122 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 123 , No. 1, March 2011 
ing season, Ismar et al. 2010a) at the Plateau 
Colony, Cape Kidnappers Gannetry, New Zealand 
(39 38' S, 177° 05' E) were equipped with GLS 
loggers (Phillips et al. 2004) on 21 February 2008 
and 12 March 2008. Deployments were timed for 
close to fledging at this location (Ismar et al. 
2010a). Devices were placed (1 = left, r = right 
leg) on two male (M-77108 [1], A08 [r] and M- 
74777 [1], 111 [r]) and two female (M-77177 [1], 
All [r] and M-74768 fl], 768 [r]) Australasian 
Gannets. The fifth bird (M-77260 [r], B60 [1]) was 
the putative male offspring of a male Cape Gannet 
(M. capensis) and a female Australasian Gannet 
hatched in 2001—2002 (Robertson and Stephenson 
2005), which had returned to breed at its natal 
colony. Gender of these banded individuals was 
ascertained from DNA samples following Daniel 
et al. (2007). The 3.5-g loggers (<0.2% of adult 
body weight) were attached with two cable ties to 
darvic PVC plastic leg bands (Rayner et al. 2008), 
and retrieved from birds A08, A77, and B60 early 
in the subsequent breeding season. All three birds 
subsequently bred in the 2008-2009 season. Birds 
M-74768 and M-74777 also returned to the Cape 
Kidnappers Plateau colony to breed in 2008, but 
had lost the GLS devices from the plastic bands. 
Geolocator data from the three retrieved 
devices were processed following Phillips et al. 
(2004). Times of sunrise and sunset were 
estimated from light records, and converted to 
location estimates using TransEdit and Bird- 
Tracker software (British Antarctic Survey, Cam¬ 
bridge, UK). Transitions associated with poor 
quality light curves were identified during pro¬ 
cessing, and the resulting positions were exclud¬ 
ed, as appropriate, after visualization in a GIS. 
Only longitudes were available around equinoxes, 
when daylength is similar throughout the world. 
The primary areas frequented by tracked birds 
were mapped as kernel density plots with the 
Spatial Analyst extension of ArcMap 9.3 (ESRI 
Inc. 1999) applying an ITRF 2005 geographic 
coordinate system (Fig. 1). Density contours were 
set to display the top 80% of spatial use around 
the Cape Kidnappers Gannetry, New Zealand, and 
at Australian wintering areas. A conservative 
estimate of the distances covered during migration 
was calculated for each bird, based on core areas 
of the kernel estimates (Cape Kidnappers Gan¬ 
netry, New Zealand; and top 20% of the kernel 
densities on the Australian/Tasmanian coast) with 
minimum distances between landmarks calculat¬ 
ed, assuming direct travel over water. Estimated 
FIG. 1. Migration routes of (A) a male and (B) a female 
adult Australasian Gannets (Moms serrator), and (C) a 
male hybrid between M. serrator and M. capensis between 
breeding seasons 2007-2008 and 2008-2009 at Cape 
Kidnappers Gannetry, New Zealand; colored symbols: 
highest quality positions obtained using geolocators; dashed 
lines: approximate migration routes inferred, lines darker 
with advancing time. 
migration paths between breeding and wintering 
areas were plotted for orientation, but should be 
treated as approximations, given the typical 
measurement errors of —186-202 km recorded 
