Field Identification

Blyth’s Reed Warbler Acrocephalus dumetorum 덤불개개비, Baengnyeong Island, October 2023

Nial Moores, Birds Korea, October 31st 2023.

(All images copyright of Nial Moores / Birds Korea and not to be used or re-posted without permission).

Blyth’s Reed Warbler Acrocephalus dumetorum 덤불개개비, Baengnyeong Island, October 22nd 2023.

Baengnyeong is a big and densely vegetated island – far too large to cover properly in a day. The island’s geography, however, does help to concentrate birds and most of the better patches of habitat are obvious – wetlands or taller trees or open fields in the southwest or northeast where birds tend to arrive or depart; or higher hill peaks used by soaring birds. Some other productive habitat patches look rather less promising in satellite images.  One such area is a valley on the edge of Jinchon in the northeast. Although the valley is 600m or more inland and prone to disturbance, it is more or less aligned north-south, with rice-fields and fallow fields edged by bushes and banks of trees. Most mornings in spring or autumn, there is high turnover through this part of the island – and highlights in this valley over the years have included Green-backed Flycatcher Ficedula elisae, two Asian Koel Eudynamys scolopaceus together and Korea’s first (and so far only) Buff-throated Warbler Phylloscopus subaffinis.

In the early hours of October 19th, a low pressure system crossed the Yellow Sea, and winds swung to the northwest. These winds strengthened substantially by the 20th, gusting up to 50 km/hr on the island with an airflow reaching deep to the west and northwest, before starting to weaken during the afternoon of the 21st.

At dawn on 22nd, conditions were clear and calm, and many birds were vocalizing in the valley including a late singing Grey-backed Thrush Turdus hortulorum and rarer still, a Red-breasted Flycatcher Ficedula parva. Waiting for better views of the flycatcher, I heard a series of unfamiliar deep, well-spaced “Chak” notes coming from a nearby area of tangled bushes and trees – similar in structure to the typical “bamboo-snapping” call of Japanese Bush Warbler Horornis diphone, but louder, sharper and less slurred, more like an Oriental Reed Warbler Acrocephalus orientalis but without any of these well-spaced calls running into each other.

Vocalisations recorded on smart phone, and boosted in Audacity. Nothing can quite capture the thrill and anxiety of hearing an unknown call, and glimpsing a completely unremarkable-looking brown bird disappearing deep into cover…

Initial, brief views of the bird in the dappled shade of early morning revealed a plain looking, unstreaked, medium-sized, brownish warbler, with a long-looking tail and bill and a strikingly plain-looking face. Although views were very brief, the tail sides looked whiteish, suggesting Booted Warbler Iduna caligata 쇠덤불개개비. The bird then disappeared into deeper cover, with the calls gradually becoming more and more distant. 

Initial views suggested pale outer tail feathers – among warblers recorded in Korea only shown by e.g. Lesser Whitethroat Curruca curruca, which would show strong grey tones to the head, and Booted Warbler Iduna caligata 쇠덤불개개비, which would look rather plain-faced…
The bird was remarkably plain-faced, with a very short supercilium, only reaching to the eye and not beyond it (ruling out species like Paddyfield Warbler Acrocephalus agricola 북방쇠개개비 and Manchurian Reed Warbler A. tangorum 우수리개개비), and an obvious pale eye-ring. The lack of well-defined dark eye-stripe but presence of a dark loral spot and very weakly darker brown behind the eye combined with a long, angular, sloping forehead and dagger bill were nonetheless more indicative of a reed warbler species than a species like Booted Warbler 쇠덤불개개비.

A very long twenty minutes passed, before the calls started again close by, with the bird foraging, mostly concealed, both in low tangles and also 2-3 meters up in the lower branches of trees.

Better views of the bird over the next three hours revealed that the suggestion of white in the outer tail feathers was more likely an illusion caused by some wear or bleaching combined with long (and quite baggy), pale undertail coverts wrapping up around the tail sides. These long undertail coverts narrowed the search down further: this was indeed an Acrocephalus reed warbler.  

Side views confirmed that the undertail coverts were rather long (a feature of reed warblers), and that the tail sides were not obviously paler, ruling out species like Booted Warbler. The bird looked quite short-winged, with short primary projection, i.e., the length of the exposed primaries was only about 50% of the length of the exposed tertials with only six primary tips visible (the still unrecorded Eurasian Reed Warbler A. scirpaceus typically has obviously longer primary projection than this, and shows 7 or 8 primary tips on the folded wing, making the wingtip look longer and rather less “blunt”).

The lack of a darker brow quickly ruled out species like Black-browed Acrocephalus bistrigiceps and Manchurian Reed A. tangorum; and ‘scope views suggested obvious emarginations –asymmetrical narrowing of the outer webs of the primary feathers – which were confirmed on p3 and p4 in the images.

Technical literature often focuses on the relative length and spacing of the ten primaries in warblers as p (primary)1 to p10, with p1 stacked lowest in the primaries of the folded wing, and p10 stacked highest. In Blyth’s Reed Warbler, p1 more or less reaches or falls short of the tips of the primary coverts (indicated as a red line, drawn below their lower edge for clarity). In addition, in Blyth’s Reed Warbler there are obvious emarginations (asymmetrical narrowing) in p3 and p4, which start to the right of the yellow lines. For a useful look at emarginations in vagrant Blyth’s Reed Warbler please see this wonderfully helpful blog post in the UK and the links from it.

These emarginations combined with the dark-looking dull plumage, lack of pale tips to the primaries, relatively short primary projection, and call appeared to rule out still-unrecorded species like Eurasian Reed A. scirpaceus and Marsh Warblers A. palustris, and left only two real options: Blyth’s Reed Warbler A. dumetorum and Blunt-winged Warbler, A. concinens, a rather poorly-known species yet to be recorded in Korea.

The bird actively foraged low down and along tree branches, stretching to glean insects from the underside of leaves. In the lowest image, the relative lengths of p1 (falling short of the tips of the coverts) and p2 (more or less equal in length to p6), and the emarginations on p3 and p4 are all shown clearly.

Based on some recordings of Blyth’s Reed and a few of Blunt-winged Warblers on Xeno-canto, and the long look to the bill, the most likely candidate was Blyth’s Reed Warbler. This is a species which is already included on the Birds Korea checklist on the basis of a bird photographed in woodland on Chilbal Islet in Shinan County by Park Chong Uk and Kim Yang Mo on August 30th 2018. Their identification was based on a review of images by Peter Kennerley and Richard Porter – which oddly did not include any reference to Blunt-winged Warbler. Calls of the Chilbal bird were either not heard or were not recorded; and the relative primary lengths seem a little challenging to interpret in the images, though the long-looking bill and fairly strong facial pattern were considered to be strongly indicative of Blyth’s Reed Warbler.

The question remained: was it possible to confidently exclude Blunt-winged Warbler, so that the Baengnyeong bird could be identified as Blyth’s Reed Warbler? I shared several images with Professor Phil Round in Thailand, an expert authority on Blunt-winged Warbler and other Acrocephalus. He very kindly and helpfully confirmed that, to his eyes, the images “tick all the boxes” for Blyth’s Reed Warbler, with the tip of “p2 roughly = p6”; with p1 “almost certainly shorter than the longest primary coverts”; and with a dark loral spot. In addition, he summarized the main differences from Blunt-winged as: (1) usually longer p1 in Blunt-winged than suggested in the images of the Baaengnyeong bird; (2) much shorter p2 in Blunt-winged, with p2 usually = p8 or p9; (3) plainer face in Blunt-winged; and (4) shorter bill in Blunt-winged. Now we know what to look for!

Habitat and behaviour also seemed to lend (weak additional) support to this identification too – as Blunt-winged is more often associated with wetland and grassland. The Baengnyeong bird could easily have moved into more wetland type habitat, but instead foraged actively in among trees and bushes – behaviour apparently often shown by Blyth’s Reed Warbler.

Habitat patch used by the Blyth’s Reed Warbler, Jinchon, Baengnyeong Island, October 22nd-26th

Reference to Kennerley & Pearson’s “Reed and Bush Warblers” (published in 2010 by Helm) enable aging as a bird born this year, as the irides are grey-brown.

Because of potential interest in the record, on October 22nd I shared an image of the bird labelled as Blyth’s Reed Warbler on Facebook and also listed the record with an image on eBird. However, perhaps unsurprisingly (considering just how plain-looking this bird was!) no-one requested any details of the exact location – even though the bird remained in the same trees and bushes at least until October 26th when I left the island.

To summarise: this is the second national record of Blyth’s Reed Warbler in the Republic of Korea. The species was identified on the basis of the call (recorded on a standard cell phone); on details of the head pattern; and especially through details of the primaries, visible in digiscoped images. Support for the identification was provided further by the habitat: few reed warbler species actively prefer to forage in woodland and woodland edge.

Finally, my thanks once more to Prof. Phil Round; and also to Dr Park Jong-Gil for sharing details of the 2018 record (and additional fascinating discussion on warbler ID!) .

More from Field Identification