Brooks and Stouffer • GRASSLAND BIRDS IN PINE SAVANNAS 
71 
TABLE 4. Candidate models used to model Henslow’s Sparrow occurrence. Plantl-Plant5 are principal components of 
the plant species guilds. Gram = graminoid, Herb = herbaceous structure. Wood = woody understory structure, Tree = 
tree density, CV = coefficient of variation, SCMU8 = Scleria muhlenbergii , OBL = obligate wetland indicator status, and 
FACW = facultative wetland indicator status. 
Model 
AICc 
AAICc 
Wi 
P GIC 
Herb density CV + SCMU8 
25.76 
0.00 
0.06 
0.39 
Herb density CV + Plant 1 + Plant5 
31.74 
5.99 
0.04 
0.24 
Herb density CV + Plant 1 + Herb density CV*Plantl 
31.85 
6.09 
0.04 
0.24 
Herb density CV + Plant 1 
32.44 
6.68 
0.04 
0.22 
Herb density CV + Plant5 
32.60 
6.84 
0.04 
0.22 
Herb density CV 
33.54 
7.78 
0.04 
0.19 
Herb PC + Plant4 + Herb PC*Plant4 
33.81 
8.05 
0.04 
0.18 
Herb cover 
34.49 
8.73 
0.04 
0.16 
Herb density CV + Plant 1 4- Plant4 
35.17 
9.41 
0.04 
0.14 
Herb density CV + Gram OBL 
35.61 
9.85 
0.04 
0.13 
Herb density CV + Gram FACW 
35.79 
10.03 
0.04 
0.12 
Herb density CV + Plant4 
Herb PC + Plant 1 + Plant4 + Herb PC*Plantl + 
36.00 
10.24 
0.04 
0.11 
Herb PC*Plant4 
36.87 
11.11 
0.03 
0.08 
Herb PC 
37.48 
11.72 
0.03 
0.06 
Wood PC + Herb PC 
37.76 
12.00 
0.03 
0.05 
Herb PC + Plant 1 
37.89 
12.13 
0.03 
0.05 
Herb density 
38.87 
13.12 
0.03 
0.01 
Null 
39.26 
13.50 
0.03 
0.00 
Wood PC 
39.73 
13.97 
0.03 
-0.02 
Herb PC + Plant4 
39.89 
14.14 
0.03 
-0.02 
Herb PC + Tree PC 
39.92 
14.16 
0.03 
-0.02 
Herb PC + Plant 1 + Herb PC*Plantl 
40.20 
14.44 
0.03 
-0.04 
Herb PC + Plant 1 + Plant4 
40.43 
14.67 
0.03 
-0.04 
Wood PC + Herb PC + Tree PC 
40.46 
14.70 
0.03 
-0.05 
Wood PC + Plant 1 + Plant2 
40.61 
14.85 
0.03 
-0.05 
Tree PC 
41.51 
15.75 
0.03 
-0.09 
Wood PC + Plant2 
41.89 
16.14 
0.03 
-0.10 
Wood PC + Tree PC 
42.19 
16.43 
0.03 
-0.11 
Global 
46.52 
20.76 
0.02 
-0.31 
birds maintain territories throughout the year (Cox 
and Jones 2009). Damage from Hurricane Katrina 
in 2005 disproportionately affected mature upland 
stands in DSNF, possibly improving upland stand 
habitat quality for Bachman’s Sparrows. The 
hurricane thinned canopies, simultaneously pro¬ 
ducing downed tree crowns and upturned root 
balls. Bachman’s Sparrows sing from downed 
crowns and may use root balls for escape; absence 
of these features reduces breeding season occu¬ 
pancy by Bachman’s Sparrows (Brooks and 
Stouffer 2010) and could affect winter habitat use. 
Henslow’s Sparrows were detected only in bog 
and RCW stands. These sparrows, in the Gulf 
Coastal Plain, seem to prefer some grassland 
habitats over others. Plentovich et al. (1999), 
working in pitcher plant bogs and upland pine 
stands in Alabama, only found Henslow’s Spar¬ 
rows in bogs and transition zones between bog 
and upland pine habitats. Tucker and Robinson 
(2003) also found Henslow’s Sparrows using 
small bogs in upland pine habitat in winter along 
the Gulf Coast. Other studies, however, have 
found high densities of Henslow’s Sparrows in 
upland longleaf pine habitats (Carrie et al. 2002, 
Johnson 2006, Palasz et al. 2010), but many of 
these sites were well-managed with fire and not 
adjacent to bogs. The primary reason Henslow’s 
SpaiTOws avoided upland habitats in DSNF is 
probably the lack of a dense, continuous herba¬ 
ceous layer, reflected by inclusion of the herba¬ 
ceous density CV in the best habitat model, even 
in stands regularly managed with fire. The 
majority of the upland stands we sampled had 
experienced only one growing season since fire, a 
condition that should favor Henslow’s Sparrows 
(Carrie et al. 2002, Bechtoldt and Stouffer 2005). 
These stands had an herbaceous layer, but it was 
