SHORT COMMUNICATIONS 
627 
METHODS 
Study Area. —Free-ranging flocks of Carolina 
Chickadees and Tufted Titmice were studied at 
eight locations in Madison County. Kentucky 
(37 4 i' 58" N. 84° 16' 20" W) from 5 January to 
27 February 2008. Study sites included private- 
residences in = 7) and a public campground (// - 
li and were separated by a minimum distance of 
1.5 km. A feeding station, if not already present at 
a site, consisting of a 1-nr section ot plywood was 
placed 1 m above ground at each site in December 
2007. All feeding stations were regularly stocked 
with black-oil sunflower seeds. 
Predator Presentations. —We used study skins 
of five raptors in our experiment that differed in 
potential threat they posed to parids (Templeton 
et al. 2005. Soard and Ritchison 2009. Courier and 
Ritchison 2010). Common predators of parids 
(Gaddis 1980. Grubb 1998. Roth and Lima 2007). 
including Eastern Screech-Owl (Megascops asio), 
Sharp-shinned Hawk, and Cooper’s Hawk were 
considered high-threat predators (Templeton et al. 
2005). Great Homed Owls (Bubo virginianus) and 
Red-tailed Hawks ( Buteo janiaicensis ), species 
that rarely, if ever, prey on parids, were 
considered low-threat predators (lempleton et al. 
2005, Nocera and Ratliffe 2010). An empty 
platform (Baker and Becker 2002. lempleton 
et al. 2005) and a study skin of a Ruffed Grouse 
(Bonasa umbel Ins: a non-predutory bird) were 
used as control presentations. 
Trials were conducted at each location from 5 
January to 27 February 2008 during 0900 to 1400 
hrs EST with at least 48 hrs between successive 
trials. Each trial was conducted by JRC and began 
by placing a randomly-selected raptor (or control) 
in a life-like position on a platform I m above 
ground and I m from a feeding station. Specimens 
were initially covered with a white sheet during a 
5-min pre-presentation period while the observer 
sat 5 m away. The pre-presentation period was 
intended to acclimate birds to the observer s 
presence and ensure chickadees and titmice would 
remain near a feeding station sufficiently long for 
an experiment to be completed. The observer then 
walked to the feeder, removed the sheet to expose 
the raptor or control, and returned to the 
observation site to monitor the behavior of 
chickadees and titmice for the 5-min trial period. 
The observer noted during each trial: (1) the 
number of chickadees and titmice present, (2) the 
distance of closest approach by a titmouse and 
chickadee, and whether a chickadee or a titmouse 
(3) first approached within 5 m ol the raptor or 
control, and (4) spent more time within 5 m of the 
predator. Mean approach distances were estimated 
( ± 0.5 m) using the width of the predator platform 
(1 m). distance of platform to feeding station 
(I m). and the distance between the observer and 
the feeding station as references. 
Statistical Analyses. —We calculated a mean 
distance of closest approach for chickadees and 
titmice in each of the eight flocks during the two 
control trials, the two low-threat predator trials, 
and the three high-threat predator trials. We 
ascertained the percentage of control, low-threat, 
and high-threat trials where titmice and chicka¬ 
dees in each flock were first to approach within 
5 m and remained within 5 m ol controls oi 
predators the longest. We compared the mean 
distance of closest approach by chickadees and 
titmice, and the percentage ol trials where each 
species was first to approach and remain within 
5 m of controls and predators using Wilcoxon 
tests for paired samples. All tests were two-tailed. 
We used Kruskul-Wallis tests to examine possible 
differences in flock si/.e during control, low- 
threat. and high-threat predator trials. We used the 
Statistical Analysis System for all analyses (SAS 
Institute 2004). Values are presented as means ± 
SE. 
RESULTS 
The mean ± SE number of chickadees present 
during trials was 3.2 ± 0.2 (range = 1-6) and the 
mean number of titmice present was 2.6 ± 0.1 
(range = 1-5). Mean numbers of chickadees 
(Kruskal-Wallis '/ji = 0.1. P = 0.93) and titmice 
(Kruskal-Wallis ti = 0,5, P = 0.80) present 
during control, low-threat, and high-threat trials 
did not differ. Chickadees and titmice were 
equally likely to he first to approach within 5 m 
of controls or predators in response to control (Z 
= 1 . 4 . p = 0.08), low-threat (Z = 0.2, P = 0.41), 
and high-threat (Z = O L P = 0.50) treatments. 
Titmice and chickadees did not differ in tendency 
to remain within 5 m ot controls (Z = 0.1, P — 
0.50). However, titmice were more likely to 
remain within 5 m during trials with both low- 
threat (Z = 3.2. P = 0.0008) and high-threat (Z = 
3.0, P = 0.0015) predators. Titmice in the eight 
flocks tested spent more time within 5 m than 
chickadees during an average of 81.2 ± 9.0% ot 
all trials with low-threat predators and 79.4 ± 
9.1% of high-threat trials. 
