were 
detected at 76 of the 240 stops during either the 
middle (65 Cerulean Warblers detected) or late 
time period (71 Cerulean Warblers detected); 44 
stops had detections in both time periods (18% of 
stops). 
Developed land cover increased and agriculture 
and cover decreased (Table 3) from the middle to 
late period at 76 stops where Cerulean Warblers 
were detected, although both land covers were 
less abundant than for all stops. We found no 
c ange in deciduous/mixed or coniferous forest 
and none of the fragmentation metrics differed 
between the middle and late time periods 
Average number of Cerulean Warbler detections 
per stop decreased more markedly than in the all 
stops analysis and the maximum number detected 
approached a significant decline (P = 0.10). 
NLCD Analysis for 1992 vs. 2001.—The j 
deciduous/mixed forest and coniferous forest land 
cover types decreased, whereas the non-forest 
type increased at 1,375 stops on 28 routes 
(Table 4). Forest patch size increased, whereas 
core forest, forest-forest edge density, and forest- 
nonforest edge density decreased. Cerulean War¬ 
blers were detected at 14% of stops in 1990-1994 
and 17% in 1999-2003. Average and maximum 
Cerulean Warbler detections per stop were not 
different, although counts increased slightly and 
approached significance (/> = 0.11). 
Cerulean Warblers were detected at 344 of 
1,375 stops in both time periods. Coniferous and 
deciduous/mixed forests decreased at these stops 
(Table 4), while non-forest land cover increased 
from 15 to 25%. Maximum size of forest patch 
and forest-nonforest edge density increased. 
1 orest-forest edge density decreased, and amount 
of core forest did not change. The mean and 
maximum number of Cerulean Warblers detected 
increased. 
DISCUSSION 
Historic Habitat and Cerulean Warbler Changes: 
1967/1971 vs. 1982. —Several land covers changed 
at 68 stops along the two BBS routes for which we 
analyzed aerial photographs from the early to 
middle time periods. The increase in deciduous/ 
mixed forest and decline in agriculture land cover 
