Year-round distribution of white-chinned petrels from South Georgia: Relationships with oceanography and fisheries

Abstract

The white-chinned petrel Procellaria aequinoctialis is a medium-sized procellariiform with a circumpolar subAntarctic breeding distribution. Feeding during both day and night, and often competing aggressively for bait, offal and discards, it has the highest incidental mortality rate of any seabird in Southern Ocean longline fisheries. Although still abundant, the limited census data suggest rapid population declines. Using geolocators, the movements of 10 white-chinned petrels from South Georgia were tracked for 226–664 days, which in combination with previous satellite-tracking provided the first comprehensive description of migration routes and year-round distribution of this species from any site. All birds migrated to Patagonian Shelf and shelf-break waters, concentrating in highly productive areas east of the River Plate estuary and to a lesser extent on the open shelf off central Argentina. Two birds traveled initially to the southern Patagonian Shelf but then moved in mid-winter to the Humboldt Current (Chile), before returning directly to South Georgia. One bird adopted this strategy in two winters, and was consistent in timing of return migration to South Georgia, but not of arrival off Chile. Despite the distance (>2000 km), birds returned to feeding sites on the Patagonian Shelf for all pre-laying exodus, and most incubation, trips. In contrast, most chick-rearing trips were to the local shelf, central Scotia Sea or South Orkney Islands, on average only 610 km from the colony. The distribution of white-chinned petrels overlapped with several major fisheries, many of which are known or suspected to have high rates of seabird bycatch. Until this issue is addressed, the status of the white-chinned petrel population at South Georgia should be viewed with considerable concern.

Full text

Page 1

Year-round distribution of white-chinned petrels from
South Georgia: Relationships with oceanography and fisheries
Richard A. Phillips*, Janet R.D. Silk, John P. Croxall, Vsevolod Afanasyev
British Antarctic Survey, Natural Environment Research Council, High Cross, Madingley Road, Cambridge CB3 0ET, United Kingdom
A R T I C L E I N F O
Article history:
Received 15 July 2005
Received in revised form
25 October 2005
Accepted 31 October 2005
Available online 15 December 2005
Keywords:
Bycatch
Conservation
Incidental mortality
Migration
Winter philopatry
A B S T R A C T
The white-chinned petrel Procellaria aequinoctialis is a medium-sized procellariiform with a
circumpolar subAntarctic breeding distribution. Feeding during both day and night, and
often competing aggressively for bait, offal and discards, it has the highest incidental mor-
tality rate of any seabird in Southern Ocean longline fisheries. Although still abundant, the
limited census data suggest rapid population declines. Using geolocators, the movements
of 10 white-chinned petrels from South Georgia were tracked for 226–664 days, which in
combination with previous satellite-tracking provided the first comprehensive description
of migration routes and year-round distribution of this species from any site. All birds
migrated to Patagonian Shelf and shelf-break waters, concentrating in highly productive
areas east of the River Plate estuary and to a lesser extent on the open shelf off central
Argentina. Two birds traveled initially to the southern Patagonian Shelf but then moved
in mid-winter to the Humboldt Current (Chile), before returning directly to South Georgia.
One bird adopted this strategy in two winters, and was consistent in timing of return
migration to South Georgia, but not of arrival off Chile. Despite the distance (>2000 km),
birds returned to feeding sites on the Patagonian Shelf for all pre-laying exodus, and most
incubation, trips. In contrast, most chick-rearing trips were to the local shelf, central Scotia
Sea or South Orkney Islands, on average only 610 km from the colony. The distribution of
white-chinned petrels overlapped with several major fisheries, many of which are known
or suspected to have high rates of seabird bycatch. Until this issue is addressed, the status
of the white-chinned petrel population at South Georgia should be viewed with consider-
able concern.
? 2005 Elsevier Ltd. All rights reserved.
1. Introduction
There is increasing conservation concern about unsustain-
ably high levels of bycatch in many marine fisheries, with
several previously abundant species of turtles, sharks and
seabirds experiencing serious population declines as a direct
result of incidental mortality (Spotila et al., 2000; Baum
et al., 2003; Lewison et al., 2004; Gerrodette and Forcada,
2005). This includes 19 of the 21 extant albatrosses or Diom-
edeidae (making this the most globally threatened of any
bird family), and several members of the related Procellarii-
dae (BirdLife International, 2004a). In recognition of the
problems facing these otherwise long-lived, extreme k-se-
lected seabirds, the international Agreement on the Conser-
vation of Albatrosses and Petrels was signed in Australia by
11 countries in 2001, and since ratified by eight nations. To
improve the status of albatrosses and petrels, signatories
(including flag states of vessels operating in waters used
by nominated species) are obliged to take specific measures
to protect critical habitat, control detrimental non-native
0006-3207/$ - see front matter ? 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biocon.2005.10.046
* Corresponding author: Tel.: +44 1223 221 610; fax: +44 1223 221 259.
E-mail address: raphil@bas.ac.uk (R.A. Phillips).
B I O L O G I C A L C O N S E R VAT I O N 1 29 ( 2 00 6 ) 33 6 –3 47
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/biocon

Page 2

species, support research and, most importantly, reduce
incidental mortality.
Awareness of bycatch issues has led to the use of tracking
data to try and identify where and when the greatest poten-
tial exists for negative interactions between albatrosses, pet-
rels, and fisheries (Nel et al., 2000, 2002b; Anderson et al.,
2003; BirdLife International, 2004b; Cuthbert et al., 2005).
Included in the species covered by the recent agreement,
and listed as Vulnerable by IUCN, is the white-chinned petrel
Procellaria aequinoctialis (BirdLife International, 2004a). It has a
circumpolar distribution, with breeding confirmed at eight
archipelagoes, but the few census data suggest rapid popula-
tion decreases (Woehler, 1996; Berrow et al., 2000a; Nel et al.,
2002a). There is evidence for reduced breeding success and
some historical extirpations on islands with introduced pre-
dators, but no long-term reductions in productivity elsewhere
(Marchant and Higgins, 1990; Berrow et al., 2000a). White-
chinned petrels forage by both night and day, are aggressive
in competing for fishing bait, offal and discards, and have a
disproportionately high chance of being hooked in relation
to the number attending vessels (Barnes et al., 1997; Weimers-
kirch et al., 2000). They are also abundant, and consequently
have the unfortunate distinction of the highest bycatch rates
of any seabird in Southern Ocean fisheries (Barnes et al., 1997;
Weimerskirch et al., 2000; Nel et al., 2002c; Ryan et al., 2002).
Although many benefit from the ready accessibility of dis-
cards and offal, which can form a major dietary component
(Jackson, 1988; Catard et al., 2000), incidental mortality cur-
rently represents an extreme threat to long-term population
viability.
Satellite-tracking studies have provided a good indication
of the at-sea distribution of breeding white-chinned petrels
from the Crozet Islands and South Georgia during chick-rear-
ing, and to a limited extent during incubation (Weimerskirch
et al., 1999; Berrow et al., 2000b; Catard et al., 2000). In con-
trast, there is little information on winter distribution beyond
observations that densities increase in northern subAntarctic
and subtropical regions (Marchant and Higgins, 1990; Olmos,
1997); furthermore, the age and provenance of the many birds
caught in fisheries is largely unknown. The South Georgia
archipelago holds the world?s largest breeding population
(up to two million pairs), with data from Bird Island indicating
a decrease of 28% in only two decades (Berrow et al., 2000a).
During chick-rearing, most of the distribution is in well-regu-
lated waters under the jurisdiction of the Commission for the
Conservation of Antarctic Marine Living Resources (CCAMLR),
and bycatch levels in these sectors are minimal (CCAMLR,
2004). In contrast, four tracks from incubating adults, and lim-
ited data from two birds dispersing after breeding failure, sug-
gest that adults might make much greater use of Patagonian
Shelf waters at other times of year (Berrow et al., 2000b). If
confirmed, this would be of considerable concern, as fishing
effort and bycatch rates are very high in many longline and
trawl fisheries in the region (Schiavini et al., 1998; Stagi
et al., 1998; Tuck et al., 2003).
The aims of the current study were to (i) identify the key
wintering areas used by white-chinned petrels from South
Georgia; (ii) extend the sample of tracks obtained during the
early breeding season, when distribution is poorly known;
(iii) examine individual variability in site preference; (iv)
determine
selection, and any temporal variation therein; (v) quantify
spatio-temporal overlap with fisheries with a view to focusing
conservation and management.
theoceanographic factorsunderlyingsite
2.Materials and methods
Initial deployments of geolocators or Global Location Sensors
(GLS-I loggers: Afanasyev, 2004) were carried out on 5–7 Febru-
ary 2003 at Cobblers Mound, Bird Island, South Georgia
(54?000S, 38?030W). Loggers attached to plastic leg rings (total
mass 10.5 g; <1% of adult body mass) were deployed on the
tarsi of 35 adult white-chinned petrels each extracted from
a different burrow. Burrows were marked with wooden
stakes, and were visited nine times in total during incubation
in the two following breeding seasons in order to retrieve de-
vices. Most study birds were weighed when recaptured, as
was a random sample of control adults from adjacent bur-
rows. These and other data are provided as means ± SD, un-
less indicated otherwise.
Breeding status was unknown at the time of deployment
(and study birds were left undisturbed thereafter), but could
usually be inferred from the subsequent pattern of colony
attendance derived from the light data. As initial deployments
took place during the post-brood period, most were non-
breeders or failed breeders. However, one individual that
continued to make regular visits to its burrow (inferred from
periods of darkness during the evening prior to dusk, or morn-
ing after dawn), and remained primarily in the South Georgia
region until mid April, was almost certainly a successful bree-
der. In the second summer, light records indicated periods of
8–16 days of continual darkness (i.e., time spent in the bur-
row), alternating with similar time spans when days and
nights were of normal duration and birds were located at
sea. This was assumed to represent a regular incubation sche-
dule, as it was only apparent during late November – early Jan-
uary (corresponding to the incubation period: Hall, 1987), and,
in January 2004, an unhatched egg was found in the burrow of
one bird exhibiting this characteristic attendance pattern.
Light data from retrieved loggers were processed subse-
quently according to Phillips et al. (2004). Sunset and sunrise
times were estimated from thresholds in the light curves, lat-
itude derived from day length, and longitude from the time of
local midday with respect to Greenwich Mean Time and Julian
day, providing two locations per day. During processing, loca-
tions derived from light curves with obvious interruptions
around sunset and sunrise, or that required unrealistic flight
speeds (>35 km h?1sustained over a 48 h period) were identi-
fied and later excluded if appropriate. In addition, locations
were unavailable for variable periods around the equinoxes
when it is difficult or impossible to estimate latitude (Wilson
et al., 1992; Hill, 1994). Validated data were smoothed twice
(Phillips et al., 2004, 2005).
Birds were considered to be wintering from the onset of
the outward migration in February–April, to their return the
following October. Key wintering areas were identified by gen-
erating kernel density maps in a Lambert equal-area azi-
muthal (south pole) projection using the ESRI?ArcGISTM
Spatial Analyst Density tool, with a cell size of 50 km and a
search radius of 200 km (Croxall et al., 2005; Phillips et al.,
B I O L O G I C A L C O N S E R VAT I O N 1 2 9 (2 0 06 ) 3 36 –3 4 7
337

Page 3

2005). White-chinned petrels at South Georgia lay between 13
November and 10 December, with a mean of 22 November
(Hall, 1987). Locations of tagged birds indicated that they each
made a long trip to the Patagonian Shelf immediately prior to
this period, corresponding to the pre-laying exodus, under-
taken by both sexes (Marchant and Higgins, 1990). For these
and incubation trips, fixed start and finish locations were
added to data files that corresponded to the dates of depar-
ture from, and return to, the burrow at Bird Island (evident
from direct examination of the light records).
Satellite-telemetry data were also available for birds
tracked using 20 or 30 g Platform Terminal Transmitters dur-
ing incubation (n = 4 trips from four birds) and chick-rearing
(n = 19 trips from six birds) from the same site in 1996/1997
and 1997/1998 (Berrow et al., 2000b). Unlikely locations were
removed using an iterative filter, assuming a maximum flight
speed of 90 km h?1(McConnell et al., 1992). For all satellite-tag
and geolocator trips, the duration, maximum range (furthest
distance reached from the colony), cumulative travel distance
along the route, and speed (cumulative travel distance/dura-
tion) were calculated. Note that for the geolocator data, these
distances and speeds areonly approximate, as the mean accu-
racy ± SD of each fix after smoothing was likely to be around
169 ± 104 km (Phillips et al., 2004). As there was considerable
light level interference in data files from one pre-laying exo-
dus and one incubation trip, and routes were therefore un-
clear, these two tracks were excluded from detailed analysis.
To compare areas of high and low utilization during the
winter, kernel density contours were overlaid on maps of
bathymetry(at1/12?
resolution,
www.gfdl.noaa.gov/products/vis/data/datasets/TerrainBase.
html), and monthly composites of remote-sensed sea surface
characteristics: sea surface temperature in ?C, and Chloro-
phyll a concentrations in mg m?3(Aqua MODIS mapped prod-
ucts at 9 km cell resolution, downloaded from http://
oceancolor.gsfc.nasa.gov).Remote-sensed
months were visualized, but quantitative comparison with
bird distribution was restricted to two representative mid
winter months (May and July 2003).
The distribution of tracked birds was also compared with
the reported number of hooks set within 5? grid squares in
1998 or 1999 (the two recent years with best data coverage)
by major longline fleets operating in overlapping and adjacent
fishing grounds. Spatially explicit pelagic longline data re-
ported to the International Commission for the Conservation
of Atlantic Tuna and Commission for the Conservation of
Southern Bluefin Tuna, and demersal longline data reported
to the Commission for the Conservation of Antarctic Marine
Living Resources and the Falklands Fishery Agency for 1998
were provided by G. Tuck (details in Tuck et al., 2003). Data
on pelagic effort during the winter by domestic and leased
Brazilian vessels in 1998 were provided by F. Olmos. Total
demersal effort by the Argentine fleet in 1999 was assigned
to six northern and six southern 5? squares based on maps
and the proportion of effort targeted to each region (Schiavini
et al., 1998; Faveroet al., 2003). Total demersal effort during the
winter by Chilean artisanal and commercial fleets in 1999 was
assigned to, respectively, three and four adjacent 5? squares
within the main operational areas (data provided by C. More-
no). No detailed information was available on demersal or pe-
availablefrom http://
data fromall
lagic Uruguayan, or Brazilian demersal fisheries. Effort was
not temporally stratified by season south of Brazil on the Pat-
agonian Shelf, as tracked birds visited this area, and hence
were vulnerable to bycatch, for most of the year.
3. Results
3.1.Retrieval details and device effects
Of the initial 35 study burrows, two could not be relocated,
three had collapsed or were blocked by ice, and five were
unoccupied in the following season. In total, 11 (31%) of the
loggers were retrieved, nine in summer 2003/2004 (one of
which failed to download) and two in summer 2004/2005
(one of which provided data for 226 days only because of bat-
tery failure). Birds were therefore tracked for a mean of 366
days (range 226–664 days). Of the 3872 locations obtained dur-
ing the winter outside of the equinox periods, 14.2% and 1.8%
were excluded subsequently because of light level interfer-
ence in the data file, or unrealistic flight speeds, respectively,
providing a total of 3251 validated locations. There was no sig-
nificant difference in mean mass between a sample of birds
with loggers, and controls (means of 1279 ± 0.10 g, n = 6, and
1300 ± 0.11 g, n = 11, respectively; T-test, t15= 0.40, P = 0.7).
3.2.Breeding status
Of the 10 birds tracked, the nine failed breeders/non-breeders
at the initial capture all departed from the South Georgia/Sco-
tia Sea region within 1–2 weeks of device deployment (Table
1). In contrast, with the exception of a single trip to the
shelf-slope east of the Falkland Islands in late March, the suc-
cessful breeder remained for the next 2.5 months in the South
Georgia/Shag Rocks/Scotia Sea region, finally departing on 19
April. This was very close to the mean chick fledging date at
Bird Island (mean 21 April, range 9 April to 9 May: Hall,
1987). In the following season (summer 2003/2004), four birds
almost certainly bred, and one definitely did not. The status
of the remainder was uncertain, as loggers were retrieved
(n = 4) or had stopped recording (n = 1) before the usual onset
of laying. The individual tracked for two winters bred in both
the second and third summers.
3.3. Wintering strategies
All of the failed breeders/non-breeders began their outward
migration from South Georgia at very similar times (5–18 Feb-
ruary, n = 9; Table 1), reaching the Patagonian Shelf usually
within 1–2 days. The bird tracked for two winters departed
on similar dates in both years. During winter, the petrels were
distributed over a wide area on the Patagonian Shelf and
shelf-slope from Tierra del Fuego to south-eastern Brazil,
the western region of the Argentine Basin, and a more re-
stricted area on the Pacific coast of Chile (Fig. 1). Core areas
(40% kernels) were: (i) east of the River Plate (Rı ´o de la Plata)
estuary off northern Argentina, Uruguay and southeastern
Brazil (centred on 35?S); (ii) east of Bahı ´a Blanco and Peninsula
Valde ´z on a north–south axis at c.61?W, extending to offshore
of Capo Tres Puntas; (iii) west of Isla de Chiloe ´ (Chile) between
40?S and 45?S, centred around 76?W.
338
B I O L O G I C A L C O N S E R VAT I O N 1 29 ( 2 00 6 ) 33 6 –3 47

Page 4

Cores overlapped only for subsets of individuals (Fig. 2). Six
birds concentrated their activities off northern Argentina,
Uruguay and Brazil, and two spent the majority of their time
in the Patagonian Shelf south of 40?S. Two others initially
used this area but then moved to the Chilean coast where
they remained for 4–5 months before returning directly to
Bird Island at the start of the following season. The bird
tracked for two winters used this strategy in both years, but
was not consistent in the timing of the switch (first arrival
in the Drake Passage was on 9 May 2003 and 1 June 2004;
the other bird was there on 13 June 2003), or in the time taken
to travel from there to the core wintering area off Isla de Chi-
loe ´ (two days in 2003, and >2 weeks in 2004). All birds returned
to South Georgia within a few weeks of each other (4–26 Octo-
ber; Table 1). Return dates of the two birds from Chile were in
the middle of this range, and for the bird tracked for two sea-
sons were highly consistent in consecutive years (Table 1).
3.4. Relationship with oceanography
Core areas in the southwest Atlantic were relatively warm
and highly productive (high Chlorophyll a), shallow continen-
tal shelf regions (mean depth 123 m within the 20%, and
470 m within the 20–40% kernel regions; Table 2). Average
water depth increased, and sea surface temperature and
Chlorophyll decreased, in increasingly peripheral areas. In
contrast, the 40% kernel in the Pacific was situated on the
shelf slope over much deeper water (mean 2745 m). Water
temperatures there were intermediate, and productivity low-
er, than in the northern and southern kernels on the Patago-
nian Shelf. More peripheral kernel contours in the Pacific
were little different from the core in terms of water depth
and Chlorophyll, but were slightly cooler. Between May and
July, Chlorophyll a and sea surface temperature decreased
within the 40% kernels in both the southern Patagonian Shelf
and the Pacific. By July, surface temperature was slightly high-
er, but Chlorophyll lower, in the 40% kernel in the Pacific.
Therefore although reduced water temperatures and produc-
tivity possibly triggered departure from the southern Patago-
nian Shelf, it was not possible to identify why birds moved to
the Chilean coast rather than further north in the Atlantic.
3.5. Summer distribution
The nine birds tracked beyond mid November 2003 all per-
formed a pre-laying exodus, departing from South Georgia
on average on 31 October, 4–36 days after their initial return
to the colony, on a trip that usually lasted 2–3 weeks (Table
1). Return dates were more consistent (all between 15 and
24 November). Birds traveled on a north-westerly axis from
South Georgia, with the exception of two birds that initially
flew west to the Burdwood Bank before turning north
(Fig. 3(a)). The main destinations were within a large area
extending from the core used by wintering birds to the east
of Peninsula Valde ´z, to the north-east and south-west along
the outer Patagonian Shelf and shelf-slope, but with some
trips closer to the coast (around Peninsula Valde ´z and Punta
Mogotes). In most cases, this was not within the 40% kernel
of that individual during the winter (Fig. 2). On average, each
pre-laying exodus extended to a mean maximum range of
Table 1 – Timing of winter migration and characteristics of the pre-laying exodus of white-chinned petrels tracked using geolocators from South Georgia in 2003–2005
Ring
number
Status
Winter
Wintering area
Timing of migration
Pre-laying exodus
Summer
2002/2003
Summer
2003/2004
Summer
2004/2005
Departure date
Return date
Start date
End date
Duration
(days)
Distance
(km)
Speed
(km day?1)
Max. range
(km)
4178
F
U
–
First
PS, HC
16-Feb-2003
14-Oct-2003
27-Oct-2003
15-Nov-2003
19
5930
310
2160
4216
F
U
–
First
PS
16-Feb-2003
16-Oct-2003
5-Nov-2003
20-Nov-2003
15
6090
410
2040
6751
F
Br
–
First
PS
5-Feb-2003
23-Oct-2003
27-Oct-2003
24-Nov-2003
28
–
–
–
6756
F
U
–
First
PS
18-Feb-2003
18-Oct-2003
31-Oct-2003
15-Nov-2003
15
6600
440
2300
6759
F
Br
–
First
PS
14-Feb-2003
4-Oct-2003
9-Nov-2003
17-Nov-2003
8
4850
610
2250
6766
Br
U
–
First
PS
19-Apr-2003
26-Oct-2003
31-Oct-2003
19-Nov-2003
19
6130
320
2000
6769
F
Br
–
First
PS
8-Feb-2003
14-Oct-2003
28-Oct-2003
16-Nov-2003
19
6750
360
2060
6016
F
U
–
First
PS
8-Feb-2003
–
–
–
–
–
–
–
6773
F
F
–
First
PS
16-Feb-2003
9-Oct-2003
30-Oct-2003
15-Nov-2003
16
5950
370
2330
6758
F
Br
Br
First
PS, HC
18-Feb-2003
18-Oct-2003
1-Nov-2003
19-Nov-2003
18
8560
480
2270
Second
PS, HC
8-Feb-2004
19-Oct-2004
30-Oct-2004
19-Nov-2004
20
–
–
–
Mean
First
–
19-Feb-2004
15-Oct-2003
31-Oct-2003
17-Nov-2003
17.4
6360
410
2180
U, unknown; F, failed breeder/non-breeder, Br, Breeder, PS, Patagonian Shelf, HC, Humboldt Current.
B I O L O G I C A L C O N S E R VAT I O N 1 2 9 (2 0 06 ) 3 36 –3 4 7
339

Page 5

2180 km, involving a cumulative travel distance of 6360 km
(equivalent to 410 km per day; Table 1).
Nine geolocator tracks were obtained from four birds dur-
ing incubation in November 2003 – January 2004, comple-
mentingfourtripsmade
November – December 1996 and 1997 (Fig. 3(b)). With the
exception of a single trip in 2003 to oceanic waters (>3000 m
deep) to the north of South Georgia at 45–50?S, all were to
the Patagonian Shelf. Destinations were often close to the
coast in areas east of Capo Tres Puntas around 47?S, 62?W
(five trips by three birds), Peninsula Valde ´z (two trips) and Ba-
hı ´a Grande (one trip), and immediately north-east of the Falk-
lands (one trip). One was to the shelf-slope around 59?W, and
two were more northerly (to the core wintering area east of
the River Plate estuary, or to the outer shelf and shelf-slope
by satellite-taggedbirdsin
around 41?S, 57?W). There was no correspondence between
the destination of an incubation trip in 2003/2004 and the
40% kernel used by the bird during the winter. On average,
the eight trips to the Patagonian Shelf took 11 days, and in-
volved a 5000 km journey to a maximum range of 2100 km,
similar to trip statistics calculated from satellite-tracks (Table
3). Considering geolocator trips to the Patagonian Shelf only,
incubation trips were significantly shorter than pre-laying
exoduses in time and distance (T-tests, t14= 2.9, P < 0.02 and
t14= 2.6, P < 0.05, respectively), but comparable in terms of
maximum range and travel speed (T-tests, t14= 0.70 and
t14= 0.61, respectively, both P > 0.4).
During chick-rearing, birds foraged predominantly over
the local shelf and shelf-slope, and south to the South Orkney
Islands (17 trips), with a small proportion (2 trips; 11%) to the
Fig. 1 – (a) Density distribution of non-breeding white-chinned petrels from South Georgia in winter 2003 (n = 10 birds) and (b)
regional oceanography and location of places mentioned in the text. The density contours encompass 20–90% of the total
distribution.
340
B I O L O G I C A L C O N S E R VAT I O N 1 29 ( 2 00 6 ) 33 6 –3 47

Page 6

Patagonian Shelf, either to the shelf-break around 42?S or to
north of the Falkland Islands (Berrow et al., 2000b). Trips with-
in the South Georgia region were relatively short in duration,
travel distance and range, but involved faster travel speeds
than those undertaken during the pre-laying exodus or incu-
bation (Table 3).
3.6. Overlap with fisheries
Overlap between the distribution of white-chinned petrels
during winter 2003 and that of fisheries operating in the
South American sector in 1998 and 1999, was clearly substan-
tial: in almost every 5? square, the fishing effort was in excess
of one million hooks during periods when the birds were at
risk (Fig. 4). The biggest fisheries within the core areas were
demersal, with the exception of waters east of Uruguay and
Brazil where there was a large pelagic effort by Brazilian
and Taiwanese fleets. In this context, it is important to note
the absence of data on magnitude and spatial distribution
of Uruguayan fishing effort, as both pelagic and demersal ef-
fort by this fleet within the region of the northern 40% kernel
on the Patagonian Shelf (Fig. 1) is undoubtedly very high (Stagi
et al., 1998). During the pre-laying exodus and incubation,
birds were likely to come into most contact with the Argen-
tine demersal longline fleet, whereas during chick-rearing
the bird distribution largely overlapped with longline fisheries
in CCAMLR waters (Figs. 3 and 4).
4.Discussion
This is the first study of white-chinned petrels from any col-
ony to provide a detailed description of migration routes
Fig. 2 – Core areas (40% kernel contours) of individual white-chinned petrels in winter 2003 (n = 10 birds).
Table 2 – Bathymetric and oceanographic characteristics within the kernel density distribution of wintering
white-chinned petrels in 2003
Kernel Shelf area Water depth (m)Sea surface temp. (?C) Chlorophyll a conc. (mg m?3)
MayJuly MayJuly
Atlantic
0–20%North
South
All
North
South
All
–
–
–
145 ± 264 (60–1583)
76 ± 50 (2–152)
123 ± 222 (60–1583)
1012 ± 1211 (30–3994)
102 ± 159 (30–1528)
470 ± 902 (30–3992)
995 ± 1396 (180–4933)
1959 ± 1996 (98–5545)
2142 ± 2162 (300–5933)
15.8 ± 2.1 (9.0–20.8)
12.0 ± 2.0 (9.1–16.0)
14.5 ± 2.7 (9.1–20.8)
16.9 ± 3.6 (8.5–22.0)
11.1 ± 2.7 (6.4–16.3)
13.5 ± 4.2 (6.6–22.0)
11.4 ± 4.4 (6.0–22.2)
11.2 ± 4.1 (6.1–22.3)
11.9 ± 5.2 (3.6–22.6)
11.8 ± 1.6 (7.5–16.9)
8.6 ± 1.3 (7.0–11.4)
10.7 ± 2.2 (7.0–17.5)
14.0 ± 3.3 (6.6–19.5)
8.1 ± 1.7 (5.1–11.5)
10.5 ± 3.8 (5.1–19.5)
8.9 ± 4.1 (3.9–20.1)
8.8 ± 4.1 (3.9–20.2)
9.7 ± 5.1 (3.9–20.8)
6.37 ± 6.88 (0–44)
3.23 ± 2.03 (1–14)
5.05 ± 5.49 (0–40)
2.91 ± 4.25 (0–46)
1.79 ± 1.79 (021)
2.16 ± 2.92 (0–41)
1.21 ± 2.91 (0–65)
0.76 ± 3.01 (0–65)
0.71 ± 2.52 (0–65)
5.79 ± 6.42 (0–35)
0.69 ± 0.32 (0–2)
3.80 ± 5.42 (0–35)
2.95 ± 4.3 (0–31)
0.84 ± 1.41 (0–30)
1.60 ± 2.99 (0–31)
0.78 ± 2.41 (0–58)
0.58 ± 2.34 (0–41)
0.76 ± 2.47 (0–52)
20–40%
40–60%
60–70%
70–80%
Pacific
0–40%
40–60%
60–70%
70–80%
–
–
–
–
2745 ± 1291 (2–4085)
2268 ± 1553 (150–4216)
2990 ± 1250 (610–4356)
3062 ± 1306 (300–4493)
13.0 ± 0.9 (10.9–14.7)
12.7 ± 1.2 (10.0–16.0)
11.3 ± 2.1 (6.4–15.7)
10.8 ± 2.4 (6.1–15.7)
10.8 ± 0.7 (9.1–12.1)
10.3 ± 1.0 (7.3–12.1)
9.6 ± 1.4 (7.2–12.6)
9.4 ± 1.9 (4.0–12.7)
0.76 ± 0.74 (0–5)
0.87 ± 1.17 (0–16)
0.77 ± 1.61 (0–30)
0.92 ± 2.57 (0–49)
0.33 ± 0.26 (0–6)
0.46 ± 0.52 (0–8)
0.40 ± 0.49 (0–8)
0.39 ± 0.57 (0–17)
Data are means ± SD with the range in parentheses.
B I O L O G I C A L C O N S E R VAT I O N 1 2 9 (2 0 06 ) 3 36 –3 4 7
341

Page 7

Fig. 3 – Routes of white-chinned petrels tracked using geolocators (squares) or satellite-transmitters (open circles) from South
Georgia during (a) the pre-laying exodus in 2003 (n = 8 trips), (b) incubation in 1996/1997 and 1997/1998 (n = 4 trips), and 2003/
2004 (n = 9 trips from four birds), and (c) chick-rearing in 1998 (n = 19 trips from six birds). Fixed points corresponding to dates
of departure and return to the colony were added to each geolocator track, and data were smoothed twice (see Section 2).
342
B I O L O G I C A L C O N S E R VAT I O N 1 29 ( 2 00 6 ) 33 6 –3 47

Page 8

and wintering destinations and to identify foraging areas
used during the pre-laying exodus. The accuracy of geoloca-
tors is poorer than that of satellite-transmitters and there will
have been some inflation of ranges (Phillips et al., 2004). Also,
sample sizes were limited to 8–10 birds tracked at each stage.
However, the main conclusions, which are based on large-
scale distribution patterns, including clear differences among
individuals, are probably robust. There were no apparent del-
eterious effects of device attachment: neither abrasion asso-
ciated with the plastic leg ring, any significant reduction in
body mass, nor effect of handling on timing of migration in
the bird tracked for two years. This compares favourably with
attachment of (heavier) satellite-transmitters, which often in-
creases mean foraging trip duration (Phillips et al., 2003). The
recovery rate of deployed devices (31%) was not particularly
high, but this is unsurprising as birds were of unknown status
when captured and some could have been prospectors with
little or no burrow fidelity. In addition, only 73% of white-
chinned petrels at South Georgia breed in successive years
(Berrow et al., 2000a), and the probability of recapturing a bird
that had lost or divorced its partner, skipped breeding, or
changed burrow would have been very low.
4.1.General distribution – winter
Wintering white-chinned petrels from South Georgia made
extensive use of the Patagonian Shelf, from Tierra del Fuego
to as far north as 25–30?S, with most activity concentrated
in the more northerly region. Some individuals (20%) also
made substantial use of a smaller area of steep shelf-slope
in the Humboldt Current off southern Chile. Many spectacled
petrels P. conspicillata (which breed only on Inaccessible Is-
land, Tristan da Cunha) also winter off Brazil, but remain fur-
ther north than white-chinned petrels in the warmer waters
of the Brazil Current (Olmos, 1997). Black petrels P. parkinsoni
(which breed in New Zealand), migrate to the Pacific coast
of South and Central America, but to subtropical and tropical
waters north of 5?S (Pitman and Ballance, 1992). Latitudinal
and habitat specialization therefore limits competition for re-
sources amongst these similarly-sized and closely-related
species, all of which have broad diet preferences (Marchant
and Higgins, 1990; Berrow and Croxall, 1999). This parallels
the spatial segregation apparent amongst non-breeding alba-
trosses on both South American coasts (Nicholls et al., 2002;
Spear et al., 2003). One albatross with a wintering distribution
Table 3 – Characteristics of foraging trips by white-chinned petrels from South Georgia tracked using geolocators in 2003/
2004, and satellite-transmitters in 1996/1997–1997/1998
StageBreeding
season
Device
type
DestinationSample size
(trips, birds)
Duration
(days)
Distance
traveled (km)
Mean speed
(km day?1)
Maximum
range (km)
Incubation 2003/2004Geolocator Patagonian
Shelf
Oceanic
Waters
8, 4 11.5 (8–16)4970 (2740–6240) 437 (342–545) 2100 (1695–2710)
Incubation2003/2004 Geolocator 1, 1 103570360 1170
Incubation1996/1997,
1997/1998
Satellite-tag Patagonian
Shelf
4, 414.1 (11.7–15.1) 5569 (4602–6386) 394 (333-427)1982 (1492–2195)
Chick-rearing 1997/1998 Satellite-tag Patagonian
Shelf
2, 210.4 (9.7–11.2)5973 (5947–5999) 577 (534–621)1922 (1756–2087)
Chick-rearing 1997/1998 Satellite-tag South Georgia/
Scotia Sea
17, 6 4.5 (2.0–8.0)2539 (1169–5208) 587 (373–860)614 (302–1114)
Data are means with the range in parentheses.
Fig. 4 – Kernel density contours of wintering white-chinned petrels in 2003 overlaid on major reported longline fishing effort
(number of hooks set per year) in 5 · · 5? ? squares in the region in 1998/1999. For origin of fisheries data, see text.
B I O L O G I C A L C O N S E R VAT I O N 1 2 9 (2 0 06 ) 3 36 –3 4 7
343

Page 9

that is remarkably similar to that of the white-chinned petrel,
particularly off Chile, is the Northern royal albatross Diomedea
sanfordi (Nicholls et al., 2002). However, the latter is much lar-
ger and presumably exhibits prey preferences and foraging
habits that are sufficiently non-overlapping with white-
chinned petrels to permit co-existence.
4.2. General distribution – early breeding season
The Patagonian Shelf was also the primary destination during
all pre-laying exoduses and most incubation foraging trips
(Figs. 3 and 4). During a typical pre-laying exodus, birds trav-
eled an estimated cumulative distance of 6360 km to a maxi-
mum range of 2180 km from the colony (Table 1). Patagonian
Shelf trips during incubation were similar in terms of mean
maximum range and travel speed, but shorter in duration
and distance (Table 3). Note that although the use of geoloca-
tion results in lower accuracy and sampling frequency, trip
statistics were broadly comparable to those obtained using
satellite-tracking (Table 3). During the early breeding season,
birds from South Georgia therefore routinely travel 2000 km
to feed, underlining the exceptional capacity for long-dis-
tance foraging by this species (Weimerskirch et al., 1999). By
comparison, mean trip duration, total flight distance and
maximum range are lower during chick-rearing, both at
South Georgia (this study), and at Crozet, where the adults
mix long (mean 8.5 days) trips mainly to deep and relatively
distant Antarctic waters, with short (mean 1.2 days) to the
Crozet shelf-break (Catard et al., 2000).
4.3.Individual variation in distribution and timing
The 10 tracked birds showed considerable individual variation
in wintering distribution (Fig. 2). Other recent tracking studies
also suggest a high degree of large-scale individual wintering
site preferences in large procellariids (Weimerskirch and Wil-
son, 2000; Croxall et al., 2005; Phillips et al., 2005). The only
white-chinned petrel tracked in two winters adopted the
same (comparatively rare) strategy in both years of utilizing
both the Patagonian Shelf and Humboldt Current, but showed
no consistency in date or duration of transit between the re-
gions. Similarly, high fidelity to primary wintering regions
(although not to staging areas) and considerable flexibility
in terms of timing of movements within the winter is seen
in the black-browed albatross (Phillips et al., 2005).
In contrast, there was much less variation in the dates of
departure and first return to the breeding colony, and partic-
ularly in the return from the pre-laying exodus, even amongst
those birds that did not lay. All failed birds began their out-
ward migration in February (mean 13 February), close to the
mean departure date of failed and deferring black-browed
albatrosses in the same year (14 February 2003: Phillips
et al., 2005). These species feed on many of the same prey,
particularly Antarctic krill Euphausia superba, and the striking
coincidence in departure date may indicate a regional reduc-
tion in prey availability around that time. The high synchrony
in return date probably has less to do with current food sup-
ply than the selective advantages of avoiding usurpation of
burrows by prospectors or late breeders, and any delay to
breeding in a species that requires >5 months from laying to
chick fledging and must avoid blockage of burrows by ice at
the onset of winter. Incidentally, the mean arrival date of
the tracked birds was three weeks later than the first sight-
ings reported at South Georgia in previous seasons, but con-
forms very well to the first peak in burrow occupancy levels
determined from playback (Berrow, 2000). Similarly, the tim-
ing of the pre-laying exodus corresponds to the drop in bur-
row occupancy from late October to mid November (Berrow,
2000).
4.4.Distribution in relation to environment
The comparison between core and peripheral portions of the
wintering distribution indicates that white-chinned petrels
concentrate their activities in the most productive waters in
the region (Table 2). The most highly utilized areas in the
southwest Atlantic were relatively shallow waters, whereas
in the Pacific they were located on the steep shelf slope. The
latter falls within the southern portion of the cold, nutrient-
rich upwelling of the Humboldt Current. Here, South Pacific
temperate water flowing east along the northern edge of the
subAntarctic Front meets the Chilean coast between 40? and
50?S, before being deflected north. South of 15?S, maximum
upwelling occurs during the austral winter, when the Hum-
boldt Current apparently extends to 150–300 km offshore
(Spear et al., 2003), which fits well with the observed distribu-
tion. Although birds were rarely tracked north of 37?S, in
some years white-chinned petrels occur in low densities off
Chile as far north as 30?S, always in association with cold
water upwelling (Weichler et al., 2004).
The northern portion of the distribution in the southwest
Atlantic was east of the River Plate, particularly off Uruguay.
This region is predominantly under the influence of the cold
Falklands Current (which meets the warmer, southward flow-
ing Brazil Current slightly further north at the subtropical
convergence), and is also enriched by outflow from the River
Plate. The confluence region is characterized by high eddy
variability and high productivity (Olmos, 1997). The Brazil
Current is the western limb of the South Atlantic tropical gyre
and carries subtropical waters of c.12.5 ?C in mid winter into
the region (Sanderson et al., in press). Hence, the mean sea
surface temperatures of 11.8 and 14.0 ?C in the 20% and 20–
40% kernels indicate that the northeastern portion of the
white-chinned petrel distribution extends to some degree
into subtropical waters.
During the pre-laying exodus and some incubation trips,
the white-chinned petrels targeted the Patagonian Shelf-
break rather than the open shelf. Here they take advantage
of the Shelf-break Front, which is a narrow transition region
between subAntarctic and shelf waters. The front is charac-
terized in the summer by frequent Chlorophyll a maxima
within the zone of upwelling created when the Falkland Cur-
rent strikes the shelf break, carrying nutrient-rich waters to-
wardsthesurface(Sanderson
chlorophyll concentrations along the shelf break are not evi-
dent north of 35?S which is presumably why the white-
chinned petrels from South Georgia rarely extend trips any
further north during the early breeding season.
Although the cores tended to be warmer than peripheral
parts of the distribution within the same geographic region,
etal., in press). High
344
B I O L O G I C A L C O N S E R VAT I O N 1 29 ( 2 00 6 ) 33 6 –3 47

Page 10

there was considerable variation in absolute values, suggest-
ing no restriction to a specific thermal regime. Lack of pro-
nounced habitat preference in terms of temperature and
water depth is fairly typical of the species: birds from Crozet
feed in cold oceanic waters, the nearby neritic slope, and
warm upwellings in subtropical areas (Weimerskirch et al.,
1999; Berrow et al., 2000b; Catard et al., 2000). Hence, relative
prey abundance rather than habitat specialization must be
the key determinant of distribution. This is perhaps unsur-
prising given the broad range of prey in the diet, and the
capacity to dive to 5–6 m routinely, indicative of a highly
opportunistic as well as wide-ranging predator (Huin, 1994;
Berrow and Croxall, 1999; Catard et al., 2000).
4.5. Relationships with fisheries
As the foraging areas targeted by white-chinned petrels in-
clude several large, highly productive coastal and upwelling
regions, this inevitably results in overlap with fisheries
(Fig. 4). In some cases, for example in the large coastal and
deep sea trawl fisheries off Uruguay, the increased availability
of offal and discards will benefit white-chinned petrels con-
siderably, with negligible associated mortality (Stagi et al.,
1998). Most of the time, however, the relationship is much less
advantageous, and it is possible to a limited extent to esti-
mate the scale of resultant incidental mortality.
In Brazilian waters, the main fisheries involve trawl,
demersal, and pelagic (tuna, shark, and particularly swordfish
Xiphias gladius) longlining (Neves and Olmos, 1998; Olmos
et al., 2000). During the 1990s, Brazilian demersal and pelagic
longliners had bird capture rates, respectively, of 0.10–0.32,
and 0.09–1.35 birds/1000 hooks, and collectively killed an esti-
mated 378–1070 (mean 724), and 2165–4004 (mean 3084)
white-chinned petrels per year (Olmos et al., 2000). Further
south, Argentinian demersal longliners deploy 29 million
hooks annually, and observed bycatch rates for these vessels
between 1999 and 2001 of 0.02–0.26 birds/1000 hooks, equate
to a mean of 230 (maximum 1400) white-chinned petrels
killed per year (Favero et al., 2003). Within the Chilean
Exclusive Economic Zone, although overall bycatch rates of
white-chinned petrels are fairly low (<0.05 birds/1000 hooks),
because of the scale of the effort directed at austral hake
Merluccius australis and Patagonian toothfish Dissostichus
eleginoides by artisanal vessels (>28 million hooks/year since
1999), an average of 1269 ± 783 white-chinned petrels are
killed each year (Moreno et al., 2006). The Chilean industrial
fleet also targets these fish further south, but catch few
white-chinned petrels (although
albatrosses).
Incidental mortality elsewhere is difficult to determine. Off
Uruguay, rates of seabird bycatch in longline fisheries for tuna
Thunnus spp., hake Merluccius hubbsi and rays remain extre-
mely high, albeit having declined in the tuna fishery from
4.7 to 1.7 birds/1000 hooks from 1994 to the late 1990s (Stagi
et al., 1998; Stagi and Vaz-Ferreira, 2000). The eight Uruguayan
longliners operating in national and international waters set
12,000 hooks on average per day (Stagi and Vaz-Ferreira,
2000). Assuming for illustrative purposes that white-chinned
petrels are vulnerable for seven months of the year (equating
to an overall effort of >2 million hooks), that the overall by-
many black-browed
catch rate is 1.7 birds/1000 hooks, and species composition
is the same as in the Brazilian pelagic longline fishery (48%
white-chinned petrels: Olmos et al., 2000), a total of 1630
would be killed every year. There is also a major Taiwanese
fishing effort in the area, in the order of 0.25–1.0 million,
and 1–4 million hooks in each of two 5? grid squares (Fig. 4).
Although there has been no dedicated study on Taiwanese
vessels in the southeast Atlantic, a report by one captain of
the death of 100 birds in one day off Uruguay (Ryan and
Boix-Hinzen, 1998) implies very high bycatch rates by some
ships. Finally, there is extensive illegal fishing activity for tuna
off Uruguay (Tuck et al., 2003), and these vessels are unlikely
to use any mitigation.
If the numbers killed in each of the Uruguayan, Taiwanese
and illegal fisheries is of the same order as the crude estimate
for Uruguay (which is less than that predicted for the Brazil-
ian pelagic fleet), overall annual white-chinned petrel mortal-
ity in fisheries off South America would be a minimum of
10,000 birds. This is actually less than the total number killed
per year by hake and tuna longliners off South Africa during
the 1990s, or by regulated Patagonian toothfish longliners
around Crozet and Kerguelen in recent years (Barnes et al.,
1997; Ryan et al., 2002; CCAMLR, 2004; Delord et al., in press).
Given the likely scale of the problem off South America, it is
clearly imperative that action be taken to regulate fisheries
in overlap areas better. This applies particularly if there is a
strong (male) sex bias in incidental mortality rates (Ryan
and Boix-Hinzen, 1999; Nel et al., 2002c). Overall capture rates
of seabirds have gradually declined in many areas with the
introduction of better observer coverage and improved miti-
gation, and around South Georgia has been virtually elimi-
nated by effective implementation of mitigation measures
and restriction of longline fishing to the winter period (Klaer
and Polacheck, 1997; Nel et al., 2002c; Croxall and Nicol,
2004). However, mitigation includes restriction of line setting
to hours of darkness, which is less effective for white-chinned
petrels that are active by both night and day (Murray et al.,
1993; Weimerskirch et al., 2000; Nel et al., 2002c).
A focus of conservation efforts should therefore be to in-
crease pressure on the South American flag states and Tai-
wan, as well astheRegional
Organisations in the southwest Atlantic to address the prob-
lem of seabird bycatch. Seasonal or area closures are unli-
kely to be economically viable or enforceable given the
scale of the issue, and possibly ineffective as effort may sim-
ply be displaced into adjacent waters. It is therefore prefera-
ble to encourage well-regulated licensing procedures with
properly trained, independent observers on board to pro-
mote and ensure compliance with best practise mitigation
measures, combined with improved policing by fishery pa-
trol vessels and, where appropriate, enhanced catch docu-
mentation schemes, to reduce levels of pirate fishing. One
economic incentive for fishermen would be a reduced rate
of bait loss during line setting, as white-chinned petrels fre-
quently dive for, and then return, baited hooks to the sur-
face (which also increases vulnerability of the surface-
feeding albatrosses). Unless these measures are introduced,
as for so many albatrosses and petrels in the Southern
Ocean, the outlook for the white-chinned petrel is gloomy
indeed.
FisheriesManagement
B I O L O G I C A L C O N S E R VAT I O N 1 2 9 (2 0 06 ) 3 36 –3 4 7
345

Page 11

Acknowledgements
We are very grateful to all fieldworkers (Simon Berrow, Ben
Phalan and Isaac Forster in particular), for assistance with
instrument deployment and retrieval at Bird Island, to Dirk
Briggs for technical support, and James Fox for retrieving
the data from the failed logger. Andrew Fleming kindly
accessed and provided the remote-sensed datasets. Geoff
Tuck, Fabio Olmos and Carlos Moreno kindly provided data
on fishing effort and information on fisheries practices, and
two anonymous referees provided helpful comments on the
manuscript.
R E F E R E N C E S
Afanasyev, V., 2004. A miniature daylight level and activity data
recorder for tracking animals over long periods. Memoirs of
National Institute of Polar Research 58, 227–233 (Special issue).
Anderson, D.J., Huyvaert, K.P., Wood, D.R., Gillikin, C.L., Frost, B.J.,
Mouritsen, H., 2003. At-sea distribution of waved albatrosses
and the Gala ´pagos Marine Reserve. Biological Conservation
110, 367–373.
Barnes, K.N., Ryan, P.G., Boix-Hinzen, C., 1997. The impact of
the hake Merluccius spp. longline fishery off South Africa
on Procellariiform seabirds. Biological Conservation 82,
227–234.
Baum, J.K., Myers, R.A., Kehler, D.G., Worm, B., Harley, S.J.,
Doherty, P.A., 2003. Collapse and conservation of shark
populations in the Northwest Atlantic. Science 299, 389–392.
Berrow, S.D., 2000. The use of acoustics to monitor
burrow-nesting white-chinned petrels Procellaria aequinoctialis
at Bird Island, South Georgia. Polar Biology 23, 575–579.
Berrow, S.D., Croxall, J.P., 1999. The diet of white-chinned petrels
Procellaria aequinoctialis, Linnaeus 1758, in years of contrasting
prey availability at South Georgia. Antarctic Science 11,
283–292.
Berrow, S.D., Croxall, J.P., Grant, S.D., 2000a. Status of
white-chinned petrels Procellaria aequinoctialis Linnaeus 1758,
at Bird Island, South Georgia. Antarctic Science 12, 399–405.
Berrow, S.D., Wood, A.G., Prince, P.A., 2000b. Foraging location and
range of white-chinned petrels Procellaria aequinoctialis
breeding in the South Atlantic. Journal of Avian Biology 31,
303–311.
BirdLife International, 2004a. IUCN Red List of Birds. CD Rom. Vol.
BirdLife International, Cambridge.
BirdLife International, 2004b. Tracking ocean wanderers: the
global distribution of albatrosses and petrels. In: Results from
the Global Procellariiform Tracking Workshop, 1–5 September,
2003, Gordon?s Bay, South Africa. BirdLife International,
Cambridge.
Catard, A., Weimerskirch, H., Cherel, Y., 2000. Exploitation of
distant Antarctic waters and close shelf-break waters by
white-chinned petrels rearing chicks. Marine Ecology Progress
Series 194, 249–261.
CCAMLR, 2004. Report of the twenty-third meeting of the
Scientific Committee of the Commission for the Conservation
of Antarctic Marine Living Resources. Commission for the
Conservation of Antarctic Marine Living Resources, Hobart.
Croxall, J.P., Nicol, S., 2004. Management of Southern Ocean
resources: global forces and future sustainability. Antarctic
Science 16, 569–584.
Croxall, J.P., Silk, J.R.D., Phillips, R.A., Afanasyev, V., Briggs, D.R.,
2005. Global circumnavigations: tracking year-round ranges of
non-breeding albatrosses. Science 307, 249–250.
Cuthbert, R., Hilton, G.D., Ryan, P.G., Tuck, G.N., 2005. At-sea
distribution of breeding Tristan albatrosses Diomedea
dabbenena and potential interactions with pelagic longline
fishing in the South Atlantic Ocean. Biological Conservation
121, 345–355.
Delord, K., Gasco, N., Weimerskirch, H., Barbraud, C., Micol, T., in
press. Seabird mortality in the patagonian toothfish longline
fishery around Crozet and Kerguelen islands, 2001–2003.
CCAMLR Science.
Favero, M., Khatchikian, C.E., Arias, A., Rodriguez, P.S., Can ˜e ´te, G.,
Mariano-Jelicich, R., 2003. Estimates of seabird by-catch along
the Patagonian Shelf by Argentine longline fishing vessels,
1999–2001. Bird Conservation International 13, 273–281.
Gerrodette, T., Forcada, J., 2005. Non-recovery of two spotted and
spinner dolphin populations in the eastern tropical Pacific
Ocean. Marine Ecology Progress Series 291, 1–21.
Hall, A.J., 1987. The breeding biology of the white-chinned petrel
Procellaria aequinoctialis at South Georgia. Journal of Zoology
212, 605–617.
Hill, R.D., 1994. Theory of geolocation by light levels. In: Le Boeuf,
B.J., Laws, R.M. (Eds.), Elephant Seals: Population Ecology,
Behaviour and Physiology. University of California Press,
Berkeley, CA, pp. 227–236.
Huin, N., 1994. Diving depths of white-chinned petrels. Condor 96,
1111–1113.
Jackson, S., 1988. Diets of the white-chinned petrel and sooty
shearwater in the Southern Benguela region, South Africa.
Condor 90, 20–28.
Klaer, N., Polacheck, T., 1997. By-catch of albatrosses and other
seabirds by Japanese longline fishing vessels in the
Australian Fishing Zone from April 1992 to March 1995. Emu
97, 150–167.
Lewison, R.L., Freeman, S.A., Crowder, L.B., 2004. Quantifying the
effects of fisheries on threatened species: the impact of
pelagic longlines on loggerhead and leatherback sea turtles.
Ecology Letters 7, 221–231.
Marchant, S., Higgins, P.J., 1990. Handbook of Australian, New
Zealand and Antarctic birds, vol. 1. Oxford University Press,
Melbourne.
McConnell, B.J., Chambers, C., Fedak, M.A., 1992. Foraging ecology
of southern elephant seals in relation to bathymetry and
productivity of the Southern Ocean. Antarctic Science 4,
393–398.
Moreno, C.A., Arata, J.A., Rubilar, P., Hucke-Gaete, R., Robertson,
G., 2006. Artisanal longline fisheries in southern Chile: lessons
to be learned to avoid incidental seabird mortality. Biological
Conservation 127, 27–36.
Murray, T.E., Bartle, J.A., Kalish, S.R., Taylor, P.R., 1993. Incidental
capture of seabirds by Japanese southern bluefin tuna longline
vessels in New Zealand waters, 1988–1992. Bird Conservation
International 3, 181–210.
Nel, D.C., Nel, J.L., Ryan, P.G., Klages, N.T.W., Wilson, R.P.,
Robertson, G., 2000. Foraging ecology of grey-headed
mollymawks at Marion Island, southern Indian Ocean, in
relation to longline fishing activity. Biological Conservation 96,
219–231.
Nel, D.C., Ryan, P.G., Crawford, R.J.M., Cooper, J., Huyser, O.A.W.,
2002a. Population trends of albatrosses and petrels at
sub-Antarctic Marion Island. Polar Biology 25, 81–89.
Nel, D.C., Ryan, P.G., Nel, J.L., Klages, N.T.W., Wilson, R.P.,
Robertson, G., 2002b. Foraging interactions between
wandering albatrosses Diomedea exulans breeding on Marion
Island and long-line fisheries in the southern Indian Ocean.
Ibis 144, E141–E154 (on-line).
Nel, D.C., Ryan, P.G., Watkins, B.P., 2002c. Seabird mortality
in the Patagonian toothfish longline fishery around the
Prince Edward Islands, 1996–2000. Antarctic Science 14,
151–161.
346
B I O L O G I C A L C O N S E R VAT I O N 1 29 ( 2 00 6 ) 33 6 –3 47

Page 12

Neves, T., Olmos, F., 1998. Albatross mortality in fisheries off the
coast of Brazil. In: Robertson, G., Gales, R. (Eds.), Albatross
Biology and Conservation. Surrey Beatty and Sons, Chipping
Norton, pp. 214–219.
Nicholls, D.G., Robertson, C.J.R., Prince, P.A., Murray, M.D., Walker,
K.J., Elliott, G.P., 2002. Foraging niches of three Diomedea
albatrosses. Marine Ecology Progress Series 231, 269–277.
Olmos, F., 1997. Seabirds attending bottom long-line fishing off
southeastern Brazil. Ibis 139, 685–691.
Olmos, F., Bastos, G.C.C., Neves, T., 2000. Estimating seabird
bycatch in Brazil. Marine Ornithology 28, 141.
Phillips, R.A., Xavier, J.C., Croxall, J.P., 2003. Effects of satellite
transmitters on albatrosses and petrels. Auk 120, 1082–1090.
Phillips, R.A., Silk, J.R.D., Croxall, J.P., Afanasyev, V., Briggs, D.R.,
2004. Accuracy of geolocation estimates for flying seabirds.
Marine Ecology Progress Series 266, 265–272.
Phillips, R.A., Silk, J.R.D., Croxall, J.P., Afanasyev, V., Bennett, V.J.,
2005. Summer distribution and migration of nonbreeding
albatrosses: individual consistencies and implications for
conservation. Ecology 81, 2386–2396.
Pitman, R.L., Ballance, L.T., 1992. Parkinson?s petrel distribution
and foraging ecology in the eastern Pacific: aspects of an
exclusive feeding relationship with dolphins. Condor 94,
825–835.
Ryan, P.G., Boix-Hinzen, C., 1998. Tuna longline fisheries off
southern Africa: the need to limit seabird bycatch. South
African Journal of Science 94, 179–182.
Ryan, P.G., Boix-Hinzen, C., 1999. Consistent male-biased seabird
mortality in the Patagonian toothfish longline fishery. Auk 116,
851–854.
Ryan, P.G., Keith, D.G., Kroese, M., 2002. Seabird bycatch by tuna
longline fisheries off southern Africa, 1998–2000. South
African Journal of Marine Science 24, 103–110.
Sanderson, E.W., Campagna, C., Coppolillo, P.B., Croxall, J.P., Piola,
A.R., Strindberg, S., in press. Translating the landscape species
approach to the sea: an example from the Patagonian Shelf
large marine ecosystem. Environmental Conservation.
Schiavini, A., Frere, E., Gandini, P., Garcı ´a, N., Crespo, E., 1998.
Albatross-fisheries interactions in Patagonian shelf waters. In:
Robertson, G., Gales, R. (Eds.), Albatross Biology and
Conservation. Surrey Beatty and Sons, Chipping Norton, pp.
208–213.
Spear, L., Ainley, D.G., Webb, S.W., 2003. Distribution, abundance
and behaviour of Buller?s, Chatham Island and Salvin?s
albatrosses off Chile and Peru. Ibis 145, 253–269.
Spotila, J.R., Reina, R.R., Steyermark, A.C., Plotkin, P.T., Palasdino,
F.V., 2000. Pacific leatherback turtles face extinction. Nature
405, 529–530.
Stagi, A., Vaz-Ferreira, R., 2000. Seabird mortality in the waters of
the Atlantic Ocean off Uruguay. Marine Ornithology 28, 148.
Stagi, A., Vaz-Ferreira, R., Marı ´n, Y., Joseph, L., 1998. The
conservation of albatrosses in Uruguayan waters. In:
Robertson, G., Gales, R. (Eds.), Albatross Biology and
Conservation. Surrey Beatty and Sons, Chipping Norton, pp.
220–224.
Tuck, G.N., Polacheck, T., Bulman, C.M., 2003. Spatio-temporal
trends of longline fishing effort in the Southern Ocean and
implications for seabird bycatch. Biological Conservation 114,
1–27.
Weichler, T., Garthe, S., Luna-Jorquera, G., Moraga, J., 2004. Seabird
distribution on the Humboldt Current in northern Chile in
relation to hydrography, productivity, and fisheries. ICES
Journal of Marine Science 61, 148–154.
Weimerskirch, H., Wilson, R.P., 2000. Oceanic respite for
wandering albatrosses. Nature 406, 955–956.
Weimerskirch, H., Catard, A., Prince, P.A., Cherel, Y., Croxall, J.P.,
1999. Foraging white-chinned petrels Procellaria aequinoctialis
at risk: from the tropics to Antarctica. Biological Conservation
87, 273–275.
Weimerskirch, H., Capdeville, D., Duhamel, G., 2000. Factors
affecting the number and mortality of seabirds attending
trawlers and long-liners in the Kerguelen area. Polar Biology
23, 236–249.
Wilson, R.P., Ducamp, J.J., Rees, G., Culik, B.M., Niekamp, K., 1992.
Estimation of location: global coverage using light intensity. In:
Priede, I.M., Swift, S.M. (Eds.), Wildlife Telemetry: Remote
Monitoring and Tracking of Animals. Ellis Horward,
Chichester, pp. 131–134.
Woehler,E.J.,1996.ConcurrentdecreasesinfivespeciesofSouthern
Ocean seabirds in Prydz Bay. Polar Biology 16, 379–382.
B I O L O G I C A L C O N S E R VAT I O N 1 2 9 (2 0 06 ) 3 36 –3 4 7
347