The Journal of Wildlife Management; DOI: 10.1002/jwmg.21284

Research Article

Declining Survival of Black Brant From
Subarctic and Arctic Breeding Areas
ALAN G. LEACH,1 Department of Natural Resources and Environmental Science, University of Nevada, Reno, 1664 North Virginia Street, Reno,
NV 89557, USA
DAVID H. WARD, U.S. Geological Survey, Alaska Science Center, 4210 University Drive, Anchorage, AK 99508, USA
JAMES S. SEDINGER, Department of Natural Resources and Environmental Science, University of Nevada, Reno, 1664 North Virginia Street,
Reno, NV 89557, USA
MARK S. LINDBERG, Institute of Arctic Biology and Department of Biology and Wildlife, University of Alaska Fairbanks, Fairbanks, AK 99775, USA
W. SEAN BOYD, Science and Technology Branch, Environment and Climate Change Canada, Pacific Wildlife Research Centre, RR1-5421 Robertson
Road, Delta, BC V4K 3N2, Canada
JERRY W. HUPP, U.S. Geological Survey, Alaska Science Center, 4210 University Drive, Anchorage, AK 99508, USA
ROBERT J. RITCHIE, ABR, Environmental Research & Services, P. O. Box 80410, Fairbanks, AK 99708, USA

ABSTRACT Since the mid 1990s, the number of black brant (Branta bernicla nigricans; brant) nests on the

Yukon-Kuskokwim Delta (YKD), Alaska, USA, the historically predominant breeding area of brant, has
declined steadily. This has caused researchers and managers to question if arctic breeding populations can
compensate for the reduction in brant nests on the YKD. An important component of the assessment of brant
population dynamics is having current estimates of first-year and adult survival. We banded brant at 4
locations in Arctic Alaska and western Canada, and at 1 location in the subarctic, the Tutakoke River (TR)
colony on the YKD, 1990–2015. We used joint live and dead mark-recapture models to estimate first-year
and adult ( 1 yr old) survival of brant. We also used band recovery rates from a Brownie model to assess
temporal trends in band recovery rates of adult brant. First-year survival of brant hatched at TR declined from
approximately 0.60 to <0.20 and, although first-year survival generally was higher for goslings marked in the
Arctic, their survival declined from approximately 0.70 in the early 1990s to  0.45 in the 2010s. Annual
survival of adult females decreased from an average of 0.881 (95% CI ¼ 0.877–0.885) to 0.822 (95%
CI ¼ 0.815–0.829) at TR and from 0.851 (95% CI ¼ 0.843–0.860) to 0.821 (95% CI ¼ 0.805–0.836) in the
Arctic, from 1990 to 2014. Band recovery rates of adults generally were <1.25% until the last several years of
study, when they reached  3.5%. Although the current harvest rates may be partially additive to natural
mortality, we do not believe that harvest is the main influence on the declines in survival. The general decline
in survival rates of brant breeding across a large geographic area may be influenced by a reduction in the
quality of migration and wintering ground habitats. We suggest an analysis of seasonal survival of brant to test
the hypothesis that declining habitat quality on wintering or spring migration areas is reducing survival. Our
results suggest that the number of breeding pairs at TR will continue to decline and also brings into question
the ability of arctic breeding populations to grow at a rate necessary to offset the declines on the YKD.
Researchers should continue to closely monitor survival and harvest rates of brant, and assess methods
currently used to monitor their abundance. Ó 2017 The Wildlife Society.
KEY WORDS band recovery rates, Branta bernicla nigricans, Brownie models, joint live and dead mark-recapture,
population dynamics, Seber reporting rate.

The breeding range of black brant (Branta bernicla nigricans;
brant) extends from northeastern Russia east to the North
Slope of Alaska and the western Canadian Arctic, and south
to the Yukon-Kuskokwim Delta (YKD) of southwestern
Alaska (Sedinger et al. 1993, Ward et al. 2005, Lewis et al.
2013). The YKD was the primary brant production area as

Received: 12 November 2016; Accepted: 17 April 2017
1

E-mail: aleach@cabnr.unr.edu

Leach et al.

 

Long-Term Survival of Brant

recently as the early 1990s, supporting >75% of the overall
breeding population (Sedinger et al. 1993). Since the mid
1990s, the number of brant nests at the 4 major brant
breeding colonies and low-density breeding congregations of
the YKD have declined by approximately 34% (Stehn et al.
2011, Wilson 2014). Available data indicate that the
numbers of breeding brant may be increasing on some
breeding areas of the North Slope of Alaska; however, to
date, the number of brant nests documented in the Arctic
(Ritchie et al. 2015) does not appear adequate to offset losses
on the YKD. Management decisions for brant populations
1

 are currently based on estimates derived from the mid-winter
survey, which after a long-term decline, have increased
substantially since approximately 2005 (Olson 2015).
Negative trends in nest counts on the YKD and current
suggestions that the numbers of brant nests at monitored
sites in the Arctic are inadequate to offset these declines seem
to conflict with increasing mid-winter counts and have
created uncertainty about the abundance of black brant.
In addition to accurate abundance estimates, having current
demographic information on populations is important for
understanding population trajectory (Koons et al. 2007).
Data on vital rates from brant populations breeding in the
Arctic are relatively limited and survival has not been
estimated since the 1990s (Sedinger et al. 2002, Ward et al.
2004, Hines and Brook 2008). Likewise, demographic rates
of brant breeding on the YKD have not been assessed since
2007 (Sedinger and Nicolai 2011). Although progress has
been made in demographic estimation and modeling in
recent years, unbiased, precise estimation of population vital
rates often are difficult to obtain for animals with long
distance migratory behaviors and wide ranging breeding
concentrations (Griffith et al. 2016). Several past assessments
of long-term annual survival of brant have used band
recovery only models (Brownie parameterization; Sedinger
et al. 2007, Hines and Brook 2008, Sedinger and Nicolai
2011). However, band recovery rates of brant were low in the
1990s and early 2000s (<1.5%; Sedinger et al. 2007, Hines
and Brook 2008, Sedinger and Nicolai 2011), often leading
to imprecise estimates and an inability to estimate annual
variation in survival (Sedinger and Nicolai 2011). Because
brant are gregarious and congregate in only a few coastal
lagoons during the non-breeding season (Pacific Flyway
Council 2002; Ward et al. 2005, 2009), researchers have been
able to collect large numbers of live encounters of colormarked brant throughout the annual cycle (Ward et al. 2004,
Sedinger et al. 2011). This provides the opportunity to
compile live encounter data collected at breeding, wintering,
and migration areas with band recovery data from hunters to
improve the precision of survival estimates by using joint live
and dead mark-recapture models (Barker 1997).
Despite the relatively poor precision of survival estimates
from Brownie band recovery models in the case of brant,
these models provide estimates of band recovery rate
(Brownie et al. 1985). If band reporting rates (i.e., the
proportion of banded birds harvested by hunters that are
reported) can also be estimated (Zimmerman et al. 2009),
then band recovery rates provide an estimate of harvest rate
(Sedinger et al. 2007). In contrast, joint live and dead markrecapture models implemented in Program MARK use the
Seber parameterization of band reporting rates (i.e.,
proportion of banded individuals that died and were
subsequently reported during an interval [r]; Seber 1970).
Although, Brownie band recovery rates can be calculated
from Seber’s r (ft ¼ rt [1   St] where ft ¼ Brownie band
recovery rate and St ¼ survival), confounding near the end of
our study period for joint live-dead models made Brownie
parameterization more informative for our questions about
harvest.
2

We had 2 objectives: estimate true survival of goslings and
adult brant marked at 1 colony on the YKD and 4 breeding
areas from the Arctic, 1990–2015, and assess temporal
variation of band recovery rates and proportion of annual
mortality attributable to harvest. We expected that first-year
survival of brant from the YKD would be low compared to
the 1990s (Sedinger and Nicolai 2011). We reasoned that if
survival rates of brant in the Arctic remained high during our
study, this may indicate that long-term declines in survival of
goslings hatched on the YKD could be influenced solely by
local factors (e.g., forage quantity during brood-rearing
[Sedinger et al. 2001a]). Alternatively, declines in survival
rates of brant across a large portion of their range may
indicate common factors could be affecting demographics of
the black brant population at large, such as degradation of
migration habitat.

STUDY AREA
We marked brant at the subarctic Tutakoke River colony
(TR) on the YKD (Fig. 1), a colony that once represented
20% of the entire black brant breeding population (Sedinger
et al. 1993). The TR colony consists of nesting meadows
primarily composed of Ramensk’s sedge (Carex ramenskii)
and grazing lawns primarily composed of Hoppner’s sedge
(C. subspathacea) generally located <1 km from, but
occasionally inundated by, high tides from the Bering Sea
(Lindberg et al. 1997, Lindberg and Sedinger 1998). We
have continuously marked and collected reencounter data on
color-marked brant breeding at TR since 1986 (Sedinger
et al. 2008, Sedinger and Nicolai 2011).
Our sample of marked brant from Arctic areas came from 4 of
the largest known nesting areas, including the Colville River
Delta and numerous small colonies throughout the Kuparuk
oil field to the Sagavanirtok River area (Stickney and Ritchie
1996) on the North Slope of Alaska, and the Tuktoyaktuk
Peninsula-Liverpool Bay region of the mainland and on Banks

Figure 1. Locations of bandings and summer encounters (circles), and nonbreeding season resighting locations (triangles) of black brant, 1990–2015.
The Journal of Wildlife Management

 

9999()

 Island and Victoria Island, Northwest Territories, Canada
(Fig. 1; Sedinger et al. 1993, Cotter and Hines 2001, Pacific
Flyway Council 2002, Hines and Brook 2008).
In Arctic Alaska and western Canada, brant generally nest
in small aggregations (tens of nests) within 2 km of the coast
on islands of river deltas, lakes, or small ponds, but some
nests may occur up to 27 km inland (Ritchie et al. 2000,
Cotter and Hines 2001). The nesting islands are typically low
lying and those close to the coast (e.g., outer deltas) may be
tidally influenced. Habitat of nesting sites is dominated by
 1 of the following communities: grass (Gramineae), sedge
(Carex spp.), dwarf birch (Betula nana), and dwarf willow
(Salix spp.; Armstrong 1996, Hines and Brook 2008).
The wintering range of brant extends coastally from
southwest Alaska to northwest Mexico where they use
embayments containing their preferred food, eelgrass
(Zostera marina; Ward et al. 2005, 2009). During our study,
most ( 70%) of the brant population wintered in Mexico
primarily within 3 lagoons of the central Pacific coast of Baja
California ( 60%) and to a lesser extent (10%) along the
Gulf of California coasts of Sonora and Sinaloa (E. J.
Mallek, U.S. Fish and Wildlife Service, unpublished data;
https://www.fws.gov/alaska/mbsp/mbm/waterfowl/reports.
htm, accessed 26 Jul 2016). The remainder of the wintering
population outside of Mexico (30%) was distributed among
Alaska (15%), Washington and British Columbia (5%), and
California (10%; Olson 2015). Generally, brant from the
YKD winter in Mexico and more southerly winter locations,
whereas a somewhat larger proportion of birds from the
Arctic may winter in Alaska and more northerly locations
(Ward et al. 2009). In recent years, a large portion of the
brant population has undergone a northward shift in
wintering distribution, resulting in a large number of brant
wintering in Izembek Lagoon, Alaska (Ward et al. 2009).
We recorded live resights at 3 wintering lagoons along the
central Pacific coast of Baja California. We made observations
at Bahia San Quintin during winters 1991–1993, 1997–2006,
and 2011–2013 and at Lagunas Ojo de Liebre and San
Ignacio, during winters 1999–2001 (Ward et al. 1997,
Lindberg et al. 2007, Sedinger et al. 2011). From spring
1991–2015, we recorded resights of brant at wintering and
migration areas in southeastern Vancouver Island and
Boundary Bay, British Columbia, Canada (Reed et al.,
1998a,b; Smith et al. 2012). We also periodically collected live
resights of brant during migration and wintering at Izembek
Lagoon, Alaska (Ward et al. 2004, 2009); Humboldt (Lee
et al. 2007) and Morro (Lindberg et al. 2007) bays, California;
and Puget Sound, Washington, USA (Fig. 1).

METHODS
Field Methods
We captured brant by herding them into pens during the adult
remigial molt between 14 July and 11 August, 1986–2015; we
also captured 34 brant at their nest in June and early July
(Sedinger et al. 1997, Hupp et al. 2017). Following capture, we
marked brant with a unique metal band issued by the Bird
Banding Lab, United States Geological Survey, and a uniquely
Leach et al.

 

Long-Term Survival of Brant

engraved plastic tarsal band, and recorded bands from
recaptures. We sexed brant using cloacal examination and
aged them as goslings (local), second-year, or after secondyear using plumage characteristics (Bellrose 1980, Owen
1980). For analysis, we combined second-year and after
second-year brant into 1 adult group to reduce model
complexity and acknowledge previous results indicating
survival is similar between these age classes (Sedinger et al.
2007, Sedinger and Nicolai 2011). We marked most brant
with corrosion-resistant bands (i.e., Monel or stainless steel);
however, 620 and 8 brant marked in Canada (1995–2000) and
TR (1986), respectively, were released with aluminum bands.
We believe inclusion of these aluminum bands, which are
more likely to be lost, resulted in no bias in survival estimates
because they were a small proportion of our sample and most
encounters were collected through live resights of the color
bands (Table S1, available online in Supporting Information).
Because arctic banding drives can include molting flocks of
failed breeders that may nest outside the Arctic (Bollinger
and Derksen 1996), we limited our sample of adults to those
captured in a brood drive ( 1 gosling captured) to reduce the
number of failed and non-breeding birds from elsewhere.
Banding efforts on the North Slope of Alaska occurred from
1990–2000 and 2011–2015, and banding efforts in western
Canada occurred from 1990 to 1999 (Table S1).
Prior to banding each summer, we recorded live resights of
color-marked brant on breeding areas using 20–50  spotting
scopes (Sedinger et al. 2008). We included all summer
encounters in our analysis (not just breeding encounters) to
improve the precision of our survival estimates. We obtained
band recovery information of brant harvested (i.e., how
obtained code ¼ 01) from August 1990 to May 2016 from the
United States Geological Survey Bird Banding Laboratory in
Patuxent, Maryland, USA. Our recovery window for each year
was 1 August–31 July. We included only brant that were colormarked in our analysis. All handling of animals was approved
by Institutional Animal Care and Use committees at the
United States Geological Survey-Alaska Science Center,
University of Alaska Fairbanks and the University of Nevada,
Reno (most recent protocol number 00056).
Joint Live-Dead Recovery Models
We used the joint live and dead encounters model (Barker
1997) in Program MARK (White and Burnham 1999) with
the primary goal of estimating survival and Seber reporting
rates (i.e., the probability of reported mortality; Alisauskas et al.
2012) of gosling and adult brant (Barker 1997). We sorted
individuals into 2 groups: those marked at subarctic and those
marked in arctic locations (Lindberg et al. 1998). We included
2 age classes in our analysis: gosling and adult brant, and
allowed goslings to graduate to adults after their second
summer. Though most individual capture histories began with
initial banding during summer, we included 2,953 brant
marked at TR before 1990 whose capture histories began with
their first live summer reencounter during or after 1990.
The joint live-dead mark recapture model estimates a large
number of parameters including true survival probability from
time t to t þ 1 (St), probability of encounter at a breeding area
3

 given alive at t (pt), probability a mortality between t and t þ 1
is reported (rt), probability of being resighted given alive
between t and t þ 1 (Rt), probability that an individual was
resighted before death between t and t þ 1 (R0 t), probability of
being available for encounter during summer given encounter
in year t   1 (Ft), and probability of being available for encounter
during summer given not being encountered in year t   1 (F 0 t).
We fixed F 0 ¼ 0.0 so that F approximated 1   permanent
emigration, although we do not report these estimates because
we included summer encounters of individuals away from
their colony of initial marking and long-term encounter data
was not available for individual sites in the Arctic.
The number of parameters estimated by the Barker model;
the fact that we had 2 age classes, 2 sexes, and 2 banding
locations; our large sample (55,912 individuals); and longterm encounter effort (26 years) made running all possible
models computationally prohibitive (Townsend et al. 2016).
Fortunately, there was a wealth of previous demographic
analyses that we could reference to create informed a priori
models of area- and sex-specific variation in permanent
emigration (Lindberg et al. 1998, Sedinger et al. 2008),
summer detection rates (Sedinger et al. 2001b), and survival
(Ward et al. 1997, 2004; Sedinger et al. 2002, 2007; Hines
and Brook 2008; Sedinger and Nicolai 2011).
We approached modeling as a 2 step process: first, using our
most general structure for S, which included a full interaction
among breeding areas, age classes, and year, we found the
optimal structure (i.e., converged and produced the lowest
corrected Akaike’s Information Criterion [AICc] score) for p, r,
R, R0 , and F; and second, we kept that optimal structure and
built a small set of models to explore variation in survival
(Table 1). Because brant do not start breeding until 2 years of
age, we built models testing for differences in p between
second-year and after-second-year brant (Sedinger et al.
2001b). For models that included full annual variation in
survival, we had to constrain the last 2 years to be the same;
likewise, we constrained the last 3 probabilities of reported
mortality (r), to achieve model convergence. We marked few
goslings (<40) at TR in 2001 and 2003 because of poor
production caused by egg predation by arctic foxes (Vulpes
lagopus), which resulted in poor precision of estimates for these
cohorts; thus, we constrained first-year survival of the
2001–2002 and 2003–2004 TR cohorts to be the same. We
used AICc corrected for overdispersion (^c ¼ 1.11; ^c was
obtained by Sedinger et al. [2007] through the median ^c
approach) to evaluate support among models with different
structures on age class, breeding area, and annual variation
(including trends) on survival (Burnham and Anderson 2002).
We built all models using the program R (R Core Team 2014)
package RMark (Laake 2013) and we used the high
performance computing system at the University of Nevada,
Reno, to run all models. We did not release brant from arctic
breeding areas from summer 2001–2010; thus, we fixed all
parameters for these cohorts to 0.0.
Brownie Models
We used Brownie dead recovery models implemented with
RMark (Laake 2013) to estimate band recovery rates. We
4

Table 1. Model selection results for joint live-dead (Barker 1997) markrecapture models used to explain variation in first-year and adult survival
(S) of black brant released from arctic breeding areas (north slope of Alaska
and western Canadian Arctic) and the Tutakoke River colony,
southwestern Alaska, USA, 1990–2015. We report the number of model
parameters (K), Akaike’s Information Criterion corrected for sample size
and overdispersion (QAICc; c^ ¼ 1.11) of model i compared with the best
model (DQAICc), the weight of each model in the candidate set (Akaike
weight; wi), and Quasi Deviance (QDev) of each model.
Survival structure
S(area   age   yr þ sex)
S(area   age þ yr þ sex)
S(area   age   trend þ sex)
S(area   age þ trend þ sex)
S(area   age þ sex)
S(area þ age þ sex)
S(.)

K

DQAICc

wi

QDev

209
141
121
118
117
116
113

0.00
540.00
762.39
1,133.86
1,701.22
1,918.67
5,627.48

1.00
0.00
0.00
0.00
0.00
0.00
0.00

127,054.1
127,730.5
127,993.0
128,370.5
128,939.8
129,159.3
132,874.1

Variables: age ¼ represents 2 age classes (gosling and adult); area ¼ arctic
breeding areas and Tutakoke River colony, yr ¼ full annual variation,
trend ¼ fitted annual trend. All models included the same structure for
summer encounter probability (p), probability of reported mortality (r),
resighting probability during winter or spring (R), and breeding area
fidelity (F). Probability of mortality after being resighted alive during
winter or spring (R0 ) and probability of being available for encounter during
summer given not being available for encounter the previous year (F 0 ) were
fixed to 0.0.

limited this analysis to only individuals released as adults
because previous analyses reported little variation in recovery
rates between age classes (Sedinger et al. 2007, Sedinger and
Nicolai 2011). We fixed S and band recovery rate (f ) to 0.0
for the 2001–2010 cohorts of our arctic group because we did
not release individuals from this area during this period. We
compared models with variation between breeding areas in
survival and band recoveries and included models with full
annual variation and annual trends in our candidate set
(Table 2). If there was support for >1 model (i.e., DQAICc
<2.0) we model averaged estimates over the full model set.

RESULTS
Joint Live-Dead Recovery Models
Our analysis included 22,509 brant released as goslings and
18,772 brant released as adults from TR and 7,508 goslings
and 7,123 adult brant from the Arctic. In step 1 of the
modeling process, our best model included a summer
encounter rate (p) structure with additive effects of age class
and banding area, and an interaction between banding area
and full annual variation (Table S2, available online in
Supporting Information); an interaction between age class
and banding area, plus the additive effect of year on
probability of reported mortality (r); additive effects of
banding area and year on R; and an interaction between
banding area, age class, and sex on F. Models that included
age, banding area, or annual variation on R0 failed to
converge or did not improve model support; therefore, we
fixed this parameter to 0.0 (Nicolai et al. 2014).
In the second step of our modeling process, the model
structure of S with a full interaction between banding area, age
class, and year, plus an additive effect of sex on survival was the
only model with support (wi ¼ 1.00; Table 1). Males generally
b S (males) ¼  0.13; 95% CI ¼  0.16 to
had lower survival (b
The Journal of Wildlife Management

 

9999()

 Table 2. Model selection results for Brownie dead recovery models used to
explain variation in true survival (S) and band recovery rates (f ) of adult
black brant released from arctic breeding areas (north slope of Alaska and
western Canadian Arctic) and the Tutakoke River colony, southwestern
Alaska, USA, from 1990–2015. We report the number of model
parameters (K), Akaike’s Information Criterion corrected for sample size
and overdispersion (QAICc; c^ ¼ 1.11) of model i compared with the best
model (DQAICc), the Akaike weight (wi), and Quasi Deviance (QDev) of
each model.
Model structure

K

DQAICc

wi

QDev

S(trend) f (area þ yr)
S(.) f (area þ yr)
S(area) f (area þ yr)
S(area) f (yr)
S(trend) f (yr)
S(.) f (yr)
S(trend) f (area   yr)
S(.) f (area   yr)
S(area) f (area   yr)
S(trend) f (.)
S(area) f (.)
S(.) f (.)

29
28
29
28
28
27
54
53
54
3
3
2

0.00
3.68
5.09
7.87
15.68
16.25
23.00
26.68
26.92
228.40
263.86
279.73

0.80
0.13
0.06
0.02
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00

1,077.1
1,082.8
1,082.2
1,086.9
1,094.8
1,097.3
1,049.9
1,055.6
1,053.8
1,357.5
1,393.0
1,410.9

Variables: area ¼ arctic breeding areas and Tutakoke River colony, yr ¼ full
annual variation, trend ¼ fitted annual trend.

 0.10) than females; however, because year was modeled
additively to sex (i.e., annual patterns were similar) we only
report estimates for female survival. Estimates of first-year
survival of female goslings from TR were high ( 0.60) in the
early 1990s and have since declined substantially, with 4 of the
5 gosling cohorts released at the end of our study (2009–2013)
having first-year survival rates <0.20 (Fig. 2; Table S3,
available online in Supporting Information). Female goslings
marked at arctic breeding colonies generally had higher firstyear survival than those marked at TR; however, their survival
declined in the late 1990s, and by the end of the study was
 0.45, down from approximately 0.70 in the early 1990s
(Fig. 2).
Parameter estimates from our model with an annual trend on
survival indicated that survival of adult females declined from
0.881 (95% CI ¼ 0.877–0.885) to 0.822 (95% CI ¼ 0.815–
b ¼  0.021; 95% CI ¼  0.024 to  0.017) and
0.829) at TR (b
from 0.851 (95% CI ¼ 0.843–0.860) to 0.821 (95% CI
b ¼  0.010; 95% CI ¼  0.016
¼ 0.805–0.836) in the Arctic (b
to  0.003) from 1990 to 2014. Estimates from our model with
full annual variation indicated that much of the decline in adult
survival occurred since 2008 (Fig. 2). For female goslings,
parameter estimates from our model with an annual trend
indicated that first-year survival declined from 0.624 (95%
CI ¼ 0.610–0.638) to 0.188 (95% CI ¼ 0.171–0.205) at TR
b ¼  0.086; 95% CI ¼  0.092 to  0.079) and from 0.717
(b
(95% CI ¼ 0.691–0.742) to 0.404 (95% CI ¼ 0.344–0.467) in
b ¼  0.057; 95% CI ¼  0.072 to  0.043) from
the Arctic (b
1990 to 2014.
Probability of reported mortality was greater for adults than
goslings and generally increased during our study period
(Fig. 3). Live resighting rates during the non-breeding
b ¼  0.099; 95%
season were lower for brant marked at TR (b
CI ¼  0.136 to  0.062) than for those from the Arctic, and
generally declined during the study period (Fig. 4).
Leach et al.

 

Long-Term Survival of Brant

Figure 2. Estimates of gosling (first-year) and adult survival ( 95% CI;
only estimates for females shown) of black brant banded at the Tutakoke
River colony in southwestern Alaska, USA, and arctic breeding areas,
1990–2014. Estimates are from a joint live and dead (Barker 1997) markrecapture model, which included an interaction between age class, banding
area, and year on survival.

Brownie Models
This analysis included 15,819 after-second-year brant
banded at TR and 7,123 after-second-year brant banded
at Arctic breeding areas, 1990–2015. Our best supported
model included an annual trend on survival and the additive
effects of banding area and full annual variation on recovery
rates (Table 2). Parameter estimates from our best performing model indicated adult survival declined during our study
b ¼  0.023; 95% CI ¼  0.042 to  0.004), similar
period (b
to results from our joint live and dead model.
Brant banded at TR had band recovery rates that were
b f (TR) ¼ 0.247; 95% CI ¼ 0.130–0.365) than those
higher (b
banded in the Arctic (Fig. 5). Except for several years in the
late 1990s, band recovery rates of brant were generally
<0.0125; however, since 2009 band recovery rates have
generally been >0.0125 (Fig. 5). Fall-spring 2014–2015 had
the highest recovery rates of the study, varying from 0.035
(95% CI ¼ 0.029–0.043) for adults banded at TR to 0.028
(95% CI ¼ 0.023–0.034) for adults banded in the Arctic
(Fig. 5).

Figure 3. Estimates of the probability of reported mortality ( 95% CI) of
gosling and adult black brant from the Tutakoke River colony and arctic
breeding areas from 1990–2014. Estimates are from a joint live and dead
(Barker 1997) mark-recapture model, which included an interaction
between age class and banding area and an additive effect of time on
probability of reported mortality.
5

 Figure 4. Estimates of live resighting rates during the non-breeding season
( 95% CI) for black brant from the Tutakoke River colony and arctic
breeding areas from 1990–2015. Estimates are from a joint live and dead
(Barker 1997) mark-recapture model, which included the additive effects of
banding area, and year on resighting rates.

DISCUSSION
First-year and adult survival of brant banded across their
breeding range generally declined between 1990 and 2014.
For birds banded at TR, the decline in first-year survival we
report is consistent with previous work (Sedinger and Nicolai
2011). However, the decline in adult female survival (from
0.88 to 0.82), estimated from our joint live-dead recovery
model with an annual trend, is previously unreported. Even
small declines in adult survival are concerning, because
growth rates of goose populations are most sensitive to
changes in this vital rate (Cooch et al. 2001). Our trends in
adult survival differed from Sedinger and Nicolai (2011),
who reported weak support for a positive trend in adult
survival at TR from 1986 to 2007; however, we do not believe
our results are contradictory because much of the decline in
adult survival we report appears to have occurred since 2008
(Fig. 2). Because of the previously high survival of breeders at
TR, researchers have pointed solely to low recruitment rates
as the primary influence on declining colony size (Sedinger
et al. 2007, 2016; Sedinger and Nicolai 2011). The current
rates of first-year survival suggest that recruitment will
remain below replacement levels at TR and because of the

Figure 5. Estimates of band recovery rates ( 95% CI) for adult black brant
from the Tutakoke River colony and arctic breeding areas, 1990–2016.
Estimates are from a Brownie model, which included an annual trend on
survival and an additive effect of banding area and year on band recovery
rates.
6

decline in adult survival, recruitment will likely lag further
behind replacement levels than previously reported (Sedinger
et al. 2016). Given the declines in first-year and adult
survival, the number of breeding pairs at TR will likely
continue to decline.
Our finding of declining first-year and adult survival of
brant banded in the Arctic was previously unreported in the
literature. Our results indicated that average survival of adult
female brant banded in the Arctic declined from 0.85 to 0.82
from 1990 to 2014. Our average estimate of survival for adult
females from the Arctic in the 1990s (0.85) is lower than
previous estimates from this period (0.88–0.94; Sedinger
et al. 2002, Hines and Brook 2008). This difference in
survival could result from the fact we combined adults
marked throughout the Arctic into 1 group, including
previously unanalyzed data from brant marked on the North
Slope of Alaska in the mid to late 1990s, and excluded many
failed or non-breeding adults. Our average estimates of firstyear survival of female brant banded in the Arctic from 1990
to 1996 (i.e., prior to the decline after 1997) were 0.68, which
is similar to the estimates of Hines and Brook (2008) who
reported first-year survival of 0.69 from brant marked in the
western Arctic of Canada from 1991 to 1999. Additionally,
our results support the conclusion of Ward et al. (2004) that
first-year survival of brant goslings from the Arctic is
generally greater than those from the YKD (Fig. 2). This
higher survival is likely related to the superior growth rates of
goslings in the Arctic compared to the YKD (Sedinger et al.
2001a; Hupp et al. 2017). The declining first-year and adult
survival for brant banded in the Arctic raises questions about
whether populations nesting in the Arctic can offset losses in
the nesting population on the YKD. However, estimates of
additional vital rates (e.g., nest success, breeding probability,
recruitment) from arctic populations are needed to more fully
assess the health of these populations. Along with these vital
rates, an accurate estimate of the number of breeding pairs in
the Arctic would aid our understanding of the productivity
potential and contribution of these northern populations to
the overall brant population.
The recent increase in band recovery rates of brant was an
unexpected result because previous studies indicated that
harvest rates of brant were at historically low levels in the
1990s (Sedinger et al. 2007, Hines and Brook 2008, Sedinger
and Nicolai 2011). When adjusted for the band reporting
rate of harvested North American geese (0.73; Zimmerman
et al. 2009) and averaged across banding regions, the average
harvest rate of brant increased from 0.9% for the 2000–2002
hunting seasons to 3.3% for the 2013–2015 hunting seasons.
The probabilities of reported mortality (Fig. 3) indicated that
the proportion of adult mortality accounted for by harvest
doubled from 6.1% for the 2000–2002 hunting seasons to
12.8% for the 2011–2013 hunting seasons. The current
harvest rates of brant are still lower than the maximal
recorded harvest rate of approximately 10% for adults during
the 1950s–1960s (Sedinger et al. 2007, Hines and Brook
2008), but could be high enough to be at least partially
additive to natural mortality (Sedinger et al. 2007). We
acknowledge that our analysis of band recovery rates did not
The Journal of Wildlife Management

 

9999()

 account for whether bands had toll-free numbers inscribed or
were address only (mostly applied prior to 1996). Regardless,
the increases in band recovery rates we report occurred more
than a decade after toll-free bands became commonplace;
thus, the increases in band recoveries we observed were likely
influenced primarily by increased harvest rates, not increased
reporting rates. It is unclear if these higher harvest rates
resulted from an increase in the total brant harvest, or
relatively stable harvest and a declining population. Annual
estimates of brant harvest based upon the Migratory Bird
Harvest Information Program (HIP) are highly variable
(Olson 2015). Therefore, accurate estimates of brant harvest
(including subsistence harvest) are needed to fully understand the role of harvest on brant population dynamics.
We interpret the low probability of reported mortality of
goslings (Fig. 3), especially those marked at TR, as an
indication that non-harvest mortality is greater for goslings
than adults, similar to lesser snow geese (Chen caerulescens
caerulescens; Francis et al. 1992). The lower probability of
reported mortality of goslings is likely related to their higher
mortality than adults prior to arriving at fall staging or
wintering areas (Ward et al. 2004), and because of this a
smaller proportion of goslings are available to hunters for
harvest. Given the magnitude and long-term nature of the
decline in first-year survival, the recent increase in harvest
rate may, at most, explain only a small proportion of the
decline in first-year survival.
The range-wide reduction in first-year and adult survival we
reported in this study suggests common factors (i.e.,
degradation of the quality of wintering and migration habitat)
may be affecting demographics of brant. For example, Bahia
San Quintin, Mexico, a previously high-quality habitat for
brant, based on the breeding probability of birds wintering
there (Sedinger et al. 2011), experienced long-term declines in
eelgrass abundance (Ward et al. 2003). Additionally, stopover
times of brant at Vancouver Island, British Columbia, have
declined, which could indicate reduced profitability of this site
resulting from reduced forage or increased disturbance from
humans and bald eagles (Haliaeetus leucocephalus; Smith et al.
2012). Black brant are one of the few nearctic goose species that
rarely forage in agricultural fields, and are especially vulnerable
to degradation of wintering and migration embayments
(Cottam et al. 1944, Ward et al. 2005). Finally, the effects of
the recent increase in the number of brant wintering in
Izembek Lagoon on brant demographics and the amount of
eelgrass available to brant during spring migration, remains
unknown.
Complementing our work here, analyses conducted on the
long-term trends in age ratios of brant staging at Izembek
Lagoon, Alaska, where the entire black brant population
stages in fall, reported a decline in the proportion of juveniles
in recent years (D. H. Ward, U.S. Geological Survey,
unpublished data). Declining age ratios, combined with our
finding of a decline in survival rates of brant, suggest that the
overall brant population could be declining. In contrast, the
mid-winter surveys, which are flown from Mexico to Alaska
and are used to make decisions about the management of
brant populations, showed a decline in brant numbers prior
Leach et al.

 

Long-Term Survival of Brant

to 2005 but have been stable or increasing since (Olson
2015). It is imperative that the discrepancy between trends in
demographic estimates and survey data be reconciled.

MANAGEMENT IMPLICATIONS
Because of the positive trend in counts of brant estimated
from mid-winter inventories (Olson 2015), brant have not
recently been a species of management concern. However,
the widespread declines in survival we report here, combined
with declines in the number of brant nests on the YKD
(Stehn et al. 2011), and recent declines in fall age ratios
(D. H. Ward, U.S. Geological Survey, unpublished data)
suggest that brant may not be as abundant as the winter
population assessment indicates. We suggest an independent, retrospective analysis of brant population trends, for
instance with Lincoln’s estimator (Alisauskas et al. 2011,
2014) or mark-resight estimators (Sanders and Trost 2013).
Continued banding of brant on the YKD, in northern
Alaska, and at least periodic banding in the western
Canadian Arctic is important to long-term brant population
monitoring. Continued generation of precise survival
estimates for brant (and potentially mark-resight estimates
of population size) will require ongoing efforts to collect live
resights during the non-breeding period (Kendall et al. 2013,
Penaloza et al. 2014); therefore, we encourage surveys to
obtain resightings of color-marked brant at wintering and
migration locations, such as Izembek Lagoon, Alaska;
Parksville-Qualicum and Boundary Bay, British Columbia;
Humboldt Bay, California; and Bahia San Quintin, Mexico.
Researchers should also consider assessing long-term trends
in eelgrass abundance on key wintering and migration areas
and estimate seasonal survival of brant since the early 1990s
to test the hypothesis that a decline in the quantity and
quality of wintering or migration habitat may be contributing
to the decline in brant survival.

ACKNOWLEDGMENTS
Any use of trade or product names is for descriptive purposes
only and does not constitute endorsement by the U.S.
Government. We are especially thankful to J. Hines, E. Reed,
and M. Robertson for sharing their banding and recapture
records for western Canada. We are grateful to the numerous
field assistants and band resighters, and R. and D. Wheeler and
staffs at Exportadora de Sal, Kuyima, Universidad Aut 
onoma
de Baja California-Instituto de Investigaciones Oceanol 
ogicas, and San Diego, Humboldt, and Izembek National
Wildlife Refuges for providing logistical support and
assistance. We thank R. Swanston for collecting band resights
in Boundary Bay, British Columbia. We also thank
Washington and Oregon Departments of Fish and Game
for providing support of the resighting of brant bands. P.
Martin led efforts west of Milne Point. M. Irinaga, L.
Gullingsrud, and staff from the Yukon Delta National
Wildlife Refuge provided much appreciated logistical field
support at TR. We thank all field technicians and graduate
students who have helped with field work at TR since 1984. R.
Alisauskas, E. Reed, and an anonymous reviewer provided
constructive comments on a previous draft; their contributions
7

 are greatly appreciated. Data files associated with this paper
are available at http://doi.org/10.5066/F76971SZ. Mexico
resighting efforts were funded by the U.S. Geological Survey,
Alaska Science Center; U.S. Fish and Wildlife Service,
Migratory Bird Management Region 7; Ducks Unlimited de
Mexico; Ducks Unlimited; the Arctic Goose Joint Venture;
and the North American Wetland Conservation Fund.
Funding sources for band resighting in Canada were
Environment and Climate Change Canada, and The Nature
Trust of British Columbia, and the Arctic Goose Joint
Venture. Studies at TR were funded by the U.S. Geological
Survey, Alaska Science Center; U. S. Fish and Wildlife Service,
Migratory Bird Management Region 7; Ducks Unlimited; the
Morro Bay Brant Group; P. Jebbia (in memory of M.
Shepherd); and the National Science Foundation (OPP
9214970, DEB 9815383, OPP 9985931, OPP 0196406, DEB
0743152, DEB 1252656). ARCO Alaska, (now ConocoPhillips Alaska) provided funding and logistics support for
banding by ABR, on the Central Arctic Coastal Plain in Alaska
(east of Milne Point). A.G.L was supported by Great Basin
Cooperative Extension Unit (CESU) agreement number
G11AC20324 with the U.S. Geological Survey, Alaska
Science Center.

LITERATURE CITED
Alisauskas, R. T., T. W. Arnold, J. O. Leafloor, D. L. Otis, and J. S. Sedinger.
2014. Lincoln estimates of mallard (Anas platyrhynchos) abundance in North
America. Ecology and Evolution 4:132–143.
Alisauskas, R. T., K. L. Drake, J. H. Caswell, and D. K. Kellett. 2012.
Movement and persistence by Ross’s geese (Chen rossii) in Canada’s arctic.
Journal of Ornithology 152(Supplement 2):S573–S584.
Alisauskas, R. T., R. F. Rockwell, K. W. Dufour, E. G. Cooch, G.
Zimmerman, K. L. Drake, J. O. Leafloor, T. J. Moser, and E. T. Reed.
2011. Harvest, survival, and abundance of midcontinent lesser snow
geese relative to population reduction efforts. Wildlife Monographs
179:1–42.
Armstrong, T. 1996. Effects of research activities on nest predation in
Arctic-nesting geese. Journal of Wildlife Management 60:265–269.
Barker, R. J. 1997. Joint modeling of live-recapture, tag-resight, and tagrecovery data. Biometrics 53:666–677.
Bellrose, F. C. 1980. Ducks, geese and swans of North America. Third
edition. Stackpole, Harrisburg, Pennsylvania, USA.
Bollinger, K. S., and D. V. Derksen. 1996. Demographic characteristics of
molting black brant near Teshekpuk Lake, Alaska. Journal of Field
Ornithology 67:141–158.
Brownie, C., D. Anderson, K. Burnham, and D. Robson. 1985. Statistical
inference from band recovery data-a handbook. Volume 156. U.S.
Department of the Interior, Fish and Wildlife Service Resource Publication,
Washington, D.C., USA.
Burnham, K. P., and D. R. Anderson. 2002. Model selection and inference: a
practical information-theoretic approach. Second edition. SpringerVerlag, New York, New York, USA.
Cooch, E., R. F. Rockwell, and S. Brault. 2001. Retrospective analysis of
demographic responses to environmental change: a lesser snow goose
example. Ecological Monographs 71:377–400.
Cottam, C., J. Lynch, and A. Nelson. 1944. Food habits and management of
American sea brant. Journal of Wildlife Management 8:36–56.
Cotter, R. C., and J. E. Hines. 2001. Breeding biology of brant on Banks
Island, Northwest Territories, Canada. Arctic 54:357–366.
Francis, C. M., M. H. Richards, F. Cooke, and R. F. Rockwell. 1992. Longterm changes in survival rates of lesser snow geese. Ecology 73:1346–1362.
Griffith, A. B., R. Salguero-Gomez, C. Merow, and S. McMahon. 2016.
Demography beyond the population. Journal of Ecology 104:271–280.
Hines, J. E., and R. W. Brook. 2008. Changes in annual survival estimates for
black brant from the Western Canadian Arctic, 1962–2001. Waterbirds
31:220–230.

8

Hupp, J. W., D. H. Ward, K. R. Hogrefe, J. S. Sedinger, P. D. Martin, A. A.
Stickney, and T. Obritschkewitsch. 2017. Growth of black brant and lesser
snow goose goslings in northern Alaska. Journal of Wildlife Management
81:In press.
Kendall, W. L., R. J. Barker, G. C. White, M. S. Lindberg, C. A.
Langtimm, and C. L. Penaloza. 2013. Combining dead recovery, auxiliary
observations and robust design data to estimate demographic parameters
from marked individuals. Methods in Ecology and Evolution 4:828–835.
Koons, D. N., R. R. Holmes, and J. B. Grand. 2007. Population inertia and
its sensitivity to changes in vital rates and population structure. Ecology
88:2857–2867.
Laake, J. L. 2013. RMark: an R interface for analysis of capture-recapture
data with MARK. Alaska Fisheries Science Center Report 2013-01.
NOAA, National Marine Fisheries Service, Seattle, Washington, USA.
Lee, D. E., J. M. Black, J. E. Moore, and J. S. Sedinger. 2007. Age-specific
stopover ecology of black brant at Humboldt Bay, California. Wilson
Journal of Ornithology 119:9–22.
Lewis, T. L., D. H. Ward, J. S. Sedinger, A. Reed, and D. V. Derksen. 2013.
Brant (Branta bernicla). in P. G. Rodewald, editor. The birds of North
America online. Cornell Lab of Ornithology Ithaca, New York, USA.
Lindberg, M. S., and J. S. Sedinger. 1998. Ecological significance of broodsite fidelity in black brant: spatial, annual, and age-related variation. Auk
115:436–446.
Lindberg, M. S., J. S. Sedinger, D. V. Derksen, and R. F. Rockwell. 1998.
Natal and breeding philopatry in a black brant, Branta bernicla nigricans,
metapopulation. Ecology 79:1893–1904.
Lindberg, M. S., J. S. Sedinger, and P. L. Flint. 1997. Effects of spring
environment on nesting phenology and clutch size of black brant. Condor
99:381–388.
Lindberg, M. S., D. H. Ward, T. L. Tibbitts, and J. Roser. 2007. Winter
movement dynamics of black brant. Journal of Wildlife Management
71:534–540.
Nicolai, C. A., J. S. Sedinger, D. H. Ward, and W. S. Boyd. 2014. Spatial
variation in life-history trade-offs results in an ideal free distribution in
black brant geese. Ecology 95:1323–1331.
Olson, S., compiler. 2015. Pacific Flyway data book, 2015. U.S. Department
of Interior, Fish and Wildlife Service, Division of Migratory Bird
Management, Vancouver, Washington, USA.
Owen, M. 1980. Wild geese of the world. Batsford, London, England.
Pacific Flyway Council. 2002. Pacific flyway management plan for Pacific
brant. U.S. Fish and Wildlife Service, Portland, Oregon, USA.
https://www.fws.gov/migratorybirds/pdf/management/focal-species/
BlackBrant.pdf. Accessed 6 Sep 2016.
Penaloza, C. L., W. L. Kendall, and C. A. Langtimm. 2014. Reducing bias
in survival under nonrandom temporary emigration. Ecological Applications 24:1155–1166.
R Core Team. 2014. R: a language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Austria.
Reed, E. T., E. G. Cooch, F. Cooke, and R. I. Goudie. 1998a. Migration
patterns of black brant in Boundary Bay, British Columbia. Journal of
Wildlife Management 62:1522–1532.
Reed, E. T., E. G. Cooch, R. I. Goudie, and F. Cooke. 1998b. Site fidelity of
black brant wintering and spring staging in the Strait of Georgia, British
Columbia. Condor 100:426–437.
Ritchie, R. J., R. M. Burgess, T. Obritschkewitsch, A. K. Prichard, and L. B.
Attanas. 2015. Surveys for nesting and brood-rearing brant and lesser
snow geese, Barrow to Fish Creek, Alaska, 2014. ABR, Inc., Fairbanks,
Alaska, for North Slope Borough, Department of Wildlife Management,
Barrow, Alaska. USA.
Ritchie, R. J., R. M. Burgess, and R. S. Suydam. 2000. Status and nesting
distribution of lesser snow geese, Chen caerulescens caerulescens, and brant,
Branta bernicla nigricans, on the western Arctic Coastal Plain, Alaska.
Canadian Field Naturalist 114:395–404.
Sanders, T. A., and R. E. Trost. 2013. Use of capture-recapture models with
mark-resight data to estimate abundance of Aleutian cackling geese.
Journal of Wildlife Management 77:1459–1471.
Seber, G. A. F. 1970. Estimating time-specific survival and reporting rates
for adult birds from band returns. Biometrika 57:313–318.
Sedinger, J. S., N. D. Chelgren, M. S. Lindberg, T. Obritchkewitch, M. T.
Kirk, P. Martin, B. A. Anderson, and D. H. Ward. 2002. Life-history
implications of large-scale spatial variation in adult survival of black brant
(Branta bernicla nigricans). Auk 119:510–515.
The Journal of Wildlife Management

 

9999()

 Sedinger, J. S., N. D. Chelgren, D. H. Ward, and M. S. Lindberg. 2008.
Fidelity and breeding probability related to population density and
individual quality in black brent geese Branta bernicla nigricans. Journal of
Animal Ecology 77:702–712.
Sedinger, J. S., M. P. Herzog, B. T. Person, M. T. Kirk, T. Obritchkewitch,
P. P. Martin, and A. A. Stickney. 2001a. Large-scale variation in growth
of black brant goslings related to food availability. Auk 118:1088–1095.
Sedinger, J. S., C. J. Lensink, D. H. Ward, R. M. Anthony, M. L. Wege,
and G. V. Byrd. 1993. Current status and recent dynamics of the black
brant Branta bernicla breeding population. Wildfowl 44:49–59.
Sedinger, J. S., M. S. Lindberg, and N. D. Chelgren. 2001b. Age-specific
breeding probability in black brant: effects of population density. Journal
of Animal Ecology 70:798–807.
Sedinger, J. S., M. S. Lindberg, M. S. Rexstad, E. A. Rexstad, N. D.
Chelgren, and D. H. Ward. 1997. Testing for handling bias in survival
estimation for black brant. Journal of Wildlife Management 61:782–791.
Sedinger, J. S., and C. A. Nicolai. 2011. Recent trends in first-year survival
for black brant breeding in southwestern Alaska. Condor 113:511–517.
Sedinger, J. S., C. A. Nicolai, C. J. Lensink, C. Wentworth, and B. Conant.
2007. Black brant harvest, density dependence, and survival: a record of
population dynamics. Journal of Wildlife Management 71:496–506.
Sedinger, J. S., C. A. Nicolai, A. W. Van Dellen, A. G. Leach, H. M.
Wilson, and R. M. Anthony. 2016. Predation and reduced grazing interact
to reduce recruitment and population growth in black brant. Condor
118:433–444.
Sedinger, J. S., J. L. Schamber, D. H. Ward, C. A. Nicolai, and B. Conant.
2011. Carryover effects associated with winter location affect fitness, social
status, and population dynamics in a long-distance migrant. American
Naturalist 178:E110–E123.
Smith, B. D., K. R. Hagmeier, W. S. Boyd, N. K. Dawe, T. D. Martin, and
G. L. Monty. 2012. Trends in volume migration chronology in spring
staging Pacific black brant. Journal of Wildlife Management 76:593–599.
Stehn, R. A., R. M. Platte, H. M. Wilson, and J. B. Fischer. 2011.
Monitoring the nesting population of Pacific black brant. U.S. Department of Interior, Fish and Wildlife Service, Division of Migratory Bird
Management, Anchorage, Alaska, USA.
Stickney, A., and R. J. Ritchie. 1996. Distribution and abundance of brant
(Branta bernicla) on the Central Arctic Coastal Plain of Alaska. Arctic
49:44–52.

Leach et al.

 

Long-Term Survival of Brant

Townsend, A. K., E. G. Cooch, T. S. Sillett, N. L. Rodenhouse, R. T.
Holmes, and M. S. Webster. 2016. The interacting effects of food, spring
temperature, and global climate cycles on population dynamics of a
migratory songbird. Global Change Biology 22:544–555.
Ward, D. H., C. P. Dau, T. L. Tibbitts, J. S. Sedinger, B. A. Anderson, and
J. E. Hines. 2009. Change in abundance of Pacific brant wintering in
Alaska: evidence of a climate warming effect? Arctic 62:301–311.
Ward, D. H., A. Morton, T. L. Tibbitts, D. C. Douglas, and E. CarreraGonzalez. 2003. Long-term change in eelgrass distribution at Bahia San
Quintin, Baja California, Mexico, using satellite imagery. Estuaries
26:1529–1539.
Ward, D. H., A. Reed, J. S. Sedinger, J. M. Black, D. V. Derksen, and
P. M. Castelli. 2005. North American brant: effects of changes in habitat
and climate on population dynamics. Global Change Biology
11:869–880.
Ward, D. H., E. A. Rexstad, J. S. Sedinger, M. S. Lindberg, and N. K.
Dawe. 1997. Seasonal and annual survival of adult Pacific brant. Journal of
Wildlife Management 61:773–781.
Ward, D. H., J. A. Schmutz, J. S. Sedinger, K. S. Bollinger, P. D. Martin,
and B. A. Anderson. 2004. Temporal and geographic variation in survival
of juvenile black brant. Condor 106:263–274.
White, G. C., and K. P. Burnham. 1999. Program MARK: survival
estimation from populations of marked animals. Bird Study 46:
S120–S139.
Wilson, H. M. 2014. Aerial photographic survey of brant colonies on the
Yukon-Kuskokwim Delta, Alaska, 2014. U. S. Fish and Wildlife Service,
Anchorage, Alaska, USA.
Zimmerman, G. S., T. J. Moser, W. L. Kendall, P. F. Doherty Jr., G. C.
White, and D. F. Caswell. 2009. Factors influencing reporting and harvest
probabilities in North American geese. Journal of Wildlife Management
73:710–719.

Associate Editor: Adam Janke.

SUPPORTING INFORMATION
Additional supporting information may be found in the
online version of this article at the publisher’s website.

9