706 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol 123, No. 4. December 2011 
completely avoided by Cerulean Warblers (Wood 
et al. 2005). 
Cerulean Warbler abundance decreased during 
this same time period and their distribution 
became more restricted, changing from 15 to 
3% of BBS stops. Regional BBS analyses (Sauer 
et al. 2008) found Cerulean Warblers declined 
from 1967 to 1982 in two of the three physio¬ 
graphic regions that comprised our study area 
(—4.8% in Ohio Hills, —5.2% in Cumberland 
Plateau). The lack of a positive response to 
increased young forest cover in breeding areas 
suggests events or conditions during migration or 
in wintering areas may contribute to population 
declines. 
Recent Habitat and Cerulean Warbler Changes: 
1982/1985 vs. 2000/2003.— The agriculture land 
cover lost from the middle to late periods was 
replaced by developed land cover for all stops and 
the set of stops al which Cerulean Warblers were 
detected. Agricultural fields appear to have been 
converted to development instead of being aban¬ 
doned during this time period. 
Core forest increased between the middle and 
late time periods with the all-slops data. Agricul¬ 
tural fields abandoned in the early lime period and 
which were early stage deciduous/mixed forests by 
the middle time period, may have developed a 
more contiguous canopy cover by the late time 
period perhaps contributing to increases in core 
forest. Core forest is considered an important 
habitat characteristic for Cerulean Warblers (Oliar- 
nyk 1996, Hamel 2000a): however, detections 
again decreased, but less than in the early versus 
middle time period. Regional BBS analyses (Sauer 
et al. 2008) similarly found a less steep decline in 
1982-2003 in the Ohio Hills (-2.2) and the 
Cumberland Plateau (-2.1) physiographic regions. 
Cerulean Warblers were detected at 27% of all 
stops in the middle and 30% in the late time 
period; however, more were detected per stop in 
the middle than the late time period (Table 3). 
The decrease in abundance but not distribution, 
die large decrease in abundance al stops with 
Cerulean Warbler presence, but lack of change in 
broad-scale fragmentation metrics considered 
important to Cerulean Warblers suggests there 
may be other small-scale factors (e.g., canopv 
gaps/heterogeneity) influencing population trends 
Of Ceralean Warblers in breeding areas along Ihe 
oHS routes examined. 
Short-term Habitat and Cerulean Warbler 
Changes: 1992 vs. 200/.-Several NLCD land 
cover variables believed to be important to 
Cerulean Warblers declined from 1992 to 2001 
at BBS stops. Deciduous/mixed forest declined 
and was replaced by non-forest land cover, 
primarily development, in both the all stops and 
presence-only analyses (Table 4). 
The amount of deciduous/mixed forest and core 
forest was greater at stops where Cerulean 
Warblers occurred than at all stops. Cerulean 
Warbler abundance increased at presence-only 
stops and did not change at all stops despite the 
decline in dcciduous/mixed forest. This, in 
concert with the increase in edge density, suggests 
microhabitat features within large, continuous 
dcciduous/mixed forests also are important and 
provides support that, at a local scale. Cerulean 
Warblers are able to tolerate some edge habitat 
which may increase stnictural diversity in the 
canopy (Weakland and Wood 2005). The smaller 
amount of non-forest land cover at presence-only 
slops supports Cerulean Warbler avoidance of 
large-scale habitat disturbance (Wood et al. 2006). 
Cerulean Warbler abundance may have increased 
at presence-only stops despite the decrease in 
dcciduous/mixed forest and forest-forest edge 
density because their density increased in the 
remaining suitable habitat. A key habitat factor for 
Cerulean Warblers includes interior, unfragmented 
forests (Oliamyk 1996, Hamel 2000a); however, 
there was little interior forest (9% for all stops and 
13% for presence-only stops) in 2001 within 300 m 
ol BBS stops. The maximum size forest patches 
within 300 m of a BBS stop increased for presence- 
only stops and all stops in our study area (Table 4). 
and were sufficiently large to contain several 
Cerulean Warbler territories (range = 0.21 hafRoth 
2004] to 1.04 ha [Oliamyk and Robertson 1996]). 
Cerulean Warbler detections did not change 
between 1992 and 2001 across all 1,375 BBS 
stops examined (Table 4); they were detected at 
14% of all stops in 1992 and 17% in 2001. Sauer 
et al. (2008) reported declines of 4.1-6.8% during 
this time period for the three physiographic 
regions intersecting our study area. This disparity 
in trends may relate to which routes were 
included. We used data from 28 BBS routes 
because ot luck of stop-level GPS coordinates and 
routes that were run inconsistently, whereas Sauer 
et al's. (2008) trend analysis was based on 75 
routes. Additionally, we focused on the core 
breeding range of Cerulean Warblers (Fig. !). 
whereas Sauer et al. (2008) included more BBS 
routes near the periphery of the range. 
