362 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 123. No. 2. June 2011 
TABLE 1. Mean ± SE perch times, perch heights, and attack distances for hunting attempts where Loggerhead Shrikes 
in Kentucky attacked or gave up and when shrikes made successful and unsuccessful attacks. Differences are not significant 
Attack 
Give up 
Successful 
Unsuccessful 
Overall 
Perch time (min) 
Perch height (m) 
Attack distance (m) 
2.1 ± 0.1 
2.8 ± 0.1 
4.3 ± 0.2 
4.7 ± 0.7 
2.9 ± 0.4 
2.6 ± 0.2 
3.1 ± 0.2 
5.1 ± 0.4 
2.7 ± 0.4 
3.1 ± 0.2 
3.4 ± 0.3 
2.2 ± 0.1 
2.8 ± 0.1 
43 ±02 
superimposing a grid created using Arcvicw 3.2 
(ESRI. Redlands, CA, USA.). A random-numbers 
table was used to select individual grids. We 
requested permission from landowners to survey 
areas for shrikes if characteristics of a grid area 
were similar to that of a typical shrike territory 
(undeveloped and < 10% of the area wooded). We 
searched areas for '-5 hrs over a period of at least 
2 days to ascertain if shrikes were present, if 
permission was given. We also looked for signs 
that shrikes were present (e,g„ impaled prey). 
Characteristics of sites were quantified using the 
same methods described for occupied shrike 
territories if no shrikes or signs of shrikes were 
observed. 
Data Analysis. —Repeated measures analysis of 
variance (ANOVA) was used to examine the 
possible effects of perch time, perch height, and 
attack distance on the outcome of attacks and, for 
perch time and perch height, on whether a shrike 
attacked or gave up. Multivariate analysis of 
variance (MANOVA) was used to compare 
habitat characteristics within Loggerhead Shrike 
territories (hunting vs. random perch analysis), 
among all occupied territories, and between 
occupied and unoccupied areas. Stepwise dis¬ 
criminant analysis was used to interpret the 
classification results for each habitat characteris¬ 
tic. All analyses were conducted using the 
Statistical Analysis System (SAS Institute 2002). 
RESULTS 
Hunting Behavior—Vie observed 19 Logger- 
head Shrikes make 1,372 hunting attempts that 
resulted in 1,322 attacks (96.4% of attempts) from 
January to March 2005 and November 2005 to 
March 2006. Trees (/ T = 481,36.9%) and fences 
0? = 466. 34.9%) were the most commonly used 
perches. Most attacks where outcome was known 
(n = 618) were successful (n - 407, 65.9%) 
Arthropods (n = 266, 65.3%) and other inverte¬ 
brates (earthworms [suborder Lumbricina] and 
unknown invertebrates; n = 95, 23.3%) were the 
most common prey. Only eight vertebrates were 
taken as prey (4 rodents, 3 salamanders, and 1 
snake). 
We found no differences in mean perch time 
(Tr.s = 0.4, P = 0.58), attack distance { t\s ~ 
0.4, P = 0.54), or perch height (F|. 5 = 3.7, P = 
0.11; Table 1) between successful and unsuccesv 
lul attacks. Mean perch time (F, 4 = 0.2, P = 
0.66) and perch height (f 1<4 = 5.3. P = 0 . 08 ) did 
not differ for hunting attempts when shrikes 
attacked or gave up (Table 1). 
Vegetation Analysis. —Characteristics of con¬ 
tinuous perches used by Loggerhead Shrikes and 
those of randomly selected, apparently unused 
continuous perches differed (Wilks’ X = 0.48. 
7 * 9,32 — 2.5, P = 0.021). Stepwise discriminant 
analysis revealed that percent grass cover, percent 
litter cover, foliage cover <0.5 m, and vegetation 
height permitted the best discrimination between 
used and unused continuous perches. Continuous 
perches in areas used by shrikes had less litter and 
grass cover, more foliage cover <0.5 m, and 
shorter vegetation (Table 2). Classification anal 
ysis using these variables revealed 80% of used 
and 66.7% of unused sites were correctly 
classified. 
TABLE 2. Mean ± SE characteristics of vegetation 
associated with continuous perches used (n = 21) by 
Loggerhead Shrikes (n = 7) and randomly selected, 
apparently unused continuous perches (n = 21) > n 
Kentucky. Differences are significant (P = 0.021). 
Variable 
Used 
Random 
Litter depth (cm) 
0.76 ± 0.20 
0.69 ± 0.33 
Vegetation height (cm) 
7.1 ± 0.8 
9.7 ± 2.6 
Foliage cover <0.5 m 
(if of stems) 
11.0 ± 1.1 
10.0 ± 0.8 
Ground cover (%) 
Grass 
58.0 ± 7.1 
72.6 ± 6 
Herb 
9.0 ± 4.5 
1.7 ± 0.9 
Litter 
5.8 ± 2.9 
8.1 1 4.0 
Bare ground 
24.8 ± 4.8 
143 ± 3.2 ^ 
