ArticlePDF Available

Estimating the Size of the Dutch Breeding Population of Continental Black-Tailed Godwits from 2007–2015 Using Resighting Data from Spring Staging Sites


Abstract and Figures

Over the past 50 years, the population of Continental Black-tailed Godwits Limosa limosa limosa breeding of the East Atlantic Flyway has been in steep decline. This decline has previously been documented in trend analyses and six Netherlands-wide count-based population estimates, the last of which was completed in 1999. We provide an updated population size estimate and describe inter-annual fluctuations in the population between 2007 and 2015. To generate these estimates, we integrated a mark-recapture survival analysis with estimates of the densities of colour-marked individuals at migratory staging sites with known proportions of Continental and Icelandic L. l. islandica Black-tailed Godwits within a Bayesian framework. The use of these analytical techniques means that, in contrast with earlier efforts, our estimates are accompanied with confidence intervals, allowing us to estimate the population size with known precision. Using additional information on the breeding destination of 43 godwits equipped with satellite transmitters at Iberian staging areas, we found that 87% (75–95% 95% CI) of the nominate subspecies in the East Atlantic Flyway breed in The Netherlands. We estimated that the number of breeding pairs in The Netherlands has declined from 47,000 (38,000–56,000) pairs in 2007 to 33,000 (26,000–41,000) in 2015. Despite a temporary increase in 2010 and 2011, the population declined by an average of 3.7% per year over the entire period from 2007–2015, and by 6.3% from 2011–2015. We conclude that investing in an intensive demographic programme at a regional scale, when combined with targeted resightings of marked individuals elsewhere, can yield population estimates at the flyway scale.
Content may be subject to copyright.
Estimating the size of the Dutch breeding population of
Continental Black-tailed Godwits from 2007–2015
using resighting data from spring staging sites
Rosemarie Kentie1,*, Nathan R. Senner1,2, Jos C.E.W. Hooijmeijer1, Rocío Márquez-
Ferrando3, Jordi Figuerola3, José A. Masero4, Mo A. Verhoeven1, & Theunis Piersma1,5
Kentie R., Senner N.R., Hooijmeijer J.C.E.W., Márquez-Ferrando R., Figuerola
J., Masero J.A., Verhoeven M.A. & Piersma T. 2016. Estimating the size of the
Dutch breeding population of Continental Black-tailed Godwits from 2007–2015
using resighting data from spring staging sites. Ardea 114: 213–225.
Over the past 50 years, the population of Continental Black-tailed Godwits
Limosa limosa limosa breeding of the East Atlantic Flyway has been in steep
decline. This decline has previously been documented in trend analyses and six
Netherlands-wide count-based population estimates, the last of which was
completed in 1999. We provide an updated population size estimate and
describe inter-annual fluctuations in the population between 2007 and 2015. To
generate these estimates, we integrated a mark-recapture survival analysis with
estimates of the densities of colour-marked individuals at migratory staging sites
with known proportions of Continental and Icelandic L. l. islandica Black-tailed
Godwits within a Bayesian framework. The use of these analytical techniques
means that, in contrast with earlier efforts, our estimates are accompanied with
confidence intervals, allowing us to estimate the population size with known
precision. Using additional information on the breeding destination of 43 godwits
equipped with satellite transmitters at Iberian staging areas, we found that 87%
(75–95% 95% CI) of the nominate subspecies in the East Atlantic Flyway breed
in The Netherlands. We estimated that the number of breeding pairs in The
Netherlands has declined from 47,000 (38,000–56,000) pairs in 2007 to 33,000
(26,000–41,000) in 2015. Despite a temporary increase in 2010 and 2011, the
population declined by an average of 3.7% per year over the entire period from
2007–2015, and by 6.3% from 2011–2015. We conclude that investing in an
intensive demographic programme at a regional scale, when combined with
targeted resightings of marked individuals elsewhere, can yield population esti-
mates at the flyway scale.
Key words: population estimate, survival probability, mark-recapture, Bayesian
framework, trend
1Conservation Ecology Group, Groningen Institute for Evolutionary Life
Sciences (GELIFES), University of Groningen, P.O. Box 11103, 9700 CC
Groningen, The Netherlands; 2present address: Division of Biological Sciences,
University of Montana, 32 Campus Drive, Missoula, Montana, USA, 59802;
3Department of Wetland Ecology, Doñana Biological Station (EBD-CSIC), Avda.
Américo Vespucio s/n, 41092 Seville, Spain; 4Conservation Biology Research
Group, Department of Anatomy, Cell Biology and Zoology, Faculty of Sciences,
University of Extremadura, Avenida de Elvas, Badajoz 06071, Spain;
5NIOZ Royal Netherlands Institute for Sea Research, Department of Coastal
Systems and Utrecht University, P.O. Box 59, 1790 AB Den Burg, Texel, The
*corresponding author (
Estimates of population sizes are fundamental to
conservation and management issues, as they indicate
whether a species should be listed as of conservation
concern (Dawson et al. 2011, Donald & Fuller 1998,
IUCN 2016). Estimates of population sizes over multiple
years also help to document trajectories of change,
which can play a role in risk assessments (Keith et al.
2015). Yet, estimating the population sizes of wide-
spread populations is often an arduous task, and esti-
mates are often imprecise. The most common method
for estimating wader population sizes is to count con -
gregating individuals at as many locations as possible
during the same period of time each year (e.g. mid -
winter; Underhill & Prys-Jones 1994, Yates & Goss-
Custard 1991). This approach has a disadvantage,
though, as individuals can be double-counted, flocks
can be missed altogether (Rappoldt et al. 1985), and
site-use can be underestimated when turnover is high
(Ganter & Madsen 2001, Loonstra et al. 2016). A com -
mon alternative method for counting birds, are ‘atlas-
projects’. These projects, however, are often restrict ed to
estimating only a limited part of a species’ total popula-
tion size (Donald & Fuller 1998, Szabo et al. 2012).
There is yet another method to estimate population
size that does not necessitate complete coverage of a
species’ range and is grounded on a clear set of assump-
tions based on the mark-resight framework (Otis et al.
1978). This method involves a four-step process: (1)
marking individuals with unique combinations of
colour rings, (2) collecting data on the survival of these
colour-marked individuals so that the remaining
numbers of marked individuals at a given time point
can be estimated, (3) determining the density of those
colour-marked individuals at sites where they randomly
mix with others from the larger population, and (4)
dividing the total number of colour-marked birds by the
proportion of colour-marked birds observed among all
individuals (Gunnarsson et al. 2005, McClintock &
White 2012, Spaans et al. 2011). Mark-resight models
have been used to estimate the population size of
staging or stopover populations (Frederiksen et al.
2001, Lyons et al. 2015, Matechou et al. 2013), these
models are less often used to estimate total population
size (but see Gunnarsson et al. 2005, Lourenço et al.
2010b, Spaans et al. 2011), or used to estimate popula-
tion size over time (but see Ganter & Madsen 2001).
One of the limitations constraining the broader use
of this method, especially in long-term studies, is that
ARDEA 104(3), 2016
Staging Black-tailed Godwits in Extremadura, Spain, almost ready to depart to their breeding grounds (1 February 2010).
the number of marked birds alive at a given moment is
not precisely known and must be separately estimated
with a mark-recapture survival analysis (White &
Burnham 1999), unless the detection probability is
equal to one. We therefore developed a model that esti-
mates the number of marked birds alive at a given
moment in time using a Cormack-Jolly-Seber (CJS)
model. To estimate the total population size, the CJS
model was integrated with a binomial model for counts
of marked and unmarked birds. This is best done in a
Bayesian framework, which enables the likelihood esti-
mates from both models to be joined, making the final
estimates of population size more precise than they
would be if they were analysed separately (Abadi et al.
2010, Doak et al. 2005).
We employ this method to generate yearly esti-
mates of the total population size of Continental Black-
tailed Godwits Limosa limosa limosa breeding in The
Netherlands over the period from 2007–2015. As a
result of agricultural intensification, godwit reproduc-
tive success (Kentie et al. 2013, 2015, Schekkerman et
al. 2008) and breeding population size have dropped
dramatically over the past 50 years (Gill et al. 2007).
Black-tailed Godwits are now labelled as ‘near-threat-
ened’ by the IUCN (2016), even though the Icelandic
subspecies Limosa limosa islandica is still increasing in
numbers (Gill et al. 2007). The most recent estimate of
the Dutch breeding population was generated from the
1999 Dutch Breeding Bird Atlas (Hustings et al. 2002) –
a new edition is currently being compiled (Schekker -
man et al. 2012) – and that only estimate of the total
continental godwit population size is from 2009
(Lourenço et al. 2010b). Given the rates of decline
previously documented in continental godwits (Gill et
al. 2007), updating these estimates is critical to on-
going conservation efforts.
To estimate the size of the Dutch breeding popula-
tion of Black-tailed Godwits, we used birds marked on
the breeding grounds in our core study area in south-
west Friesland from 2004 onwards in combination with
subsequent resightings of these individuals throughout
The Netherlands to estimate their yearly survival prob-
abilities. To couple these with estimates of the density
of colour-marked individuals at sites with known
proportions of continental and Icelandic godwits
(Lopes et al. 2012), from 2007 onwards we counted
marked and unmarked godwits at staging areas in Spain
and Portugal. Finally, we used data gathered from indi-
vidual godwits tagged with satellite transmitters at the
same Iberian staging areas (e.g. Senner et al. 2015) to
estimate the proportion of these godwits that bred in
The Netherlands. Taken together, this allowed us to
generate precise annual estimates of the size of the
Dutch-breeding populations of continental godwits and
to assess trends that can be used to guide future godwit-
related conservation and management activities.
Study species and study areas
Continental Black-tailed godwits (hereafter: ‘godwits’,
and ‘Icelandic godwits’ when referring to the Icelandic
subspecies) are long-distance migrants that spend the
nonbreeding season in West Africa and southern Spain
(Hooijmeijer et al. 2013, rquez-Ferrando et al.
2014). During northward migration, large numbers of
godwits stage in Extremadura (39°01'N, 5°58'W) and
Doñana Wetlands (37°06'N, 6°10'W), Spain, and
coastal Portugal (38°55'N, 8°55'W), where they feed
efficiently on leftover rice kernels on agricultural fields
(Lourenço et al. 2010a, Lourenço & Piersma 2008,
Santiago-Quesada et al. 2009). From early March
Black-tailed Godwits staging in Extremadura, Spain, flying up
after foraging in a rice field near Yelbes (9 February 2016).
onwards, godwits arrive on their breeding grounds
(Lourenço et al. 2011), where during the course of
March and April they establish their territories and lay
a clutch of four eggs.
Godwits are faithful to previous breeding sites and
partners (Kentie et al. 2014). Their chicks are precocial
and leave the nest within 24 hours after hatching
(Schekkerman & Boele 2009). After the cessation of
parental care, adult godwits leave The Netherlands
from early June onwards (with unsuccessfully breeding
individuals leaving earliest; Hooijmeijer et al. 2013).
Juveniles prepare for migration slightly later, and often
gather in flocks in July and even August (Schekkerman
et al. 2014). Not all young godwits return to the
breeding grounds in their second calendar year, and
some arrive only after the breeding period, probably as
prospectors (Kentie et al. 2014). However, it is not yet
known whether these prospecting individuals are using
staging areas during the same time period as do
godwits that move on to the breeding grounds for the
entire breeding season.
As part of a long-term demographic study, we
marked individual godwits with colour rings in south-
west Friesland, The Netherlands (52°55'N, 5°5'E; Kentie
et al. 2014). Additional godwits were marked during
the breeding season elsewhere in The Netherlands.
Adults were captured on the nest and uniquely marked
with four plastic colour rings, a coloured flag, and a
numbered metal ring. Pre-fledging chicks captured at
greater than 10 days of age were large enough to wear
a colour-ring combination: these comprise 47% of the
3499 individuals used in the analysis (Table 1). Smaller
chicks were given an engraved lime flag, but were not
included in the analysis. See Kentie et al. (2013) for
more details on capture procedures.
Density samples
Starting in 2007, we monitored rice fields in Extrema -
dura, Spain, and the Tejo and Sado estuaries in
Portugal for colour-marked godwits in January and
February of each year (Lourenço et al. 2010b, Masero et
al. 2011). From late winter 2010 we also began
surveying Doñana National and Natural Park and its
surrounding area in southern Spain for colour-marked
godwits (Márquez-Ferrando et al. 2014). The numbers
of godwits occurring here at these times of year are
considerable: approximately 25,000 in Extremadura
(Masero et al. 2011), 45,000 in Portugal (Lourenço et
al. 2010b), and 28,000 in Doñana (Márquez-Ferrando
et al. 2014). During the surveys we scanned the god -
wits whose legs were clearly visible. For each scan, we
noted the number of godwits scanned and the number
of godwits with a colour-ring combination from our
colour-marking scheme. If a flock was large, we some-
times made multiple scans of the flock, but from
different vantage points to minimize the risk of noting
individuals more than once.
From 2011 onwards, management in and around
the Giganta rice fields in the Tejo area changed, leaving
the fields inundated with water throughout the godwit
staging period (J.A. Alves pers. comm.). These changes
led to greater use of the area by Icelandic godwits. Such
an influx likely ‘diluted’ the density estimates of
marked continental godwits at the site, because it was
not possible to exclude the Icelandic subspecies during
the density measurements. For this reason, in our
analyses we did not incorporate samples from the Tejo
taken after 2011. As individuals with a colour-ring
combination containing a lime flag included birds
marked at the staging sites, and therefore consisted
partly of Icelandic godwits, we excluded those as well.
Percentage of staging godwits breeding in The
In 2013–2015 we fitted satellite transmitters to 60
adult female godwits that were captured in mist nets at
nocturnal roosts within Spanish (Extremadura: n= 45)
and Portuguese (Tejo: n= 15) staging sites (see Senner
et al. 2015). We deployed solar-powered PTT-100s (9.5
g) from Microwave Technology Inc. that were attached
with a leg-loop harness made of 2 mm nylon rope; in
total, the attachment weighed c. 12 g for an average
loading factor
of 3.43 ± 0.22% (±SE) of an individual’s
mass at the time of capture. We specifically targeted
large, female godwits, as they were best able to accom-
modate the size of the transmitter. The location of each
individual during the breeding period was used to
determine the proportion of godwits staging in Iberia
that bred in The Netherlands.
Integrated model
We integrated the models estimating the number of
marked birds alive with those estimating the density of
marked birds during the staging period into one
Bayesian model. To estimate the number of marked
birds alive, we first estimated juvenile and adult
survival with an age-dependent CJS based on resight-
ings at the breeding grounds. Nearly all godwits survive
the period between staging and breeding (Senner et al.
in prep.), which justifies the use of yearly survival
estimated during the breeding period. Because we
suspected high trap-dependence when including all
godwits marked in The Netherlands outside of our core
godwit study area, we estimated survival only with
ARDEA 104(3), 2016
birds marked in our core study area in southwest
Friesland, but included resightings of those individuals
from across The Netherlands. Godwits equipped with
satellite transmitters were excluded from this analysis
because their survival rate may be lower than that of
colour-marked individuals (Hooijmeijer et al. 2014,
Senner et al. in prep.).
We first tested our mark-resighting data for
Goodness-of-Fit in U-CARE (Choquet et al. 2009).
Because we already included age structure within our
model, we only tested for capture-heterogeneity
(test2.ct; Pradel et al. 2005), which was significant
(c29= 17.8, P= 0.04). We therefore included indi-
vidual random effects for resighting probability pin our
CJS model. We allowed pto vary between years and
included an additive age effect with two age classes,
because not all godwits return in their second calendar
year (Kentie et al. 2014). For apparent survival (
), we
also included two age classes in the model and treated
year as a random effect separately for each age class.
By using year as a random effect, we were able to use
the survival estimate of the final year of the study (Kéry
& Schaub 2012). We believe that our estimate
approaches true survival for the following reasons:
godwits are highly site-faithful (Kentie et al. 2014, van
den Brink et al. 2008), before and after breeding they
forage and roost within or near our study area, and we
used resightings from across The Netherlands.
Next, we used a binomial model to estimate a
yearly proportion of marked individuals in flocks at
each of the staging areas. This proportion was then
used to estimate the total population size using the
equation Nt~ Kt/ pbandt, where Ntis the total popula-
tion size at time t, Ktis the number of marked birds
alive at time t, and pbandtis the proportion of marked
birds seen at time t. We used the number of colour-
marked godwits per year, including godwits marked in
The Netherlands outside our core study area, and
multiplied these by our yearly survival rates to estimate
the marked population in year t. Because not all second
calendar year godwits may use the staging areas at the
times we took our density samples, we corrected for the
occurrence of young marked birds. To do so, we
included within the Bayesian framework a binomial
model which estimated the proportion of second
calendar year birds resighted at the staging sites. Of
chicks marked before 2015, we resighted 26 individuals
between 2009 and 2016 which were in their second
calendar year, and 42 which were in their third
calendar year. We corrected for the mortality between
second and third calendar year birds, by using the
mean adult survival estimated by the model. Because of
the necessity to correct for mortality with age, we
abstained from using older age classes. Last, we esti-
mated the population size of Dutch-breeding godwits
by integrating the proportion of godwits with satellite
tags that bred in The Netherlands, and corrected for the
proportion of Icelandic godwits (6.5%, n= 278; Lopes
et al. 2012), both included as binomial models, and
assumed that the proportion of second calendar year
godwits passing through the staging areas were
breeding birds.
We ran JAGS (Plummer 2003) in the R statistical
platform (v. 3.2.3; R Core Team 2014) with the R2jags
package (Su & Yajima 2015) to perform Markov Chain
Monte Carlo (MCMC) simulations for parameters esti-
mation. We used uninformative priors for all parame-
ters. We ran three parallel chains of 50,000 iterations
with a burn-in of 10,000 and kept every 6th observa-
tion. We checked the R-hat for convergence of the
model (in all but one cases < 1.01, with the random
part of year dependent survival equalling 1.05).
Estimations are presented as the posterior means with a
95% credibility interval.
From 2004–2015 we colour-marked 3499 godwits, of
which we used 1891 godwits marked in southwest
Friesland to estimate yearly adult and juvenile survival
probabilities (Table 1). Adult survival was high (0.94,
0.85–1.00 95% CI) in the first year of the study, and the
Number ringed with Number ringed in
our scheme southwest Friesland
Year nadults nyoung nadults nyoung
2004 66 28 66 24
2005 57 13 57 9
2006 104 55 46 22
2007 143 102 124 46
2008 131 112 115 68
2009 218 124 134 38
2010 201 147 117 53
2011 119 189 65 41
2012 233 181 181 58
2013 350 322 267 106
2014 241 363 162 92
Table 1. Number of marked Continental Black-tailed Godwits
with our colour ring scheme (excluding those with a lime flag
colour), and number marked in southwest Friesland.
mean adult survival over the whole period was 0.85
(0.84–0.87; Figure 1). Juvenile survival probability
ranged between 0.30 in 2014 and 0.54 in 2010 with
relatively large credibility intervals (Figure 1), and the
mean juvenile survival was 0.34 (0.33–0.45). The
random year effect of juvenile survival was SD = 0.47
(0.09–1.00), and the random year effect of adult
survival was SD = 2.71 (1.71–4.31). The averaged
mean posterior resighting probability was 0.82
(0.80–0.85) for adults and 0.30 (0.21–0.39) for second
calendar year birds (Table S1). The individual resighting
random effect was SD = 1.34 (1.12–1.56). For all
parameter estimates see the Supplementary Material.
Of the 60 godwits fitted with satellite transmitters
at the Iberian staging sites, 13 transmitters or godwits
died before migration or did not migrate further, and
39 females established breeding territories in The
Netherlands. Moreover, two individuals appeared to be
Icelandic godwits. Excluding these two, 87% (75–95%)
of the tagged birds were Dutch breeding birds.
In total, we checked 420,206 godwits for colour-
rings at Spanish and Portuguese staging sites (Table 2).
The density of colour-ringed godwits increased from
1/500 godwits in 2007 to 1/77 godwits in 2015. The
fraction of godwits in their second calendar year
passing through the staging sites in January and
February was 0.53 (0.36–0.74). The estimated popula-
tion size of godwits, including Icelandic godwits using
the Iberian rice fields, increased from 115,305
(98,304–134,265) individuals in 2007 to 164,010
(128,479–209,253) in 2011, before decreasing to
81,793 (66,973–98,309) in 2015 (Figure 2). The
annual growth rate from 2007–2015 was 3.7%, with
the fastest decline occurring from 2011–2015 at 6.3%
per year. The Dutch breeding population in 2015 was
estimated at 33,140 (26,031–41,303) breeding pairs.
We estimated the total population size of Continental
Black-tailed Godwits in the East-Atlantic Flyway, by
estimating survival probabilities of colour-marked indi-
viduals and then using the density of surviving colour-
marked individuals observed at staging areas in Spain
and Portugal. We could correct for the fraction of the
islandica subspecies, which are staging on the Iberian
rice fields and fish ponds, within the model. Although
we found that 87% of continental godwits staging in
Iberia breed in The Netherlands and that adult survival
was relatively high from 2007–2015, during the course
of our study, the Dutch-breeding population declined
ARDEA 104(3), 2016
2004 2006 2008 2010 2012 2014
apparent survival probability
Figure 1. Apparent survival probability of adult and juvenile
Black-tailed Godwits obtained from the CJS in the Bayesian
model. Posterior means and 95% credibility intervals are shown.
2007 2009 2011 2013 2015
population size (x 1000)
Figure 2. Population size of Black-tailed Godwits Limosa limosa
limosa of the East-Atlantic Flyway based on colour ring densities
on the staging sites. These estimates include godwits from de
Icelandic subspecies Limosa limosa islandica using Iberian rice
fields, which we assume to be 6.5% of the population size.
Year Total birds checked
2007 136,623
2008 65,222
2009 70,196
2010 11,030
2011 16,395
2012 17,914
2013 24,952
2014 67,638
2015 10,236
Table 2. Number of Continental Black-tailed Godwits checked
for colour rings at the staging sites in Spain and Portugal.
by 3.7% per year, and this decline accelerated from
2011–2015 to a rate of 6.3% per year. In total, we esti-
mated that nearly 33,000 pairs of godwits currently
breed in The Netherlands, which is less than one third
of the number of pairs that bred there in the 1970s
(Figure 3). In combination with the already steady
declines that have occurred over the past four decades,
these recent, rapid declines suggest that drastic meas-
ures are necessary to stop the disappearance of one of
the most iconic meadow bird species breeding in The
One of the findings in our study is that in the midst
of a steady population decline, the size of the breeding
population actually increased from 2009 to 2011,
before again declining from 2011–2015. The causes of
this increase could potentially result from two separate
processes. On the one hand, the increase might have
been caused by an increasing proportion of Icelandic
godwits using Iberian rice fields, as the Icelandic
godwit population has steadily grown over the past
three decades (Gill et al. 2007). We assumed a constant
proportion of Icelandic godwits of 6.5% at Iberian
staging sites based on a DNA study of godwits caught in
the rice fields of Extremadura between 2005 and 2008
(Lopes et al. 2012). That study found no increase in the
proportion of Icelandic godwits, and the percentage
they found resembled the proportion of Icelandic
godwits in rice fields estimated on the basis of ring
resightings (10%; Alves et al. 2010, 7.7%; Masero et al.
2011). Two of the godwits we fitted with satellite trans-
mitters appeared to be Icelandic godwits (4.3%, 1.2–
14.2% CI), however, we targeted continental godwits
thus this proportion will be too low. However, if the
peak in godwit numbers in 2011 was only caused by an
increase in the proportion of Icelandic godwits at
Iberian staging sites, their proportion should have been
30% or higher. This we consider unlikely.
The increase did correspond with high reproductive
success in at least part of the population’s breeding
range in 2010 (Kentie 2015). Nevertheless, if the popu-
lation increase was explained entirely by an increase in
godwit reproduction, godwit pairs would have been
required to produce on average 0.61 chicks per year
that survived to become breeders per year in 2009 and
2010. Previous studies have found that godwit nest
success averages c. 50% (Kentie et al. 2015), with a
maximum of 69% on herb-rich meadows in 2008
(Kentie 2015), meaning that in an average year 30% of
hatched chicks would have had to survive to the next
year, or 22% in a year with high nest survival. The
highest first-year survival, from nestling to second
calendar year bird, in our research area was 24%
(Kentie 2015), which occurred in herb-rich meadows in
2010. However, in the grassland monocultures, where
the majority of godwits breed in The Netherlands, the
highest first-year survival was 14%. After 2011, the
decline in godwit numbers happened so rapidly that
only complete reproductive failures would make this
possible. We therefore conclude that a combination of
changes in the proportion of Icelandic godwits and
variation in continental godwit reproductive success is
likely to explain the temporary increase, and note that
the confidence intervals for the population estimates
between 2010 and 2012 were rather large.
We estimate that there were 33,000 breeding pairs
in The Netherlands in 2015. This estimate was based
on two other estimates. First, with information gath-
ered from godwits equipped with satellite tags at
Iberian staging sites, we estimated that 87% of the
continental population breeds in The Netherlands. This
percentage falls within the range of previous findings,
which were derived from comparisons of country-wide
counts (Table 3). Second, we estimated that adult
survival was around 85% throughout much of our
study, which corresponds closely with previous esti-
mates from other Dutch study areas (Roodbergen et al.
2008). Despite the close correspondence of these
underlying estimates with those from other published
1960 1970 1980 1990 2000 2010
number of breeding pairs (x 1000)
atlas count
this study
Figure 3. Number of breeding pairs of Continental Black-tailed
Godwits in The Netherlands, based on estimations, atlas counts,
interpolations, and this study (references: Bekhuis et al. 1987,
Bijlsma et al. 2001, Hustings et al. 2002, Mulder 1972, Osieck &
Hustings 1994, Teixeira 1979, Teunissen et al. 2005, Teunissen
et al. 2012, van Dijk et al. 2005). When available, minimum and
maximum estimation or 95% CI are plotted. The lines are fitted
with a local polynomial regression fitting (LOESS) in R, and
represent the mean and the 95% CI of the point estimates of the
number of breeding pairs.
studies, as well as the increased precision of our statis-
tical estimates in relation to previous efforts, the upper
and lower 95% CI of our estimate ranged from 26,000
to 41,000, indicating that our estimates still include
some uncertainty.
More generally, with the Bayesian mark-resighting
methodology presented here, we were able to estimate
the numbers of Black-tailed Godwits using staging sites
in Iberia with relatively tight confidence limits in most
years. The benefit of this method is that the total fly -
way population can be estimated without the necessity
of surveying individuals across the entire breeding
range of Continental godwits. For populations that are
not easily counted across breeding areas, but congre-
gate at wintering or staging sites, such as many wader
species, this method may thus have considerable
advantages (Spaans et al. 2011). Moreover, when oper-
ating a demographic monitoring programme, this
method makes it possible to track the population size
on a yearly basis with relatively low additional effort,
especially when compared with such intensive under-
takings as breeding bird atlases. Nonetheless, there are
potential drawbacks: for instance, our estimates of the
Continental Black-tailed Godwit population would
have been more precise if continental and Icelandic
godwits did not mix in Iberia, or if we had temporal
estimates of the mixture of continental and Icelandic
godwits for the whole study period.
With a breeding population of 33,000 pairs in 2015, the
Dutch godwit population has plummeted by nearly
75% since the first nation-wide estimate of 120,000
pairs in 1967 (Mulder 1972, Figure 3). In spite of this,
the agricultural grasslands of The Netherlands remain
the single most important stronghold for breeding
Continental Black-tailed Godwits in the East Atlantic
Flyway population — they breed nowhere else in such
large numbers (Piersma 1986, Thorup 2006) and are
declining just as rapidly in most other countries in
which they still breed (Gill et al. 2007, Thorup 2006).
Furthermore, previous work has identified low levels of
reproductive success to be the single most important
factor driving the decline (Schekkerman et al. 2008,
Kentie et al. 2013, Roodbergen et al. 2012), which
shows that factors operating in The Netherlands are
largely responsible for the declines that have occurred
thus far. Although enormous amounts of money and
effort have been expended to conserve continental
godwits (Kleijn et al. 2010), our findings make clear
that these have been ineffective or insufficient.
Initiatives leading to drastically improved management
are thus necessary to preserve one of the most iconic
species of the Dutch countryside before it becomes rele-
gated to a few small corners of its former range.
ARDEA 104(3), 2016
Year Percentage Source
1970 91% Mulder (1972)
1985 91% Piersma (1986)
1990 90% Hötker et al. (1991)
1995 85% Beintema et al. (1995)
2000 84% Thorup (2006)
2015 87% this study
Table 3. Percentage of the East Atlantic Flyway Continental
Black-tailed Godwits breeding in The Netherlands. To define the
flyway population, we included godwits breeding in Belgium,
United Kingdom, Germany, France, Spain, Italy, Luxembourg,
Austria, Sweden and The Netherlands.
We thank the godwit field crews of 2004–2015 for their invalu-
able assistance in the field, both in The Netherlands and abroad.
Land management organisations (It Fryske Gea and Staatsbos -
beheer) and private landowners organized in the Weidevogel -
collectief Súdwestkust generously gave permission to access
their properties. Miguel Medialdea (Veta la Palma fish farm),
José M. Abad-Gómez, Helena Silva Pinto (Reserva Natural do
Estuário do Tejo), Rui Alves (Companhia das Lezerias, S.A.),
Teresa Catry, Luisa Mendes, Alfonso Rocha and Jose Alves
helped with logistics, tagging, and access to the staging sites.
We thank Niko Groen, Pedro Lourenço, Dirk Tanger, Allert
Bijleveld, René Faber, Wim Tijsen, Bob Loos, Haije Valkema,
Gjerryt Hoekstra, Egbert van der Velde and Alice McBride for
their additional density samples at the staging sites, and many
more birdwatchers for colour-ring sightings across the flyway.
We thank Eldar Rakhimberdiev and Tamar Lok for discussions
on the Bayesian model. Ruth Howison provided help by cura-
ting and summarizing the data on tagged godwits. Tómas
Gunnarsson, Hans Schekkerman and Adriaan Dokter provided
helpful feedback that greatly improved the first submitted
version. This study was funded mainly by the former
Netherlands Ministry of Agriculture, Nature Management and
Food Safety, now subsumed in the Ministry of Economic Affairs,
by the Province of Fryslân, and by the Spinoza Premium Award
2014 from The Netherlands Organization for Scientific Research
(NWO) to TP, with some additional funding by the Prins
Bernard Cultuurfonds, the Van der Hucht Beukelaar Stichting,
BirdLife Netherlands and WWF-Netherlands through Global
Flyway Network and the Chair in Flyway Ecology, FP7-Regpot
project ECOGENES (Grant No. 264125), the NWO-TOP grant
‘Shorebirds in space’ (854.11.004) awarded to TP, ExpeER
Ecosystem Research, ‘ICTS-RBD’ to the ESFRI LifeWatch,
MINECO, and European Union Structural Funds (AIC-A2011-
0706). This work was done under license numbers 4339E and
6350A following the Dutch Animal Welfare Act Articles 9 and
Abadi F., Gimenez O., Arlettaz R. & Schaub M. 2010. An assess-
ment of integrated population models: bias, accuracy, and
violation of the assumption of independence. Ecology 91:
Alves J.A., Lourenço P.M., Piersma T., Sutherland W.J. & Gill J.A.
2010. Population overlap and habitat segregation in
wintering Black-tailed Godwits Limosa limosa. Bird Study
57: 381-391.
Beintema A.J., Moedt O. & Ellinger D. 1995. Ecologische atlas
van de Nederlandse weidevogels. Schuyt & Co, Haarlem.
Bekhuis J., Bijlsma R., van Dijk A., Hustings F., Lensink R. &
Saris F. 1987. Atlas van de Nederlandse vogels. SOVON,
Arnhem, The Netherlands.
Bijlsma R.G., Hustings F. & Camphuysen C.J. 2001. Algemene
en schaarse vogels van Nederland (Avifauna van Nederland
2). GMB Uitgeverij/KNNV Uitgeverij , Haarlem/Utrecht.
Choquet R., Lebreton J.-D., Gimenez O., Reboulet A.-M. &
Pradel R. 2009. U-CARE: Utilities for performing goodness
of fit tests and manipulating CApture–REcapture data.
Ecography 32: 1071–1074.
Dawson T.P., Jackson S.T., House J.I., Prentice I.C. & Mace G.M.
2011. Beyond predictions: Biodiversity conservation in a
changing climate. Science 332: 53–58.
Doak D.F., Gross K., Morris W.F. 2005. Understanding and
predicting the effects of sparse data on demographic
analyses. Ecology 86: 1154–1163.
Donald P.F. & Fuller R.J. 1998. Ornithological atlas data: A
review of uses and limitations. Bird Study 45: 129–145.
Frederiksen M., Fox A.D., Madsen J. & Colhoun K. 2001.
Estimating the total number of birds using a staging site. J.
Wildl. Manage. 65: 282–289.
Ganter B. & Madsen J. 2001. An examination of methods to esti-
mate population size in wintering geese. Bird Study 48:
Gill J.A., Langston R.H.W., Alves J.A., Atkinson P.W., Bocher P.,
Cidraes Vieira N., Crockford N.J., Gélinaud G., Groen N.,
Gunnarsson T.G., Hayhow B., Hooijmeijer J.C.E.W., Kentie
R., Kleijn D., Lourenço P.M., Masero J.A., Meunier F., Potts
P.M., Roodbergen M., Schekkerman H., Schroeder J.,
Wymenga E. & Piersma T. 2007. Contrasting trends in two
Black-tailed Godwit populations: a review of causes and
recommendations. Wader Study Group Bull. 114: 43–50.
Gunnarsson T.G., Gill J.A., Potts P.M., Atkinson P.W., Croger R.E.,
Gélinead G., Gardarsson A. & Sutherland W.J. 2005.
Estimating population size in Black-tailed Godwits Limosa
limosa islandica by colour-marking. Bird Study 52: 153–158.
Hooijmeijer J.C.E.W., Senner N.R., Tibbitts T.L., Gill R.E., Jr.,
Douglas D.C., Bruinzeel L.W., Wymenga E. & Piersma T.
2013. Post-breeding migration of Dutch-breeding Black-
tailed Godwits: Timing, routes, use of stopovers, and
nonbreeding destinations. Ardea 101: 141–152.
Hooijmeijer J.C.E.W., Gill R.E. Jr, Mulcahy D.M., Tibbitts T.L.,
Kentie R., Gerritsen G.J., Bruinzeel L.W., Tijssen D.C.,
Harwood C.M. & Piersma T. 2014. Abdominally implanted
satellite transmitters affect reproduction and survival rather
than migration of large shorebirds. J. Ornithol. 155:
Hötker H., Davidson N.C. & Fleet M. 1991. Waders breeding on
wet grasslands. Wader Study Group Bull. 61, suppl.
Hustings F., Vergeer J.-W. & Eekelder P. 2002. Atlas van de
Nederlandse broedvogels 1998–2000: verspreiding, aantal -
len, verandering. SOVON Vogelonderzoek Nederland, Beek-
IUCN 2016. The IUCN Red List of Threathened Species. Version
2016-1. (accessed 30/6/2016)
Keith D., Akçakaya H.R., Butchart S.H.M., Collen B., Dulvy N.K.,
Holmes E.E., Hutchings J.A., Keinath D., Schwartz M.K.,
Shelton A.O. & Waples R.S. 2015. Temporal correlations in
population trends: Conservation implications from time-
series analysis of diverse animal taxa. Biol. Conserv. 192:
Kentie R. 2015. Spatial demography of Black-tailed Godwits.
Metapopulation dynamics in a fragmented agricultural
landscape. PhD-thesis, University of Groningen, Groningen.
Kentie R., Hooijmeijer J.C.E.W., Trimbos K.B., Groen N.M. &
Piersma T. 2013. Intensified agricultural use of grasslands
reduces growth and survival of precocial shorebird chicks. J.
Appl. Ecol. 50: 243–251.
Kentie R., Both C., Hooijmeijer J.C.E.W. & Piersma T. 2014. Age-
dependent dispersal and habitat choice in Black-tailed
Godwits (Limosa l. limosa) across a mosaic of traditional
and modern grassland habitats. J. Avian Biol. 45: 396–405.
Kentie R., Both C., Hooijmeijer J.C.E.W. & Piersma T. 2015.
Management of modern agricultural landscapes increases
nest predation rates in Black-tailed Godwits (Limosa limosa
limosa). Ibis 157: 614–625.
Kéry M. & Schaub M. 2012. Bayesian population analysis using
WinBUGS: a hierarchical perspective, Academic Press.
Kleijn D., Schekkerman H., Dimmers W.J., van Kats R.J.M.,
Melman D. & Teunissen W.A. 2010. Adverse effects of agri-
cultural intensification and climate change on breeding
habitat quality of Black-tailed Godwits Limosa l. limosa in
the Netherlands. Ibis 152: 475–486.
Loonstra A.H.J., Piersma T. & Reneerkens J. 2016. Staging dura-
tion and passage population size of Sanderlings in the
western Dutch Wadden Sea. Ardea 104: 49–61.
Lopes R.J., Alves J.A., Gill J.A., Gunnarsson T.G., Hooijmeijer
J.C.E.W., Lourenço P.M., Masero J.A., Piersma T., Potts P.M.,
Rabaçal B., Reis S., Sánchez-Guzman J.M., Santiago-
Quesada F. & Villegas A. 2012. Do different subspecies of
Black-tailed Godwit Limosa limosa overlap in Iberian
wintering and staging areas? Validation with genetic
markers. J. Ornithol. 154: 35–40.
Lourenço P.M. & Piersma T. 2008. Stopover ecology of Black-
tailed Godwits (Limosa limosa limosa) in Portuguese rice
fields: a guide on where to feed in winter. Bird Study 55:
Lourenço P.M., Mandema F.S., Hooijmeijer J.C.E.W., Granadeiro
J.P. & Piersma T. 2010a. Site selection and resource deple-
tion in Black-tailed Godwits Limosa l. limosa eating rice
during northward migration. J. Anim. Ecol. 79: 522–8.
Lourenço P.M., Kentie R., Schroeder J., Alves J.A., Groen N.M.,
Hooijmeijer J.C.E.W. & Piersma T. 2010b. Phenology,
stopover dynamics and population size of migrating Black-
tailed Godwits Limosa limosa limosa in Portuguese rice plan-
tations. Ardea 98: 35–42.
Lourenço P.M., Kentie R., Schroeder J., Groen N.M., Hooijmeijer
J.C.E.W. & Piersma T. 2011. Repeatable timing of northward
departure, arrival and breeding in Black-tailed Godwits Limosa
l. limosa, but no domino effects. J. Ornithol. 152: 1023–1032.
ARDEA 104(3), 2016
Lyons J.E., Kendall W.L., Royle J.A., Converse S.J., Andres B.A.
& Buchanan J.B. 2015. Population size and stopover dura-
tion estimation using mark-resight data and Bayesian
analysis of a superpopulation model. Biometrics 72:
Márquez-Ferrando R., Figuerola J., Hooijmeijer J.C.E.W. &
Piersma T. 2014. Recently created man-made habitats in
Doñana provide alternative wintering space for the threat-
ened Continental European Black-tailed Godwit population.
Biol. Conserv. 171: 127–135.
Masero J.A., Santiago-Quesada F., Sánchez-Guzmán J.M.,
Villegas A., Abad-Gómez J.M., Lopes R.J., Encarnação V.,
Corbacho C. & Morán R. 2011. Long lengths of stay, large
numbers, and trends of the Black-tailed Godwit Limosa
limosa in rice fields during spring migration. Bird Conserv.
Int. 21: 12–24.
Matechou E., Morgan B.J.T., Pledger S., Collazo J.A. & Lyons
J.E. 2013. Integrated analysis of capture–recapture–
resighting data and counts of unmarked birds at stop-over
sites. J. Agric. Biol. Environ. Stat. 18: 120–135.
McClintock B.T. & White G.C. 2012. From NOREMARK to
MARK: Software for estimating demographic parameters
using mark–resight methodology. J. Ornithol. 152:
Mulder T. 1972. De Grutto (Limosa limosa (L.)) in Nederland:
Aantallen, verspreiding, terreinkeuze, trek en overwin-
tering. Bureau van de K.N.N.V., Hoogwoud, The Nether -
Osieck E. & Hustings F. 1994. Rode lijst van bedreigde soorten
en blauwe lijst van belangrijke soorten in Nederland.
Technisch rapport 12, Vogelbescherming Nederland, Zeist.
Otis D.L., Burnham K.P., White G.C. & Anderson D.R. 1978.
Statistical inference from capture data on closed animal
populations. Wildlife Monogr. 62: 3–135.
Piersma T. 1986. Breeding waders in Europe. Wader Study
Group Bull. 48, Suppl.: 1–116.
Plummer M. 2003. JAGS: A program for analysis of Bayesian
graphical models using gibbs sampling. Proceedings of the
3rd International Workshop on Distributed Statistical
Computing (DSC 2003). Vienna, Austria.
Pradel R., Gimenez O. & Lebreton J.D. 2005. Principles and
interest of GOF tests for multistate capture-recapture
models. Anim. Biodivers. Conserv. 28: 189–204.
R Core Team 2014. R: A language and environment for statis-
tical computing. R Foundation for Statistical Computing,
Vienna, Austria.
Rappoldt C., Kersten M. & Smit C. 1985. Errors in large-scale
shorebird counts. Ardea 73: 13–24.
Roodbergen M., Klok C. & Schekkerman H. 2008. The ongoing
decline of the breeding population of Black-tailed Godwits
Limosa l. limosa in The Netherlands is not explained by
changes in adult survival. Ardea 96: 207–218.
Roodbergen M., van der Werf B. & Hötker H. 2012. Revealing
the contributions of reproduction and survival to the
Europe-wide decline in meadow birds: review and meta-
analysis. J. Ornithol. 153: 53–74.
Santiago-Quesada F., Masero J.A., Albano N., Villegas A. &
Sánchez-Guzmán J.M. 2009. Sex differences in digestive
traits in sexually size-dimorphic birds: insights from an
assimilation efficiency experiment on Black-tailed Godwit.
Comp. Biochem. Physiol. A 152: 565–568.
Schekkerman H. & Boele A. 2009. Foraging in precocial chicks
of the black-tailed godwit Limosa limosa: Vulnerability to
weather and prey size. J. Avian Biol. 40: 369–379.
Schekkerman H., Teunissen W. & Oosterveld E. 2008. The effect
of mosaic management on the demography of Black-tailed
Godwit Limosa limosa on farmland. J. Appl. Ecol. 45:
Schekkerman H., Van Turnhout C., van Kleunen A., van Diek H.
& Altenburg J. 2012. Naar een nieuwe vogelatlas; achter-
gronden van de veldwerkopzet. Limosa 85: 133.
Schekkerman H., Gerritsen G.J. & Hooijmeijer J.C.E.W. 2014.
Jonge grutto's in Nederland in 2014: een aantalsschatting
op basis van kleurringdichtheden. Sovon Vogelonderzoek
Nederland, Nijmegen.
Senner N.R., Verhoeven M.A., Abad-Gómez J.M., Gutiérrez J.S.,
Hooijmeijer J.C.E.W., Kentie R., Masero J.A., Tibbitts T.L. &
Piersma T. 2015. When Siberia came to the Netherlands: the
response of Continental Black-tailed Godwits to a rare
spring weather event. J. Anim. Ecol. 84: 1164–1176.
Spaans B., van Kooten L., Cremer J., Leyrer J. & Piersma T.
2011. Densities of individually marked migrants away from
the marking site to estimate population sizes: a test with
three wader populations. Bird Study 58: 130–140.
Su Y.-S. & Yajima M. 2015. R2jags: Using R to Run 'JAGS'. R
package v. 0.5-6.
Szabo J.K., Fuller R.A. & Possingham H.P. 2012. A comparison
of estimates of relative abundance from a weakly structured
mass-participation bird atlas survey and a robustly designed
monitoring scheme. Ibis 154: 468–479.
Teixeira R.M. 1979. Atlas van de Nederlandse broedvogels.
Vereniging tot Behoud van Natuurmonumenten in Neder -
land, ‘s-Gravenland.
Teunissen W., Altenburg W. & Sierdsema H. 2005. Toelichting op
de gruttokaart van Nederland 2004. SOVON Vogelonder -
zoek Nederland, Beek-Ubbergen. Altenburg & Wymenga
ecologisch onderzoek bv, Veenwouden, SOVON-onder -
zoeks rapport 2005/04, A&W-rapport 668.
Teunissen W.A., Schotman A.G.M., Bruinzeel L.W., ten Holt H.,
Oosterveld E.O., Sierdsema H., Wymenga E. & Melman
T.C.P. 2012. Op naar kerngebieden voor weidevogels in
Nederland. Werkdocument met randvoorwaarden en
handreikeningen. Alterra, Wageningen, Alterra-rapport
2344; Sovon Vogelonderzoek Nederland, Nijmegen, Sovon-
rapport 2012/21; Altenburg & Wymenga ecologisch onder-
zoek, Feanwâlden, A&W-rapport 1799.
Thorup O. 2006. Breeding waders in Europe 2000. International
Wader Studies 14. International Wader Study Group, UK.
Underhill L.G. & Prys-Jones R.P. 1994. Index numbers for water-
bird populations. 1. Review and methodology. J. Appl. Ecol.
31: 463–480.
van den Brink V., Schroeder J., Both C., Lourenço P.M.,
Hooijmeijer J.C.E.W. & Piersma T. 2008. Space use by
Black-tailed Godwits Limosa limosa limosa during settle-
ment at a previous or a new nest location. Bird Study 55:
van Dijk A.J., Dijksen L., Hustings F., Koffijberg K., Schoppers J.,
Teunissen W., van Turnhout C., van der Weide M.J.T.,
Zoetebier D. & Plate C.L. 2005. Broedvogels in Nederland in
2003. SOVON-monitoringsrapport 2005/01. SOVON Vogel -
onderzoek Nederland, Beek-Ubbergen.
White G.C. & Burnham K.P. 1999. Program MARK: survival esti-
mation from populations of marked animals. Bird Study 46:
Yates M.G. & Goss-Custard J.D. 1991. A comparison between
high water and low water counts of shorebirds on the Wash,
east England. Bird Study 38: 179–187.
tijdens de voorjaarstrek elk jaar de fractie Grutto’s met kleur-
ringen geschat. We hebben daarbij gecorrigeerd voor het
percentage IJslandse Grutto’s Limosa l. islandica (waarvan de
aantallen toenemen) dat gebruikmaakt van dezelfde pleister-
plaatsen. Door deze getallen te combineren kon een populatie-
schatting worden gemaakt en kon tevens de nauwkeurigheid
van die schatting worden aangegeven met betrouwbaarheidsin-
tervallen. Met behulp van Grutto’s die op de pleisterplaatsen
werden uitgerust met satellietzenders, kwamen we erachter dat
87% (75–95%) van de West-Europese populatie in Nederland
broedt. Deze nieuwe schatting laat bovendien zien dat het
aantal broedparen in Nederland vanaf 1967 met 75% is afge-
nomen en dat de snelheid van de afname in de periode
2011–2015 sneller ging dan in de vier jaar daarvoor. We laten
met deze analyse zien dat met behulp van een intensief regio-
naal ring- en monitoringprogramma in combinatie met het
verzamelen van terugmeldingen van geringde vogels elders,
betrouwbare schattingen kunnen worden gemaakt van popula-
ties op de schaal van een hele trekroute. Aangezien eerder
onderzoek aantoonde dat het lage broedsucces de oorzaak van
de achteruitgang is, heeft Nederland als het belangrijkste broed-
gebied voor West-Europese Grutto’s de grootste verantwoorde-
lijkheid binnen Europa voor het in stand houden van deze popu-
Corresponding editor: Adriaan Dokter
Received 24 August 2016; accepted 28 December 2016
Hoewel we weten dat het aantal Grutto’s Limosa limosa limosa
in West-Europa de laatste 50 jaar hard achteruit is gegaan,
stamt de laatste Nederlandse populatieschatting nog uit 1999.
In dit artikel presenteren we nieuwe populatieschattingen voor
de gehele continentale gruttopopulatie van West-Europa en
voor de Nederlandse broedpopulatie afzonderlijk. Deze schat-
tingen laten zien dat de Nederlandse populatie tussen 2007 en
2015 met 3,7% per jaar is afgenomen van 47.000 (95%-
betrouwbaarheidsinterval: 38.000–56.000) tot 33.000 (26.000
– 41.000) broedparen in 2015. Om tot deze schatting te komen,
hebben we gegevens afkomstig van Grutto’s met kleuringen en
zendertjes geïntegreerd in één analyse in een Bayesiaans
statistisch model. Allereerst hebben we voor 2007 tot en met
2015 door middel van een mark-recapture overlevingsanalyse
het aantal nog in leven zijnde Grutto’s met kleurringen geschat.
Daarna hebben we op pleisterplaatsen in Spanje en Portugal
ARDEA 104(3), 2016
Parameter*posterior mean SD 2.5% CI 97.5% CI R-hat
phi adult 2004–2005 0.939 0.038 0.854 0.997 1.001
phi adult 2005–2006 0.865 0.038 0.786 0.936 1.001
phi adult 2006–2007 0.871 0.037 0.797 0.940 1.001
phi adult 2007–2008 0.834 0.032 0.770 0.894 1.001
phi adult 2008–2009 0.826 0.029 0.767 0.883 1.001
phi adult 2009–2010 0.830 0.025 0.780 0.878 1.001
phi adult 2010–2011 0.847 0.023 0.800 0.890 1.001
phi adult 2011–2012 0.848 0.022 0.803 0.890 1.001
phi adult 2012–2013 0.869 0.018 0.833 0.903 1.001
phi adult 2013–2014 0.831 0.016 0.798 0.862 1.001
phi adult 2014–2015 0.828 0.036 0.764 0.904 1.004
phi juv 2004–2005 0.355 0.076 0.205 0.505 1.002
phi juv 2005–2006 0.370 0.095 0.181 0.565 1.001
phi juv 2006–2007 0.421 0.083 0.268 0.602 1.001
phi juv 2007–2008 0.330 0.068 0.199 0.461 1.003
phi juv 2008–2009 0.379 0.058 0.268 0.496 1.001
phi juv 2009–2010 0.427 0.071 0.296 0.578 1.002
phi juv 2010–2011 0.543 0.084 0.391 0.714 1.002
phi juv 2011–2012 0.323 0.069 0.188 0.454 1.001
phi juv 2012–2013 0.440 0.062 0.327 0.568 1.001
phi juv 2013–2014 0.383 0.048 0.29 0.479 1.002
phi juv 2014–2015 0.298 0.085 0.144 0.461 1.006
padult 2005 0.901 0.038 0.813 0.960 1.001
padult 2006 0.858 0.039 0.772 0.923 1.001
padult 2007 0.816 0.040 0.730 0.887 1.001
padult 2008 0.789 0.035 0.715 0.853 1.001
padult 2009 0.717 0.035 0.646 0.782 1.001
padult 2010 0.795 0.027 0.739 0.844 1.001
padult 2011 0.691 0.030 0.630 0.749 1.001
padult 2012 0.732 0.028 0.676 0.785 1.001
padult 2013 0.871 0.017 0.836 0.903 1.001
padult 2014 0.907 0.014 0.878 0.932 1.002
padult 2015 0.957 0.029 0.891 1.000 1.005
p2nd calendar year 2005 0.411 0.110 0.215 0.638 1.001
p2nd calendar year 2006 0.311 0.081 0.172 0.485 1.001
p2nd calendar year 2007 0.248 0.064 0.141 0.388 1.001
p2nd calendar year 2008 0.215 0.051 0.128 0.327 1.001
p2nd calendar year 2009 0.156 0.037 0.093 0.237 1.001
p2nd calendar year 2010 0.219 0.047 0.137 0.320 1.001
p2nd calendar year 2011 0.140 0.032 0.086 0.209 1.001
p2nd calendar year 2012 0.165 0.038 0.101 0.247 1.001
p2nd calendar year 2013 0.327 0.058 0.221 0.449 1.001
p2nd calendar year 2014 0.412 0.065 0.290 0.540 1.002
p2nd calendar year 2015 0.648 0.179 0.348 0.997 1.005
prop marked birds 2007 0.002 0.000 0.002 0.002 1.001
prop marked birds 2008 0.003 0.000 0.003 0.003 1.001
prop marked birds 2009 0.003 0.000 0.003 0.004 1.001
prop marked birds 2010 0.004 0.001 0.003 0.005 1.001
Table S1. Parameter estimates (mean, standard deviation (SD), credibility interval (2.5% CI and 97.5% CI) and R-hat) from the
Bayesian population model.
Parameter*posterior mean SD 2.5% CI 97.5% CI R-hat
prop marked birds 2011 0.004 0.001 0.003 0.005 1.001
prop marked birds 2012 0.005 0.001 0.004 0.007 1.001
prop marked birds 2013 0.009 0.001 0.007 0.010 1.001
prop marked birds 2014 0.011 0.000 0.010 0.012 1.001
prop marked birds 2015 0.016 0.001 0.013 0.018 1.001
prop marked birds 2016 0.013 0.001 0.011 0.016 1.001
prop 2nd calendar year birds 0.533 0.097 0.361 0.742 1.001
prop Icelandic Godwits 0.065 0.015 0.039 0.096 1.001
prop Dutch Godwits 0.866 0.050 0.753 0.948 1.001
total population 2007 115305 9162 98304 134265 1.001
total population 2008 109806 9467 92480 129539 1.001
total population 2009 118629 9336 101376 138187 1.001
total population 2010 151309 24363 110442 205736 1.001
total population 2011 164010 20735 128479 209253 1.001
total population 2012 141498 15179 114424 174234 1.001
total population 2013 109621 8356 94275 127185 1.001
total population 2014 106690 5167 96944 117249 1.001
total population 2015 81793 8251 66973 99309 1.003
breeding pair Netherlands 2007 46717 4650 37922 56103 1.001
breeding pair Netherlands 2008 44490 4685 35800 54168 1.001
breeding pair Netherlands 2009 48066 4785 39026 57679 1.001
breeding pair Netherlands 2010 61304 10546 43431 84692 1.001
breeding pair Netherlands 2011 66451 9320 50376 86506 1.001
breeding pair Netherlands 2012 57332 7067 44572 72306 1.001
breeding pair Netherlands 2013 44415 4319 36264 53188 1.001
breeding pair Netherlands 2014 43228 3352 36461 49724 1.001
breeding pair Netherlands 2015 33140 3894 26031 41303 1.002
mean padults 0.821 0.013 0.795 0.847 1.001
mean p2nd calendar year 0.296 0.047 0.208 0.392 1.002
mean phi adults 0.854 0.008 0.838 0.869 1.001
mean phi juv 0.388 0.029 0.333 0.448 1.003
random effect phi year adults 2.706 0.659 1.705 4.312 1.001
random effect phi year juv 0.468 0.231 0.089 1.002 1.045
random effect pindividual 1.335 0.112 1.122 1.560 1.002
Deviance 9183.043 142.312 8896.685 9449.126 1.003
*pis resighting probability, phi is apparent survival probability, prop is short for proportion, juv is short for juvenile.
Table S1. Continued
... One approach was to remove all single sightings per occasion from the data; this solution was based on the reasonable assumption that it is more likely to make a wrong assignments once than multiple times (see, e.g. Kentie et al., 2016Kentie et al., , 2018Loonstra et al., 2019). Although generally correct, this data filtering also excludes correct single sightings and thus decreases precision of the estimates, and it does not guarantee that all incorrect assignments were excluded. ...
Full-text available
1. Misidentification of marked individuals is unavoidable in most studies of wild animal populations. Models commonly used for the estimation of survival from such capture‐recapture data ignore misidentification errors potentially resulting in biased parameter estimates. With a simulation study we show that ignoring misidentification in Cormack‐Jolly‐Seber (CJS) models results in systematic positive biases in the estimates of survival and in spurious declines of survival over time. 2. We developed an extended robust design capture mark‐resight (RDM) model that includes correct identification parameters to get unbiased survival estimates when resighting histories are prone to misidentification. The model assumes that resightings occur repeatedly within a season, which in practice is often the case when resightings of colour‐marked individuals are collected. We implemented the RDM model in a state‐space formulation and also an approximate, but computationally faster, model (RDMa) in JAGS and evaluated their performances by using simulated and empirical capture‐resight data on black‐tailed godwits Limosa limosa. 3. The CJS models applied to simulated data under an imperfect identification scenario data produced strongly positively biased estimates of survival. For a range of degrees of correct identification probabilities, the RDM model provided unbiased and accurate estimates of survival, reencounter and correct‐identification probabilities. The RDMa model performed well for large datasets (>25 individuals), with high resighting (>0.3) and high correct identification (>0.7) probabilities. For the empirical data the CJS model estimated average juvenile survival at 0.997% and adult survival at 0.939% and also detected a strong decline in adult survival over time at a rate of ‐0.14±0.029. In contrast, the RDMa model estimated a probability of correct identification of 0.94, annual juvenile survival at 0.234%, adult at 0.834% and less strong decline over time (‐0.046±0.016). 4. We conclude that estimates of survival probabilities obtained from data that include misidentification errors and analysed with standard CJS model are unlikely to be correct. The bias in survival increases with the magnitude of misidentification errors, which is inevitable as datasets become longer. Since misidentification due to tag misreads is common in empirical data, we recommend the use of the here presented RDM model to provide unbiased parameter estimates.
... In particular, the nominate subspecies of the black-tailed godwit (Limosa limosa limosa; hereafter 'godwit') -of which more than 90% breed in agricultural grasslands within the European Union -has experienced a dramatic decline (Gill et al., 2007). Following population declines of more than 75% over the past four decades, godwits have been labelled near threatened by the IUCN and made the focus of considerable conservation and management efforts (Kentie et al., 2016). Despite this attention, recent years have witnessed accelerated declines and record-low indices of reproductive success across the godwit range Verhoeven, Smart et al., 2021). ...
Full-text available
Abstract Maintaining the biodiversity of agricultural ecosystems has become a global imperative. Across Europe, species that occupy agricultural grasslands, such as black‐tailed godwits (Limosa limosa limosa), have undergone steep population declines. In this context, there is a significant need to both determine the root causes of these declines and identify actions that will promote biodiversity while supporting the livelihoods of farmers. Food availability, and specifically earthworm abundance (Lumbricidae), during the pre‐breeding period has often been suggested as a potential driver of godwit population declines. Previous studies have recommended increasing the application of nitrogen to agricultural grasslands to enhance earthworm populations and aid agricultural production. Here we test whether food availability during the pre‐breeding period affects when and where godwits breed. Using large‐scale surveys of food availability, a long‐term mark‐recapture study, focal observations of foraging female godwits, and tracking devices that monitored godwit movements, we found little evidence of a relationship between earthworm abundance and the timing of godwit reproductive efforts or the density of breeding godwits. Furthermore, we found that the soils of intensively managed agricultural grasslands may frequently be too dry for godwits to forage for those earthworms that are present. The increased application of nitrogen to agricultural grasslands will therefore likely have no positive effect on godwit populations. Instead, management efforts should focus on increasing the botanical diversity of agricultural grasslands, facilitating conditions that prevent hardening soils, and reducing the populations of generalist predators.
Full-text available
We present a novel formulation of a mark-recapture-resight model that allows estimation of population size, stopover duration, and arrival and departure schedules at migration areas. Estimation is based on encounter histories of uniquely marked individuals and relative counts of marked and unmarked animals. We use a Bayesian analysis of a state-space formulation of the Jolly-Seber mark-recapture model, integrated with a binomial model for counts of unmarked animals, to derive estimates of population size and arrival and departure probabilities. We also provide a novel estimator for stopover duration that is derived from the latent state variable representing the interim between arrival and departure in the state-space model. We conduct a simulation study of field sampling protocols to understand the impact of superpopulation size, proportion marked, and number of animals sampled on bias and precision of estimates. Simulation results indicate that relative bias of estimates of the proportion of the population with marks was low for all sampling scenarios and never exceeded 2%. Our approach does not require enumeration of all unmarked animals detected or direct knowledge of the number of marked animals in the population at the time of the study. This provides flexibility and potential application in a variety of sampling situations (e.g., migratory birds, breeding seabirds, sea turtles, fish, pinnipeds, etc.). Application of the methods is demonstrated with data from a study of migratory sandpipers.
Full-text available
The population of Sanderlings Calidris alba along the East Atlantic flyway has grown considerably during the last decades. Perhaps reflecting this augmented population size, increasing numbers of Sanderling have been reported to stage in the Wadden Sea during spring and autumn migration. Estimates of the numbers of Sanderlings in the Wadden Sea have previously been based on a limited number of counts that were not corrected for the turnover of individuals. In this study, we accounted for turnover using estimates of the probability that individually colour-ringed Sanderlings are still present two days after a sighting. In combination with daily counts during high tide, we estimated the total number of Sanderlings that used the island Griend and surrounding mudflats, in the western Dutch Wadden Sea, during southward passage in 2013 and 2014. We also estimated minimal staging durations of Sanderlings at Griend. Non-moulting birds were significantly heavier upon capture, which suggests that they were refuelling for long non-stop migratory flights. Winter sightings confirmed that the non-moulting Sanderlings winter in sub-Saharan Africa and that the moulting Sanderlings spent the winter in Europe or northern Africa. With an average minimal stay in the western Dutch Wadden Sea of 9 days in 2013 and 12 in 2014, non-moulting Sanderlings stayed much shorter than moulting Sanderlings which stayed for 32 days in 2013 and 36 days in 2014. Non-moulting individuals were less likely to be resighted between years. We discuss that estimates of minimal staging duration are likely underestimates of the true staging duration, and we propose that moulting Sanderlings probably complete their wing moult in the Wadden Sea. We estimated that the total number of Sanderlings using the western Dutch Wadden Sea before migration to European or African wintering areas were 27,546 (95% CI 22,739–41,449) in 2013 and 22,574 (95% CI 16,436–46,114) in 2014. This would amount to 11–14% of a total flyway population of 200,000 individuals, representing an amazing degree of concentration for what is regarded as a rather widely and thinly spread shorebird species.
Full-text available
Een methode is uitgewerkt om kerngebieden te identificeren voor weidevogels. Als gidssoort is de grutto gebruikt, implicaties voor de andere weidevogelsoorten zijn aangeduid. Als zoekgebied voor kerngebieden zijn aangeduid gebieden die voldoen aan minumumdichtheden (15 dan wel 30 bp/100 ha). Aan de hand van trendgegevens is geanalyseerd welke factoren bepalend zijn voor de aantalsontwikkeling. De resultaten hiervan zijn als randvoorwaarden gehanteerd voor de nadere invulling van de kerngebieden. Met een metapopulatiemodel is verkend aan welke ruimtelijke voorwaarden kerngebieden moeten voldoen: o.a. omvang en onderlinge afstanden, in relatie tot de ruimtelijke kwaliteit. Scenarioberekeningen zijn uitgevoerd naar verschillende ruimtelijke invullingen. Er is een handreiking opgesteld als voorbeeld hoe kerngebieden in de praktijk geidentificeerd en uitgewerkt zouden kunnen worden.
Full-text available
Population trends play a large role in species risk assessments and conservation planning, and species are often considered threatened if their recent rate of decline meets certain thresholds, regardless how large the population is. But how reliable an indicator of extinction risk is a single estimate of population trend? Given the integral role this decline-based approach has played in setting conservation priorities, it is surprising that it has undergone little empirical scrutiny. We compile an extensive global dataset of time series of abundance data for over 1300 vertebrate populations to provide the first major test of the predictability of population growth rates in nature. We divided each time series into assessment and response periods and examined the correlation between growth rates in the two time periods. In birds, population declines tended to be followed by further declines, but mammals, salmon, and other bony fishes showed the opposite pattern: past declines were associated with subsequent population increases, and vice versa. Furthermore, in these taxa subsequent growth rates were higher when initial declines were more severe. These patterns agreed with data simulated under a null model for a dynamically stable population experiencing density dependence. However, this type of result could also occur if conservation actions positively affected the population following initial declines—a scenario that our data were too limited to rigorously evaluate. This ambiguity emphasizes the importance of understanding the underlying causes of population trajectories in drawing inferences about rates of decline in abundance.
Full-text available
Capsule Conservation management of rice fields may be necessary to guarantee the availability of high quality stopover habitats. Aims To analyse habitat selection and quantify the diet composition of birds. Methods Using water level and agricultural management of the fields as variables, habitat selection was analysed by compositional analysis. Godwit diet composition was quantified by faecal analysis, and food abundance was sampled to explain the observed habitat selection. Results We found evidence of higher use of flooded and ploughed paddies than expected from their relative abundance. These fields have the highest densities of buried rice kernels, which seem to be the main food source for Black-tailed Godwits. Conclusion Currently, godwits find good foraging areas in Portuguese rice fields, feeding primarily on rice kernels that are mostly found in flooded ploughed fields. Changes in rice farming, late ploughing and predicted decreases in rainfall may lead to loss of this habitat. However, because of the man-made nature of their requirements, it should be possible to install relevant land-use practices that guarantee the availability of high quality stopover habitats.
Full-text available
Extreme weather events have the potential to alter both short- and long-term population dynamics as well as community- and ecosystem-level function. Such events are rare and stochastic, making it difficult to fully document how organisms respond to them and predict the repercussions of similar events in the future. To improve our understanding of the mechanisms by which short-term events can incur long-term consequences, we documented the behavioural responses and fitness consequences for a long-distance migratory bird, the continental black-tailed godwit Limosa limosa limosa, resulting from a spring snowstorm and three-week period of record low temperatures. The event caused measurable responses at three spatial scales - continental, regional and local - including migratory delays (+19 days), reverse migrations (>90 km), elevated metabolic costs (+8·8% maintenance metabolic rate) and increased foraging rates (+37%). There were few long-term fitness consequences, however, and subsequent breeding seasons instead witnessed high levels of reproductive success and little evidence of carry-over effects. This suggests that populations with continued access to food, behavioural flexibility and time to dissipate the costs of the event can likely withstand the consequences of an extreme weather event. For populations constrained in one of these respects, though, extreme events may entail extreme ecological consequences. © 2015 The Authors. Journal of Animal Ecology © 2015 British Ecological Society.
Bayesian statistics has exploded into biology and its sub-disciplines, such as ecology, over the past decade. The free software program WinBUGS and its open-source sister OpenBugs is currently the only flexible and general-purpose program available with which the average ecologist can conduct standard and non-standard Bayesian statistics. Bayesian Population Analysis Using WinBUGS goes right to the heart of the matter by providing ecologists with a comprehensive, yet concise, guide to applying WinBUGS to the types of models that they use most often: linear (LM), generalized linear (GLM), linear mixed (LMM) and generalized linear mixed models (GLMM). Comprehensive and richly-commented examples illustrate a wide range of models that are most relevant to the research of a modern population ecologist. All WinBUGS/OpenBUGS analyses are completely integrated in software R. Includes complete documentation of all R and WinBUGS code required to conduct analyses and shows all the necessary steps from having the data in a text file out of Excel to interpreting and processing the output from WinBUGS in R.
Bayesian statistics has exploded into biology and its sub-disciplines, such as ecology, over the past decade. The free software program WinBUGS and its open-source sister OpenBugs is currently the only flexible and general-purpose program available with which the average ecologist can conduct standard and non-standard Bayesian statistics. Bayesian Population Analysis Using WinBUGS goes right to the heart of the matter by providing ecologists with a comprehensive, yet concise, guide to applying WinBUGS to the types of models that they use most often: linear (LM), generalized linear (GLM), linear mixed (LMM) and generalized linear mixed models (GLMM). Comprehensive and richly-commented examples illustrate a wide range of models that are most relevant to the research of a modern population ecologist. All WinBUGS/OpenBUGS analyses are completely integrated in software R. Includes complete documentation of all R and WinBUGS code required to conduct analyses and shows all the necessary steps from having the data in a text file out of Excel to interpreting and processing the output from WinBUGS in R.