Himalayan Swiftlet: 1. Sighting

posted in: Swifts-Swallows | 11

On 7th January 2008 at 1047 hours, KC Tsang was witness to a number of Himalayan Swiftlets (Aerodramus brevirostris) among a flock of swifts and Barn Swallows (Hirundo rustica) hunting for insects stirred up by grass cutting activity at the grounds of Turf Club City. The number of birds hovering around was about 50.

The occurrence of Himalayan Swiftlet in Singapore has been mired in controversy since the 1990s when R. Subaraj, among others, first reported it as an uncommon passage migrant (Wang & Hails 2007). Unfortunately these observations were not authenticated with notes and thus not included in the latest Annotated Checklist of the Birds of Singapore.


Now, KC has provided hard evidence in the form of an image. He posted his image in BirdForum.Net and got a response from Sean in Hiroshima, Japan, who noted:

“…think there are four possible swiftlet species in Singapore. …Two are resident – Black-nest and Germain’s while two are migrants – Glossy and Himalayan.

“I have only seen three of the species, Black-nest, Germain’s and Himalayan, but Himalayan seems right to me for the following reasons.

“Glossy can be ruled out, as it is very small, has dark throat that contrasts with white belly and a slightly forked tail. Black-nest is darker underneath and has no notch in the tail. It also has a pale rump band, though this cannot be seen in the picture. Germain’s is possible as it is light underneath on throat and belly, but as it only has a shallow forked tail this seems unlikely. This species has a very obvious wide white rump, though again this is not viewable in the picture.

“So Himalayan with its relatively pale underparts and deeper notched tail seems most likely, They do have a paler area on the rump in some races, but never as white or obvious as Germain’s or even Black-nest.”

Most of the experienced younger birders are confident that it is Himalayan but the older ones are rather hesistant to give their outright support.

Except Subaraj, who is naturally jubilant, that he has been right all along: “Nice shot! Pity that you did not manage a shot of the upperparts as well but based on the curved, pointy wings and prominent tail notch, I would agree that this is a Himalayan Swiftlet.

“I cannot be certain of which race you have photographed though as there is no view of the upperparts and the colouration of a photo is not always the exact colour of the actual bird.

“Based on my observations over the years, the Himalayan Swiftlet is one of our commonest migrants. Most pass through on passage but birds can be seen throughout the wintering months. They can be seen anywhere in Singapore and Johor. Birds on passage can regularly be observed at various sites including MacRitchie, Changi Reclaimation, Serangoon, Tanjong Piai, Bintan or even on a boat between Singapore and the Riau Islands. At the Gap, at Frasers, hundreds can be observed moving through with Fork-tailed Swifts during passage.

“I have identified two races here; the nominate race brevirostris, which is very common on passage and in smaller numbers during the mid-winter period and the race rogersi (sometimes treated as a separate species called the Indochinese Swiftlet), which is mainly a passage migrant in small numbers.

“The nominate race appears slightly larger, is darker above with a narrow wide rump band, and a longer trail and wings. C.b.rogersi is slightly shorter winged and tailed, has a brownish wash to the upper plumage and the broader rump is buff. Both are prominently notched, in the tail.

“Most birders here live with an old-fashioned mind-set and still refuse to accept that this species can be readily identified in the field. That is why you will not find it on the NSS BG’s checklist. It has been on my personal bird checklist for nearly a decade now. Elsewhere in the world, birders have no problem identifying this species in the field. They work hard with better optics and field id guides, to resolve the field characteristics of similar species. I spent years doing the same here but the so called “powers to be” scoffed at anything that they themselves have not bothered learning to identify!

“I will continue operating outside the box and am glad that you have chosen to do so too.”

KC Tsang & R Subaraj
February 2008
(image by KC Tsang)

11 Responses

  1. Ben Li

    The two renowned ornithologist, LK Wang and Chris Hails has stated clearly in their book (annotated checklist of the birds of Singapore), that the Himalayan Swiftlet is still an unconfirmed status in Singapore. See page 171 under Appendix III – doubtful/Unconfirmed records. It it not in their main checklist. Maybe someone should write to them and ask for their comments on the photos by KC Tsang.

  2. Alan OwYong

    I can understand why Subaraj is frustrated over the ID of this swiftlet. After every Bird Race, the arbitrators threw out his team’s record for this swiftlet. The arbitrators were following the rules of the race. New species that are not in the NSS BG Check List will require strong evidence like photos and concise notes to be counted. ( Anyone can compile his own CL for their own reference or share with others. But there has to be an official CL administered by a relevant committee.)
    But why is this swiftlet not in the BG Checklist? The simple answer is that no one had submitted a record to the Record’s Committee for this swiftlet. The RC cannot just add in any bird species in the CL based on hearsay, conversations and discussions over cyberspace. A formal submission to the RC is still required if it wants to be considered for inclusion.
    KC Tsang has now submitted the first record ( date, time, location etc) supported by an excellent photograph.
    When the RC meets, his submission together with expert’s opinion will help the RC members to determined the ID.
    I understand that some overseas experts admitted that it is not easy to positively ID this swiftlet in the field.

  3. Ding Li

    While the field characteristics seem to point it to a himalayan swiftlet, caution is definitely needed. Basing a species ID on one or two feature alone is reckless. And it is more difficult in a photo than actual field observation where no comparisons could be made with surrounding swiftlets. Some of the references mention tail bifurcation of up to over 20%, but the lower range can and does overlap with some of the other swiftlets. It is highly possible and totally logical to infer the occurrence of the migratory swiftlets here, most agree that the gizz n other features appear to point to a himalayans, but as said field ID cannot be totally objective until a specimen is taken, measured and tarsus examined.

  4. See Toh

    Just like to reopen the debate. Sighted a couple of deep tail-notched swiftlets at Tanah Merah Grassland (the area behind the Newater Plant) on the afternoon of 11 Dec 2011. The swiftlets are bigger than the usual resident swiftlets (more the size of House Swift) and also darker. The images can be viewed at http://www.birdforum.net/showthread.php?t=218081. They were photographed (not by me) using Canon 300mm f2.8 lens. Unfortunately, images showing the underparts are not available.

    Cognizant to the concerns and views of the experts and experienced birdwatchers about the danger of confirming the occurence of a new specie just by photographic evidence, if the images are not Himalayan Swiftlet, what are the other possibiities? An overgrown Germain’s? Since according to Craig Robson, Black-nest Swiftlets do not display tail-notched.

  5. jytou

    The 1st and 2nd bird does not seemed to be the same swiftlet:

    1. 1st bird had rather dark brown rump with very small notch on a closed tail.

    2. 2nd bird had pretty pale whitish on rump and a stronger notch

    3. 3rd photo is a bit out of focus, may not be exactly helpful.

    The 2nd bird with such a pale, almost whitish rump, and a considerable tail notch, is potentially a Germain’s Swiftlet. It was found that the notch in Germain’s could be pretty much varying, and some are pretty deep as well, so based on tail notch itself, it doesn’t seem to be too reliable.

    The 1st bird is definitely more interesting, provided with a pretty dark brownish rump which was only very slightly paler, and therefore there should not be significant pale rumps that can be observed in the field, however the tail did not exhibit a deep tail notch that is consistently seen in Himalayan, although not exclusively.

    The swiftlet taxa in this region could still be in a serious mess. From known guides, we at least knew the Germain’s Swiftlet (previously part of Edible-nest) is the dominant species in the Malay Peninsular but only nest in sea caves on offshore islands, with no signs of breeding in any mainland caves. However, the recently swiftlet keeping industry had kept many inland in swiftlet hotels but the purity of their genetics are questionable, however generally resembling more of a Germain’s than any other species here, so usually conveniently treated as Germain’s Swiftlet (Aerodramus germani amechana). Those from north-west Malaysia northwards are said to be (Aerodramus g. germani).

    However, other than that, we also have the very similar looking Black-nest Swiftlet, said to be larger, with a weak notch (not something easy to judge in the field), the Malay Peninsular form is said to have a pale rump as well but narrower than the Germain’s, with lack of photos, I have no idea how it would really look like to have such a rump and how it would differ from Germain’s “broader” pale rump. Otherwise it is best recommend to ID them at their nest sites where their nests are black in color due to the feather contents. However, the dark-rumped “lowi” is recorded in Benom, Pahang at least (according to Robson’s).

    The species that caused dispute is the Himalayan Swiftlet, even with it being in the official checklist of West Malaysia, records from the wild are again barely accepted by the local record committees. These records comes from the mist netted birds during migration that seemed to be the best proof of their existence, or else seemed to be terribly difficult to tell, every now and then in the wintering months, I always notice some very well notched tail individuals in my garden at Ipoh but never yet dare to really place it under a confirmed record like some other birders as well because you failed to observe other features usually from these fast and rapid birds. Although said to have a generally browner rump, it is also still said to be pale, but browner than the Germain’s and consistently a strongly notched tail. However, a few subspecies could be occurring to this region every year and could definitely pose stronger difficulty as the different forms had different darkness of brown on the rump. Again they are poorly any photographic evidence in the Malaysian either to help people understand the distinction of the few confusing species.

    The next bird that had recently added to the confusion is a mysterious swiftlet collected from the limestone hills of Ipoh in the recent years, this swiftlet had no pale rump at all, which is a plain dark blackish brown all over the upperparts and generally duskier brown below, giving us a darkish swiftlet. According to measurements, it is believed to be the Mossy-nest Swiftlet that was mysteriously missing from the Malay Peninsular but found in the major Sundanese land mass that surrounds the Peninsular, showing potential of it being overlooked. With no DNA samples collection, it is still pending decision on what species it really was, but when you do meet such a swiftlet with no paleness on rump and dusky underparts, you are probably looking at this one.

    Other than these 4 species, there are some other older mysterious swiftlets collected or photographed, one of them would be a brown-rumped specimen from Tioman that clearly contrast with the regular grey-rumped Germain’s. It was believed to be an Edible-nest Swiftlet that was not found in the wild in West Malaysia but is breeding inland such as on Borneo and Java. If their ID were confirmed, they are either vagrants or released feral birds to increase breeding density by some farmers. Your brown-rumped swiftlet (1st photo) could had been this? perhaps reaching over from Sumatra?

    There were also many other odd rumps with mixture of grey and brown and etc in the domestic swiftlet hotel areas which could alert us on potential swiftlet hybridization or possible released of non native species. Those from the swiftlet hotels do show certain variations from the true wild birds and are often believed by some researchers that it may actually represent an all new domesticated species?

  6. See Toh

    Hi jytou

    According to the photographer, the first photo (with slight slight tail-notched) is taken about 1 minute earlier than the 2nd and 3rd photo (which is the same bird).

    As the photographer had shot the swiftlets in continous mode, I will ask him to show me the other photographs of that 1st bird regardless whether it is in focus or not.

    I knew it may be impossible to conclusively ID swiftlets in flight based soley on photographic evidence. But since many of my friends are keen amateur bird photographers with the latest state-of-the-art photographic equipment, why not get them to capture such seemingly different looking swiftlet images for constructive deliberation even though the discussion may still lead to nowhere.

    I must say that we should not just rely on one or two photographs but study a series of them since most photographers shoot bird in flight in continous mode though not all shots will be in focus.

    And thanks for the excellent write-up on the current state of swiftlet taxa in our region.


    • Tou Jing Yi

      hi See Toh,

      Definitely true, a series of photos are definitely a lot more helpful to study swiftlets, I did the same to a flock of swiftlets I encounter here in Ipoh, but no conclusive conclusion as evidently, seemed that lighting affects perception very strongly and my device was not even close to state of the art and could only give me bad images.

      Features that are helpful to ID swiftlets are usually not capable to be caught by cameras, for instance feather tufts on tarsus as well as the measurement differences, and a strong reason why swiftlets continues to bug us over the years. They are not the species that would perch and provide better images to study its ID.

      Since your friend can proof the 2 different looking birds come from the same bird in continuous shots, it can show us how serious lighting is affecting the effects on photography. I had personally seen potential Germain’s Swiftlets under very strong noon sunlight show almost white rumps under these conditions. It is best still leave swiftlets to the tentative ID category until we can gain any useful field characteristic that can be confidently used.

      Currently in the region, the Pale/Sand Martin pair is also underminable without measurement studies on hand, and Pintail/Swinhoe’s Snipe pair is also only positively identifiable by studying tail feathers (usually on hand), very rarely during preening of tail. The 3 Pond herons in wintering plumage are also non-distinguishable. Just some examples of the undeterminable groups in the region.

  7. See Toh

    Agreed that there is no way to ID these swiftlets besides having a bird in hand.

    Guess I could only state that based on the images captured (in particular the fairly deep tail-notched), the swiftlets displayed pro-Himalayan characteristics.

    By the way, I have attached 20 more images of the deep tail-notched swiftlets in my ID post at birdforum.net. These images were taken in better morning lighting and also they flew much lower. Image quality is thus much better this round.

    If any of the swiftlets are of interests to you, tell me which one and I will ask my photography friend to provide the entire image series he shot of that bird.

    • Tou Jing Yi

      agreed that quite some show very good notch which is pretty much likely outside the notch depth range of Germain’s.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.