ArticlePDF Available

Identification of Azores Gull

Authors:
  • Corridor cv
  • Independent Researcher

Abstract and Figures

This paper by Peter Adriaens, Peter Alfrey, Chris Gibbins and Daniel López-Velasco, deals with the identification of Azores Gull. With more than 40 photographs, the paper discusses, for the first time in detail, the identification of this difficult taxon, gives the guidelines for ID, especially useful in a vagrancy context to Europe
Content may be subject to copyright.
303
[Dutch Birding 42: 303-334, 2020]
Identication of Azores Gull
Peter Adriaens, Peter Alfrey, Chris Gibbins & Daniel López-Velasco
Yellow-legged Gull Larus michahellis is a fa-
miliar species in many European countries.
However, the identication, taxonomy and distri-
bution of its various subspecies have received
comparatively little attention and a number of im-
portant issues still remain clouded in considerable
uncertainty. Recent authoritative literature on gulls
does not agree on the number of subspecies:
Howell & Dunn (2007) recognize nominate L m
michahellis and the Atlantic subspecies L m atlantis
(hereafter atlantis) but Olsen & Larsson (2004) and
Olsen (2018) include a third one, L m lusitanius
(hereafter lusitanius) from north-western Iberia.
There are particular problems and uncertainties
associated with lusitanius, a taxon that we will
consider in some detail here. The taxon has been
poorly described, and is not recognised by many
authors (eg, Howell & Dunn 2007, Dickin son &
Remsen 2013, Gill et al 2020). Moreover, there
does not seem to be agreement on the precise
breeding ranges of the different subspecies, with
authors variously restricting atlantis to only the
Azores (Dubois 2001, Yésou 2002, Olsen 2018),
to all of the Macaronesian islands (Dwight 1922,
1925, Cramp & Simmons 1983, Grant 1986,
Garner & Quinn 1997, Jonsson 1998, AERC TAC
2003, Olsen & Larsson 2004, Howell & Dunn
2007), to the Macaronesian islands and coastal
north-western Africa (Collinson et al 2008) or all of
these locations with the coasts of north-western
Spain and Portugal sometimes included (Stegman
1934, de Knijff et al 2001, Liebers et al 2001,
Sternkopf et al 2010). Most recently, Stoddart &
McInerny (2017) highlighted uncertainties over
the distribution and taxonomic status of atlantis in
their review of records of this taxon in Britain.
This paper focuses on the identication of atlan-
tis, to which we refer as ‘Azores Gull’. Other au-
thors have sometimes referred to this taxon as
Atlantic Gull, Atlantic Yellow-legged Gull or Azo-
re an Yellow-legged Gull (eg, Stoddart & McInerny
2017; www.dutchavifauna.nl/wpvogelnamen). In
order to treat its identication in a comprehensive
way, we needed to make sense of Mediterranean
michahellis, birds on other Macaronesian islands,
as well as the gulls that occur in Portugal and
north-western Spain. Looking at this regional vari-
ation caused us to also assess vocalisations, and
the results of these analyses led us to reconsider
the taxonomy of Yellow-legged Gulls. As well as
dealing with the identication of atlantis, we
therefore offer some views on taxonomic relations
between the various populations we have ana-
lysed. However, we stress that we are eld orni-
thologists rather than taxonomists and are mainly
interested in how distinctive a certain taxon is in
the eld or, in other words, whether we can deter-
mine the geographic origin of an individual bird
with any degree of certainty. To our eyes, it is
mainly the birds from the Azores that look suf-
ciently different from nominate micha hellis to be
identiable away from their breeding range and
this is our focus. The specic aims of this paper
are: 1 to present solid identication criteria for va-
grant Azores Gull; and 2 to use these criteria, voice
and genetics to comment on taxonomic relations
between the various Yellow-legged Gull taxa and
on the vagrancy of Azores Gull.
Material and methods
Sample data
The authors have extensive eld experience with
Yellow-legged Gulls, from the Azores east to the
Black Sea coast of Georgia. Daniel López-Velasco
lives on the Cantabrian coast in Asturias, Spain,
and is thoroughly familiar with the local Yellow-
legged Gulls. This paper is based on many trips to
the Azores, including a specic one in February
2015 to study and photograph Azores Gulls, sev-
eral trips to Portugal (Porto, Peniche and Algarve
coast) and Morocco, and eastwards to Turkey and
Georgia. We also checked skins of adult birds from
the Azores, the Canary Islands, Madeira and Iberia
in the Natural History Museum at Tring, England,
and we received data from Andres Bermejo, Pim
Edelaar and Bert Saveyn who examined skins in
Estación Biológica de Doñana at Sevilla, Spain,
and Museo Nacional de Ciencias Naturales at
Madrid, Spain. In addition, we received and col-
lected numerous photographs from the Atlantic
and Iberian breeding range taken throughout the
year, including many from the Canary Islands,
Madeira and the Gibraltar area.
Photographs and our eld observations were
used to distill features for each age type that offer
most help with the identication of out-of-range
birds. For all immature age classes this was done
304
qualitatively, and accordingly, in the text that fol-
lows, we simply describe these features. Many of
the birds that we looked at were not ringed, so the
usual caveats about classifying immature birds
into the correct age class apply (see also Arizaga et
al 2019). However, it goes without saying that we
have taken utmost care when assigning birds to a
certain age type. Because adult plumage traits are
more amenable to quantitative analysis, we adopt-
ed a different approach for this age class, scoring
the primary pattern of a large sample and assessing
the frequency of different primary patterns in the
various populations. This is detailed in the follow-
ing section. We also include scores for the upper-
part grey tones of adult birds based on a Kodak
grey scale applied to museum skins.
For practical reasons, we use the classication
in cycles (rst-cycle, second-cycle and so on) to
age birds, rather than the classic system of rst-
winter, rst-summer, second-winter and so on (cf
Dutch Birding 1985), because the cycle classica-
tion ts the complex moult cycles of gulls better (cf
Howell et al 2003, Howell 2010, Adriaens &
Gibbins 2016).
Scoring system for adult birds
Using high quality photographs, we examined the
wing-tip pattern of a total of 959 adult Yellow-
legged Gulls from the Macaronesian islands,
Iberia, Morocco, Croatia, Greece, Israel and Tur-
key. Sample sizes for each region are presented in
table 1, and sample locations are shown in gure
1. Birds showing signs or remnants of immaturity
were excluded from the samples (except those
with a few thin dark streaks on primary coverts,
which can be shown by very old gulls; cf Muusse
et al 2011). All birds examined were photograph-
ed during the breeding season, except for a num-
ber of birds from the Macaronesian Islands (see
table 1). Since the latter populations are isolated
and largely sedentary (see section ‘Movements
and vagrancy’), we felt that Macaronesian birds
photographed at any time of the year could safely
be included in the sample.
Specically, we looked at the following features
that we deemed most useful for identication:
1 the length of the pale tongue (= wedge) on the
underside of the outermost primary (p10) com-
pared with the length of the feather (from tip to
primary coverts); 2 the presence and extent of a
white mirror on p9; 3 the length of the black pat-
tern on the outer web of p8 (again compared with
the length of the feather); and 4 the presence and
extent of a black pattern on p3-4. The categories
for each of these criteria are presented and illus-
trated in table 5; for topography of a gull wing, see
Olsen & Larsson 2004 (p 21). We also systemati-
cally noted the presence of thin dark streaks on the
primary coverts, but these turned out to be irrele-
vant for identication purposes.
Analysis of calls
We examined the display calls (‘long calls’) of
adult Yellow-legged Gulls from 126 recordings.
FIGURE 1 Sample locations of adult Yellow-legged Gulls Larus michahellis used for scoring of primary patterns. Green
= proportion of birds from the breeding season; blue = proportion of non-breeding birds. The size of each pie chart
reects sample size (see table 1). Azores unsp. = Azores unspecied (no exact location given).
Identication of Azores Gull
305
The number of recordings that we obtained from
each region is shown in table 2. Some recordings
were made by us (Azores, February 2015) but most
were either sent to us or were obtained from the
Xeno-Canto website (www.xeno-canto.org). Long
calls of immature birds were excluded from the
analysis. Research on calls of Yellow-legged Gulls
was published by Teyssèdre (1983, 1984). She
compared recordings of the calls of 30 Yellow-
legged Gulls from a breeding colony in País Vasco
(Basque country), northern Spain, with 30 from
Camargue, Bouches-du-Rhône, France, and eight
from Ile d’Oléron, Charente-Maritime, France. In
order to obtain results that could be compared
with her work, we adopted the same analytical
methods. Thus, we looked at the following ele-
ments of the long call on sonagrams: 1 maximum
frequency of the fundamental (FFm), ie, the lowest
harmonic (lowest layer) of each call note on the
sonagram; 2 maximum frequency of the dominant
harmonic (DFm), ie, the most visible, boldest har-
monic; 3 number of harmonics (nH) visible in the
sonagram (for each call note); and 4 prominence/
visibility of fth harmonic (H5) in the sonagram
(categorized as 0 = weak/absent, 1 = intermediate,
2 = strong/bold). Harmonics are layers of sound
that show up in a sonagram one above the other;
they are also called overtones. The more nasal a
call note, the more harmonics it will show in a
sonagram. We added one more feature to the anal-
ysis: 5 distance between the harmonics (Dist),
which was not examined by Teyssèdre (1983,
1984). This was analysed by measuring the dis-
tance between DFm and the rst harmonic above
it, although generally the distance between all har-
monics is roughly the same. Because the number
of harmonics visible in the sonogram tends to vary
with distance to the bird, only good quality record-
FIGURE 2 Long call of adult Azores Gull Larus micha-
hellis atlantis, Terceira, Azores, 18 February 2015 (Peter
Adriaens). Example of sonagram analysis.
TABLE 1 Summary of all locations and sample sizes (n) for adult Yellow-legged Gulls Larus michahellis included in
analysis of wing-tip patterns. * = non-breeding birds.
location n
Azores, Corvo 1+4*
Azores, Faial 3+1*
Azores, Flores 3*
Azores, Graciosa 1+1*
Azores, Pico 12+1*
Azores, Santa Maria 3
Azores, São Miguel 12+21*
Azores, Terceira 5+108*
Azores, unspecied 25*
subtotal Azores 37 + 164* (=201)
Canary Islands, El Hierro 2*
Canary Islands, Fuerteventura 4+11*
Canary Islands, Gran Canaria 1+10*
Canary Islands, La Gomera 1
Canary Islands, La Palma 3+5*
Canary Islands, Lanzarote 19+32*
Canary Islands, Tenerife 2+20*
Canary Islands, unspecied 3*
subtotal Canary Islands 30 + 83* (=113)
location n
Gibraltar 10
Ceuta 22
subtotal Gibraltar 32
Madeira 30+59* (=89)
Morocco (Casablanca - Agadir) 33
Portugal (Peniche) 103
Spain, Galicia 28
Spain, País Vasco 89
subtotal northern and north-western Spain 117
Spain, Andalucía 31
Spain, Catalunya 113
subtotal michahellis Spain 144
Croatia 55
Greece 22
Israel 30
Turkey 20
subtotal ‘eastern michahellis’ 127
TOTAL 959
Identication of Azores Gull
306
TABLE 2 Locations and numbers of recordings (n) of display calls of adult Yellow-legged Gulls Larus michahellis
analysed for this paper.
location n
eastern Spain 2
south-eastern France 7
Bulgaria 2
Switzerland 2
Italy 2
subtotal michahellis 15
Galicia 19
País Vasco 2
Asturias 1
subtotal northern and north-western Spain 22
Terceira 7
São Miguel 1
Santa Maria 5
Flores 19
subtotal Azores 32
location n
La Palma 1
Fuerteventura 5
Lanzarote 3
Lobos 3
subtotal Canary Islands 12
Madeira 8
Morocco 3
Berlengas 7
Peniche 7
Sagres 6
Lisboa 3
subtotal Portugal 23
Cádiz, near Gibraltar 11
TOTAL 126
TABLE 3 Kodak grey scale values of upperparts of adult Yellow-legged Gulls Larus michahellis; skins examined mainly
at Natural History Museum (Tring, England), Estación Biológica de Doñana (Sevilla, Spain) and Museo Nacional de
Ciencias Naturales (Madrid, Spain); * subadult bird.
location sample Kodak grey scale sources
size value
Azores >15 7-9 own data; Olsen & Larsson 2004
Canary Islands 7 7-8 own data; Andres Bermejo unpublished data;
Pim Edelaar & Bert Saveyn unpublished data
Madeira 2 8-9 own data
Western Sahara, Morocco 1* 7 Pim Edelaar & Bert Saveyn unpublished data
Portugal ? 6.5-8 Olsen & Larsson 2004
Galicia (north-western Spain) 15 5.5-7 own data; Andres Bermejo unpublished data;
Pim Edelaar & Bert Saveyn unpublished data
Asturias (north-western Spain) 3 6.5-7.5 own data
Cantabria (northern Spain) 1 5 Andres Bermejo unpublished data
País Vasco (northern Spain) 1* 5 Pim Edelaar & Bert Saveyn unpublished data
western Mediterranean 25 5-7 own data; Andres Bermejo unpublished data;
(nominate michahellis) Pim Edelaar & Bert Saveyn unpublished data
ings were used that were taken at close distance.
An example of the analysis is shown in gure 2.
The results are discussed in the section on ‘Voice’.
Regional variation
Adults
In this section we rst provide details of the nature
and extent of variation found across the geograph-
ic range occupied by Yellow-legged Gulls. This as-
sessment is then used in the second section as the
platform from which to approach the identication
of vagrant adult Azores Gulls.
Regional variation throughout Iberia and Atlantic
Most books and many websites only depict nomi-
nate michahellis, which has a relatively homoge-
neous appearance throughout its extensive breed-
ing range. It should be noted, however, that small
differences exist in primary pattern between west-
ern and eastern adults of this taxon. There are, for
example, differences in the pattern of p10, with
nearly 17% of a sample of 121 adults from Istan-
bul, Turkey, having a wholly white tip (ie, birds
lacked any black terminal marks on this primary)
compared with only 1% of 99 birds from Cata-
lunya, Spain. Eastern birds also tend to have longer
Identication of Azores Gull
307
TABLE 4 Measurements (in mm) of adult Yellow-legged Gulls Larus michahellis from various locations (sexes
combined). Given are mean and range.
location sample wing length tarsus length bill depth sources
size at gonys
western 388 451.4 (410-485) 68 (58-80) 18.4 (16-20.5) Andres Bermejo unpublished
Mediterranean data; Isenmann 1973, Cramp
& Simmons 1983, Carrera et al
1987, Bosch 1995
País Vasco, Spain 252 425.7 (399-471) 64.5 (55.5-80.5) 19 (14-24) Mínguez & Ganuza 1995,
Arizaga et al 2008, Galarza
et al 2008
Asturias, Spain 12 428.3 (407-446) 62.9 (58.9-68.4) 17.1 (15.2-20.1) Andres Bermejo unpublished
data
Galicia/Portugal 111 420.5 (337-454) 63.4 (55-75.4) 18.1 (11.5-20.5) Andres Bermejo unpublished
data; Cramp & Simmons 1983,
Carrera et al 1987
Morocco ? 428.5 (410-445) 63.3 (61.6-65) Andres Bermejo unpublished
data; Urban et al 1986
Canary Islands/ >30 424.8 (395-450) 64.4 (60-69.5) 18.4 (16.6-20.1) Andres Bermejo unpublished
Madeira data; Volsøe 1951, Bannerman
1963, Cramp & Simmons 1983
Azores >34 417.5 (395-438) 64 (59-68) 18.25 (16.5-20.5) Dwight 1922, 1925, Vaurie 1965
pale tongue on p10. However, these differences
make eastern birds much less like Azores Gull and
so they are not considered in further detail here.
Further studies may reveal additional differences
but they are not the focus of this paper. In the west-
ernmost part of its range, however, Yellow-legged
Gull shows more variation and there appear to be
several populations that differ in plumage and
voice (discussed later).
Structure
Measurements of adult Yellow-legged Gulls from
various populations are shown in table 4. Com-
pared with nominate michahellis, Macaro nesian,
Atlantic Iberian and Moroccan birds all average
appreciably smaller with shorter wing length and
shorter tarsus. Among the ‘Atlantic’ populations
themselves there are no real differences in struc-
ture, although birds from País Vasco appear to
show a rather deep bill by comparison, at least as
heavy as in michahellis, which is remarkable given
the smaller overall size of these birds and because
birds from Asturias, which is only slightly further
west, on average seem to have the most slender
bill of all. The sample size from the latter region is
very small, however, and no clear differences in
bill depth are evident between birds from País
Vasco and those from Galicia.
Head Streaking
Compared with nominate michahellis, adult
Yellow-legged Gulls from the Atlantic region tend
to show more head streaking in autumn. This is
especially true for Azores Gull, which regularly ex-
hibits a hood of strong streaking that is restricted to
the head and that covers the forehead, lore and
chin. Often, the malar area is distinctly streaked. In
other populations, the head streaking is usually not
so intense, although a few birds from the Canary
Islands, Madeira and even Galicia can show a very
similar strong hood, so this character seems not
restricted to the Atlantic islands. We have seen
several birds in Galicia with such a head pattern;
since their mantle colour was the same as in the
local Yellow-legged Gulls, these were probably
just local birds. In addition, a few colour-ringed
birds born in Asturias have shown rather strong
streaking not just around the eye but also on fore-
head, lore and malar area (although not the chin/
throat). However, most have shown some streak-
ing on hindneck too, making the hood look less
neatly demarcated than in Azores Gull. Such birds
were young adults when sporting such a head pat-
tern, in their fth- or sixth-cycle. Even subadult
michahellis can show a hood, though only rarely
so (plate 409). Along the Iberian Atlantic coast,
head streaking can be much more extensive than
in michahellis, reaching far down onto the lower
Identication of Azores Gull
308
404 Azores Gulls / Azorengeelpootmeeuwen Larus michahellis atlantis, adult, Terceira, Azores, 21 February 2015
(Chris Gibbins). In addition to their dark, slaty-grey upperparts and deep yellow leg colour (in breeding plumage),
adult Azores also typically show extensive black at base of wing-tip (just visible here below tertials). 405 Azores Gull /
Azorengeelpootmeeuwen Larus michahellis atlantis, adult, Azores, 15 October 2016 (Asier Aldalur). In autumn, head
streaking typically forms isolated hood that does not reach down to lower neck area and that covers entire head in-
cluding area around bill base. A few birds are already white-headed by this time of year though. 406 Azores Gull /
Azorengeelpootmeeuw Larus michahellis atlantis, adult, Azores, 14 October 2016 (Asier Aldalur). Although autumn
birds often show eye-catching head pattern, identication can be problematic since outer primaries are still growing
and primary pattern is therefore incomplete. Note, however, that dark ‘shin pads’ still remain on tarsus of this bird.
407 Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis, adult, Terceira, Azores, 18 February 2015 (Peter
Adriaens). Nearly one out of every two birds lacks grey base on outer web of p8. There is usually only one white mir-
ror, black spot on p4, and grey tongue on underside of p10 is very short.
neck, where it may even form a brown ‘shawl’
(plate 414). Some adult birds from Atlantic Iberia
may retain head streaking until February, i.e. much
longer than in michahellis and birds from the
Azores, which both often become white-headed
by November.
iriS
As already noted by Dubois (2001), the iris of adult
Azores Gull is often slightly paler than in michahell-
is, sometimes even whitish. However, some birds
in the Canary Islands, Madeira, Morocco and the
Iberian Atlantic coast can show a similarly pale
iris. Nominate michahellis shows a yellow iris,
sometimes with dark pigmentation. Differences
are (very) subtle though, and we did not score the
iris colour in our sample of birds.
upperpartS
Kodak grey scale values of the upperparts of all
groups studied are shown in table 3. Adult birds
from the Azores have clearly darker grey upper-
parts than nominate michahellis; their upperparts
can be nearly as dark as in British Lesser Black-
backed Gull L fuscus graellsii, although with a
more bluish tinge. Birds from the Canary Islands
and Madeira are similarly dark, as are many birds
from Portugal and Morocco. Birds from Galicia
Identication of Azores Gull
404
406
405
407
309
408 Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis, adult, Terceira, Azores, 18 February 2015 (Peter
Adriaens). Another bird with typical primary pattern, showing only one white mirror on each wing, black spot on p4,
very short tongue on p10 and full black outer web of p8 up to primary coverts. 409 Mediterranean Yellow-legged
Gull / Mediterrane Geelpootmeeuw Larus michahellis michahellis, (sub)adult, Port de Palma, Mallorca, Balearic
Islands, Spain, 4 November 2006 (Maties Rebassa). Relatively paler, more bluish-grey upperparts, big white primary
tips and location make this most likely nominate bird but its head pattern could certainly cause confusion with
Azores Gull L m atlantis. 410 Yellow-legged Gull / Geelpootmeeuw Larus michahellis, adult, with Lesser Black-
backed Gull / Kleine Mantelmeeuw L fuscus, Ares, Galicia, Spain, 19 November 2007 (Antonio Guttierez). Ringed
as pullus near Xove, Galicia; at time of photograph, it was young adult (in its fth-cycle). Note very extensive streak-
ing on head and neck. 411 Yellow-legged Gull / Geelpootmeeuw Larus michahellis, adult, Sagres, Portugal,
8 November 2018 (Peter Adriaens). Birds wintering in mainland Portugal can show extensive head streaking, though
it tends to be sparser and more diffuse around bill base (as can be seen on chin of this bird) and often extends down
on neck.
and Asturias, Spain, show more intermediate up-
perparts coloration, overlapping with darker ex-
amples of nominate michahellis. Contra Olsen &
Larsson (2004) and Olsen (2018), their upperparts
do not appear paler in the eld; rather, when
ringed adult michahellis turn up in Galicia, it is
often apparent that the local Yellow-legged Gulls
appear subtly darker. By contrast, adults from
Cantabria and País Vasco in northern Spain look
rather pale, not much darker than British Herring
Gull L argentatus argenteus and overlapping with
paler nominate michahellis.
primary pattern
Details of the primary pattern of adult birds
throughout the Atlantic and Iberian region are
shown in table 5. Birds from the Azores show the
most extensive black on the outer primaries. Most
of the adult Azores Gulls (94%) in our sample
lacked a white mirror on p9, most showed some
black pattern on p4 (78%) and, importantly, nearly
half showed an entirely black outer web on p8. In
adult michahellis through out the entire Medi ter ra-
nean region, only 19-26% lacked a white mirror
on p9, only 24-38% showed some black on p4
and in only one bird (0-1%) the outer web of p8
Identication of Azores Gull
408
410
409
411
310
Ca na ry Islands, while
about half of the birds from
Ma deira and the Canary Is-
lands showed some black
on p4.
Although there is consid-
erable overlap, the primary
pattern of adult Yel low-
legged Gulls from north ern
and north-western Spain
differs subtly from nomi-
nate michahellis in that
more birds (43%) lack a
white mirror on p9 and in
more birds (71% versus 49-
61%) the grey tongue on
the inner web of the outer-
most primary (p10) is short,
ie, less than one-third of the
length of the feather. An-
dres Ber me jo (in litt) found
a similar result in that 49%
of his sample of 111 adult
birds from north-western
Spain lack ed a white mirror
on p9.
It is worth noting that we
could not nd any clear
differences in primary pat-
tern between the birds
from País Vasco in our
sam ple and those from Ga-
li cia, while birds from Por-
tugal differed from the
Basque birds by being less
likely to show a white mir-
ror on p9 and a bit more
likely to show some black
pattern on p4. Garner &
Quinn (1997), quoting
Teys sèdre (1983), also
point ed out slight differ-
ences in primary pattern
between birds from Portu-
gal, País Vasco and nomi-
nate michahellis from
south ern France. However,
their notes are puzzling
since they describe the op-
posite of our results, as they report Basque birds
averaging more extensive white on p9 than micha-
hellis and tending to show a longer grey tongue on
p10. Their description was copied by Olsen &
Larsson (2004) and Olsen (2018) but we have no
way of conrming it; in our samples the Basque
was fully black up to the primary coverts. The com-
bination of these characters is useful for identica-
tion, and is discussed in the next section.
The lack of a white mirror on p9 is also a notable
feature in many adult Yel low-legged Gulls from
Ma deira and Portugal and, to a lesser extent, the
TABLE 5 Frequencies of different wing-tip features of adult Yellow-legged Gulls Larus
michahellis included in our analysis; values show percentage of sample birds
showing each feature. For sample sizes, see table 1.
Identication of Azores Gull
feature
category
pattern illustration
Azores
Canary Islands
Madeira
Berlengas
(Portugal)
Morocco
North-western
Spain
Eastern Spain
Gibraltar
Eastern
Mediterranean
p10 A
tongue >
of length of
feather
0% 0% 0% 1% 0% 0% 1% 10% 10%
B tongue 1/3-1/2
4% 33% 39% 12% 56% 28% 38% 40% 41%
C tongue < 1/3
93% 67% 61% 84% 44% 71% 61% 50% 49%
D no pale tongue
2% 0% 0% 2% 0% 1% 0% 0% 0%
p9 A no mirror
94% 61% 78% 73% 63% 43% 26% 23% 19%
B
mirror
confined to
one web
5% 23% 10% 22% 31% 33% 31% 26% 37%
C mirror on both
webs
2% 16% 11% 6% 6% 24% 43% 52% 44%
p8 A
entire outer
web black to
primar
y coverts
(PC )
42% 9% 1% 10% 0% 3% 1% 0% 0%
B
black reaches
PC only along
outer edge
17% 11% 10% 32% 13% 6% 5% 4% 1%
C
black falls up
to 1/3 of length
of feather short
of PC
41% 76% 85% 56% 75% 83% 80% 76% 74%
D
black falls
>1/3 to 1/2
short of PC
0% 4% 3% 2% 13% 7% 13% 20% 25%
E black falls >
1/2 short of PC
0% 0% 0% 0% 0% 0% 1% 0% 0%
p4 A no black
23% 45% 55% 64% 44% 78% 72% 71% 6 2%
B black spot on
only one web
73% 54% 41% 35% 50% 18% 26% 26% 33%
C both webs but
broken
3% 1% 4% 1% 6% 3% 2% 3% 5%
D complete
band
3% 0% 0% 0% 0% 1% 0% 0% 0%
p3 A no black
95% 100%
100%
100% 100% 100% 100%
100% 100%
B black spot on
only one web
5% 0% 0% 0% 0% 0% 0% 0% 0%
1/2
311
412 Yellow-legged Gull / Geelpootmeeuw Larus michahellis, adult, Sagres, Portugal, 7 December 2016 (Peter
Adriaens). This bird shows quite extensive black on outer primaries including black spot on p4 but, on outer web of
p8, black colour does not reach primary coverts. 413 Presumed hybrid gull / vermoedelijk hybride meeuw Larus,
adult, Rotterdam, Zuid-Holland, Netherlands, 31 March 2014 (Merijn Loeve). This colour-ringed bird (ringed as adult)
was breeding in harbour of Rotterdam annually in 2014-16. Its primary pattern is very similar to typical Azores Gull
L michahellis atlantis but note thin sliver of grey at base of p8. 414 Yellow-legged Gull / Geelpootmeeuw Larus
michahellis, (sub)adult, Sagres, Portugal, 27 November 2016 (Peter Adriaens). A few birds wintering in mainland
Portugal show impressive brown ‘shawl’ on neck, which, together with pale, staring iris, may even be reminiscent of
American Herring Gull L smithsonianus. 415 Putative ‘Macaronesian’ Yellow-legged Gull / mogelijke ‘Macaronesische’
Geelpootmeeuw Larus michahellis, subadult, with adult Yellow-legged Gull / Geelpootmeeuw in background, Cariño,
Galicia, Spain, 4 January 2012 (Antonio Gutierrez). This bird still showed signs of immaturity, such as extensive black
on primary coverts and alula, which means that large amount of black on outer primaries cannot be considered
diagnostic of Azores Gull L m atlantis here. Still, head pattern with restricted hood that is solidly dark in front of eye,
as well as slaty-grey upperwings clearly darker than in local Yellow-legged Gulls seem to indicate this subspecies or at
least seem to point to Macaronesian origin.
birds (n=89) tend to show a little more black on
the outer primaries than michahellis, not less (see
table 5). A sample of 26 adult birds from Cantabria
(Andres Bermejo in litt) further corroborates our
results (half of his sample birds lacked a mirror on
p9, and 85% showed only a short grey tongue on
p10). In another sample of 155 adult birds from
País Vasco (Arrizaga et al 2008), 34% were found
to lack a mirror on p9, ie, again a bit more fre-
quently than in michahellis. The problem seems to
be that Teyssèdre’s (1983) data from País Vasco was
based on only 12 birds, too few for meaningful
conclusions, but several subsequent works have
relied on these data so the ‘myth’ has perpetuated.
Our small sample of adult Yellow-legged Gulls
from Morocco (n=33) suggests that the primary
pattern may be rather similar to that of michahellis,
with 56% of these birds showing a similar, distinct
grey tongue on p10 (longer than one-third of the
length of the feather). However, the Moroccan
birds less frequently show a white mirror on p9
(absent in 63%) and more often have some black
pattern on p4 (in 56%).
The primary pattern of birds in the Gibraltar-
Identication of Azores Gull
412
414
413
415
312
416 Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis, rst-cycle, Azores, 10 October 2016 (Asier
Aldalur). In early autumn, rst-cycle birds are often less characteristic than during winter since their greater coverts are
still fresh, with barred pattern. In this respect, they do not differ from Madeiran, Canarian or even Iberian birds, which
may complicate identication. As an aside, this bird shows advanced moult: not only scapulars but also many median
and lesser coverts have been replaced. 417 Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis, rst-cycle,
Terceira, Azores, 21 February 2015 (Chris Gibbins). During winter, wing-coverts become worn and look uniformly
brown, almost oil-stained. Note also zebra-like barring on ank (above legs). 418 Azores Gull / Azorengeelpootmeeuw
Larus michahellis atlantis, rst-cycle, Terceira, Azores, 16 February 2015 (Peter Adriaens). Note dusky head with dark
forehead and distinct streaking on throat, dark scapulars, chestnut colour of wing-coverts which lack clearly barred
pattern, and breast pattern recalling that of Eurasian Scops Owl Otus scops. Like this bird, Azores can look quite
dainty and short-legged. 419 Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis, rst-cycle, Terceira,
Azores, 21 February 2015 (Peter Adriaens). Very typical bird with dusky forehead and lore, streaked throat, dark,
messy pattern on scapulars, uniformly chestnut brown wing-coverts, ‘zebra pattern’ on ank, and smooth grey-brown
colour on upper mantle.
Ceuta area appears to be identical to that of nomi-
nate michahellis but our sample from this region
was small (n=32). Our sample from western Anda-
lucía (Cádiz, Huelva) was too small (n=10) to ana-
lyse these birds separately, so we included them
with our Spanish sample from the Medi terranean
region. However, it is interesting to note that the
primary pattern of these 10 birds was clearly more
similar to michahellis than to the breeding birds
from Portugal (despite the latter being geographi-
cally closer).
In summary, if we rank the populations we have
sampled from most extensive black to least exten-
sive black on the outer primaries, the order is
roughly as follows: 1 Azores; 2 Madeira, Canary
Islands and Portugal; 3 Morocco; 4 northern and
Identication of Azores Gull
416
418
417
419
313
420 Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis, rst-cycle, with Lesser Black-backed Gull / Kleine
Mantelmeeuw L fuscus, rst-cycle, Terceira, Azores, 16 February 2015 (Peter Adriaens). This photograph allows inter-
esting comparison. Lesser Black-backed in background shows whiter lore and chin and more regular pattern on
scapulars than Azores in front. It also has fresher, darker wing-coverts with bold white fringes, and lacks vertical bar-
ring on ank. 421 Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis, rst-cycle, Terceira, Azores,
17 February 2015 (Peter Adriaens). In ight, upperwing looks almost uniformly dark, with very dark greater coverts
and inner primaries, and tail often shows broad dark band and fairly extensive dark spotting on outermost tail-feather.
This bird has unusually pale and neat pattern on scapulars. 422 Azores Gull / Azorengeelpootmeeuw Larus michahel-
lis atlantis, rst-cycle, Terceira, Azores, 21 February 2015 (Chris Gibbins). Very typical individual with plain brown
upperwings, broad dark tail band, extensive dark centre on scapulars, brown head and uniformly grey-brown collar
around neck. 423 Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis, rst-cycle, Terceira, Azores,
17 February 2015 (Peter Adriaens). Underwing is usually very dark brown. Note distinct, dense ‘zebra barring’ on rear
ank of this bird.
north-western Spain; 5 eastern Spain/Gibraltar (ie,
western Mediterranean region); and 6 eastern
Medi terranean.
moult
Yellow-legged Gulls in the Mediterranean region
have a rather early breeding season and their an-
nual complete moult is therefore also earlier than
in many other European gulls. Egg-laying starts in
early April and the primaries are moulted between
late April and early November. In the Atlantic re-
gion, however, the climate is different, and breed-
ing generally starts later, which also delays the
complete moult. In the Azores and Madeira, egg
laying starts from mid-April at the earliest, while in
Portugal, Cantabria and País Vasco it is even later,
Identication of Azores Gull
420
422
421
423
314
from late April to early May.
In Portugal and Galicia, it is not unusual to see
adult birds still completing their primary moult in
late November or even early December (own data;
Teyssèdre 1983, Pons et al 2004, Arizaga et al
2012). In Morocco, some adult birds start moult-
ing their innermost primary in late April but moult
must vary considerably, because in late November
some Yellow-legged Gulls have still not completed
the moult of the outermost primaries (own obser-
vations). In the Canary Islands, which are further
south than all other breeding areas, egg laying
starts in early April as in michahellis, and moult is
similar to this taxon (Aldalur 2016).
424 Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis, rst-cycle, Vestmannaeyjar, Iceland, 1 May 2010
(Omar Runolfsson). Vagrant bird found almost 3000 km from Azores and constituting northernmost accepted record
of this taxon. It shows classic dusky brown head (including dark lore and chin), dark, heavily marked underparts in-
cluding breast, plain chestnut brown wing-coverts and extensive dark scapular centres. Other photographs of this bird
revealed dark inner primaries, white tail with contrasting, well demarcated black band, and a few moulted wing-
coverts in right wing. 425 Yellow-legged Gull / Geelpootmeeuw Larus michahellis, rst-cycle, Madeira, 15 January
2016 (Merijn Loeve). Differs from typical Azores Gull L m atlantis in its overall slightly paler brown plumage with
paler lore, throat and breast. Lack of bold dark centre on scapulars is also good feature here. 426 Yellow-legged Gull
Larus michahellis, rst-cycle, Sagres, Portugal, 30 November 2016 (Peter Adriaens). Ringed as juvenile near Olhão,
southern Portugal. Head and underparts quite dark but note barred pattern on greater coverts and neat, regular pattern
on scapulars. 427 Yellow-legged Gull / Geelpootmeeuw Larus michahellis, rst-cycle, Sagres, Portugal, 31 October
2018 (Peter Adriaens). Another dark bird from mainland Portugal, with rather dusky face and heavily marked under-
parts, including Otus owl-like pattern on breast. Differs from typical Azores Gull L m atlantis in bold white fringes on
wing-coverts, neat, regular anchor pattern on scapulars, and more widely spaced vertical barring on rear ank.
Identication of Azores Gull
424
426
425
427
315
Identication
Characteristic features of adult Azores Gull
In this section, we focus on traits that might allow
vagrant Azores Gulls to be identied. Even though
adult Azores Gulls have the most extensive black
on the primaries of all Yellow-legged Gulls and
have rather distinctive head streaking in autumn,
birds of this age class are among the most difcult
to identify in a vagrant context, due to wide over-
lap with other populations and due to the limited
number of useful features.
In autumn, the head pattern, with strong streak-
ing creating a restricted hood and the neck being
clean white, may be the rst thing that catches the
eye. Importantly, the lore and forehead often show
dark markings, and the malar area is streaked. Still,
some birds from the other Atlantic islands and
from Atlantic Iberia can look similar, and even the
occasional (subadult) michahellis can look hooded
(cf plate 409). Therefore, as many features as pos-
sible should be considered when dealing with a
bird outside of its normal range, including upper-
parts and primary pattern. The following combina-
tion of features in the primary pattern provides a
rm basis for identication: 1 no grey tongue or
just a short one on p10 (shorter than one-third the
length of the feather); 2 no white mirror on p9;
3 fully black outer web on p8 up to primary
coverts (ie, no grey visible at base of outer web);
and 4 black spot or band on p4. This combination
was shown by 37% of the adult Azores Gulls in
our sample and by only one bird from other popu-
lations (from the Canary Islands).
An additional character worth mentioning is
that Azores Gull seems to be the only Yellow-
legged Gull in which the adults sometimes show a
black mark on p3. Nevertheless, this remains rare
in Azores Gull, shown by only 5% in our sample;
thus, one would be lucky to encounter such a bird
out of range.
When combined with head streaking, primary
pattern should help to secure identication.
However, there is something of a ‘catch 22’ in that
head streaking is only present in autumn (August-
November), when birds are actively moulting their
primaries. This means that in many birds not all
features will be visible; in autumn one or two cru-
cial primaries may be missing and, conversely,
when all primaries are in place, head streaking is
lost. This problem can make it impossible to distin-
guish a genuine Azores Gull from the odd Yellow-
legged Gull from the Canary Islands with maxi-
mum amount of black, unless a black mark on p3
is also present. Fortunately, the proportion of birds
from the Canary Islands with maximum amount of
black on primaries is very small (less than 2%) and
the population appears to be largely sedentary (see
‘Movements and vagrancy’). Another difculty is
that correct ageing is essential, because subadult
Yellow-legged Gulls can show very adult-like
plumage but with more extensive black on prima-
ries and with stronger head streaking. Still, subtle
features will usually give away the age of such
birds, such as bold black markings on the bill,
black pattern on the alula, thick black streaks on
the primary coverts or brown pattern on a few
wing-coverts. It should be noted that adult Yellow-
legged Gulls from the Atlantic region regularly
show black shaft streaks on the primary coverts but
these are only thin.
In bad light, when the colour of the upperparts
cannot be judged accurately, separation from
British Lesser Black-backed Gull can be problem-
atic but, in normal conditions, the upperparts of
adult Azores Gull will always look slightly paler
with a more bluish tinge, and in ight the wing-tip
contrasts more strongly with the rest of the upper-
wing and, particularly, underwing. Autumn head
streaking usually differs in that in Azores Gull it is
concentrated on the head and face while the neck
is clean white. The body is often a little bulkier,
more robust, and the white scapular and tertial
crescents tend to be slightly narrower than Lesser
Black-backed Gull.
Characteristic features of other ‘Atlantic’ gulls in
adult plumage
Most adult Yellow-legged Gulls from populations
other than the Azores cannot safely be told from
nominate michahellis in a vagrant context, al-
though a few birds from Atlantic Iberia show such
an extensive brown ‘shawl’ on lower hindneck
that their origin could at least be suspected (plate
410 and 414). Also, it is worth mentioning that a
handful of birds from the Canary Islands (2%),
Madeira (2%) and Portugal (6%) sported a combi-
nation of three features in the primary pattern that
was not shown by any of the adult nominate
michahellis in our samples: 1 no white mirror on
p9; 2 completely black outer web on p8; and
3 black spot or band on p4.
Hybrid problems
Adult hybrid ‘yellow-legged’ gulls have been doc-
umented in breeding colonies in Belgium and the
Netherlands (Cottaar 2004, Adriaens et al 2012,
Cottaar et al 2019). They are assumed or have
been proven to be the result of mixed breeding
between European Herring Gull L argentatus and
Identication of Azores Gull
316
428 Yellow-legged Gull / Geelpootmeeuw Larus michahellis, rst-cycle, Sagres, Portugal, 8 November 2018 (Peter
Adriaens). Ringed as pullus on Berlenga Islands, Portugal, illustrating how heavily marked tail base can be in Atlantic-
Iberian region. Note that separation of such birds from Lesser Black-backed Gull L fuscus in the eld can be nearly
impossible. 429 Yellow-legged Gull / Geelpootmeeuw Larus michahellis, rst-cycle, Galicia, Spain, 6 December 2015
(Pierre-André Crochet). Dark individual but with barred pattern on greater coverts, distinct white fringes on wing-
coverts, and regular pattern on scapulars. 430 Mediterranean Yellow-legged Gull / Mediterrane Geelpootmeeuw Larus
michahellis michahellis, rst-cycle, Sagres, Portugal, 26 November 2016 (Peter Adriaens). Ringed as pullus near
Malaga, southern Spain, and should therefore be considered nominate michahellis. Note dainty and elongated struc-
ture with short legs though, quite similar to Lesser Black-backed Gull L fuscus. Note also dark shin pads on tarsus.
431 Lesser Black-backed Gull / Kleine Mantelmeeuw Larus fuscus, rst-cycle, Daytona Beach, Florida, USA,
11 February 2019 (Will Chateld-Taylor). Bird showing some similarities to Azores Gull L michahellis atlantis, such as
extensive streaking on head including forehead, lore and throat, and Otus owl-like pattern on breast. There are clear
differences too, however, like barred greater coverts, distinct white fringes on many wing-coverts, regular pattern on
scapulars, and more widely spaced barring on rear ank.
Lesser Black-backed Gull or between Yellow-
legged and either of these two species, with
Yellow-legged x Lesser Black-backed Gull being
the most frequent one. This hybrid combination
can be problematic with regard to the identica-
tion of adult Azores Gull, since such hybrids have
shown a mantle colour intermediate between that
of the two parent species, bright yellow legs, bright
red orbital ring, large red gonys spot and extensive
black on the outer primaries. Hybrids have also
been documented backcrossing with Lesser Black-
backed and successfully rearing young. Mixed
pairs of Yellow-legged and Lesser Black-backed
have also been reported from the Berlengas,
Portugal, and Galicia (Paterson 1997) but descrip-
tions or photographs of the offspring of these birds
Identication of Azores Gull
428
430
429
431
317
432 Lesser Black-backed Gull / Kleine Mantelmeeuw Larus fuscus, rst-cycle, Portimão, Portugal, 10 October 2018
(Carl Baggott). Bird ringed as pullus in Longford, Ireland. It is included here as example of pale end of variation: very
pale new scapulars, whitish head with dark eye mask, one newly moulted median covert, and worn tertials and inner
greater coverts make this bird extremely similar to Yellow-legged Gull L michahellis. It has retained a few juvenile
scapulars but that can also be the case in Cantabrian or Portuguese Yellow-legged Gulls at this date. 433 Azores Gull /
Azorengeelpootmeeuw Larus michahellis atlantis, early second-cycle, São Miguel, Azores, 10 August 2005 (Pierre-
André Crochet). In summer of their second calendar-year, many Azores acquire uniform dark brown plumage without
any grey on upperparts. Head is still very brown and typically darkest in front of eye. Note also extensive dark centres
on scapulars, as well as zebra-like barring just above legs. Some birds show even darker greater coverts than this indi-
vidual, lacking any pale barring. Other populations of Yellow-legged Gull L michahellis do not normally become this
dark and uniform. 434 Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis, second-cycle, Terceira, Azores,
20 February 2015 (Chris Gibbins). When moult to second winter plumage has nished, plumage is variable and shows
lots of overlap with other populations of Yellow-legged Gull L michahellis. However, birds like this one, showing iso-
lated brown belly patch in combination with isolated hood of head streaking and uniformly dark greater coverts (at
most with minimal white peppering along edges), can be assumed to be from Azores population. Note also dark ‘shin
pads’ in this bird. 435 Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis, second-cycle, Terceira, Azores,
20 February 2015 (Peter Adriaens). Very typical individual with plain dark brown greater coverts, isolated brown
belly patch, concentrated streaking on head (heaviest in front of eye and on throat), dark ‘shin pads’ on tarsus, and dark
slaty-grey upperparts.
Identication of Azores Gull
432
434
433
435
318
436 Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis, second-cycle, Terceira, Azores, 17 February 2015
(Peter Adriaens). Note belly patch and extensive head streaking. Greater coverts form rather solid dark wing-panel.
437 Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis, second-cycle, Terceira, Azores, 18 February 2015
(Chris Gibbins). Another bird showing typical combination of features on head, belly and greater coverts. Inner
greater coverts show only minimal white peppering, while outer ones are solidly dark. Underwing very dark, and tail
showing extensive black band. 438 Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis, second-
cycle, Terceira, Azores, 21 February 2015 (Chris Gibbins). Brown greater coverts of this bird show minimal white pep-
pering. 439 Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis, second-cycle, Terceira, Azores,
17 February 2015 (Peter Adriaens). Note same features as in plate 434-438, including solid dark panel on greater
coverts.
have not been published. In addition, mixed
breeding between Yellow-legged Gull and Arme-
nian Gull L armenicus has been proven in central
Turkey (Liebers & Helbig 1999), and the hybrid
offspring may theoretically resemble Azores. How-
ever, Armenian Gull tends to show a dark iris, and
nominate Yellow-legged Gull has rather dull yel-
low iris, so hybrids can be expected to show a
slightly darker eye colour than adult Azores.
Armenian has a restricted, eastern range in the
Western Palearctic, and its offspring is probably
not very likely to reach the Atlantic region. Mixed
breeding of Yellow-legged Gull with Caspian Gull
L cachin nans and European Herring Gull is known
from Poland (Neubauer et al 2010) but such hy-
brids do not show dark grey upperparts nor exten-
sive black on outer primaries and so do not really
match Azores.
We examined the primary pattern of 37 adult
hybrid gulls breeding in colonies in Belgium and
the Netherlands and found that none showed the
characteristic combinations of features discussed
for ‘Atlantic’ Yellow-legged Gulls. All showed
some grey at the base of the outer web of p8 and/
Identication of Azores Gull
436
438
437
439
319
440 Yellow-legged Gull / Geelpootmeeuw Larus michahellis, second-cycle, Canary Islands, 16 January 2010 (Eric
Didner). From this angle, this bird with its isolated brown belly patch, isolated hood and very dark underwing looks
identical to Azores Gull L m atlantis. Other photographs, however, revealed distinct white fringes to outer greater
coverts, as well as rather bold white spotting on inner. For this reason, we feel that it is more likely local bird than
vagrant from the Azores, although we admit that difference is subtle here. 441 Yellow-legged Gull / Geelpootmeeuw
Larus michahellis, second-cycle, Madeira, 16 January 2016 (Merijn Loeve). Although dark head of this bird might sug-
gest Azores Gull L m atlantis, it clearly differs from identiable types in plate 433-438 because neck and breast are
heavily marked and brown pattern on belly is not isolated (as streaking continues up to breast). Note also white barring
on one inner greater covert, as well as broad white tips to outer ones. 442 Yellow-legged Gull / Geelpootmeeuw Larus
michahellis, second-cycle, Sagres, Portugal, 30 November 2016 (Peter Adriaens). Bird ringed in its rst-cycle near
Olhão, Portugal, in December 2015. In autumn or early winter of their second-cycle, many Portuguese birds show
only limited grey on upperparts. This, together with their rather brown underparts can make them quite similar to
European Herring Gull L argentatus but they usually show bold, rounded dark spots on lesser coverts (as in this bird,
on outer feathers). 443 Yellow-legged Gull / Geelpootmeeuw Larus michahellis, second-cycle, Sagres, Portugal, 5
November 2018 (Peter Adriaens). Rather dark bird, which overlaps with some Azores Gulls L m atlantis in plumage,
although that taxon usually shows less white in wing-coverts. Still, birds like this are one of reasons why only second-
cycle Azores of type shown in plate 434-439 can be safely identied in vagrant context.
or a white mirror on p9. One bird from the Nether-
lands wearing a white leg ring with black inscrip-
tion 9.8 (plate 413) looked most similar to Azores
Gull but still sported a thin sliver of grey at the
base of the outer web of p8. Only a few of these
hybrids have been documented in autumn and
winter so far but those that exhibited head streak-
ing differed from Azores in that streaking extended
onto the neck. However, one colour-ringed bird
from the Netherlands (green AH6) already sported
Identication of Azores Gull
440
442
441
443
320
444 Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis, third-cycle, Azores, 10 October 2016 (Asier
Aldalur). In vagrant context, identication of third-cycle Azores is generally not possible but birds with blackish bill,
isolated hood of head streaking, dark slaty-grey upperparts and dark ‘shin pads’ on tarsus can readily be assumed to
be from Macaronesian islands. This combination of features is not common, however. 445 Azores Gull /
Azorengeelpootmeeuw Larus michahellis atlantis, third-cycle, Azores, 15 October 2016 (Asier Aldalur). Note blackish
bill, very dark hood, whitish iris and dark brown greater coverts of this bird. Aged by adult-like inner primaries.
446 Yellow-legged Gull / Geelpootmeeuw Larus michahellis, third-cycle, Lanzarote, Canary Islands, 4 February 2017
(Juan Sagardia). This bird also shows combination of extensive head streaking, large amount of black on bill, slaty-grey
upperparts and dark ‘shin pads’. Iris is very pale. Compared with Azores Gulls L m atlantis in plate 444-445, loral area
and chin are whiter, and only distal half of bill is black. Other Canarian and Madeiran birds are trickier though, and
suggest that there may be complete overlap with birds from Azores. 447 Yellow-legged Gull / Geelpootmeeuw Larus
michahellis, third-cycle, Sagres, Portugal, 20 November 2016 (Peter Adriaens). Bird with blackish bill and dark ‘shin
pads’ but with dark iris and lacking strong dark hood. Upperparts also slightly paler grey than in most Macaronesian
birds.
an immaculate white head and neck on 24
December 2001 and again on 10 January 2003
(Adriaens 2003), ie, early in winter like Azores. An
additional, relatively helpful feature is that the
white tertial and scapular crescents of hybrids tend
to be a little wider than in Yellow-legged, some-
times obviously so.
In a North American context, hybridisation be-
tween Lesser Black-backed Gull and American
Herring Gull L smithsonianus has been document-
ed from Appledore, Maine, USA (Ellis et al 2014),
and such hybrids are sometimes referred to as
Appledore Gulls’. Not much is known about this
type of hybrid, since few of them have been ringed
or studied on their breeding colonies but, as can
be expected of the offspring of two parent species
Identication of Azores Gull
444
446
445
447
321
with heavy head streaking, they seem to show ex-
tensive streaking on the neck from late autumn
(October) to February. They also appear to moult
late, with the occasional bird still growing outer
primaries in January. The primary pattern may be
helpful: in a sample of 25 adult putative hybrids
(22 on eBird.org and three that we observed in
Newfoundland, Canada), none showed the char-
acteristic combination of features that we dis-
cussed for ‘Atlantic’ Yellow-legged Gulls, and
nearly half of them showed a prominent white
tongue tip on p7 – a trait presumably inherited
from the American Herring parent, and usually ab-
sent in Azores Gull. Leg colour in these hybrids
seems to be variable, often being a strange mix of
pink and yellow but in some birds it is bright yel-
low. As in European hybrids, the white tertial and
scapular crescents may appear slightly wider and
more prominent.
First-cycle birds
First-cycle Azores Gulls look quite different to rst-
cycle michahellis, especially after the post-juvenile
moult, when the pattern of the new scapulars adds
to the difference. Contra Dubois (2001), Olsen &
Larsson (2004) and Olsen (2018), we found no
Azores that retained a largely juvenile plumage
into the winter; our observations indicate that
birds start replacing their scapulars (and wing-
coverts) in July-August and nish in October-
November, ie, slightly later than michahellis. In a
sample of 151 rst-cycle Azores from October,
150 were actively moulting their scapulars and 88
were also replacing wing-coverts. In our February
visit to the Azores, most showed a full set of new
scapulars.
First-cycle Azores Gulls show darker plumage
overall than michahellis, with dusky face, solid
brown ear-patch, solid grey-brown hindneck and
upper mantle (which may even recall American
Herring Gull), and rather dark brown breast and
belly. In fresh juvenile plumage, the greater wing-
coverts tend to be more densely barred than in
michahellis. From October onwards, these feathers
wear to a very uniform brown colour and then be-
come a typical chestnut brown, often looking as if
they are oil-stained. Unlike michahellis, these
feathers tend to lack pale fringes when no longer
fresh. Some birds show a dense pattern of dark
streaks and crossbars on the breast, almost recall-
ing that of some owls (notably Eurasian Scops Owl
Otus scops or Eastern Screech-Owl Megascops
asio). After the post-juvenile moult, the scapulars
usually show a messy, mixed pattern with various
extensive dark centres to many of these feathers,
quite unlike the regular anchor pattern seen in
rst-winter michahellis. From October onwards,
some birds show bold, ‘zebra-like’ vertical barring
on the rear ank; these ‘black-and-white’ bars
contrast against the relatively plain brown belly
and fore ank. Michahellis often shows vertical
barring too but it is generally thinner, paler, more
widely spaced and does not contrast as much.
Many Azores (but far from all) show dark ‘shins’
(dark bars across the tarsus) throughout the entire
rst-cycle, which are seen in few michahellis only
(but see plate 430). Some birds already show a
somewhat pale, amber-coloured iris during their
rst winter. In ight, rst-cycle Azores generally
shows a broader black tail band than michahellis,
and some birds even show dark barring or spotting
at the base of the outermost tail-feather (unlike
michahellis, which usually has a clean white base).
Rump and uppertail-coverts can look barred (ver-
sus spotted in michahellis). Note that nominate
michahellis typically combines solid, heart-shaped
spots on uppertail-coverts with a clean white tail
base, a combination that is less common in other
populations of Yellow-legged Gull (especially
Azores Gull) and in Lesser Black-backed Gull (see
gure 3). The underwing-coverts usually look
darker brown than in michahellis, and the inner
primaries tend to be darker too, with very little col-
our difference between inner and outer webs.
First-cycle Yellow-legged Gulls from the Canary
Islands and Madeira are generally more like
michahellis than Azores Gull, although they fre-
quently show a broad black tail band (sometimes
with extensive pattern on outermost tail-feather
too) and rather dark underwing-coverts. They usu-
ally show a pale head and neck like michahellis,
pale fringes on the wing-coverts, and contrasting,
regular anchor pattern on the post-juvenile scapu-
lars. However, a few birds from Madeira show
confusingly dark plumage with dusky head and
neck, plain tawny brown wing-coverts or dark,
messy pattern on replaced scapulars. Such birds
will still differ from typical Azores in their pale
chin and throat, lacking the prominent dark streak-
ing. Also, it is very rare to see these unusual fea-
tures combined in one bird; usually birds from
Madeira with dark, messy scapular pattern will
show pale fringes on their wing-coverts, and birds
with dark, plain wing-coverts are likely to show a
regular anchor pattern on scapulars.
First-cycle birds from Portugal are rather varia-
ble. Many are similar to michahellis but slightly
smaller, with shorter legs and more gently rounded
head; some are darker though, with dark, streaked
head, and may, in fact, look so similar to Lesser
Identication of Azores Gull
322
FIGURE 3 Comparison of tail patterns of rst-cycle Azores Gull / Azorengeelpootmeeuw Larus michahellis atlantis (rst
column; Chris Gibbins), Yellow-legged Gull / Geelpootmeeuw L michahellis (second column; all nominate michahel-
lis except for bottom bird, which was photographed in southern Portugal; Peter Adriaens) and Lesser Black-backed
Gull / Kleine Mantelmeeuw L fuscus (third column; Peter Adriaens)
Identication of Azores Gull
323
Black-backed Gull that they become practically
impossible to tell from this species in the eld.
While most have a rather michahellis-like tail pat-
tern, some do show a broad black band and exten-
sive barring on outermost tail-feathers (plate 428).
They can show boldly spotted rump and uppertail-
coverts and may also show an Otus owl-like pat-
tern of crossbars on the breast. However, even the
darkest birds still differ from typical Azores Gull in
several features: poorly streaked chin and throat,
cleaner, more regular pattern on the post-juvenile
scapulars, prominent pale fringes to the wing-cov-
erts, and thinner, less contrasting barring on rear
ank (if present). For the sake of completeness, we
should also mention that some rst-cycle birds
from Portugal show a pale plumage with neatly
barred greater coverts that is similar to European
Herring Gull – although the inner primaries are
darker. From experience, we can safely say that
trying to nd a rst-cycle European Herring Gull in
Portugal is hard work!
Like those in Portugal, rst-cycle Yellow-legged
Gulls from northern and north-western Spain are
variable, though dark individuals seem to be much
rarer, especially along the Cantabrian coast. Some
birds, even pale individuals, show a broader black
tail band than michahellis, and the outer tail-feath-
ers can be extensively barred.
An important pitfall for Azores Gull is the ap-
pearance of dark rst-cycle Lesser Black-backed
Gulls. Such birds can show a dark, densely streak-
ed head and thus suggest Azores. Size and shape
are not always helpful since there is a rather wide
overlap in structure between Yellow-legged Gulls
of the Atlantic region and Lesser Black-backed.
However, dark individuals of the latter species will
usually still differ from typical Azores in the fol-
lowing ways: 1 the face does not usually look as
dusky due to paler lore, forehead, malar area and
chin (rst-cycle Lesser Black-backed tends to have
whitish feathers around the bill base, and its head
pattern looks streaked rather than solidly dark);
2 even when worn, the juvenile wing-coverts and
scapulars generally retain white fringes and there-
fore do not give an oil-stained impression (many
birds show prominent pale barring on the greater
coverts, which is something that Azores mainly
shows in fresh juvenile plumage); 3 the breast is
often paler than in Azores; a few birds show the
Otus owl-like barred pattern but most look
streaked or show V-shaped dark marks; 4 after the
post-juvenile moult, the scapulars may show large
dark centres but usually in a neatly arranged, regu-
lar pattern; in dark individuals, replaced scapulars
and wing-coverts tend to show a greyish or dark
beige base colour; 5 only relatively few birds show
vertical barring on rear ank and, if present, the
dark bars tend to be thin and widely spaced; also,
the belly is only rarely uniformly dark; and 6 the
post-juvenile moult is somewhat later than in
Azores and less extensive; it is rare to see birds
with replaced wing-coverts in August-September.
As an aside, it is worth noting that some Lesser
Black-backed Gulls acquire very pale, even whit-
ish scapulars in their post-juvenile moult, with
only thin dark anchors (see plate 432). Such birds,
if they are a bit bulkier than average and show a
rather strong bill, may not be distinguishable from
Ibero-Atlantic Yellow-legged Gulls, since plumage
features including tail pattern are of no real use.
Plumage wear and extent of moult will often differ
but it is not a good sign if one has to rely on moult
as the only identication feature. Moreover,
Yellow-legged from northern and north-western
Spain can retain several juvenile lower scapulars
well into their rst winter, sometimes even into
April of their second calendar-year. Also, a few
colour-ringed Portuguese birds have been docu-
mented with a nearly full set of juvenile scapulars
in mid-November (Carl Baggot in litt).
Due to their overall dark, uniform plumage,
rst-cycle Azores Gulls could conceivably be con-
fused with American Herring Gull. However, their
vent, undertail-coverts and uppertail-coverts al-
ways look contrastingly white, the underparts are
more coarsely marked, the inner primaries are
rather uniformly dark (lacking the bold dark sub-
terminal spots of American Herring), and the tail
looks more contrastingly patterned. The bill usual-
ly looks darker throughout the whole rst-cycle,
more blackish and often with a thin sliver of pink
at the base of lower mandible only. The legs often
retain dark shins all throughout the rst-cycle. The
post-juvenile moult progresses more rapidly and
often includes some wing-coverts (though a few
American Herring also appear to replace a small
number of wing-coverts during their rst winter, as
we have observed).
First-cycle Kelp Gull L dominicanus may also be
quite similar to Azores Gull and could present a
pitfall in some regions (eg, along the Atlantic coast
of Morocco (Cape Gull L d vetula), or in south-
eastern USA (nominate L d dominicanus)). Kelp,
however, has a different structure with heavier bill
and bulbous gonys. The body is bulky and the
wings are so broad that in many standing birds the
secondaries become visible below the greater
wing-coverts. Even in fresh juvenile plumage, the
face is usually whitish and contrasts against the
brown mask around the eye. During the rst-cycle,
Identication of Azores Gull
324
head and breast quickly wear to whitish. Although
the wing-coverts are rather uniformly dark, they do
not give an oil-stained impression. The tail is often
all dark. The ank lacks the barred pattern shown
by many rst-cycle Azores. After the post-juvenile
moult, the scapular pattern is variable but gener-
ally regular (not messy) and dark, sometimes al-
ready blackish.
The appearance of rst-cycle hybrids is not re-
ally known. European hybrids are not likely to
cause more problems than Lesser Black-backed
Gull, since their plumage is not likely to be darker
than this species. Four rst-cycle hybrids and back-
crosses were shown in Adriaens et al (2012); these
birds did not really suggest Azores Gull. The situa-
tion might be worse in North America, however,
where mixed pairing between American Herring
Gull and Lesser Black-backed could theoretically
produce something that resembles Azores. In this
context, the zebra-like barring on the rear ank,
sooty face, plumage wear, extent of post-juvenile
moult, as well as the colour and pattern of the
wing-coverts all become critical features. As far as
we know, only two such hybrids of known origin
have been photographed during their rst-cycle:
both had been colour-ringed on Appledore Island,
Maine, one with code F02 and one F07 (photo-
graphs on www.gull-research.org). F02 was photo-
graphed in March of its rst-cycle and looked
rather like Lesser Black-backed (although with
whitish base colour to scapulars, making it very
similar to Yellow-legged Gulls from the Ibero-
Atlantic region); F07 was photographed as a fresh
juvenile in August and appeared intermediate be-
tween both parent species. It did not really suggest
Azores since its face already looked pale, the
brown patch behind the eye was small, the tertial
centres had distinctly jagged edges, and the wing-
coverts showed broad pale fringes.
Second-cycle birds
Only a proportion of second-cycle Azores Gulls
can be identied with condence out of range.
Separation from typical Mediterranean michahellis
is rather straightforward, and in most cases they
are also rather distinct from birds belonging to
other populations. Nevertheless, there is a signi-
cant degree of overlap and only birds showing cer-
tain diagnostic combinations can be identied
safely.
The following text concentrates on second-cycle
birds over the autumn and winter period (approxi-
mately October to April) but rst we need to point
out the distinctiveness of moulting birds in sum-
mer. Formally, birds enter their second-cycle once
the innermost primary has been dropped in April-
May of their second calendar-year. The complete
moult is well advanced by July, such that most of
the feathers they had over the winter (their rst
winter of life) will have been replaced. At this time
(July-August), the patterns on these new feathers
make second-cycle Azores Gull look very dark,
heavily marked on the head and body including
upperparts, and brown overall (plate 433). The
forehead, lore and chin often remain distinctly
streaked, unlike in other Yellow-legged Gull taxa
of the same plumage type, and obvious whitish
barring on wing-coverts or scapulars is usually ab-
sent. The scapulars often show a dark, messy pat-
tern. Birds from Madeira and the Canary Islands
tend to look paler at this time of year, with whitish
face, rather streaked hindneck (instead of blotchy
or uniform brown as in Azores Gull), some whitish
barring on wing-coverts and more regular barring
on those scapulars that are not plain grey yet. Our
collection of photographs of immature birds from
the Canary Islands and Madeira from the summer
period is small, admittedly, but it suggests they do
not acquire the very dark plumage that ‘rst-sum-
mer’ Azores show. The rst British record of Azores
was of such a dark second-cycle bird, which lin-
gered for several months in Cornwall, England,
over the summer and autumn of 2008 (Elliott
2008).
In early autumn, birds undergo a partial moult
which is critical from an identication perspective,
as it is one of the things that creates the distinctive
appearance of many second-cycle Azores Gulls
over the winter period. In very simplied terms,
there are four ‘types’ of second-cycle Azores.
These types, and their frequencies, are illustrated
in gure 4: type 1 birds with a distinct hood and
belly-patch, dark brown wings and grey saddle;
based on our own observations and photographs
from the Azores in February, this type is numeri-
cally dominant (c 60% of all birds photographed);
type 2 a variable group of birds that lack both
hood and belly patch; they are typically rather
scruffy looking, with no striking features – they can
look rather like European Herring Gulls or
Cantabrian birds and so are quite different from
type 1; type 3 birds that show either a belly patch
or a hood but not both and so are rather intermedi-
ate/show mixed characters; and type 4 advanced,
clean-looking birds with bright bare parts; overall,
they resemble typical Mediterranean birds.
While we are not suggesting that Azores Gulls
have ‘morphs’ or ‘forms’ sensu stricto, grouping
them broadly in these four types is a helpful way to
approach identication of out of range birds; the
Identication of Azores Gull
325
search image for second-cycle Azores should be
type 1, as these are the only ones that show diag-
nostic combinations of features.
The following features, when combined, are di-
agnostic for second-cycle Azores Gull: 1 dark
hood of streaks restricted to the head and includ-
ing the chin and throat; 2 isolated, dark ‘Dunlin
Calidris alpina patch’ on belly; and 3 solid, dark
brown band on greater wing-coverts. The hood is
clearly demarcated from the neck, with a very
sharp border. It is separated from the dark belly
patch by an area of unmarked white on the breast.
The dark hood tends to accentuate the pale (typi-
cally steely-grey or silver) eye. The dark belly
markings appear as a solid, isolated patch. The last
important and characteristic feature of Azores is an
extensive dark panel on the wing formed by the
plain brown greater coverts. Pale notches or bar-
ring are almost absent. It is quite common for such
birds to have a few grey feathers on the inner
greater coverts (two to three feathers) but these
tend not to contrast too much with brown ones as
they are darker grey than other taxa and often have
a distinct brown cast. This extensive brown panel,
when combined with the fact that the inner prima-
ries tend to be very dark (see below) creates an
impression on ying birds that is unlike michahel-
lis, namely that of a uniformly brown upperwing,
contrasting with a grey saddle. When added to the
hood and belly patch, the combination of features
can be very distinctive.
Second-cycle Yellow-legged Gulls from the
Mediterranean should not present any serious
identication problems in relation to type 1 Azores
Gull; they are not hooded and do not have a belly
patch. The problem comes with certain birds from
Identication of Azores Gull
0
10
20
30
40
50
60
70
1 Hooded 2 Scruffy 3 Intermediate/mixed 4 Advanced
Percentage of sample
Azores (n=331) Canary Islands (n=70) Madeira (n=42)
FIGURE 4 Frequency of different plumage types in second-cycle Azores Gull Larus michahellis atlantis relative to birds
from Canary Islands and Madeira. Sample of Azores Gull consists of total of 331 birds photographed in February;
samples of birds from Canary Islands and Madeira were taken from photographs supplied to us by numerous photo-
graphers, all from winter period (November-March). Photographs included here are all examples drawn from sample
from Terceira, Azores, in February 2015 (Chris Gibbins). 1 classic type of atlantis with both hood and ‘Dunlin patch’;
2 resembles European Herring Gulls L argentatus or some Yellow-legged Gulls L michahellis from north-western
Spain; 3 shows either hood or ‘Dunlin patch’ but not both; 4 matches Yellow-legged Gulls from other populations,
appearing unstreaked, and typically having bright bare parts and much grey on upperparts.
326
the other Atlantic islands and to a lesser degree
some from Portugal and north-western Spain. As
shown by gure 4, type 2 and 3 birds occur in
populations in the Azores and Canary Islands with
equal frequency and so such birds are not those to
look for. The relative frequency of type 1 and 4
birds is, however, reversed: around half of all birds
from the Canary Islands are type 4 (‘advanced’)
whereas only c 5% of Azores are like this, and
while most from the Azores are type 1, birds with
such a pattern are rare amongst populations from
the Canary Islands. The sample from Madeira is
interesting, as it suggests that birds from here are a
little more like Azores, in that advanced types are
less frequent than on the Canary Islands and
Azorean-like type 1 birds are a little more frequent.
Samples sizes are, however, rather small and it
would be good to investigate this intermediacy in
greater detail with a better sample. Nevertheless,
the overriding message is that c 1-2% of second-
cycle birds from the Canary Islands and Madeira
will have both a hood and belly patch and so it is
critical that the greater covert pattern of a suspect-
ed vagrant matches Azores.
In addition, there are some other plumage fea-
tures shown by many Azores Gulls that can be
used to support identication, though as they are
shared with some birds from other populations
they are not in themselves critical either way (ie,
the presence of a hood, belly patch and dark wing-
panel would override these).
Upperwing
The inner primaries have a strong brown cast on
most Azores Gulls. Some have a slight grey cast,
but these feathers are only rarely as contrastingly
clean grey as seen in typical Mediterranean micha-
hellis. Brown inner primaries can be shown by the
other populations of Yellow-legged Gull, however.
Underwing
Very often looks uniform and dark overall, strik-
ingly so for anyone familiar with Mediterranean
michahellis. Nevertheless, the pattern is extremely
variable in all populations and should not be used
as a rm identication criterion.
Bare parts
Dark areas on the front of the tarsus are a feature of
rst-cycle Azores Gull and in fact these ‘shin pads’
are regularly also present in second-cycle birds.
However, they occur in other Yellow-legged Gull
populations of this age class too so are of no real
signicance for identication.
Other features
Other features that we looked at, such as the pres-
ence of a pale tongue on p10, tail pattern, bill col-
our, pattern of longest undertail-coverts, number
of brown tertials, and the number of grey greater
coverts all appeared too variable to be of any use
in the identication process.
The appearance of typical, type 1 Azores Gulls is
distinctive and should easily rule out Lesser Black-
backed Gull, which also shows slightly darker,
slaty-grey upperparts from its second-cycle on.
Hybrids in this plumage are, again, poorly known
but are unlikely to show the distinctive combina-
tion of features described above. A second-cycle
backcross from Zeebrugge, West-Vlaanderen, Bel-
gium, looked identical to Lesser Black-backed
(Adriaens et al 2012), and a second-cycle hybrid
Lesser Black-backed x American Herring Gull
from Appledore Island appeared intermediate be-
tween its parents (Ellis et al 2014). This was the
bird with code F02 that was also photographed
during its rst-cycle (discussed earlier). In its
second-cycle, it showed extensive brown spotting
on the underparts (including breast), somewhat
paler head, dense brown barring on uppertail-
coverts, pale inner primaries, and some obvious,
pale vermiculations on greater coverts – all unlike
the typical type 1 Azores.
Third- and fourth-cycle birds
Subadult Azores Gulls (third- and fourth-cycle) are
the most difcult age group to identify, since their
primary pattern has not fully developed yet and
their plumage features overlap completely with
birds from the Canary Islands and Madeira. On
current knowledge, most birds of this age class
cannot be identied safely, at least not to the exact
island group. Some (a small minority – see below
for percentages) appear distinct from the mainland
populations, although more research may be
need ed into the variation of the subspecies lusita-
nius.
In autumn (August-December), many third-
cycle Azores Gulls show distinct dark head streak-
ing forming a well-demarcated hood, while the
neck is often mainly white. However, such a hood
is also shown by some birds from the Canary
Islands, Madeira, Galicia and even the odd nomi-
nate michahellis (plate 409). By winter, the head
becomes whiter due to a partial body moult.
On the Macaronesian Islands as a whole, a
small proportion of subadult birds show an eye-
catching combination of features, namely: 1 dark
streaking on head, including forehead, lore and
Identication of Azores Gull
327
chin, well demarcated from the white neck; 2 dark
grey upperparts (slightly darker than michahellis);
3 blackish bill; and 4 dark ‘shin pads’. This combi-
nation was shown by c 9% of the subadult Azores
Gulls in our photograph collection, 9% of birds
from Madeira and 5% of birds from the Canary
Islands. We did not nd it in the other Yellow-
legged Gull populations. A few birds with all of
these features combined have been photographed
in Galicia; since they differed from local birds in
not just one but multiple features, it seems justied
to consider them vagrants from Macaronesia.
Trying to decide which birds with the above
combination are from the Azores and which from
Canary Islands/Madeira is probably impossible. In
the latter, at most half of the bill seems to be black-
ish, while in birds from the Azores this colour aver-
ages more extensive, but this is a subtle difference
and may be more variable than our data suggests.
Note also that an extensively blackish bill can be
found in all Yellow-legged Gull populations, in-
cluding subadult nominate michahellis.
Voice
Teyssèdre (1983, 1984) demonstrated clear differ-
ences in the long call of adult Yellow-legged Gulls
from País Vasco compared with nominate micha-
hellis. Simply put, the birds from País Vasco she
recorded sounded much more like European
Herring Gull. Compared with michahellis, the
number of harmonics in the sonagrams of Basque
birds was consistently lower, the fundamental was
clearly higher, and the fth harmonic was absent
or weak. In other words, the calls sounded clearer,
higher-pitched, and far less nasal than in micha-
hellis. We had access to only few recordings from
northern Spain (two from País Vasco and one from
Asturias) but these corroborate the results from
Teyssèdre (1983, 1984): their sonagrams show
only four to ve visible harmonics, an FFm of 1.2-
1.6, an empty space where the fth harmonic
should be, and more vertical distance between the
harmonics (Dist = 0.9-1.2). In addition, their DFm
is equal to FFm.
Table 6 shows the results of our sonagram analy-
sis for all other populations of Yellow-legged Gull.
The long call of michahellis is characterized by a
low fundamental, a high number of harmonics (up
to 14), a clear H5, and little vertical distance be-
tween the harmonics. The DFm is only rarely equal
to FFm. The call sounds fairly deep, nasal, and
quite like Lesser Black-backed Gull.
In the other populations, the long call sounds
more or less intermediate between that of
michahellis and birds from País Vasco. Birds from
the Azores and Canary Islands have rather variable
long calls, some sounding similar to michahel lis
and others sounding higher pitched and less nasal,
so recalling Basque birds or even European
Herring Gull. Birds from Galicia and Portugal
(Berlengas, Peniche, Lisbon, Sagres) sound similar
to birds from País Vasco (ie, high FFm, few har-
monics, large Dist), though a bit more variable,
sometimes slightly deeper or slightly more nasal
(eg, more prominent H5 in birds from Portugal).
The same is true for birds from Madeira, surpris-
ingly. Birds from south-western Spain (Cádiz re-
gion, close to Gibraltar) sound very similar to
michahellis, although slightly higher pitched (high-
er FFm value). The sample from Morocco is prob-
ably too small to allow rm conclusions (only
three recordings) but it may be worth mentioning
that, although these few birds have rather high
pitch, they sound as nasal as michahellis (mean nH
= 8.67).
In terms of the core focus of this paper, the over-
all message from this analysis is that the long call
of adult Azores Gull offers no staring clues for
identication, but that it is interesting to consider
the regional variation in calls throughout the
whole Ibero-Atlantic region from a taxonomic
standpoint.
TABLE 6 Results of sonagram analysis of long calls in various Yellow-legged Gull Larus michahellis populations; see
section Material and methods for meaning of abbreviations in headings. Given are mean and range (for H5 only
mean).
location FFm DFm FFm = DFm nH H5 (mean) Dist
Galicia 1.27 (0.7-1.7) 1.60 (1.1-3.7) 63% 5.37 (3-9) 0.37 0.93 (0.4-1.5)
Portugal 1.19 (0.6-1.7) 1.54 (1.2-2.4) 57% 6.70 (4-10) 1.09 0.84 (0.4-1.5)
Madeira 1.49 (1.2-1.8) 2.35 (1.3-3.9) 50% 6.25 (5-9) 0.75 0.74 (0.5-1.0)
Azores 1.08 (0.5-1.7) 1.77 (1.1-3.9) 38% 7.38 (4-13) 1.03 0.63 (0.4-1.1)
Morocco 1.20 (0.8-1.5) 1.67 (1.2-2.5) 33% 8.67 (5-13) 0.67 0.63 (0.4-1.0)
Canary Islands 0.97 (0.6-1.7) 1.76 (1.3-2.3) 8% 7.75 (5-10) 1.17 0.61 (0.5-0.8)
south-western Spain 0.92 (0.7-1.6) 1.53 (1.3-1.8) 9% 8.36 (6-11) 1.18 0.60 (0.4-1.0)
nominate michahellis 0.66 (0.5-1.1) 1.45 (1.1-2.8) 7% 8.47 (6-14) 1.60 0.54 (0.4-0.7)
Identication of Azores Gull
328
Discussion
Taxonomic relations between Yellow-legged Gull
taxa
It is prudent to use multiple lines of evidence to
help assess the taxonomic relations between the
gull populations included in this paper. Thus, the
following discussion attempts to link our analysis
of phenotypic traits and vocalisations to studies of
population genetics. Interestingly, several genetic
studies have been carried out on Iberian and
Atlantic Yellow-legged Gulls but, due to different
methodology and different samples, they have
reached somewhat different conclusions and have
not (yet) led to a well resolved phylogenetic tree.
Liebers et al (2001) and Sternkopf et al (2010)
examined mitochondrial DNA (mtDNA) of all
populations of Yellow-legged Gull except for those
from the Canary Islands. According to their data,
all Ibero-Atlantic populations share many haplo-
types with nominate michahellis but these haplo-
types occur with different frequencies. Some gene
ow was found between these two groups, how-
ever. All Ibero-Atlantic populations differ signi-
cantly from each other in mtDNA, with the excep-
tions being Madeira clustering with Morocco, and
the Azores clustering with Galicia. Birds from
Portugal (sampled from the Berlengas) have dis-
tinct mtDNA, different from Galicia. Gene ow
between birds from Galicia and Mediterranean
michahellis seems limited. Hence, the mtDNA
does not reect the current taxonomic borders of
the different populations. Birds from the Azores
appear to be genetically more closely related to
birds from Galicia and even to michahellis from
the Mediterranean, rather than to birds from
Madeira and Morocco. MtDNA data suggests a
signicant divide between ‘northern atlantis
(Azores, Galicia) and ‘southern atlantis’ (Madeira,
Morocco). Birds from Gibraltar are considered
michahellis; their mtDNA does not differ from
birds from the French Mediterranean coast, al-
though it does show signicant differences from
the populations in Italy and Malta.
Pons et al (2004) compared birds from northern
Spain (País Vasco and Galicia) with birds from the
Mediterranean. They found no signicant differ-
ences in mtDNA between these populations but
some limited differences in nuclear DNA (nDNA).
They suggested recognizing the Atlantic Iberian
population as a distinct subspecies, L m lusitanius
(after Joiris 1978).
Arizaga et al (2006) looked at nDNA of Yellow-
legged Gulls from northern Spain (Gipuz koa,
Biscay and Asturias) and compared it with birds
from the Balearics (nominate michahellis). Their
results suggested little genetic variation between
these four populations, those from País Vasco
(Gipuzkoa and Biscay) being most similar and
grouping together, and the other two populations
showing more variability. Only the birds from
Gipuzkoa differed signicantly in their nDNA
from Balearic michahellis. They concluded that
there is still substantial gene ow along the north-
ern coast of Spain.
More extensive work on genetics is currently be-
ing done by a team of Spanish and French re-
searchers. The results are still very preliminary and
may be subject to change. They have not been
published yet but were presented at the 2018
International Gull Meeting in Bulgaria (Arizaga
2018). Their sample includes birds from the Medi-
terranean region, northern and north-western
Spain, Portugal, south-western Spain (including
Gibraltar), Morocco, Canary Islands, Madeira and
Azores, with analysis focusing on both mtDNA
and nDNA. The preliminary results suggest that
three groups can be distinguished on the basis of
nDNA: a Mediterranean group, a north ern Spanish
group (from País Vasco to Galicia), and a
Macaronesian group, which also includes the
Atlantic coast of Morocco. Birds from Portugal and
south-western Spain (including Gibral tar) seem to
have rather intermediate DNA and could be as-
signed to either the Mediterranean or northern
Spanish group.
Genetics are only one part of the puzzle though,
and we hope that they can be combined with our
data on phenotype (not just of adult but also im-
mature birds) and voice. The common message
from these three lines of evidence is that variation
appears to be largely clinal and so it is difcult to
draw the line between the various populations.
Our personal interpretation of all the current data
on Ibero-Atlantic birds is that: 1 the birds in the
Azores are the most distinct population of Yellow-
legged Gull, at least in plumage; 2 there is a sepa-
rate population of Yellow-legged Gulls along the
Atlantic coast of Spain and Portugal, with relative-
ly distinct adult plumage, voice, biometrics and
nDNA; this group shows clinal variation, ranging
from adult birds with relatively paler upperparts
and most distinctive long call in País Vasco and
Cantabria to adult birds with relatively darker upp-
perparts and more variable long call in Portugal;
3 birds in the Gibraltar area appear very similar to
nominate michahellis; and 4 birds in Morocco, the
Canary Islands and Madeira all seem rather inter-
mediate and so are difcult to assign to any sub-
specic group. While the nDNA of the latter three
Identication of Azores Gull
329
populations appears very similar to that of Azores
Gull, the mtDNA and phenotypic appearance
paint a somewhat different picture and make it dif-
cult to include Azores Gull with the other two
Maca ro nesian and the Moroccan populations.
None of these populations seem distinct enough
from michahellis to warrant full species status but a
classication into three subspecies certainly makes
sense to us: 1 nominate michahellis in the Medi-
terranean region and further north and east up to
and including Gibraltar and possibly south-west-
ern Spain (Mediterranean Yellow-legged Gull);
2 atlantis conned to the Azores (Azores Gull);
and 3 a third subspecies in Atlantic Iberia. Such a
classication raises the question of what name to
use for the latter. The name lusitanius is available
but it was used to describe the Portuguese popula-
tion (Joiris 1978), which differs subtly from the
Cantabrian/Basque birds in morphology, nDNA
and voice. The description by Joiris (1978) was
very brief, lacking in detail, partly incorrect, and
did not include photographs or drawings. It was
based on eld observations only, not skins. No
type specimen was collected. Still, this name is the
rst one used to describe this local population and
therefore has priority. We could consider lusitanius
a subspecies showing clinal variation in the colour
of the upperparts of adult birds and in voice along
the Iberian Atlantic coast. A specimen like MNCN
2557 from Suances, Cantabria, preserved at Museo
Nacional de Ciencias Naturales (Madrid, Spain)
could then become the type. The type specimen of
the subspecies atlantis was collected on Fayal,
Azores, in March 1922 (Dwight 1922).
The other populations (group 4) are difcult to
assign to any subspecies since in morphology, ge-
netics and voice they appear to be intermediate.
Should they go with Azores Gull or with lusitanius?
Olsen (2018) states that birds from Madeira and
the Canary Islands are ‘almost identical’ to nomi-
nate Yellow-legged Gull. Our analysis shows that
they are rather intermediate in plumage and voice,
and this accords broadly with the genetic data. It is
therefore hard to make a strong case either way
about where they should be placed. Our personal
view is that these birds may simply be part of one
or two intergradation zones, like for example
many European Herring Gulls in the Netherlands,
Denmark and Germany can be considered inter-
grades between argenteus and argentatus.
Other taxonomic hypotheses are possible too, of
course, and we welcome other views and interpre-
tations of the data, especially from taxonomists.
We look forward to further results of the colour-
ringing projects and the genetic research that are
currently being undertaken by French, Portuguese
and Spanish ornithologists.
Movements and vagrancy
On the one hand, our analysis of the appearance
and voice of Yellow-legged Gull populations has
not shed any strong new light on their taxonomic
relations – the evident complexity and resulting
uncertainties in the purely genetic studies pervade
interpretation of the phenotypic and vocalisation
data we have presented. On the other hand, we
have been able to isolate a set of ‘safe’ features that
can be used to identify Azores Gull, and point to
others which, although often used to support iden-
tication, do not stand up to critical scrutiny. In
turn, these safe and unsafe features allow us to say
something about reports of out-of-range Azores.
Although ringing data suggest that the Atlantic
and western Iberian Yellow-legged Gulls are gen-
erally resident, there are indications that some
birds may wander, especially those from the
Azores population. In the Azores, 332 pulli have
been colour-ringed since 2017, and none have
been observed outside of this island group so far
(Aldalur 2019). However, Moore (1996) reported a
total of ve rst-cycle Yellow-legged that followed
sea-going vessels for considerable distances be-
tween the Azores and mainland Portugal over the
course of eight years. One bird followed the ship
all the way from the Azores to Madeira; another
stayed with the ship for 1120 km when it returned
from the Azores to Portugal. One bird followed the
ship for 670 km from Madeira towards the
Portuguese mainland, and another one did the
same for 720 km. In addition, in September 2012
a rst-cycle Azores Gull was photographed stand-
ing on the bow of a research vessel 1600 km
south-west of the Azores. In the Canary Islands, at
least 350 pulli have been colour-ringed on Gran
Canaria since 2010 (Gutierrez 2012, Aldalur
2016), and 233 pulli have been colour-ringed
since 2017 (Aldalur 2019). Several of these birds
has been observed outside of the Canaries, but
only in Dakhla Bay, Western Sahara. These were
all immature birds (Xabier Remirez pers comm).
The Atlantic Iberian populations generally re-
main in the Iberian Peninsula all year, with imma-
ture birds (particularly rst-cycle) covering some-
what greater distances than adults. However, col-
our-ring projects have shown that long distance
movements are possible. For example, birds from
the Berlengas, Portugal, have been seen in Britain
(Gloucester, England; second-cycle), western
France (Île d’Oléron, second-cycle) and Morocco
(Oued Souss, two rst-cycle birds) (Morais et al
Identication of Azores Gull
330
1998). Birds ringed in Galicia have been reported
from Morocco (two), Isle of Wight, England, and
Den Helder, Noord-Holland, Netherlands (de
Juana & García 2015). Yellow-legged Gulls ringed
in País Vasco have been observed in Maine-et-
Loire (ve, all ages; Fossé 2019) and near Paris,
France (adult; Thibaut Chansac pers comm), as
well as in the Greater London area, England (third-
cycle; Peter Alfrey). Birds from Tarifa, southern
Spain, have been recovered in Morocco (four, in-
cluding one as far south as the Agadir region) and
Madeira (Cuenca & Delgado 2014).
Until 2019, there have been 15 records of
Azorean Yellow-legged Gull’ in Ireland, three in
Britain, one in mainland Spain and one in Iceland
(Hobbs & Irish Rare Birds Committee 2016, Hud-
son & Rarities Committee 2016, Gil-Velasco et al
2017; Omar Runolfsson pers comm). All of these
birds have been accepted provisionally as ‘show-
ing characters of the Azorean form’, without ex-
cluding the possibility of an origin from the Canary
Islands or Madeira. Nine of the Irish records con-
cerned adult birds, often returning to the same
area for several years; the others were of subadult
individuals. The British records included one sec-
ond-cycle and two subadult birds, one of which
was also observed in adult plumage in subsequent
years. We did not have access to the full documen-
tation of these Irish and British records and there-
fore cannot comment extensively on them. How-
ever, as already mentioned, the British second-
cycle bird (Elliot 2008) looks like a typical, dark
‘rst-summer’ type Azores Gull to us. The 2009
Oxfordshire bird certainly appeared to be of the
Macaronesian type when it rst appeared as a sub-
adult (Lewington 2009) with its extensively black-
ish bill, streaked hood, dark shins and dark grey
upperparts, which is typical of Macaronesian birds
in general but does not allow pinpointing the exact
origin (see the section on identication of third-
cycle birds). This bird returned to the Midlands as
an adult in subsequent years (Hudson & Rarities
Committee 2016, Holt & Rarities Committee
2017) but we have not seen detailed photographs
of its primary pattern. The Spanish bird was also a
subadult of the Macaronesian type (and hence not
necessarily Azorean). The Icelandic bird was a
rst-cycle and looks like a typical Azores Gull to
us. Since this bird has not been ofcially published
yet, a photograph of it is shown here (plate 424).
Furthermore, we have seen photographs of several
promising candidate Azores from mainland Portu-
gal, eg, a second-cycle bird (‘type 1’) from Lagoa
on 2 January 2010, and a rst-cycle from Olhão on
18 December 2013, and also one possible candi-
date from north-western Spain, a rst-cycle bird
from A Coruña on 11 October 2004, which is cur-
rently being assessed by the Spanish rarities com-
mittee. Among the European records, the high pro-
portion of (near-)adult birds is somewhat surpris-
ing, rstly because data seems to suggest that it is
the immature birds that are more likely to wander,
and secondly because these are generally the
more difcult plumages in which to identify Azores
– although it is true that a few subadults have a
rather striking appearance and can at least be sus-
pected to be of Macaronesian origin. In most cases
though, identication of adult and subadult birds
is complicated by extensive variation in lusitanius.
Based on colour-ringed birds from Galicia and a
few (unringed) nominate michahellis showing
strong head streaking (see section on regional vari-
ation), it is clear that not every hooded bird is a
true Azores, so care should be taken not to put too
much weight on this character alone. Also,
Portuguese adults can show as much black on the
upperside of the primaries as Macaronesian birds.
In North America, Yellow-legged Gull has been
considered a nearly annual vagrant to Newfound-
land since 1995 (Howell et al 2014). Most records
refer to adult birds, probably including returning
individuals, with an additional few records of sub-
adults. All are believed to be Azores Gulls. How-
ever, the photographs that we have seen of these
birds do not show the full suite of characteristic
features of this taxon, and we therefore cannot con-
rm their precise origin. In fact, we cannot even be
sure that they are from the Macaronesian islands
and not from Portugal, Galicia or Morocco.
Elsewhere in North America, there have been two
well-documented records of adult Yellow-legged
from Quebec, Canada, and one from Washington
DC, USA (Wilds & Czaplak 1994), but again the
documentation does not allow positive identica-
tion as Azores in our opinion. As in Europe, the
lack of records of younger birds is puzzling (al-
though some may have been overlooked). Quite a
few rst-cycle and second-cycle Yellow-legged
candidates have been documented from Florida,
USA, and, especially, Texas, USA (c 35 individuals)
but in the photographs that we have seen of these
birds we nd it impossible to exclude Lesser Black-
backed Gull or, in some cases, hybrid Lesser Black-
backed x American Herring Gull.
Summary and conclusions
Not all Azores Gulls can be identied safely. Our
data suggest that around one third of adult birds
have a combination of wing-tip features that make
them identiable (absent or very short tongue on
Identication of Azores Gull
331
p10, no mirror on p9, fully black outer web on p8
reaching primary coverts, and black on p4). Birds
with these features, darker upperparts and, at the
right time of year, a hood of streaks should prove to
be from the Azores. The dark plumage of rst-cycle
birds has been known for some time, and we pro-
vide some additional specic features to support
descriptions from earlier literature. We estimate
that approximately 50% of rst-cycle Azores can
be identied in a vagrant context. Second-cycle
birds are problematic, especially because of vari-
ability and resulting overlap between birds from
the various Macaronesian islands; the result is that
only birds sporting a hood, ‘Dunlin patch’ and ap-
propriately dark greater wing-coverts are safe to
identify as Azores. For this age class, we would
estimate that around 30-40% can be identied.
Despite our best efforts, we have been unable to
nd any safe features for third-cycle (and subadult)
Azores (although it is possible to recognize some
birds of this age category as belonging to the
Macaronesian population in general).
Olsen (2018) presented the most recent synthe-
sis of large white-headed gulls. He gives Azores
Gull a separate chapter and states that ‘in all plum-
ages they are different’ (to Yellow-legged Gull).
Although in fact no new detailed empirical analy-
sis of its identication has been published since
Dubois (2001), momentum seems to have been
building in both the published literature and on-
line for treating Azores as a full species. However,
the evidence we have presented here (that the ma-
jority of adult and subadult birds appear not to be
safely identiable) rather contradicts this popular
consensus.
Our scoring system for the wing-tip pattern of
adult gulls was deliberately kept simple, to ease
eld application. This simplication may mean
that it is rather too coarse to detect more subtle
differences. Thus, while for the moment we sug-
gest retaining Azores Gull as a subspecies of
michahellis (albeit the most distinctive one), we
recognize that more detailed work may add a dif-
ferent perspective.
We realize that readers probably expect identi-
cation criteria not only for Azores Gulls but also
for birds from the Macaronesian population as a
whole. However, this is fraught with difculty, not
only because the Canarian and Madeiran popula-
tions are much more similar to lusitanius and nom-
inate michahellis, but also because the variation in
lusitanius and Moroccan birds is extensive and still
not fully known. This is an important point. While
quite a few Azores Gulls – especially immatures –
are distinctive and can be identied with con-
dence even in a vagrant context, the same cannot
be said of birds from the Canaries/Madeira. We
have already indicated that Portuguese adults can
show as much black on the upper hand as birds
from Macaronesia, and that they can sport quite
heavy head streaking and dark grey upperparts. In
their rst-cycle, some Portuguese, Galician and
probably Moroccan birds too show an overall very
dark brown plumage, rendering the identication
of rst-cycle Canarian/Madeiran birds unsafe.
Only birds combining many distinctive features,
such as plain brown wing-coverts (looking ‘oil-
stained’), sooty face, streaked throat, dark breast,
‘zebra barring’ just above the legs, dark and messy
scapular pattern, very dark underwing and plain
dark inner primaries, can be identied as
Macaronesian, but such distinctive birds should be
from the Azores anyway. In second-cycle, birds of
type 1 (hooded, and with isolated ‘Dunlin patch’)
should be the focus of attention, since such birds
seem to be very rare on the continent, if they occur
at all. We have seen only one example, at Sagres,
Portugal, on 27 November 2016, and this may
have been a vagrant from Macaronesia. Type 1
birds are rather common in the Azores, making up
60% of second-cycle birds there, but not too many
(perhaps only 30%) show the plain dark greater
coverts needed to separate them from Canarian/
Madeiran birds. If the pattern of the greater coverts
can be disregarded for the Macaronesian popula-
tion as a whole, then quite a lot of Macaronesian
second-cycle birds can be identied in a vagrant
context. More research into the variation of
Moroccan birds and lusitanius is needed, especial-
ly in colour-ringed birds of known origin. For third-
cycle birds, we have already suggested that a com-
bination of blackish bill, dark hood, dark grey up-
perparts and dark ‘shin pads’ should indicate a
Macaronesian origin, but that this combination is
shown by only 5-9% of this age class.
Acknowledgements
We wish to thank Hein van Grouw of Natural History
Museum at Tring, London, and Carlos Urdiales of the ver-
tebrate collection of Estación Biológica de Doñana –
CSIC for allowing access to skins of Yellow-legged Gulls.
Our warmest thanks also go to César Álvarez Laó, Pim
Edelaar and Bert Saveyn, who were very kind and helpful
to examine skins for us. We received valuable informa-
tion from Juan Arizaga, Andres Bermejo, Inês Catry, Luis
Gordinho, Alfredo Herrero and Xabier Remirez. We
were sent important series of photographs by Asier
Aldalur, Carl Baggott, Richard Bonser, Albert Cama,
Derek Charles, Will Chateld-Taylor, Alex Colo rado,
Pierre-André Crochet, Eric Didner, Daan Drukker, Javier
Elorriaga, Antonio Gutierrez, Jeff Higgott, Gareth Knass,
Identication of Azores Gull
332
Merijn Loeve, Joaquin López, Ola Marklund, Jose
Manuel Méndez García, Dominic Mitchell, Fernando
Pereiras, Maties Rebassa, Ricardo Rodríguez, Omar
Runolfsson, Juan Sagardia, Lou Salo mon, Yeray Semi-
nario, Luc Teugels, Alfredo Valiente, Pieter Vantieghem
and Thijs Valkenburg. Sound-recordings were obtained
from Pere Josa Anguera, Pierre-André Crochet, José
Ramón Gómez Gonzalez, Antonio Gutierrez, Harry
Hussey, Magnus Robb and Darryl Spittle. Pierre-André
Crochet reviewed the section on taxonomy and provided
helpful comments. Finally, many Dutch Birding editors
plowed through a draft version of this paper and made
important corrections. Many thanks to all.
Samenvatting
Herkenning van azorengeelpootmeeuw Hoewel Geel-
poot meeuw Larus michahellis een vertrouwde soort is
voor veel Europese vogelaars, is er nog steeds veel on-
duidelijkheid over het aantal ondersoorten en hun ver-
spreiding. Is Azorengeelpootmeeuw L m atlantis be-
perkt tot de Azoren of komt deze zelfs tot op het conti-
nent voor? Hoe is die in het veld te herkennen? Is het
een ondersoort of aparte soort? Bestaat er zoiets als een
ondersoort L m lusitanius? Tot welke ondersoort beho-
ren de vogels in Marokko?
Op deze vragen probeert dit artikel een antwoord te
geven maar de nadruk ligt op de herkenning van
Azorengeelpootmeeuw. Onze conclusie is dat alleen
de vogels van de Azoren voldoende verschillen van de
overige populaties Geelpootmeeuwen om in het veld
te kunnen worden herkend. Volgens onze steekproef
heeft ongeveer een derde van de adulte Azorengeelpoot-
meeu wen een uniek handpenpatroon, gekenmerkt
door de combinatie van de (vrijwel) volledig zwarte
basis van p10, geen witte spiegel op p9, de volledig
zwarte buitenvlag van p8 (tot tegen de handpendekve-
ren) en de zwarte tekening op p4. Exemplaren met dit
handpenpatroon en erg donkergrijze bovendelen kun-
nen met zekerheid tot de Azorenpopulatie gerekend
worden, zeker als ze ook nog eens de typische, zwaar
gestreepte kopkap tonen na de zomerrui. C 50% van
de eerstejaarsvogels is te herkennen aan het erg don-
kere verenkleed met effen donkere vleugeldekveren
die met enige verbeelding zelfs onder de olie lijken te
zitten, een donker gestreept gezicht (inclusief kin en
keel), rommelig donker patroon op de schouderveren,
zwaar getekende borst, brede zwarte staartband en
vaak ook enige brede zwart-witte, verticale bandering
op de achterank (net boven de poten). Van de twee-
dejaarsvogels is 30-40% te herkennen aan een donker
buikschild, gestreepte kopkap en effen donkere grote
dekveren. Bij derdejaarsvogels is er te veel overlap in
uiterlijk met de populaties op de Canarische Eilanden
en Madeira om een vogel met zekerheid als Azoren-
geelpootmeeuw te bestempelen.
Op basis van verenkleed, afmetingen, geluid en
DNA vormen de vogels van de Azoren weliswaar een
opvallende groep maar wellicht niet uniek genoeg
voor volwaardige soortstatus. Verder lijkt ook een af-
zonderlijk taxon aanwezig langs de noordelijke en
westelijke kust van het Iberisch Schiereiland, van
Baskenland tot het uiterste zuiden van Portugal. Deze
populatie kan als ondersoort L m lusitanius worden be-
schouwd, hoewel het taxon slechts heel summier werd
beschreven.
Geelpootmeeuwen in het zuidwesten van Spanje en
in Gibraltar en Ceuta tonen vooral kenmerken van no-
minaat L m michahellis, hoewel meer onderzoek hier
zeker wenselijk is. De populaties van de Canarische
Eilanden, Madeira en Marokko vertonen veel interme-
diaire kenmerken tussen atlantis en lusitanius/micha-
hellis, en lijken ons moeilijk in te delen bij een be-
paalde ondersoort.
Het artikel gaat verder nog in op de dispersie van de
verschillende ondersoorten op basis van ringgegevens
en waarnemingen (ver) buiten het normale
verspreidings gebied.
References
Adriaens, P 2003. Hybrid gulls breeding in Belgium.
Website: https://tinyurl.com/y5z6xqk5.
Adriaens, P & Gibbins, C 2016. Identication of the
Larus canus complex. Dutch Birding 38: 1-64.
Adriaens, P, Vercruijsse, H J P & Stienen, E W M 2012.
Hybrid gulls in Belgium – an update. Br Birds 105:
530-542.
AERC TAC 2003. AERC TAC’s taxonomic recommen-
dations. Website: www.aerc.eu.
Aldalur, A 2016. Sobre las Gaviotas patiamarillas de
Canarias (Larus michahellis atlantis) y las islas de
Macaronesia. Website: https://tinyurl.com/y3fvt5tx.
Aldalur, A 2019. El balance del anillamiento de
Gavio tas patiamarillas atlánticas de las islas Azores
y Canarias, el tercer año de estudio. Website: htt-
ps://tinyurl.com/y6gwnpbv.
Arizaga, J 2018. The Yellow-legged Gull in the Basque
region: current studies, future perspectives. Inter-
national Gull Meeting, Russe, Bulgaria.
Arizaga, J, Alda, F, Doadrio, I & Herrero, A 2006.
Análisis de la dispersión de la gaviota patiamarilla
(Larus michahellis lusitanius) en Guipúzcoa. Inédito.
Socie dad de Ciencias Aranzadi.
Arizaga, J, Aldalur, A, Herrero, A & Galicia, D 2008.
Sex differentiation of Yellow-legged Gull (Larus
michahellis lusitanius): the use of biometrics, bill
morphometrics and wing tip coloration. Waterbirds
31: 211-219.
Arizaga, J, Aldalur, A, Cuadrado, J F, Diez, E, Goi koet-
xea, J, Herrero, A, Jauregi, J I, Laso, M & Sanchez,
J M 2012. Parámetros reproductores de la gaviota
patiamarilla Larus michahellis lusitanius Naumann
1840 en Gipuzkoa. Munibe (Ciencias Naturales-
Natur Zientziak) 60: 167-174.
Arizaga, J, Herrero, A & Aldalur, A 2019. Seasonal pat-
tern of plumage colour in Yellow-legged Gulls at
Bay of Biscay, Spain. Dutch Birding 41: 166-174.
Bannerman, D A 1963. Birds of the Atlantic Islands 1.
A history of the birds of the Canary Islands and of
Identication of Azores Gull
333
the Salvages. Edinburgh & London.
Bosch, M 1995. Sexual size dimorphism and determi-
nation of sex in Yellow-legged Gulls. J Field Ornithol
67: 534-541.
Carrera, E, Trias, J, Bermejo, A, de Juana, E & Varela, J
1987. Etude biométrique des populations ibériques
et nord-africaine du Goéland leucophée. L’Oiseau
et RFO 57: 32-38.
Collinson, M, Parkin, D T, Knox, A G, Sangster, G &
Svensson, L 2008. Genetic relationships among the
different races of Herring Gull, Yellow-legged Gull
and Lesser Black-backed Gull. Br Birds 94: 523-
528.
Cottaar, F 2004. Geelpootmeeuwcomplex van IJmui-
den: in hoeverre zijn geelpootmeeuwen echte Geel-
poot meeuwen? Dutch Birding 26: 36-42.
Cottaar, F, Kleinwee, M & Verbeek-Cottaar, J 2019.
Gemengd broedpaar Zilvermeeuw x Kleine Mantel-
meeuw in IJmuiden in 2017. Limosa 92: 95-99.
Cramp, S & Simmons, K E L (editors) 1983. The birds
of the Western Palearctic 3. Oxford.
Cuenca, D & Delgado, D 2014. La colonia de Gaviota
patiamarilla (Larus michahellis) de la Isla de Tarifa:
evolución y dispersión postjuvenil. Rev Soc Gad
Hist Nat 8: 5-10.
Dickinson, E C & Remsen Jr, J V (editors) 2013. The
Howard and Moore complete checklist of the birds
of the world. Fourth edition, vol 1: Non-passerines.
Eastbourne.
Dubois, P J 2001. Atlantic Islands Yellow-legged Gulls,
an identication gallery. Birding World: 14: 293-
304.
Dutch Birding 1985. Terminologie voor verenkleed en
rui. Dutch Birding 7: 1-5.
Dwight, J 1922. Description of a new race of the
Lesser Black-backed Gull, from the Azores. Am Mus
Novit 44: 1-2.
Dwight, J 1925. The Gulls (Laridae) of the world: their
plumages, moults, variations, relationships and dis-
tribution. Bull Am Mus Nat Hist 52: 63-401.
Elliot, M 2008. The Azorean Atlantic Gull in Cornwall.
Birding World 21: 462-466.
Ellis, J C, Bogdanowicz, S M, Stoddard, M C & Clark,
L W 2014. Hybridization of a Lesser Black-backed
Gull and Herring Gulls in eastern North America.
Wilson J Ornithol 126: 338-346.
Fossé, A 2019. Données de baguage. Goéland leuco-
phée Larus michahellis ssp. (Yellow-legged Gull).
Website: https://tinyurl.com/y5yqvdlf.
Galarza, A, Hidalgo, J, Ocio, G & Rodríguez, P 2008.
Sexual size dimorphism and determination of sex in
Atlantic Yellow-legged Gulls Larus michahellis lusi-
tanius from northern Spain. Ardeola 55: 41-47.
Garner, M & Quinn, D 1997. Identication of Yellow-
legged Gulls in Britain, part 1. Br Birds 90: 25-62.
Gill, F, Donsker, D & Rasmussen, P (editors) 2020.
IOC world bird list (version 10.2). Website: www.
worldbirdnames.org.
Gil-Velasco, M, Rouco, M, Ferrer, J, García-Tarrasón, M,
García-Vargas, F J, Gutiérrez, A, Hevia, R, López, F,
López-Velasco, D, Ollé, A, Rodríguez, G, Sagardía, J
& Salazar, J A 2017. Observaciones de aves raras en
España 2014. Ardeola 64: 161-235.
Grant, P J 1986. Gulls: a guide to identication.
Second edition. Calton.
Gutierrez, A 2012. Michahellis de Canarias. Website:
https://tinyurl.com/y4d85wg2.
Hobbs, J & Irish Rare Birds Committee 2016. Irish rare
bird report 2015. Irish Birds 10: 391-428.
Holt, C & Rarities Committee 2017. Report on rare
birds in Great Britain in 2016. Br Birds 110: 562-
631.
Howell, S N G 2010. Molt in North American birds.
New York.
Howell, S N G & Dunn, J 2007. Gulls of the Americas.
Boston.
Howell, S N G, Corben, C, Pyle, P & Rogers, D I 2003.
The rst basic problem: a review of molt and plum-
age homologies. Condor 105: 635-653
Howell, S N G, Lewington, I & Russell, W 2014. Rare
birds of North America. Princeton.
Hudson, N & Rarities Committee 2016. Report on rare
birds in Great Britain in 2015. Br Birds 109: 566-
631.
Isenmann, P 1973. Biometrische Untersuchungen an
der gelbfüßige Silbermöwe Larus argentatus micha-
hellis aus der Camargue. Vogelwarte 27:16-24.
Joiris, C 1978. Le Goéland argenté portugais (Larus ar-
gentatus lusitanius), nouvelle forme de Goéland ar-
genté à pattes jaunes. Aves 15: 17-18.
Jonsson, L 1998. Yellow-legged gulls and yellow-leg-
ged Herring Gulls in the Baltic. Alula 3: 74-100.
de Juana, E & Garcia, E 2015. The birds of the Iberian
Peninsula. London.
de Knijff, P, Denkers, F, van Swelm, N D & Kuiper, M
2001. Genetic afnities within the Herring Gull
Larus argentatus assemblage revealed by AFLP
Geno typing. J Mol Evol 52: 85-93.
Lewington, I 2009. Identication of the Azorean
Atlan tic Gull in Oxfordshire. Birding World 22:
459-462.
Liebers, D & Helbig, A 1999. Phänotypische Charak-
teri sierung und systematische Stellung der Arme-
nien möwe Larus armenicus. Limicola 13: 281-321.
Liebers, D, Helbig, A J & de Knijff, P 2001. Genetic
differentiation and phylogeography of gulls in the
Larus cachinnans-fuscus group (Aves: Chara drii-
formes). Mol Ecol 10: 2447-2462.
Mínguez, E & Ganuza, J 1995. Biometrías de la
Gaviota Patiamarilla (Larus cachinnans) nidicante
en Gipuz koa. Chioglossa. Vol Esp 1: 31-34.
Moore, C C 1996. Ship-attending movements of Atlan-
tic Yellow-legged Gulls in Portuguese waters. Dutch
Birding 18: 17-18.
Morais L, Santos, C & Vicente, L 1998. Population in-
crease of yellow-legged gulls Larus cachinnans
breeding on Berlenga Island (Portugal), 1974-1994.
Sula 12: 27-37.
Identication of Azores Gull
334
Muusse, M, Muusse, T, Buijs, R-J, Altenburg, R, Gib-
bins, C & Luijendijk, B-J 2011. Phenotypic charac-
teristics and moult commencement in breeding
Dutch Herring Gulls Larus argentatus & Lesser
Black-backed Gulls Larus fuscus. Seabird 24: 42-
59.
Neubauer, G, Faber, M & Zagalska-Neubauer, M
2010. Yellow-legged Gull in Poland: status and sep-
aration from yellow-legged Herring Gull and hy-
brids. Dutch Birding 32: 163-170.
Olsen, K M 2018. Gulls of the world. A photographic
guide. London.
Olsen, K M & Larsson, H 2004. Gulls of Europe, Asia
and North America. Second edition. London.
Paterson, A M 1997. Las aves marinas de España y
Portugal. Barcelona.
Pons, J-M, Crochet, P-A, Thery, M, & Bermejo, A 2004.
Geographical variation in the Yellow-legged Gull:
introgression or convergence from the Herring
Gull? J Zool Sys Evol Res 42: 245-256.
Stegmann, B 1934. Ueber die Formen der grossen
Möwen (subgenus Larus) und ihre gegenseitigen
Bezie hungen. J Ornithol 82: 340-380.
Sternkopf, V, Liebers-Helbig, D, Ritz, M S, Zhang, J,
Helbig, A J & de Knijff, P 2010. Introgressive hy-
bridization and the evolutionary history of the her-
ring gull complex revealed by mitochondrial and
nuclear DNA. BMC Evol Biol 10: 348.
Stoddart, A & McInerny, C 2017. The ‘Azorean Yellow-
legged Gull’ in Britain. Br Birds 110: 666-674.
Teyssèdre, A 1983. Etude comparative de quatre popu-
lations de Goélands argentés à pattes jaunes
d’Europe occidentale. L’Oiseau et RFO 53: 43-52.
Teyssèdre, A 1984. Comparaison acoustique de Larus
argentatus argenteus, L. fuscus graellsii, L. cachin-
nans (?) michahellis et du Goéland argenté à pattes
jaunes cantabrique. Behaviour 88: 13-33.
Urban, E K, Fry, C H & Keith, S (editors) 1986. The
birds of Africa 2. Princeton.
Vaurie, C 1965. The birds of the Palearctic fauna.
A systematic reference. Non-Passeriformes. London.
Volsøe, H 1951. The breeding birds of the Canary
Islands 1. Introduction and synopsis of the species.
Vidensk Meddel Dansk Naturhist Foren 113: 1-152.
Wilds, C & Czaplak, D 1994. Yellow-legged Gulls
(Larus cachinnans) in North America. Wilson Bull
106: 344-356.
Yésou, P 2002. Systematics of Larus argentatus-cach-
innans-fuscus complex revisited. Dutch Birding 64:
271-298.
Peter Adriaens, Brusselsesteenweg 31, 9050 Gentbrugge, Belgium (p_adriaens@yahoo.com)
Peter Alfrey, Little Oak, 23 Wood Street, Hackbridge, Surrey, CR4 4JT, UK
(littleoakgroup@btinternet.com)
Chris Gibbins, School of Environmental and Geographical Sciences, University of Nottingham,
Malaysia, 43500 Selangor, Malaysia (christopher.gibbins@nottingham.edu.my)
Daniel López-Velasco, C/Pablo Laloux, 10, 6 izda D Salinas, 33450 Asturias, Spain
(dskater20@gmail.com)
Identication of Azores Gull
... According to several genetic, morphological, observational and phenological studies (e.g. Olsen & Larson 2004;Pons et al. 2004;Howell & Dunn 2007;Adriaens et al. 2020), the 'Macaronesian Yellow-legged Gull' comprises a subspecies L. m. atlantis. There is controversy about the range of L. m. atlantis because some authors consider this subspecies to only breed in the Azores (Dubois 2001;Yésou 2002;Olsen 2018), while others report that the subspecies breeds in the rest of the Macaronesian islands (Cramp & Simmons 1983;Olsen & Larson 2004;Howell & Dunn 2007), and even the northwestern African coast (Collinson et al. 2008). ...
Technical Report
Full-text available
In the eighteenth report of the Portuguese Rarities Committee of the Portuguese Society for the Study of Birds (SPEA), relating mainly to the year 2021, 395 records of 153 species were analysed for the referred period of which 97% were accepted and refer to all the regions of the country considered (mainland Portugal, autonomous regions of the Azores and Madeira). For the whole of the territory of Portugal, five (5) new species were approved for the period considered in this report: White-winged Scoter Melanitta deglandi (Azores), American Kestrel Falco sparverius (Azores), Horned Lark Eremophila alpestris (mainland Portugal), Grey Catbird Dumetella carolinensis (Azores) and Black-faced Bunting Emberiza spodocephala (mainland Portugal). Referring exclusively to the autonomous region of Azores, the first records of Rook Corvus frugilegus, and for the autonomous region of Madeira, the first records of Tundra-Bean Goose Anser serrirostris, Lesser Scaup Aythya affinis, Little Stint Calidris minuta. Pallid Harrier Circus macrourus, Mediterranean Short-toed Lark Alaudala rufescens, Western Bonelli’s Warbler Phylloscopus bonelli, Trumpeter Finch Bucanetes githagineus, were accepted.
Article
Full-text available
This new issue of the 'Eléments de bibliographie ornithologique marocaine' brings together a selection of articles dealing with the avifauna of Morocco, including the Spanish enclaves of Ceuta and Melilla. As in the previous issues, we have separated those dealing specifically with this country from those of more general scope but also concerning Morocco. An informative summary follows some of them, especially when the title does not clearly express the link with Morocco. A third paragraph presents a selection of recent works relating to other nearby countries (Spain and the Canary Islands, Portugal, Libya, Tunisia, Algeria and Mauritania in particular), which are directly related to Moroccan avifauna. It should be remembered that the Moroccan ornithological bibliography available at the beginning of 2022 was presented in 'Oiseaux du Maroc - Birds of Morocco' (Bergier et al. 2022. Société d'Études Ornithologiques de France, Paris. 648 pp).
Article
Full-text available
Genetic analyses of the relationships among different forms of Lesser Black-backed Gull Larus fuscus, Herring Gull L. argentatus and Yellow-legged Gull L. cachinnans are beset with difficulties, associated with apparently low levels of genetic divergence, inbreeding, and the highly unstable nature of gull populations. This short review summarises and discusses a new attempt to tackle this problem (de Knijff et al. 2001). While the results do not resolve the 'gull problem', they have provided much new evidence on the extent of genetic variation among different individuals of the same taxon, which will be crucial to any future consideration of gull taxonomy.
Article
Full-text available
Geelpootmeeuwen in Polen: status en verschillen met Zilvermeeuw met gele poten en hybriden De status van Geelpootmeeuwen als broedvogel in Polen wordt besproken. De meeste broedgevallen van Geelpootmeeuw Larus michahellis betreffen gemengde paren, vaak met Pontische Meeuw L cachinnans, terwijl zuivere paren op slechts drie locaties in Polen zijn vastgesteld. Het gebrek aan geschikt broedhabitat en aan partners van de eigen soort is waarschijnlijk de bepalende factor voor de hoge hybridisatiefrequentie. Het onderscheiden van Geelpootmeeuwen en geelpotige Zilvermeeuwen L argentatus is eenvoudig wanneer maten genomen worden en een constante, vrijwel diagnostiche combinatie van vleugelpuntkenmerken gebruikt wordt. Veldherkenning van Geelpootmeeuw in Centraal-Europa is echter aanzienlijk moeilijker dan in andere delen van het continent door het voorkomen van hybriden met Zilvermeeuw en, vooral, Pontische Meeuw, die beide algemeen zijn (en die ook onderling kruisen). Deze hybriden zijn variabel in uiterlijk maar hebben doorgaans intens gekleurde naakte delen en veel zwart in de vleugelpunt. Ze zijn ook intermediair in formaat en structuur en missen vaak de donkere irisvlekken karakteristiek voor Pontische. Zulke hybriden kunnen erg veel op zuivere Geelpootmeeuw lijken en verkeerd gedetermineerd worden. In Centraal-Europa is het daarom wenselijk om een broedgeval van een op een Geelpootmeeuw lijkende vogel goed te documenteren.
Article
Full-text available
RESUMEN El objetivo del presente artículo es estudiar la reproducción de la gaviota patiamarilla Larus michahellis lusitanius Naumann, 1840 en Gipuzkoa. En particular, analizamos (1) el tamaño de puesta (número de huevos/nido), (2) la fecha de eclosión, (3) el número de huevos eclosionados/nido y (4) la proporción de huevos eclosionados en relación al número de huevos en el nido, considerando dos zonas que, a priori, presentan diferente tipo de sus-trato de nidificación. Para ello, durante un periodo de 40 días (16.05.2011-24.06-2011) la colonia de Ulia se visitó a diario. En promedio, se registró un tamaño de puesta de 3 huevos/nido, un número de huevos eclosionados de 3/nido y un porcentaje de eclosión de 88,4%. La fecha media de eclosión suce-dió en la segunda quincena de Mayo. Asimismo, se registraron algunas dife-rencias entre las dos zonas en que se dividió la colonia (tamaño de puesta inferior y porcentaje de huevos eclosionados superior en una zona que pre-sentó más roca y vegetación de mayor porte en relación a otra zona con menos rocas así como más hierba). Discutimos esto en un contexto de zoni-ficación de la colonia de estudio en el marco de áreas de carácter óptimo o subóptimo en términos reproductivos. • PALABRAS CLAVE: Golfo de Vizcaya (Cantábrico), eclosión, láridos, tamaño de puesta, nidificación, productividad, Ulia.
Article
Full-text available
We used feathers and skeletal measurements, white and black areas at the wing tip and bill morphometrics of Yellow-legged Gulls (Larus michahellis lusitanius) in order to test for possible sex-differences. Overall, 157 individuals from the eastern Bay of Biscay (N Spain) were measured, and the sex determined in 155 individuals, by means of DNA-analyses. All feathers and skeletal-associated measurements, except the distance between each primary (P1 to P5) feather and the wing tip in a folded wing varied between the sexes, with males being larger than females. Sexual selection is discussed to be the major cause explaining these differences. A discriminant function is provided to separate sexes. By contrast, both wing tip patterns of coloration and bill morphology did not vary between sex classes.