804 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 123, No. 4. December 2011 
oriented valleys of the Basin and Range province 
would facilitate ground-based tracking using 
conventional telemetry methods. Study site loca¬ 
tions and sampling periods were staggered with 
work occurring at the most northern site in 
September and at the most southern site in late 
October. Study sites were in the southwestern 
portion of the breeding range of Amphispiza belli 
ssp. nevadensis (Johnson and Marten 1992). Sage 
sparrows from these areas generally depart on 
migration from mid-September to mid-November 
(Littlefield 1990). 
Predominant vegetation at all sites was black 
greasewood; salt desert shrubs (Atriplex spp.) and 
Wyoming big sagebrush comprised minor por¬ 
tions of shrub communities. Sage Sparrows were 
observed eating greasewood seeds and many 
individuals captured for marking had inkweed 
(Suaeda spp.) stains on their bills. Inkweed has 
been associated with Sage Sparrows in wintering 
areas (Meents et al. 1982). 
We captured Sage Sparrows by flushing them 
into mist nets and use of audio playbacks, 
classified birds as adults (AHY) or hatch year 
(HY), and attached 0.55-g transmitters (Advanced 
Telemetry Systems Inc.. Isanti. MN. USA) using a 
figure-8 harness (Rappole and Tipton 1991). 
Transmitters were --3.0% of bird mass and were 
below the recommended maximum of 5% (Cac- 
camise and Hedin 1985). We monitored presence 
of birds and triangulated at least one location 
daily from 0630 to 1900 hrs MST using hand-held 
Yagi antennae. Up to three birds were monitored 
each night by placing data-logging receivers 
(Lotek Wireless Inc., Newmarket, ON. Canada) 
and Yagi antennae in their immediate (<100 m) 
vicinity at dusk. 
We recorded Global Positioning System coor¬ 
dinates of observer, time of observation, and 
compass bearing for each telemetry location. We 
estimated point locations of birds using maximum 
likelihood methods and 95% confidence interval 
error ellipses (LOAS 4.0,3.1; Ecological Software 
Solutions LLC 2004). We calculated a minimum 
convex polygon (MCP) for each bird after 
discarding triangulations with error ellipse axes 
>250 m or biangulations (bird locations based on 
2 telemetry points) outside the MCP of all other 
observations for each individual. Point locations 
and MCPs were imported into a Geographic- 
Information System (ArcGIS 9.2; ESRI Inc. 
2008) for additional mapping, including area 
calculations and daily movement statistics based 
on distance between daily sequential observations. 
We characterized Sage Sparrows as having 
migrated if they were not located after 2 days 
within a 35-km search radius. 
We obtained weather data from the nearest 
weather station (Pueblo Valley barometric pres¬ 
sure = KREO, 121 km; Pueblo Valley all other 
measurements = FLS03. 46 km; Silver State 
Valley = KWMC. 48 km; Carson Sink = 
CSNSK, 18 km) compiled in the MesoWest data 
archive (Horel 2007). Hourly measurement 1 'were 
summarized to daily mean, maximum, and 
minimum temperatures, mean relative humidity, 
mean barometric pressure (Pressure Mean Sea 
Level, corrected for elevation), and mean wind 
speed. Wind direction data were retained in an 
hourly format. We used principal component 
analysis (PCA) and factor loadings to characterize 
weather conditions of days when we lost radio 
contact with any bird and for 1-2 days before and 
after the assumed departure day. 
RESULTS 
We radiomarked 38 Sage Sparrows (Pueblo 
Valley = 2 AHY, 3 HY; Silver State Valley = 4 
AHY. 8 HY; Carson Sink = 5 AHY, 16 HY). We 
recorded 477 bird locations of which 455 were 
retained: 380 based on triangulations, 68 biangu¬ 
lations. and 7 single observations when a bird was 
observed. Mean number of total observations per 
bird was 13.5 (range = 1—39) and mean number 
of daily observations was 1.7 (range = 1-6) Our 
longest tracking period for an individual spanned 
18 days. Mean transmitter/receiver distance was 
266.5 m; we were able to detect signals at 
distances up to 1.5 km when listening from local 
high terrain. Mean observation error (mean length 
of max and min error ellipse axes) for retained 
locations was 42.9 m (range = 0-235 m). 
Three radio-marked birds died during the study: 
two (1 AHY and 1 HY) appeared to have been 
preyed upon by raptors. A radio transmitter from 
one HY w r as recovered within a regurgitated pellet 
consistent in size with those produced by Short¬ 
eared Owls (Asia flammeus ) or Northern Hamers 
(Circus cyaneu.s) (C. D. Marti, pers. comm.); both 
species were observed regularly during the study. 
A third HY died from an unknown cause. 
Range Size and Movements. —We calculated 
range sizes for 35 of 38 marked Sage Sparrows 
after filtering data to remove individuals with 
<3 days of observation or <5 triangulated 
locations. Typical daily movements were <1 ^ 
