I'iATUHML HISTORY MUSEUM LIBRARY \ 2 NOV 2013 ISSN 0007-0335 British Birds Established 1907, incorporating The Zoologist, established 1843 Published by BB 2000 Limited, trading as ‘British Birds’ Registered Office: c/o Chappell Cole & Co, Heritage House, 34B North Cray Road, Bexley, Kent DA5 3LZ www.britishbirds.co.uk Editorial Roger Riddington Spindrift, Eastshore, Virkie, Shetland ZE3 9JS Tel: 01950 460080 editor@britishbirds.co.uk ‘News & comment’ material to Adrian Pitches adrianpitches@blueyonder.co.uk Subscriptions & administration Hazel Jenner 4 Harlequin Gardens, St Leonards on Sea, East Sussex TN37 7PF Tel & fax: 01424 755155 subscriptions@britishbirds.co.uk Design & production Mark Corliss m.corliss@netmatters.co.uk Advertising Mathew Hance, Digital Spring Ltd, Adam House, 7-10 Adam Street, The Strand, London WC2N 6AA Tel: 020 7520 9326 BBAdsales@digital-spring.co.uk Guidelines for contributors See www.britishbirds.co.uk British Birds Editorial staff Roger Riddington (Editor), Caroline Dudley, Peter Kennerley Editorial Board Dawn Balmer, Ian Carter, Richard Chandler, Martin Collinson, Mark Holling, Chris Kehoe, Robin Prytherch, Nigel Redman, Roger Riddington, Brian Small, Steve Votier Rarities Committee Adam Rowlands (Chairman), Chris Batty, Chris Bradshaw, Paul French, Martin Garner, Nic Hallam, James Lidster, Richard Millington, Mike Pennington, Richard Schofield, Steve Votier Secretary Nigel Hudson, Carn Ithen, Trench Lane, Old Town, St Mary’s, Scilly TR21 OPA; secretary@bbrc.org.uk Notes Panel Angela Turner (Chair), Will Cresswell, Ian Dawson, Jim Flegg, Ian Newton, Malcolm Ogilvie Annual subscription rates Individual subscriptions: UK - £53.00 Overseas (airmail) - £60.00 Libraries and agencies - £99.00 Back issues available from www.britishbirds.co.uk or the subscriptions office. Printed by Hastings Printing Company British Birds is owned and published by BB 2000 Limited, the directors of which are John Eyre (Chairman), Jeremy Greenwood, Mark Holling, Conor Jameson, Ciaran Nelson, Ian Packer, Adrian Pitches and Richard Porter. BB 2000 Limited is wholly owned by The British Birds Charitable Trust (registered charity No. 1089422), whose trustees are Richard Chandler, Jeremy Greenwood, Ian Newton and Peter Oliver. Directors and trustees are volunteers who draw no remuneration. Copyright: When submitting articles, letters, commentary, text, photographs, artwork, figures or images (the ‘Copyright Work’) to the Editor, you are agreeing to grant to British Birds a perpetual, irrevocable, non-exclusive, royalty-free, copyright licence to use, edit, alter, adapt, translate, copy, publish, continue to publish or republish the Copyright Work (and/or an edited, adapted or translated version of it or part of it) in all forms, formats and media (including, but not limited to, print, digital and electronic forms) anywhere in the world. You must ensure that by submitting a Copyright Work that you are not infringing the Copyright of any other person. By submitting a Copyright Work you are warranting that you are the Copyright Work owner and that you have the right to grant the non-exclusive licence described above. For the avoidance of doubt, the Author/ Artist shall remain the owner of the Copyright Work. Front-cover photograph: Male Subalpine Warbler Sylvia cantillans (alt. Western Subalpine Warbler S. inornata ), Avila, Castile and Leon, Spain, May 2011. Roger Tidman/FLPA THE VIKING ED PRO STUNNING CLARITY, LIGHTWEIGHT DESIGN WITH OUTSTANDING BUILD QUALITY* THE ED PRO, LIGHTWEIGHT, HIGH PERFORMANCE SSMCICIIIAR WITH v - DESIGN The new 8x42 & 10x42 ED Pro takes the Viking binocular range to a whole new level. With lifelike colour reproduction and stunning edge to edge clarity thanks to the high performance ED glass. Ergonomically designed to give a comfortable feel built on a lightweight magnesium alloy body. With outstanding Japanese build quality this high performance binocular is a serious contender amongst the very best alternatives. BirdLife INTJER^AtitOSiU. PREVENTING EXTINCTIONS Viking Viking Optical Limited, Btyth Road, Halesworth Suffolk IP19SEN, UK Tel: 01986 875315 email: sales@vikingopticai.co.uk For more information & availability please visit www.vikjn30ptical.co.uk/edpro wwt.org.uk/nwbwf Optical & digital manufacturers, guest speakers, free workshops, exhibitors & thousands of wintering waterfowl. WWT Martin Mere, Burscough, Ormski Lancashire L40 OTA T 01704 895181 - SQ@wm 5PORT OPTICS PREMIERE DEALER Sapphire ED jjjjf Extra-low dispersion (ED) glass for optimum clarity Q Dielectric coating (DC) to increase light reflectivity Q Fully multi-coated (FMC) to produce sharp images HI Hi-res phase corrected (PC) BAK 4 roof prisms d Water repellent coating on objective lenses fp Precision focus wheel and dioptre adjustment Of Lightweight rubber coated magnesium alloy body Q Stay-on lens covers for maximum lens protection G2 Twist-up eye cups with position stops for eye relief Q Tan leather hard carry case and a padded neck strap 2H Nitrogen purged - waterproof and fog proof £2 Hawke Worldwide warranty for added protection from £359.99 Endurance 0s" mmw. *#" y Dielectric coating (DC) to increase light reflectivity d Fully multi-coated optics produce sharp images §2 BAK 4 pom prisms for intense colour and contrast U| Dual focus knob to achieve ultra fine focusing 88 Body rotation for viewing in virtually any position Q§ Stay-on soft scope cover for maximum protection §2 Twist-up eye cup for eye relief and pull out sunshade fJJ Supplied with zoom eye piece (upgrades on pg.31) ifl Digi-scope compatible for use with camera (pg,30) U§ Nitrogen purged - waterproof and fog proof £§ Hawke Worldwide warranty for added protection 16-48x68 £349.99 20 - 60 x 85 £399.99 London camera exchange 1 5 The Square, Winchester 01962866203 winchester (9 LCEgroup.co.uk Experiencing nature, with impressive clarity. For longer than ever before. This is the moment we work for. zeiss.de/sportsoptics VICTORY HT. The brightest premium binoculars in the world. Experience a revolution in observation optics: the VICTORY HT. Thanks to its innovative optical concept with SCHOTT HT-glass it achieves light transmission levels of up to 95% which means you can continue observing for longer periods. Equally impressive is the ergonomic Comfort- Focus-Concept, with its intuitive operation and quick, precise focussing, as well as the extremely robust design with new Double-Link-Bridge. Available in 8x42 & lOx 42. www.zeiss.com/sportsoptics 01223401525 We make it visible. ZEISS British Birds Volume 106 • Number I I • November 2013 646 BB eye Richard Chandler 648 News and comment Adrian Pitches 65 I Subalpine Warbler variation and taxonomy Lars Svensson 669 Further comments on the ageing and sexing of the Red-flanked Bluetail Magnus Hellstrom and Gabriel Norevik 678 Treswell Wood: 40 years of integrated population monitoring Chris du Feu 690 Notes 693 Reviews 696 Recent reports The Subalpine Warbler used to be a relatively straightforward rarity for British birders. Over the past couple of decades, the more obvious exam- ples of ‘Eastern Subalpine Warbler’ and, latterly, ‘Moltoni’s Warbler’ have been sifted out by sharp-eyed observers. In this month’s BB, however, Lars Svensson sets out the case for a three-way split; assuming that BOURC follows these recommendations, it means that all vagrant Subalpine Warblers will suddenly need a lot more care. Sound recordings and photo- graphs from every angle might be the minimum requirement for anything other than a ‘classic’ male if the record is not to become mired in ‘either/or’ land. Two decades ago that would have seemed an impossible level of detail but now it seems well within the bounds of possibility. The paper on ageing and sexing of Red-flanked Bluetails, from workers at one of Sweden’s top observatories, is another contribution that we might not have expected to see in BB 20 years ago. Red-flanked Bluetail was an extraordinary rarity in 1993, as the massive twitch for the bird at Winspit, on the Dorset coast, in the late autumn of that year showed. Now, despite the stats in last month’s BBRC report showing that 2012 was a poor year, this bird is a feasible target on the east coast in autumn. The last of the main papers in this issue, last year’s Bernard Tucker lecture, shows what can be done by combining ringing with other fieldwork in a well-planned study of a single site rather than a particular species or group. Chris du Feu gives a flavour of just some of the results of 30-odd years of work in one small wood in the East Midlands. It’s another issue of BB that emphasises the diversity of ways in which we can enjoy birds. Roger Riddington sO FSC MIX Paper from responsible sources FSC* C022506 British Birds aims to: •$» provide an up-to-date magazine for everyone interested in the birds of the Western Palearctic; publish a range of material on behaviour, conservation, distribution, ecology, identification, movements, status and taxonomy as well as the latest ornithological news and book reviews; ♦> maintain its position as the journal of record; and ♦> interpret scientific research on birds in an easily accessible way. © British Birds 2013 88 eye Thinking big The ‘Big Idea’ was one of those crazy dreams that are usually dismissed instantly, either as impossible, totally impractical or simply downright mad! The dream was to combine two not-so-adjacent nature reserves and add a considerable area of the surrounding farm- land, thus multiplying the original area almost eight-fold to recreate a significant extent of wet fenland. The initial response of most people was either that the visionaries had delusions of grandeur or, more pro- saically, to ask where all the money might come from? In the mid 1990s, the Big Idea was first mooted by Adrian Colston, then Director of Operations at the Wildlife Trust for Bedford- shire, Cambridgeshire and Northampton- shire, who was much influenced by the major restoration project to create the Dutch ‘mega reserve’ of Oostvaardersplassen. He managed to persuade the Wildlife Trust to take the Big Idea seriously, and the Great Fen Project (as it is now known) was born. Adrian moved on, but the seeds had been sown and the Great Fen Project was developing a life of its own. The initial doubters, myself included, have been shown that it is indeed a grand idea and is far from delusionary! The Great Fen Project (see www.greatfen. org.uk), at the western edge of the Fens in Cambridgeshire, is big thinking for both the present and the future. It takes the two important National Nature Reserves, Wood- walton Fen (208 ha) and Holme Fen (266 ha), as its nuclei, and hopes to draw its even- tual boundaries far beyond these, to make a total area of 3,700 ha (37 square kilometres). As one of the largest nature restoration proj- ects in the country, it is both inevitable and, of course, essential that a number of bodies are involved, all with different roles but all co-operating in the overall scheme. These bodies include, as might be expected, Natural England and the Wildlife Trust; and, less immediately obviously, though just as impor- tant, the Environment Agency, Huntingdon- shire District Council and the Middle Level Commissioners, the last being responsible, 646 inter alia , for flood protection in the area. The two nuclei reserves represent almost the last remaining remnants of how the fen- lands would have been before the major drainage of the area commenced in the early seventeenth century, though even these have been modified by the hand of man. Wood- walton Fen is an absolute delight - a substan- tial area of woodland, meres, open damp grassland, reedbeds and waterways - while footpaths provide good access. To the first- time visitor it is completely unexpected, lying hidden in the vast, flat expanse of the largely featureless fenland. The reserve boasts a regular list of about 170 bird species, including Marsh Harrier Circus aeruginosas , Hobby Falco subbuteo , and all the expected reedbed and damp -woodland warbler species, together with more than 400 species of flowering plants. The other reserve is Holme Fen, owned by Natural England, which includes the largest Silver Birch Betula pendula woodland in lowland Britain, areas of heath grassland and a small fragment of raised bog, the most south- easterly in England. Although most of this area of the Fens was drained successfully by the 1850s, Holme Fen proved difficult to drain completely so that it became a nature reserve almost by default. Part of it includes the lowest area in Britain, 2.75 m below sea level. The site of the largest of the former meres of the ancient fenland, Whittlesey Mere, lies at the north end of the Great Fen Project area. With the drainage of this Mere, as recently as 1851, went the Swallowtail Papilio machaon and the English race of the Large Copper Lycaena dispar butterfly, along with many species of the local flora. The nine- teenth-century naturalist Charles Rothschild was aware of these losses to our natural her- itage. Seeing the speed with which drainage was transforming the Fens to arable farm- land, and with it the loss of the natural fenland wildlife, he purchased Wicken Fen in 1899 - which he subsequently presented to the National Trust - and later, in 1910, he purchased Woodwalton Fen, which became a © British Birds 106 • November 2013 • 646-647 BB eye reserve in 1919. Rothschild built a bungalow - on stilts to avoid flooding - to use as a base for his scientific work at Woodwalton Fen. It is still being used, and it provides a memorial and a direct connection to a far-sighted natu- ralist to whom we owe a considerable debt, not just for the preservation of Woodwalton Fen but as a pioneer of nature reserves in Britain. With the drainage of the fenlands came the ‘shrinkage’ of the local peat soils; consoli- dation, oxidation and deflation all combined to result in a significant lowering of the general ground surface. Ground lowering measured at the Flolme Post, which was installed in 1850 to record the anticipated shrinkage, is in excess of 4 m. Ground lowering is evident to the visitor at Wood- walton Fen, where the reduced extent of shrinkage at the reserve has resulted in it being significantly higher than the sur- rounding farmland; this height difference makes the reserve more difficult to keep wet. This problem provides an excellent example of earlier co-operation between two of the Great Fen partners, when, some 40 years ago, a scheme was developed to prevent water leaking from Woodwalton Fen to the surrounding, lower-lying farmland. The perimeter of the Fen was sealed with an impervious clay cut-off, dug down through the remaining peat to the underlying clay, which stopped the majority of the leakage. In return for carrying out this work, the Middle Level Commissioners were allowed to store floodwater in the reserve, thus simultane- ously keeping the reserve wet while pre- venting the flooding of the surrounding farmland. Unfortunately it now appears that floodwater storage may be having a delete- rious effect on the reserve, and this problem must be addressed, perhaps using other, less sensitive areas within the Great Fen Project area for the storage of floodwater. The Great Fen has truly caught the public’s imagination. In 2008, a record- breaking grant of £7. 2m from the Heritage Lottery Fund, together with matching funding from a large number of other donors, enabled the purchase of a large estate in the north of the Great Fen area. As a result of these and other grants and donations the present-day statistics are impressive: the part- ners now own 55.5% of the project area, and by the end of 2013 41% (1,519 hectares) will be managed for conservation. Not there yet, but at this rate it will not be long! These are the hard facts, but what is equally encouraging is what is being achieved on a broader front under the umbrella of nature conservation. The Royal Institute of British Architects held a competition to design the Great Fen Visitor Centre, their second most successful competition ever, with 201 entries, and now fundraising is underway to construct the winning design. Education has for some time been an important element of the Wildlife Trust’s work, with much co- operation with local schools, and the Great Fen Project is continuing and extending this. There is a strong volunteer base, helping with such activities as fundraising and guided walks, while volunteer rangers assist visitors at weekends. The local economy is being devel- oped with the assistance of the local authority, improving the tourist infrastruc- ture, commissioning footpaths, cycle routes, and offering local business opportunities. Although the project has not been without some controversy - the main criticism being that it will potentially drown high-quality arable land - the result is that the Great Fen Project is rapidly developing into a land- scape-scale nature reserve. Within its bound- aries visitors will find a landscape comparable to the ancient fenland. With good luck and good management its sheer size will allow space for wildlife to develop, and to attract such potential breeding species as Common Crane Grus grus. Surely it will prove to be an exciting area for nature, enjoyed by both the local community and all who love wide open spaces, and will come to be seen as a flagship landscape-scale reserve and a fitting memorial to those pioneering nineteenth-century naturalists who started the nature conservation movement. Richard Chandler What do you think? Join the debate at www.britishbirds.co.uk/category/editorials British Birds 106 • November 2013 • 646-647 647 Arnoud van den Berg News and comment Compiled by Adrian Pitches Opinions expressed in this feature are not necessarily those of British Birds New species of owl discovered in Oman In another groundbreaking piece of research by The Sound Approach team, a new Strix owl has been identified, not on a remote Indonesian island but in the Middle East. The team has described its findings in Dutch Birding, where it gives the new taxon the name Omani Owl Strix omanensis. In late March this year, Magnus Robb and Rene Pop spent seven nights studying Pallid Scops Owls Otus brucei at Al Jabal A1 Akhdar, the central part of the Al Hajar Mountains in northern Oman. Having located the Scops Owls in a dry wadi, they spent several nights recording their calls. But one night, MR detected through his headphones a totally different owl call that sounded unlike that of any other Palearctic owl species. He likened its four hoots to the opening bars of the famous wedding march from Richard Wagner’s opera Lohengrin. (The new species could indeed have been called ‘Here Comes The Bride’ Owl!) The compound structure of its hooting suggested a Strix owl but one with a voice even deeper than Ural Owl S. uralensis. On their final night in Oman, MR and RP made good-quality recordings of the Strix owl calls but failed to see the bird before they had to leave for the airport. Within a month, MR was back in Oman and back in the wadi for another week of nocturnal vigils, this time with Arnoud van den Berg. On their fifth night, playback of the Strix call finally lured the mystery owl into their torch beams. It had the shape of a Strix owl, lacking ear tufts and appearing slightly smaller-headed than a Tawny Owl S. aluco. Since it was facing them, only its face and underparts could be seen, which looked rather featureless except for the presence of a dark breast band, and AvdB noted obvious longi- tudinal stripes. Both observers were convinced that this was not the other Middle Eastern desert Strix , Hume’s Owl S. butleri (which occurs in southern Oman), and the capture of good-quality images became the next priority. But this had to wait for a third expedition to the Omani desert, in late May, when AvdB returned with Cecilia Bosman. By this time even night-time temperatures were exceeding 40°C! It took a further week of playback and climbing up the lower walls of the wadi before AvdB obtained the first photograph. He subsequently obtained an excellent series of perched and flight shots of the mystery owl. He and CB returned to Oman again in July to obtain further photographic and audio evidence. Sonograms of the owl’s calls were clearly dif- ferent from all Hume’s Owl recordings. The team then searched all available museum collections for specimens of their ‘Omani Owl’ but could find none. There were no Strix speci- mens at all in the museum in Muscat, Oman, and only three specimens of Hume’s Owl in the NHM at Tring. All three differed in the tones of their upperparts but all three showed hori- zontal bars on the under- parts, a feature not shown by the Omani owls. The team concluded: ‘Vocalisations and plumage characteristics demonstrate that the owls sound- recorded and photo- graphed in the Al Hajar Mountains of Oman repre- sent a new species, which we propose to name Strix 648 © British Birds 106 • November 2013 • 648-650 News and comment omanensis , Omani Owl.’ This is the first ‘non-cryptic’ bird species to be discovered in the Western Palearctic (using the widely recognised modern definition of the WP to include the Arabian Peninsula) for 40 years, the last being Algerian Nuthatch Sitta ledanti in 1973. Record season for Bald Ibises in Morocco The largest fully wild population of the Critically Endangered Bald Ibis Geronticus eremita has had its second most successful breeding season on record, with the number of breeding pairs at its highest since surveys began in the 1980s. The colonies at Souss- Massa National Park and nearby Tamri, in southwest Morocco, fledged 148 young, bringing the total population at the end of the breeding season to 443 birds. Once widespread in North Africa and Europe, the Northern Bald Ibis survives in two disjunct populations. Well to the east of the Moroccan birds is the semi- captive population at Birecik in Turkey, and south of that a tiny remnant population at Palmyra, Syria. Management and conservation of the Moroccan population is supervised by SEO/BirdLife (BirdLife in Spain) in conjunction with the High Commission for Water, Forests and Desertification Control and GREPOM (BirdLife in Morocco). The Tamri colony commenced breeding in early February. In contrast to 2012, when breeding at Tamri failed completely, possibly because of low rainfall, 60 pairs fledged 71 young. In Souss Massa National Park breeding did not begin until the first week of March. One subcolony of six pairs was prevented from nesting by disturbance from a Lanner Falcon Falco biarmicus , which flew so regu- larly over the nest ledge that the birds abandoned it. The remaining 53 pairs produced 77 fledglings. 382. Adult and juvenile Bald Ibis Geronticus eremita, Morocco. The success may be due partly to better weather, including 200 mm of rainfall between September 2012 and April 2013, which improved prey availability. ‘Our monitoring shows that the Northern Bald Ibis population at Souss-Massa contains the largest number of breeding pairs recorded since conserva- tion of this species began in 1993, and certainly since the first surveys of the species in the early 1980s,’ said SEO/BirdLife’s Jorge Fernandez Orueta. ‘Only in 2004 was the number of fledged young higher, and if it were not for the territorial behaviour of the Lanner Falcon, this year would probably have exceeded it.’ Red-backed Shrikes nest on Dartmoor again Closer to home it’s been another successful breeding season for the Red-backed Shrikes Lanius collurio that first colonised Dartmoor in 2010, as a pair fledged two young. Kevin Rylands from the RSPB said: ‘This is now the fourth year the shrikes have returned to Dart- moor, but they failed to breed successfully last year, probably due to the awful weather. A lone male visited the previous breeding site in May this year but failed to find a mate. Fortunately, a pair was found at a new site in June and this bodes well for the future of the species on Dartmoor.’ Red-backed Shrikes were driven to extinction in the UK at the end of the last century and egg-col- lecting remains a major threat. Wildlife Crime Officer PC Josh Marshall, of Devon and Cornwall Police, has been directly involved in the protection scheme since the birds first bred on Dartmoor in 2010. He explained: ‘As in previous years, we used a combination of volunteers, staff and sophisticated wildlife surveillance equipment as part of site pro- tection and monitoring. Although it’s been hard work, the efforts have been rewarded with two youngsters fledged. We are particularly grateful to the volunteers involved and to Devon Birds for funding some of the cameras used on site.’ British Birds 106 • November 2013 • 648-650 649 BirdLife International News and comment 2,000 Saker chicks reared in artificial nests An extraordinary 2,000 Saker Falco cherrug chicks have hatched in 2013 as a result of an artificial nesting programme in Mongolia. The project, which is run by the Environment Agency of Abu Dhabi (EAD) in partnership with Mongolia’s Min- istry of Nature, Environment and Tourism, was launched in 2010 with the aim of increasing the wild Saker population. EAD, on behalf of the UAE Government, signed an agreement with the Mongolian Govern- ment to build 5,000 artificial nests in the Mongo- lian steppes to increase the world’s population of Sakers, which, of course, has suffered significantly because of the demand for falconers’ birds fuelled by oil-rich Middle Eastern states. Indeed, this project is an acknowledgment that, for the past 20 years, the UAE has been the main destination for thousands of falcons caught and sold illegally for hefty sums on the black market. (Neigh- bouring Kazakhstan is estimated to lose up to 1,000 Sakers per year to Middle Eastern falconers.) EAD reported that 3,700 chicks have been hatched since the project was first launched. This year the project has also addressed the problem of Sakers being electrocuted by power lines by adding insulation covers near nest-sites. Razan Khalifa A1 Mubarak, Secretary General of EAD, said: ‘This initiative was introduced to promote sustainable breeding practices and to provide birds with safe and secure breeding envi- ronments in a bid to boost global population numbers. I am happy to report that we have also built up the capabilities of local biologists and have incorporated an educational programme in schools in Mongolia, as well as two schools in Abu Dhabi.’ Middle East Falcon Research Group www.savethesaker.com RSPB has a new (TV presenter) President RSPB members have elected BBC TV presenter Miranda Krestovnikoff as the Society’s new Presi- dent. She follows in the footsteps of fellow TV star Kate Elumble, who was elected in October 2009 and who also followed another TV presenter, Julian Pettifer. Ms Krestovnikoff is best known for her appear- ances on The One Show and as one of the original Coast team members on BBC2. In her new role, Ms Krestovnikoff aims to increase support for the RSPB. She wants to celebrate British wildlife and raise awareness of the wildlife we have on our doorstep. In particular she wants to enthuse children about nature, and help adults learn how to enthuse their children. Mike Clarke, RSPB Chief Executive, said: ‘The start of Miranda’s RSPB Presidency comes at a time when nature is facing huge challenges. To meet these challenges we need to grow support for nature by inspiring more people to enjoy it, care about it and save it. Miranda brings a natural ability to enthuse others about nature and we are pleased to welcome her as our new President.’ In praise of... The Sound Approach To find one previously undescribed species within the boundaries of the Western Palearctic may be regarded as fortunate; to discover two within the space of five years is nothing short of miraculous, as Oscar Wilde might have observed. These are the achievements of the Anglo-Dutch collective The Sound Approach, whose mission is to turn birdwatchers into ‘bird listeners’. The formidable sound-recording talents of Mark Constantine, Marcus Robb and Arnoud van den Berg are backed up by the expertise of identifica- tion consultants Killian Mullarney, Dick Forsman and Rene Pop. The team has produced some of the most innovative bird books of recent years (for example, The Sound Approach to Birding and Petrels Night and Day) and along the way described Monteiro’s Storm-petrel Oceanodroma monteiroi from the Azores (previously a ‘cryptic’ population of Madeiran Storm-petrel O. castro) and now the Omani Owl. In addition, AvdB played a decisive role in the 2009 rediscovery of Turkish Brown Fish Owl Bubo zeylonensis semenowi, which may yet prove to be another new species. As the team modestly observes: ‘For us to have been involved in such events twice in five years suggests that there is still much to discover in the WP, especially during the night. . .’ For extended versions of many of the stories featured here, and much more, visit our website www.britishbirds.co.uk 650 British Birds 106 • November 2013 • 648-650 Subalpine Warbler variation and taxonomy Lars Svensson Abstract The current knowledge of variation within the Subalpine Warbler complex is summarised. A new morphological character, regarding differences in the tail pattern, is described here, which will help in the diagnosis of taxa away from the breeding grounds. This new character is central to the proposal of a three-way split of the complex: Western Subalpine Warbler Sylvia inornata , Moltoni’s Warbler S. subalpina and Eastern Subalpine Warbler S. cantillans (see Svensson 2013). Based on current knowledge of morphological characters and confirmed by genetic sampling, it is established that the first British record of Moltoni’s Warbler was a male on St Kilda, Outer Hebrides, on 1 3th June 1 894. The Subalpine Warbler Sylvia cantillans as traditionally constituted breeds across much of the Mediterranean region, from Iberia and Morocco in the west to western Turkey and northwest Libya in the east, including southern France, large parts of Italy and southeast Europe (e.g. Shirihai et al. 2001). Four subspecies are generally recog- nised, with the nomenclature of most current handbooks and checklists giving us nominate cantillans (Iberia, France, the northwest corner of Italy and, after an apparent gap in © British Birds 106 • November 2013 • 651-668 the distribution, in central and south Italy), inornata (Morocco, northern Algeria, Tunisia, northwest Libya), albistriata (the northeast corner of Italy, Balkans including Greece, southern Bulgaria, western Turkey) and subalpina (often referred to as moltonii, but see below; in Mallorca, Corsica, Sardinia and much of the north Italian mainland). The taxonomic status of all these populations will be discussed below. Hereafter, when the taxon names inornata , albistriata and sub- alpina are used alone they refer to the 651 Richard Johnson Lars Svensson © Natural History Museum, Tring Svensson Avelianeous *Fa m Color *Wood Brown Bisffy Brown Natal Brown Olive- Brown mentioned breeding ranges, whereas the name cantillans is generally replaced by either ‘western cantillans ’ (mainly Iberia and France) or ‘eastern cantillans ’ (central and south Italy). MoltonPs Warbler is different It is now well established that the taxon sub- alpina , usually called Moltoni’s Warbler, is quite distinct. Orlando (1937, 1939) noted this in Sardinia, and gave a quite detailed account of Moltoni’s Warbler and the known variation among Italian taxa. Based on studies in the Balearics, Gabriel Gargallo (1994) reported the different male plumage and call note of Moltoni’s Warbler. Shirihai et Pole Vinaceous-Fawn Tilley I- Buff Light Vinaceous-Fawn *Vsnaceous-Buff Vinaceous-Fawn y- 383. A male Moltoni’s Warbler Sylvia subalpina next to plate XL in Ridgway ( 1 9 1 2), showing that the underparts colour in this species is closest to some combination of ‘Vinaceous’ (‘Vinaceous-Buff’ or ‘Vinaceous-Fawn’) or to ‘Avelianeous’. al. (2001) treated Moltoni’s Warbler as a dis- tinct subspecies and suggested that it could warrant allospecies status (they were unaware of its parapatric breeding range with eastern cantillans) based on male plumage, its dif- ferent moult strategy, phenology (later arrival in spring), call and a 3.7% difference in the mitochondrial-DNA cytochrome-^ gene. Mattia Brambilla and colleagues (2008b, 2010) confirmed that Moltoni’s Warbler differs substantially in the mitochondrial cytochrome-^ gene. Moltoni’s Warbler is known to breed sympatrically with Subalpine Warbler (eastern cantillans) in northern Italy, apparently without mixing (Brambilla et al. 2008a), which is the strongest evidence for Moltoni’s being a species in its own right. Moltoni’s Warbler differs in male plumage by being paler and more buff-pink below, not rusty- orange or brick- red. The pink colour is also more uniform across the throat, breast, belly and flanks, with lower areas of the underparts being only subtly paler than the throat and breast (and lacking the clear contrast between the darker chest and whitish belly and pale flanks seen in eastern cantillans and albistriata) . The underparts coloration in male Moltoni’s Warbler is closest to one of ‘Vinaceous-Buff’, ‘Vinaceous-Fawn’ or ‘Avelianeous’ in plate XL of Ridgway (1912) - see plate 383 - (or 652 British Birds 106 • November 2013 • 651-668 Subalpine Warbler variation and taxonomy ‘Vinaceous’ in plate XXVII - not shown), but in everyday English would best be described as buffish-pink, salmon-pink or ‘old rose1. The call of Moltoni’s Warbler differs clearly from that of all other Subalpine Warbler populations, being a brief metallic or buzzing, dry trill, zerrr or trrrt , recalling Wren Troglodytes troglodytes or (perhaps) Long-tailed Tit Aegithalos caudatus. The call of both western cantillans and inornata is a monosyllabic, short, dry clicking tec or tett, rather like Lesser Whitethroat S. curruca. The call of eastern cantillans is very similar but slightly fuller chep or trep , almost like the call of Red-necked Phalarope Phalaropns lobatus or Radde’s Warbler Phylloscopus schwarzi (pers. obs.). The call of albistriata is similar to that of eastern cantillans but is usually (though not invariably) quickly doubled, te- ret or che’rep. See also published sonograms of the different calls (Gargallo 1994; Shirihai et al. 2001). The song of Moltoni’s Warbler is also dif- ferent. With practice it is possible to hear it as faster, slightly higher-pitched and more insect-like; it is buzzing or scratchy and hard in tone and it may even vaguely recall the song of European Serin Serinus serinus , a comparison that does not readily come to mind when listening to song of western can- tillans, which more closely resembles that of Linnet Carduelis cannabina (with its slightly slower pace, frequent insertion of the tett call and more clear and varied structure). The buzzing call is also frequently incorporated within the song of Moltoni’s. In addition to differences in morphology and vocalisations there is genetic evidence that Moltoni’s is different. Brambilla et al. (2008b) showed that Moltoni’s Warbler differs from western cantillans by 4.3%, and from albistriata by 4.4% in the mitochondrial cytochrom e-b gene. Interestingly, the greatest genetic difference, 5.0%, was found when Moltoni’s Warbler was compared with eastern cantillans. This level of reciprocal genetic divergence between taxa is strongly suggestive of reproductive isolation or a long period of isolation, consistent with separate species status, although there are no firm rules for this. Nonetheless, a difference of 4-5% is a strong indication of separation at the species level, and taken together with other consistent and diagnosable differences in morphology, vocalisations or behaviour it strengthens the argument considerably (though note that so far only mitochondrial DNA has been analysed). Initial problems with the Moltoni’s split Three factors caused initial doubts over the splitting of Moltoni’s Warbler from the remaining Subalpine Warbler forms, even though the arguments in favour were in many respects convincing. hirst was an assumption (now known to be false) that all mainland Italian breeders were cantillans. Lor example, in June 1998, I noted ‘Subalpine Warblers’ in Tuscany (south of Sienna) calling with a Wren-like trill; I con- cluded (wrongly as it turned out) that there was variation in the call types of cantillans (whether western or eastern). Soon after, Lestari et al. (2002) drew attention to the fact that Moltoni’s Warbler was not confined to the large islands in the western Mediterranean Sea (part of the Balearics, Corsica and Sardinia) but bred also in the north Italian mainland. In 1998, I was also unaware of an earlier pub- lished report of this fact (Cova 1976). More recently, Brambilla et al. (2006) described an ambitious mapping project covering much of mainland Italy (676 sites visited), mainly during 2000-04, which firmly established that Moltoni’s Warbler is widespread in Tuscany and Emilia Romagna and also in parts of Piedmont, Lombardy and coastal Liguria. Second, published photographs, including those in Shirihai et al. (2001), appeared to show a much wider morphological variation in the male plumage of both Moltoni’s Warbler and cantillans than was reassuring, contradicting written claims in the same book. In particular, pi. 68: 4 is confusing, claiming to show an adult male Moltoni’s yet the bird resembles a western cantillans. Others which are almost as puzzling are pi. 62: 3 (a first-summer male cantillans being pale and salmon-pink like a Moltoni’s) and pi. 67: 3 (first-summer male albistriata with poorly developed brick-red throat that resembles Moltoni’s). An initial search of museum specimens seemed to confirm this large variation. Third, it seemed that splitting Moltoni’s British Birds 106 • November 2013 • 651-668 653 Svensson Warbler left the remainder of the Subalpine Warbler complex in a less than perfect order. Brambilla et al. (2008b) noted that what was left when Moltoni’s Warbler and albistriata were removed, in other words the subspecies cantillans in a broad sense, was polyphyletic and formed two distinct clades - one in Spain and France (western cantillans ) and one in southern Italy (eastern cantillans ) - and postulated that ‘a further subdivision into two branches... representing two dif- ferent (allo)species, could be expected.’ Support for such a split was deemed insuffi- cient based on the data available, and the question was left unresolved - in spite of the fact that these two populations (of cantillans) differed genetically by as much as 3.7%, almost on the same level as Moltoni’s Warbler differed from either. The first concern soon proved unfounded. There is apparently no variation in call types. Moltoni’s Warbler and cantillans (western or eastern) call differently, and do so consis- tently. The birds encountered in Tuscany giving a trilling call were simply Moltoni’s Warbler, not cantillans ; I just did not examine the male well enough to realise this. The second issue, the seemingly large vari- ation in the male plumage of Moltoni’s Warbler, was set to rest by more careful exam- ination of all available specimens of Moltoni’s Warblers in several museum collections (table 1), plus some limited but useful fieldwork in Mallorca, and Emilia Romagna and Cam- pania, Italy, together with local workers (see acknowledgments). These studies provided a sample of 43 males of Moltoni’s Warbler in 2CY spring plumage or older, which showed clearly that it is possible to recognise this form by the colour of its underparts - invari- ably salmon-pink, usually slightly paler and more uniform. The darkness of the colour can vary somewhat, a few males being a little more deeply saturated than others, but the colour shade invariably lacks any orange and is quite consistent and possible to distinguish after some practice. The darkest birds are closer to pale albistriata than to cantillans, owing to the absence of any orange element in the red ( albistriata averaging slightly darker, more brick-red, and less orange). Some of the variation evident from study specimens in museum collections proved to be due to misidentifications, seemingly often caused by reliance on locality rather than morphology. A few were quite obvious, while others were proven wrong in a targeted test of 16 selected specimens which were DNA sampled using the cytochrom e-b gene by Martin Irestedt. Once these labels had been corrected, a much more consistent pattern in terms of plumage colours emerged (e.g. plate 384). The apparently large variation in the plates in Table I . Number of specimens of the various taxa within the Subalpine Warbler complex examined. Nomenclature follows Svensson (2013). NHM = Natural History Museum, Tring; AMNH = American Museum of Natural History, New York, USA; MNHN = Museum National d’Histoire Naturelle, Paris, France; NRM = Naturhistoriska Riksmuseet, Stockholm, Sverige; NBC = Naturalis Biodiversity Center, Leiden, Netherlands (incl. collections previously kept in Amsterdam); ZFMK = Zoologisches Forschungsmuseum Alexander Koenig, Bonn, Germany; ZMC = Zoologisk Museum, Copenhagen, Denmark. inornata iberiae subalp (Iberia, France) ina cantillans (C & S Italy) albistriata NHM 11 8 23 8 9 2 2 1 25 19 AMNH 15 4 5 1 12 11 2 0 29 5 MNHN 8 4 8 2 7 0 0 0 16 3 NRM 2 1 0 1 4 3 0 0 11 4 NBC 1 0 4 3 11 3 3 0 5 2 ZFMK 1 1 0 0 3 3 0 0 2 0 ZMC 1 0 1 0 0 0 3 1 4 1 Live birds 0 0 0 0 7 2 8 1 0 0 Totals Total no. of 39 18 41 15 Cf Cf 243, of $ 9 94, grand total 337. 53 24 18 3 92 34 654 British Birds 106 • November 2013 • 651-668 Subalpine Warbler variation and taxonomy 384. Misidentifications of museum specimens can alter the perception of normal variation within any one taxon of the Subalpine Warbler complex. This photo shows a series at NHM,Tring, meant to represent Moltoni’s Warbler S. subalpina and implying extensive variation in darkness and colour shade. When the DNA of the two birds at the left-hand end (Formentera, Balearics, 19th April 1914, NHM 1914.8.3 1.38, and Bordighera, N Italy, 9th April 1872, NHM 1 875.5.5.4 1 , both with rather dark and orange-tinged red underparts) was sequenced, they proved to be Western Subalpine Warblers S. /nornota.The remaining birds in the series are true Moltoni’s Warblers and show less variation. Shirihai et al. (2001) may be at least partly due to reproductive shortcomings or in some cases an unusual bird. In pi. 68: 4, 1 suspected a misidentification, but G. Gargallo, the pho- tographer, assured me that this bird was a genuine Moltoni’s Warbler, a dark-pigmented one but with correct pinkish hue when handled. Nonetheless, the outcome in the printed book is misleading. The caption tries to convince the reader that the colour of the underparts of this bird fits a desired pattern (‘note lack of pure orange pigmentation), yet the red on throat and breast of this bird appears distinctly orange-tinted. Vaurie (1954) did not recognise Moltonis Warbler because he felt that the underpart coloration of males was too variable, ren- dering Moltoni’s insufficiently distinct. It seems obvious that his view was caused by several misidentified specimens in museums, identifications which he did not challenge, or by migrants collected within the ranges of other taxa in the complex and hence wrongly labelled. In terms of the third issue, the genetic data presented in Brambilla et al. (2008b) seemed robust. There is obviously consider- able genetic distance between Moltoni’s Warbler and all remaining populations of the Subalpine Warbler, but also substantial dif- ference between western and eastern cantil- lans. Only between eastern cantillans and albistriata is the distance less obvious (1.7%). The logical solution to this situation was first hinted at by Brambilla et al. (2008b) and later (independently) by Hadoram Shirihai (pers. comm. 2011): to split Moltoni’s Warbler from the remaining Subalpine Warbler forms, but also to divide the latter in two. Thus Western Subalpine Warbler would be western cantillans and inornata , whereas Eastern Subalpine Warbler would be albis- triata and eastern cantillans. This arrange- ment will be discussed in more detail below. British Birds 106 • November 2013 • 651-668 655 Lars Svensson © Natural History Museum, Tring Lars Svensson © AMNH, New York Svensson 385. Male plumages of the three species of the Subalpine Warbler complex. - a. Western Subalpine Warbler Sylvia inornata. Note rufous-orange colour of much of the underparts, and that the flanks are nearly as dark as throat and breast. All three are of the European subspecies iberiae. — b. Moltoni’s Warbler S. subalpina. Note buffish-pink (‘old rose’) colour of much of underparts and very little contrast between flanks and throat/breast. - c. Eastern Subalpine Warbler S. cantillans. Note slightly darker and more brick-red colour of throat and upper breast compared with Western Subalpine Warbler, and clearly paler flanks that create a contrast between darker chest and paler belly. White submoustachial stripes are often broader and more prominent than in the other two species. The two left-hand birds are of subsp. albistriata from the Balkans, whereas the right-hand bird is of subsp. cantillans from Sicily. The logic of three species Although more fieldwork remains to be done, particu- larly in northern Italy, the logic of a three-way split of the Subalpine Warbler complex is already compelling based on available data. The degree of genetic differ- ence has already been mentioned; see Brambilla et al. (2008b) for details. The haplotype network recovered in their study (p. 465, fig. 2) shows three major clades, not two, and what Brambilla et al. call ‘southern cantil- lans' (defined here as eastern cantillans) was linked genetically with albistriata of the Balkans and Turkey, not with western cantillans of France and Spain. The plumage patterns concur with the genetic evi- dence. In Western Subalpine Warbler the throat, breast and flanks (and sometimes even upper belly) are more uniformly dark red-orange (‘Russet’ or ‘Tawny’ in Ridgway 1912, pis. XIV and XV, respec- tively), on average slightly more yellowish and with less contrast between breast and flanks compared with Eastern Subalpine Warbler, in particular com- pared to subsp. albistriata (plate 385). Tail pattern differs too (see pp. 660-663), with Western Subalpine Warbler having a short and square white tip on one or a few outer tail feathers inside the outermost. The tail in albistriata has a narrow, often long and pointed, white wedge on the inner web of the penultimate feather (T5), and sometimes also similar but shorter wedges on one or two adjacent feathers next to this. Interestingly, eastern cantillans has the same (or at least very similar) plumage to that of albistriata both regarding underpart coloration (males) and tail pattern (both sexes). The slight and average plumage differences between these two will be treated below and are at subspecies level. Morphologi- cally, eastern cantillans groups clearly with albistriata , not with Western Subalpine Warbler. This finding 656 British Birds 106 • November 2013 • 651-668 Subalpine Warbler variation and taxonomy Fig. I. Map showing the approximate ranges of the three species in the Subalpine Warbler complex. The five taxa are shown in different colour shades. Note that cantillans overlaps with subalpino in northern Italy, with dots signifying localities with sympatry. Breeding range in Tunisia south of the northwest corner is scant or irregular, but probably poorly known. Based on Brambilla et al. 2006, Shirihai & Svensson in prep, and other available maps and range statements in the literature. provides the elements for diagnosis, which were lacking to Brambilla et al. (2008b) and allows a more robust split of the eastern group from the western one. The evidence thus suggests that the Sub- alpine Warbler should be treated as three dif- ferent species. The ensuing ranges are shown on the map (fig. 1), based on information available in Brambilla et al. (2006) and in several of the references cited in this paper. Nomenclature There are a number of controversial issues regarding the nomenclature of the Subalpine Warbler complex once it is split into three species. These have been addressed elsewhere (Svensson 2013), but a summary here may be helpful. Moltoni’s Warbler, named after a renowned Italian ornithologist, has been known alternatively as subalpina and moltonii. The latter name, given by Orlando (1937), seems practical at first sight, being the same as the English name. However, there are nomenclature rules to follow and, as Baccetti et al. (2007) pointed out, the older subalpina is valid and has priority. In Temminck’s plate 6, no. 2 (Temminck 1820a; see fig. 2, overleaf), a painting of what appears to be a male subalpina (on the plate marked as ‘female’, but this is questionable) shows an adult bird with lead-grey crown and mantle and well-marked white tips to the tail feathers but with pink, not orange-red, underparts. Temminck stated clearly that the bird on that plate shows the type specimen, sent to him by Bonelli and that it was the only known specimen at the time. There can be no doubt that Temminck’s plate refers to this taxon and not to cantillans. The Temminck type specimen was described (1820b) as having ‘a beautiful vinaceous colour’ below. The adjective used (Fr. ‘vineuse’, vinaceous) is exactly the same as Ridgway (1912) used for such a pink colour (plate 383). All other Sub- alpine Warbler populations in male plumage have a more orange-brown or reddish colour and thus it seems very likely that Temminck was struck by the unusual and attractive pink colour of the male Moltoni’s Warbler when he described subalpina. With a split also of the remainder of the Subalpine Warbler complex into Western and Eastern Subalpine Warblers, the scientific names of both need to be revisited. It is questionable which form Pallas’s (1764) name cantillans referred to, since Pallas gave the type locality as ‘Italy’ and the original type specimens, a male and a female, are long since lost ( fide Baccetti et al. 2007, and requests to relevant museums). Both Western and Eastern Subalpine Warblers breed in Italy, but the former is comparatively rare (perhaps only 100-200 pairs), occurring in a small area of Liguria and Piedmont near the British Birds 106 • November 2013 • 651-668 657 Svensson i . Bec-im *r faneth#, u 7i/. j'/f&afyw' ,/vA 3.U. ('iffrro/e .m#& ■ Fig. 2. Temminck’s rendering of Moltoni’s Warbler Sylvia subalpina (‘Bec-fin subalpin’) in his ‘Planches coloriees’ (Temminck 1 820a). The bird appeared in plate 6 and is captioned by Temminck as ‘female’, although the amount of lead-grey on head and mantle and pink below seem to indicate a male. It is stated to be in spring plumage, and the fresh plumage with brown-edged wing-feathers then fits Moltoni’s Warbler best; Moltoni’s frequently has a complete winter moult, while the two other species have a complete post-breeding moult. Elsewhere, Temminck (1820b) stated that this depicts the only specimen he had access to at the time, so the plate obviously depicts the lost type. The pink colour on a spring bird supports the identification as Moltoni’s, and the tail pattern (with square white tips on several of the outermost feathers) eliminates Eastern Subalpine Warbler (see main text). French border, whereas the latter is wide- spread and common in large parts of central and southern Italy, including Sicily, so it is highly likely that Pallas was referring to Eastern Subalpine Warbler. For clarity I have thus designated a neotype collected on northwest Sicily, which restricts the type locality (Svensson 2013); this means that the Eastern Subalpine Warbler becomes Sylvia cantilla ns (Pallas 1764). Since the Eastern Subalpine Warblers breeding in the Balkans and Turkey are subtly but consistently larger than Italian breeders, and since the brick-red throat and breast of males is on average a little colder and darker, the lead-grey upperparts are darker, the call is slightly different and there are some differences in mtDNA, these groups merit separate sub- species status. The name albistriata (C. L. Brehm, 1855) has long been in use for the birds breeding in the Balkans and Turkey and no change is required; the full name of this sub- species becomes Sylvia cantillans albistriata. The lectotype from Brehm’s collection is kept in AMNH, New York. Finally, the nomencla- ture of the Western Sub- alpine Warbler also needs to be adjusted. There is apparently no genetic study of the complex that includes the North African subspecies inor- nata , but both mor- phology and vocalisations strongly indicate a close relationship between inor- nata and breeders in Iberia and France. These two are therefore best seen as subspecies of the same species. With the name cantillans now given to the Eastern Subalpine Warbler, no scientific name is available for the west European breeders of Western Subalpine Warbler. The name for the species thus becomes Sylvia inornata Tschusi, 1906; for breeders in Iberia and France, I pro- posed the new name iberiae and designated a holotype for it, an adult male collected in Madrid (Svensson 2013). The complex can thus be summarised as follows: Western Subalpine Warbler Sylvia inornata (Tschusi 1906) S. i. inornata (Tschusi 1906, N Tunisia). Breeds in Morocco, N Algeria, Tunisia, locally in northwest Libya. Differs from European breeders only in its slightly warmer, more yellow-tinged colours below in both sexes, appreciable only when series are compared. Breeders in southern Iberia tend towards this subspecies in being slightly more yellowish- red, but differences slight and best kept included in nominate. Status: resident or makes local movements only. S. inornata iberiae (Svensson 2013, Madrid). Iberia, S France, parts of Liguria and Piedmont. Returns rather early in spring, usually from late March to mid April. Winters S Morocco, C and S Algeria and presumably from Senegal and Mauritania east to Niger. 658 British Birds 106 • November 2013 • 651-668 Subalpine Warbler variation and taxonomy Moltoni’s Warbler Sylvia subalpina (Temminck 1820) Monotypic ('Italy, near Turin’; fide Bonelli, in Temminck & Laugier 1820b). Mallorca and Cabrera in Balearics, Corsica, Sardinia, Emilia Romagna, Tuscany and parts of Pied- mont, Lombardy and coastal Liguria. Returns late in spring, usually mid April to mid May, probably due to, unlike other taxa, extensive pre-breeding moult in late winter. Presumed winter range is in W Sahel, recorded in N Nigeria and N Cameroon and on migration in southern Algeria; winter range probably extends at least to northern Nigeria and Niger. On migration common in N Algeria, Tunisia and S Italy. Eastern Subalpine Warbler Sylvia cantillans (Pallas 1764) S. c. cantillans (Pallas 1764, ‘Italy’ (but later restricted to ‘Licuzza, Sicily’ (Svensson 2013)). Sicily, Campania, Puglia and north to at least Umbria and Marche but probably sparsely and also locally in Tuscany and Emilia Romagna. Returns in spring from mid March to mid April. Presumably winters in W Sahel but details poorly known owing to difficulty of separating the two subspecies away from breeding range. S. cantillans albistriata (C. L. Brehm 1855, ‘Egypt’ [migrant]). Trieste, S Slovenia, Croatia and much of Balkans, Greece, S Bulgaria, W Turkey. There is an isolated possible breeding record from SW Romania (Weber 1994). Differs from cantillans by slightly larger size and longer wing (but tail similar, thus pro- portionately marginally shorter); in male plumage slightly darker and ‘colder’ brick-red throat and upper breast ( cantillans very slightly more orange-red in direct comparison of series) and darker lead-grey from crown to mantle ( cantillans a little paler and more neutral grey). In addition there are differences in call (see above) and in mtDNA (Brambilla et al. 2008b). For migration habits and winter range see also preceding sub- species; albistriata is known through collected specimens to winter in Niger, Nigeria and Chad and to winter or migrate as far west as C Morocco (Mansour, High Atlas, 28th March, DNA tested) and in S and C Algeria. Claimed by Misonne (1974) to winter in large numbers in SW Egypt near Libyan and 386. Two adult male Eastern Subalpine Warblers Sylvia c. cantillans collected in Sicily, the right-hand specimen being the neotype of this taxon, collected in Ficuzza, northwest Sicily, on 23rd May 1906 (the date visible on the label is given incorrectly as ‘March’). Sudanese borders, but not substantiated. Some reports of winter records in Sudan are presumed to refer to ‘French Sudan’ ( = Niger), but genuine records (NB risk of con- fusion with Menetries’s Warbler S. mystacea) are probably restricted to northwest corner of Sudan. Nikolaus (1987) reported only four records from Sudan, all in the northwest (and no specimens exist in larger museums from S Egypt or Sudan). Passes through N Algeria, Tunisia, Libya, N and C Egypt and Levant in spring and to a lesser extent through the east of this region in autumn. Birds are also known to pass through SE France in spring (Brambilla et al. 2012). British Birds 106 • November 2013 • 651-668 659 Svensson HUGH WHISTLER, F ; s B TICEHURST COLL: C ^tkeOUojU. & oJi , ^ T&wdgi ^ <; . cc Lomiii* QaJtt 387. Typical tail pattern of Eastern Subalpine Warbler Sylvia cantillans albistriata (all examples from the Balkans). Note narrow white wedge inside shaft of penultimate feather (T5), present in adults of both sexes and often also in first-years (d).The size and number of wedges on the outer tail feathers varies somewhat individually, some having quite a small wedge on T5 only (a), others a very extensive wedge on T5 and smaller wedges also onT4 andT3 (e). Juveniles and first- winters with retained outer tail feathers have either a diffuse hint of a wedge or none at all. - a. adult 9 Durazzo, Albania, 9th April 1935, NHM 1 949W. 1 . 1 2062; very small wedges, still narrow and pointed in shape. - b. adult cf Mostar, Bosnia-Herzegovina, 2 1 st April 1 929, NHM 1 949W. 1 . 1 2070; deep and narrow wedges onT5 (note that bothT6 feathers are more or less hidden). - c. adult 9 Durazzo, 1 0th April 1931, NHM 1 94 1 .5.30.5969; typical adult appearance. - d. first-summer 9 Durazzo, 9th July 1930, NHM 1941.5.30.4698; typical first-summer pattern with broad but poorly marked wedge sullied greyish. - e. adult Cattaro, Dalmatia, Croatia, 4th April 1901, MNHN 1 960. 1 060; extreme amount of white in outertail. All photographs by the author with permission from NHM,Tring (a-d), and MNHN, Paris (e). Differences in the tail pattern During work on a forthcoming handbook (Shirihai 8c Svensson in prep.), I examined series of all subspecies of the Subalpine Warbler complex in various museums. While doing so, I discovered an important plumage difference, present in both sexes, between (i) Western Subalpine Warbler and Moltoni’s Warbler - both of which have a small and more or less square white tip to the second- outermost tail feather, or T5 - and (ii) Eastern Subalpine Warbler (both cantillans and albistriata) - which has a long, narrow and pointed white wedge inside the shaft of T5. The white wedge of Eastern is invariably present in adults and in first-year birds that have replaced the juvenile T5, and some have even a smaller white wedge on T4 (or on both T3 and T4). In juveniles and first-years that have not yet moulted any outer tail feathers, such wedges are less distinct and can be completely missing - although in many young birds there is already a well-developed or at least diffuse wedge or hint of a paler oblong patch or stripe inside the shaft of T5. 660 British Birds 106 • November 2013 • 651-668 Subalpine Warbler variation and taxonomy 388. Typical tail pattern of Eastern Subalpine Warbler Sylvia cantillans cantillans (all examples from S Italy). Note same pattern as in albistriata (plate 387) with narrow white wedge onT5, sometimes on both T5 and T4. - a. first- summer cf Ficuzza, Sicily, 23rd May 1906, NHM 1 909. 1 1 . 1 8.50 (neotype of this taxon). - b. adult cf Palermo, Sicily, May 1 870, NHM 1 907. 1 2.20.627. - c. adult cf Palermo, 1 4th April I939.AMNH 708854. - d. first-summer $ Conza, Avellino, Campania, 2nd May 2012, live bird. - e. adult cf Bari, Puglia, 1 0th April 1884, MCZR (Rome) 10596. Photographs a-d by the author, a-c with permission from NHM, Tring, d by AMNH, New York. Picture e by Andrea Corso with permission from Museo Civico di Zoologia, Rome. The description of the tail pattern in Shirihai et al. (2001) (fig. 9, p. 379), intended to refer to all taxa, shows a typical pattern for Eastern Subalpine Warbler but is wrong for Western Subalpine and Moltoni’s. The two tail-pattern types and their moderate varia- tions are shown here with a series of photo- graphs in plates 387-390. Note that it is crucial to age any bird first before examining the tail pattern. In spring, the unmoulted outer tail feathers of first-summer birds will be worn and bleached juvenile ones and are likely to show no wedges or at least less devel- oped wedges that are not distinct and not pure white. This plumage character should prove useful when separating trapped migrants in spring and autumn, although it will not help to discriminate female Western Subalpine and Moltoni’s Warblers. Brambilla et al. (2012) noted broad-front migration in spring of most subspecies, with albistriata occurring as far west as France, and used mitochondrial DNA to discriminate between iberiae (French breeders) and nominate cantillans (southern Italian breeders), whereas of course the examination of the tail-feather pattern will provide a much cheaper and faster method for diagnosis. Occasionally the tail pattern may be less typical or even ambiguous. A few examples of less typical variations are shown in plate 391. It is prudent to check both sides of the tail before forming an opinion on any less typical birds, and the underpart coloration of males should also be factored in. Fuckily, atypical birds seem to be very rare. British Birds 106 • November 2013 • 651-668 661 Svensson 389 a. 389 b. 389 d. 189 British Birds 106 • November 20 1 3 • 65 -668 662 Subalpine Warbler variation and taxonomy 389. Typical tail pattern of Moltonis Warbler Sylvia subalpina. Note small and basically square white tip (less often bluntly wedge-shaped tip, as in c) toT5, present in adults of both sexes and often also in first-year birds. Note also less extensive white onT6 compared with both Western and Eastern Subalpine Warblers, often with white restricted to edges and tip (but variable and some overlap in this character). Ageing in spring of Moltoni s Warbler often difficult owing to typically more extensive pre-breeding moult than in related taxa. - a. cf Genoa, Italy, no date, NHM 1886.7.8.236. — b. cf Liguria, Italy, summer, NHM 1 898.9. 1 . 1 5 1 . — c. cf Serra d’Alfabia, Mallorca, 3rd June 2011, live. — d. cf Tunis, 24th April, NHM 1939.12.9.745 (labelled 'inornata' presumably based on locality, but species established by DNA). — e. first- summer cf St Kilda, Outer Hebrides, 1 3th June 1 894, NHM 1 90 1 . 1 .4. 1 . All photographs by the author with permission from NHM,Tring (a, b, d, e). 390. Typical tail pattern of Western Subalpine Warbler Sylvia inornate, a-c of subsp. iberiae (Iberia, France), d and e of inornata (NW Africa). Note usually small and basically square white tip to T5 in adults of both sexes and often also in first-years. If anything, Western Subalpine has on average more white onT6 than Moltonis Warbler but is similar to Eastern Subalpine. White tips onT5 andT4 average slightly more extensive in inornata than in iberiae, but much overlap. - a. adult cf Chamonix, France, 5th April 2000, MNHN 2000.2027. - b. adult cf Madrid, 24th May 193 1, NHM 1934. 1.1.249. — c. adult cf Coto Donana, Spain, 13th March 1910, NHM 1 934.1.1. 4755. - d. adult cf Senegal, I Ith September 1958, MNHN 1962.777. — e.first-summer(?) cf Azrou, Lesser Atlas, Morocco, 1 0th May 1 9 1 9, NHM 1 9 1 9. 1 2. 1 1 . 1 67. Photographs a-d by the author, e by Mark Adams with permission from MNHN, Paris (a, d) and NHM.Tring (b, c, e). 39 1 . Extreme variation and problem birds in terms of the tail-pattern character. - a. Presumed iberiae adult cf near Marseilles, SE France, 13th May 1917, NHM I965M I402I.T5 has small wedge rather than square tip, although the wedge is blunt and rather diffuse. Flanks dark rufous-orange with little difference from colour of chest, thus supporting the identification as iberiae rather than migrant albistriata. — b. 9 albistriata Mostar, Bosnia- Herzegovina, 2 1 st April 1 929, NHM 1 94 1 .5.30.46 1 7. Very similar to previous, with small wedge on T5, but wing of 65.5 mm in a female excludes other taxa. — c. ad $ Durazzo, Albania, 1 0th April 1935, NHM 1 949W. 1 . 1 2063. On locality this should be albistriata, but tail pattern strongly indicates either subalpina or iberiae. Should be DNA sequenced. — d. Presumed first- summer 9 albistriata, Durazzo, I Ith April 193 1, NHM 1 949W. 1 . 1 2060. Very poorly developed, vaguely paler stripe inside shaft of leftT5 only, and on right side of tail newly moulted T5 has no wedge. A very difficult bird, but wing length of 63 mm for a first-year 9 supports identification as albistriata. — e. first-summer cf subalpina Italy, spring, NHM 1 843. 1 2. 1 . 1 0. Pinkish underparts, so surely a subalpina. Maximum development of a whitish wedge on T5 to be seen in this species. All photographs by the author © NHM,Tring. ; UGH WHISTLER, f.1 S. 391 b. ■ :sr^ British Birds 106 • November 2013 • 651-668 663 Svensson Other differences between the taxa There are only slight differences in structure and size among the three proposed species, or the five taxa in the complex (table 2). Eastern Subalpine Warbler averages slightly larger and longer-winged than the other two species ( albistriata in turn slightly larger than cantillans ), and has a slightly longer tarsus. Western Subalpine Warbler has a slightly shorter and blunter wing than the other two species, although the difference from Moltoni’s Warbler is subtle. The outermost short primary (PI) is on average 1.0-1. 5 mm longer than the tip of the longest primary- covert in Western Subalpine, equal to the longest primary-covert in Moltoni’s and 0. 5-1.0 mm shorter in Eastern Subalpine Warbler. The primary projection, expressed as the distance between wing-tip and either the innermost primary (P10) or outermost secondary (SI) is larger in Eastern than Western Subalpine Warbler, while Moltoni’s Warbler is intermediate (table 2). There is overlap between all measurements, but extreme measurements can often support an identification, especially if several are used in combination. The characteristic pink-buff colour of the underparts of male Moltoni’s Warbler has already been mentioned. Males of the remaining taxa are more similar, but slight average differences can be found. Male iberiae is reddish-orange, inornata more yellowish-orange, albistriata colder and rather dark brick-red, and cantillans very similar to albistriata or a shade warmer orange-red, tending towards iberiae. Single birds can rarely be told on colour alone (except for Moltoni’s Warbler). It is often easier to use the presence or absence of a contrast between throat and chest against whitish belly. All Eastern Subalpine Warblers have this contrast, although it is somewhat variable in prominence and on average most obvious in albistriata. The yellowish tinge to the underparts of male inornata , noticeable when series are compared with series of iberiae , is also present in females. This difference in colour is the main reason for recognising this sub- species. In all other respects, it is similar to cantillans , although a closer study of genetics and vocalisations remains to be performed. Adult male Eastern Subalpine Warblers have, on average, a broader and more promi- nent white submoustachial compared with the typically narrower submoustachial of both Western Subalpine and Moltoni’s Warbler. There is individual variation, however, and paying too much attention to the prominence of the white stripe can lead Table 2. Biometrics of the various taxa within the Subalpine Warbler complex. Nomenclature follows Svensson (20 1 3). Mean values in mm. (n) = sample size; wing, range = variation within sample;T/Wx 1 00 = tail/wing ratio; bill (S) = bill length to skull; P I ±PC = tip of short outermost primary (PI) in relation to tip of longest primary-covert (PC); P I