334 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 123. No. 2. June 2011 
Chickens. “Coursing” was a low, slow glide over 
the landscape in search of prey, typical of 
Northern Harriers. “Perch" was noted any lime 
a raptor landed in view of the camera, and a “fly¬ 
by" was when the raptor paid no apparent 
attention to prairie-chickens as it flew over. 
Responses of prairie-chickens to raptors and 
ravens were categorized as no response, squat, 
partial flush, or flush. We considered a raptor/ 
raven-prairie-chicken encounter as eliciting “no 
response" if we could not observe any alteration 
in behavior of the grouse when the encounter 
occurred. “Squat” consisted of seeking cover or 
flattening against the ground. “Partial flush” 
consisted of at least one but not all prairie- 
chickens flushing from the lek, and “flush” 
occurred when every grouse on the lek flushed. 
We noted the lime when lekking behavior 
ceased to when at least two Lesser Prairie- 
Chickens returned to lekking behavior. Raptor 
encounter rates were calculated by dividing the 
number of raptors observed at leks by the hours of 
lek observation or recording during which prairie- 
chickens were present. We used an exact rate ratio 
test assuming Poisson counts in Program R to 
compare raptor encounter rates (R Development 
Core Team 2008, Fay 2009). 
^ e used a G-test for goodness of fit to compare 
frequency of occurrence of different raptor 
species observed at leks to the overall raptor 
“ thC arCa fol,owin g Sokal and Rohlf 
(1995). We used frequency of occurrence at leks 
by Swamson's Hawks. Northern Harriers, falcon 
species, and buteo species (not including Swain- 
son s Hawks) as observed values. Expected values 
were computed from data collected during 
rapt ° r mneys in ,he area (Behney 
2009). Only surveys conducted during February. 
March, and April were included and were pooled 
over years. The total number of each raptor 
species or species group observed during surveys 
was d.vided by the total number of all raptors 
recorded during surveys. We then multiplied this 
proportion by the total number of raptors observed 
at leks to sum to the same total as raptors seen at 
leks, while maintaining the same proportion of 
each species/species group. 
We used Chi-square tests of independence 
(Conover 1999) to assess independence among 
approach, species, and response. We also used a 
Ch.-square test to assess differences in behavioral 
response of Lesser Prairie-Chickens to raptor 
encounters compared to raven encounters We 
used a Kruskall-Wallis test (Conover 1999) lo 
compare prairie-chickens return times associated 
with different raptor species because the data were 
not normally distributed. We categorized the 
lekking season into five 2-week intervals toasts 
temporal variations in encounter rates and specie' 
dynamics: I = 9 to 22 March. 2 = 23 March to 5 
April. 3 = 6 to 19 April, 4 = 20 April to 3Mjy 
and 5 = 4 to 17 May. 
RESULTS 
Raptor Encounter Rates. — We conducted 155hrs 
of direct observation at seven leks while Lesser 
Prairie-Chickens were present between 24 February 
and 21 May 2007 and 2008. We observed 21 raptors 
and 37 ravens on or near leks while prairie-chicken- 
were present. We recorded 495 hrs of real-time 
video footage on seven leks while prairie-chicken' 
were present from 3 April through 9 June 2007aoc 
8 March through 22 May 2008, We observed 4U 
raptors and 104 ravens on or near leks while praiue- 
chickens were present during the video recordings 
There was no difference in encounter rates between 
viewing platform (direct observations = 0.14 
raptors/hr, video = 0.08 raptors/hr; P - 008) and 
the data were pooled (overall = 0.09 raptars/hn 
Study leks averaged 12.3 males (range = 6-19J- 
We did not observe any successful predation event' 
on the video or through direct observations. 
Northern Harriers (n = 30) were the ^ 
commonly observed raptor at leks followed by 
Swain son's Hawks (n =11). other buteos d r 
falcons (n = 5), and accipiters (n = 2). wea 
unable to identify five raptors to the genu.' lew 
We observed one additional Peregrine Falcon 
encounter at a lek during direct observations b- 1 
failed to note the time Lesser Prairie-Chickf ■ 
departed the lek and were unable to assess d- 
duration of lek attendance by prairie-chickens that 
morning. This encounter was included in beha 
•°ral analyses but not in encounter rate cakni- 
tions. Encounter rates peaked during the second - 
week interval and steadily decreased to * r 
lowest rates during the last interval 'Fig 
Proportions of individual raptor species or si* 1 - 11 ’ 
groups at leks diverged from that expected bi s ' 
on the observed raptor community (Xy ' J -' 
< 0.001). More Northern Harriers (contribution 1 
overall G-statistic = 28.92), fewer Swain wn ' 
Hawks (—12.08), and more falcons (8.20! n- 
observed at leks than expected based on the rap' 
community present in the area. 
