Category: Science

The Asian Koel is today one of Singapore’s most numerous urban species, but did you know that it was previously only found as a migrant?

The Singapore Birds Project recently participated in a conference organised by the Asia Research Institute, National University of Singapore (ARI, NUS). Titled “Learning from Aliens: New Directions in Environmental Humanities Research and Practice”, speakers from numerous fields extending from science, history, to art and poetry shared their perspectives on “aliens” in biodiversity settings. During the conference, Keita and Movin shared about birds in Singapore and discussed the complexities of classifying “non-native” and “native” species in the highly modified local landscape. The Asian Koel’s history in Singapore is also covered. We’ve uploaded our talk on Youtube so do check it out if you were unable to attend!

We would like to extend a massive thanks to everyone from ARI and Yale-NUS who organised this event, all the speakers for the insightful perspectives, and most importantly Anthony for inviting us. We’re also looking forward to future collaborations 🙂

ARI is organising an upcoming conference on Singapore’s Biodiversity History that you can attend for free (either in person or via zoom) on 13 and 14 July. More information can be found here! The Singapore Birds Project will not be a part of this but established speakers from various fields will be there, so do check it out!

Another week, another rare bird at Chek Jawa! The Stripe-throated Bulbul Pycnonotus finlaysoni is another bird previously recorded in Singapore, just in April last year in fact. The details were first published online two months ago and last month the Records Committee came to the conclusion that this bird, recorded at Upper Seletar Reservoir, was more likely an escapee than a wild bird. With the recent sighting at Chek Jawa, some key discussion points have changed, so we reconsidered the previous record again alongside our review of the Chek Jawa bird. 

After extensive deliberation our committee reached split decisions on both records, and with no majority of votes in favor of treating either bird as being wild, the species has been placed in Category D. Category D species are not in the Singapore Bird Checklist; they would otherwise appear in Category A, but there is reasonable doubt that the records are of truly wild birds. (The exact vote counts for the Chek Jawa bird were 4 wild, 4 escapee; and for the Upper Seletar bird, 3 wild, 5 escapee.)

There are a considerable number of points to consider in favor of both sides, and we will summarize them below. Ultimately, there are no clear answers, but we aim to use the evidence available to weigh the likelihoods of each scenario. The fact that several committee members did not vote in the same way is a feature of our robust process, and signifies the difficulty in assessing these records. The end result or categorization of a species is also by no means set in stone; rather it is open to change as and when new information becomes available. 

Feather condition

There are no indications that the feather condition is poor or unnatural. However, bulbuls are small birds, kept extensively in captivity – it is not hard to find a bulbul in a pet shop with normal feathers. 

Status in the pet trade

The Stripe-throated Bulbul is a songbird kept extensively in captivity across Southeast Asia (Chng & Eaton, 2016; Limparungpatthanakij, 2022). It is often kept for cross-breeding with Red-whiskered Bulbuls, and records from Bangkok, which is in the middle of the species’s range, are treated as escapees due to their high prevalence in the pet trade. The increasing scarcity of Red-whiskered Bulbuls in the wild due to trapping is also likely contributing to higher trapping pressure on other bulbuls with pleasant songs, including Stripe-throated – leading to more possible sources of escapee Stripe-throated Bulbuls across the region (Limparungpatthanakij, 2022). YouTube searches for this species’s Thai and Malay names reveal hundreds of hits, several ostensibly advertising their video as a means to trap wild Stripe-throated Bulbuls. Several others feature birds in cages, many of which appear to have normal feather condition. While we did mention previously that you can “find and buy nearly anything in a cage these days”, the volume of trade in this species is likely to be extremely high in comparison to the birds we’ve been getting so far (SBP, 2023). What’s important here is that more than simply being ‘present’ in the regional pet trade – which by itself does not really indicate anything – it is quite a popular species as well. Its high presence in captivity elsewhere in the region may indicate popularity locally as well; the regional pet trade is closely intertwined. Anonymous sources have also informed us that this species is locally traded.

Habitat/Location

Chek Jawa is a prime location for visitors to turn up, and the location is an argument against the escapee hypothesis. But we do have surprisingly many records of introduced bulbuls on Pulau Ubin, including Black-crested and Red-whiskered. Either these birds are adept at dispersing from our main island, or they have been kept on Ubin at some point, or there is some kind of hidden influx of wild Black-crested and Red-whiskered Bulbuls from further north. The last possibility is near impossible to verify and best treated as unlikely given these species are also highly traded in captivity, and the first two would unfortunately throw this bird’s origins into question too. 

Having said that, the location is still a point in favor of the case that the bird is wild.

Status in southern Peninsular Malaysia

eBird range map for Stripe-throated Bulbul. Here is a link to the map if it does not load immediately. This simplified map does not do a good job of illustrating how rare this species in Johor (zoom in to view the the full eBird map with hotspot pins for a clearer impression). 

Many of the species which have recently been added to our checklist, or recorded as visitors (see Table 1 below) are indeed quite scarce in Johor. Examples include the Scarlet-breasted Flowerpecker, Black-and-white Bulbul, and White-crowned Hornbill. However, the Stripe-throated Bulbul is markedly rarer than all of them in Johor, even though it is known to tolerate more degraded habitats than those species. All time eBird records for Panti show well over 100 records of Black-and-white Bulbul and Scarlet-breasted Flowerpecker and at least 10 of White-crowned Hornbill, yet only 5 of Stripe-throated Bulbul. Similar results have been observed from Gunung Bekok and Gunung Belumut in Johor. Surveys there found this species to be the second-least frequently encountered bulbul among the 15 species recorded, second only to Scaly-breasted Bulbul which has no confirmed records in Singapore (Peh et al., 2005).

Meanwhile, the Stripe-throated Bulbul is actually quite common in all sorts of habitat from Malacca northwards, so the very abrupt drop in abundance in the south suggests its dispersal outside its core range is infrequent.  This is also in contrast to the other species listed above, which are scarce to rare throughout their range. It would be more straightforward to construct a case for the species’s dispersal if it was more common, or becoming more common, in Johor, where vast areas of suitable habitat are available. At the same time, the deforestation explanation, which we believe is pushing several of these species out of Johor’s forests, does not apply as well here; the species is simply very tolerant of low-quality forest habitats, to the point that it appears to be benefiting from forest degradation across its range (Nor Hashim, 2013; Wells, 2007).

We did note, however, a recent sighting from Gunung Pulai Recreational Forest in Johor, which may indicate the early signs of a southward influx of this species. The good news is that we will probably have good indications that such an influx is occurring, with more observers at birding sites in Johor such as Panti and Lenggor. Our RC decisions are never cast in stone and constantly move with evolving evidence; please post your records on eBird or let us know if you see this species in Johor!

(a) Wang & Hails, 2007; (b) Wells, 1999; (c) Wells, 2007; (d) eBird

* Species with status “Limbo” are provisionally not accepted as wild birds in Category A but are placed in Category D, recognizing the possibility that the bird originated from captivity.

Table 1 Records of Sundaic visitors treated as very rare in our checklist, as well as other possible Sundaic visitors not included in our checklist, Jan 2022–present.

(Cover photo: Stripe-throated Bulbul at Chek Jawa on 9 May 2023. Photo credit: Raymond Siew.)

References

Chng, S. C. L., and J. A. Eaton (2016). Snapshot of an on-going trade: an inventory of birds for sale in Chatuchak weekend market, Bangkok, Thailand. BirdingASIA 25, 24–29. Link

Limparungpatthanakij, W. L. (2022). Stripe-throated Bulbul (Pycnonotus finlaysoni), version 2.0. In Keeney, B. K. & Maleko, P. N. (Eds.), Birds of the World. Cornell Lab of Ornithology, Ithaca, NY, USA. 

Nor Hashim, N. E. (2013). Comparative study of understorey birds inhabiting selected logged and virgin lowland forests. University of Malaya. Link

Peh, K. S.-H., Jong, J. de, Sodhi, N. S., Lim, S. L.-H., & Yap, C. A.-M. (2005). Lowland rainforest avifauna and human disturbance: persistence of primary forest birds in selectively logged forests and mixed-rural habitats of southern Peninsular Malaysia. Biological Conservation, 123(4), 489–505.

Singapore Birds Project. (2023). Singapore’s first wild White-crowned Hornbill. Link

Wang, L. K. & Hails, C. J. (2007). An Annotated Checklist of the Birds of Singapore. The Raffles Bulletin of Zoology, 15(Suppl), 1–179. Link

Wells, D. R. (1999). The Birds of the Thai-Malay Peninsula (Vol. 1). Academic Press, London.

Wells, D. R. (2007). The Birds of the Thai-Malay Peninsula (Vol. 2). Christopher Helm, London.

In November 2022, word of a Black-headed Ibis at Sungei Buloh Wetland Reserve was spreading through birding circles like wildfire. Over six months later, we have another species previously recorded as an escapee, this time seen at Chek Jawa – the White-crowned Hornbill Berenicornis comatus. Kiri Zhang’s quick alert meant many were able to rush down and see the bird the same day, and despite some disappointing days when it wasn’t sighted, the bird seems to be still wandering around Chek Jawa today. 

Our Records Committee’s deliberations led us to the unanimous conclusion that this was likely a wild bird, although this was by no means a straightforward decision.

The identification was never in doubt. The entirely white tail and white crown rule out all other hornbill species in the region. The bigger question then – is it more likely to be a wild bird, a genuine visitor from Malaysia, or an escapee from someone’s collection? The RC’s discussions centred around a few issues which are summarized below.

Where does this species occur usually?

This is a typical Sundaic species, ranging from the forests of Kaeng Krachan in Phetchaburi, Thailand, to Johor. It also occurs in Sumatra and Borneo although we could not find any records from the Riau Archipelago in Indonesia. David Wells does not list this species from Johor in his 1999 book, although there are several recent observations at both Panti and Lenggor – the linked checklists are just a few of several on eBird.

eBird range map for White-crowned Hornbill. Here is a link to the map if it does not load immediately.

How regularly does this bird occur in the pet trade?

Hornbills appear in the wildlife trade mainly for their casques but are occasionally kept as pets as well. The White-crowned Hornbill is not well-represented in TRAFFIC surveys. It’s highly possible, though, that the species occurs infrequently in the regional pet trade. But its mere presence in the pet trade does not provide a strong case for why this particular bird must be escaped – one search of Shopee in Indonesia reveals that you can find and buy nearly anything in a cage these days. Our committee always aims to objectively weigh the relative likelihoods when dealing with these records, and while frequent occurrence in the pet trade is a red flag, there is no evidence that this species is recorded particularly often in captivity.

So, does the location matter?

Chek Jawa is currently seeing its second influx of Sundaic visitors in two years. Last year, we had a Large Woodshrike, Ruby-cheeked Sunbirds, Scarlet-breasted Flowerpecker, and best of all, two Black-and-white Bulbuls visiting the same trees frequented by the hornbill. The location is developing a reputation for attracting these wandering birds, and looking at a map it’s not hard to see why; it could be deforestation in Johor, just natural dispersal, or a combination of the two, driving them southwards in search of new habitats, and Chek Jawa would be one of the first places they land on.

The location also means that a bird which escaped from captivity on the main island would need to cover some distance and make a sea crossing in order to reach Chek Jawa in the first place. This strengthens the argument for this bird being a wild bird. 

What about the bird’s physical condition?

Nothing about the bird’s appearance suggests it has origins in captivity. The bird is not ringed, its bare parts show no signs of unnatural wear, and the feathers are in near-perfect condition; the tail feathers are perhaps very slightly tattered, but this is quite normal for any wild bird. Like all the other points, though, this in itself does not really prove anything, as a well-kept captive bird could theoretically be in good condition too. So this fact again merely lends a bit more credence to the argument that this bird was always wild.

What about all the other past hornbill records which have been treated as escapees?

This is really a very good question, and one we don’t know the answer to. Before this White-crowned Hornbill, Singapore has actually had several records of hornbills which have been classified as escapees.

The repeated Rhinoceros Hornbill sightings between 2004 and 2009 were accompanied by a Great Hornbill, and both birds spent several years in the Eng Neo area. Black Hornbills – at one point up to six birds, including immatures – were recorded between 1986 and 2003, mostly near the Upper Seletar/Nee Soon area. And of course, the White-crowned Hornbill was itself recorded previously, at the Singapore Botanic Gardens in 1987. 

In fact, several RC members have expressed doubts over the rather liberal application of the escapee tag for many historical records – and it’s not just restricted to hornbills. A few examples of past records the RC has previously re-evaluated and finally added into the checklist include the Velvet-fronted Nuthatch, Crimson-winged Woodpecker, Orange-breasted Green Pigeon, and even Pied Kingfisher. These were generally contemporaneously treated by past records-keepers as being escapees, although they share similarities with many of the birds we consider today to be genuinely wild dispersing birds. 

Last year, as a committee, we considered the Great Hornbill record from the Bukit Timah area in 2016 and 2017 and voted by a 6-2 margin to treat the record as an escapee instead of a wild bird. Part of that could be due to the species’s ecology and perhaps higher prevalence in the pet trade – see the record page for more information – but it also may be due to some degree of wariness to apply our Chek Jawa-tinted lenses of today to the past, where details of several records are rather sparse or provide few indications of why birds were considered escapees in the first place. If there were good reasons, we may indeed never know. And so while we have reviewed some past records, we felt those cases were more clear-cut inclusions in the checklist with especially strong claims for wild provenance. 

With this, the White-crowned Hornbill is now the 429th species on the Singapore Bird List. If you haven’t gotten your tick yet, maybe this is the motivation you need to go down and see this new visitor to our shores!

White-crowned Hornbill at Chek Jawa. Photo credit: Adrian Silas Tay

White-crowned Hornbill at Chek Jawa. Photo credit: Goh Cheng Teng

White-crowned Hornbill at Chek Jawa. Photo credit: Raghav Narayanswamy

White-crowned Hornbill at Chek Jawa. Photo credit: Jayden Kang

White-crowned Hornbill at Chek Jawa. Photo credit: Lim Qian Xun

References

Chan, Y. M., Chan, M., & Wee, Y. C. (2008). Aberrant behaviour of a female Great Hornbill and a female Rhinoceros Hornbill. Nature in Singapore, 1, 31–34. Link

Wells, D. (1999). Birds of Thai-Malay Peninsula. Academic Press.

On 1 November 2022, excitement spread through our local birding community when Art Toh shared his sighting from Sungei Buloh Wetland Reserve of a white ibis donning a black head. Birders headed down to get a glimpse of the bird which thankfully stayed for several days, allowing many to observe it. The Singapore Bird Records Committee took some time to evaluate the record and we assessed that this individual is likely to be Singapore’s first documented individual of a wild Black-headed Ibis Threskiornis aethiopicus.

How was the bird identified?

There are several species similar to the Black-headed Ibis such as the African Sacred Ibis T. aethiopicus, Australian Ibis T. molucca and Malagasy Sacred Ibis T. bernieri. These four species were previously all thought to be one species, then called the Sacred Ibis, until works including the evaluation of their plumage differences established that they were distinct from one another (Lowe & Richards, 1991).

The range restricted Malagasy Sacred Ibis can be differentiated from the other three by its white iris among other features. However, the Australian Ibis and African Sacred Ibis look quite similar to the Black-headed Ibis. Differences between them include their bill size, which can sometimes be difficult to assess in the field. How should we identify the Sungei Buloh bird then?

The magic is in the small details: one of the key features to look out for is the primary pattern – the black markings are restricted to tips of the three outermost primaries in Black-headed Ibis, unlike that in the African and Australian Ibis that are more extensive. Additionally, the tertial feathers in Black-headed Ibis is grey, as opposed to black in African and Australian Ibis.

Here are some images showing these features on the African Sacred Ibis (extensive primary markings: example 1, example 2 and example 3; black tertials: example 1 and example 2) and Australian Ibis (extensive primary markings example 1 and example 2; black tertials: example 1 and example 2).

Comparing the “Sungei Buloh Ibis” with photographs of known Black-headed Ibises shows that the features fit this species perfectly instead.

Is the Sungei Buloh Black-headed Ibis a wild bird or an escapee?

Now that we’ve identified the bird, then comes the question of its provenance. Interestingly, the Sungei Buloh bird is actually not the first Black-headed Ibis seen in Singapore! This species used to be sighted locally back in the ~80s but were then thought to be escapees from the Jurong Bird Park (Wang & Hails, 2007). Today, the Mandai Wildlife Reserve makes an effort to ring all birds, and there are no free-flying Black-headed Ibises as far as we understand. Sungei Buloh’s Black-headed Ibis had no rings on its leg, making it unlikely that it originated from the Jurong Bird Park.

Where could the bird have come from, then? The Black-headed Ibis has a patchy range and inhabits wetlands from South Asia all the way to Indonesia. In Thailand, they are resident in the southern parts and disperse in the winter months (Treesucon & Limparungpatthanakij). Slightly closer in Peninsula Malaysia, there might have been wild breeding populations in the past, but if any were extirpated nearly a century ago (Wells, 1999; Puan et al., 2020). There are now only very occasional sightings along the landmass (such as this record). Meanwhile in Sumatra and Java, the populations used to be very robust but have declined dramatically in the past few decades (Iqbal & Hasudungan, 2012). The birds there are, however, thankfully still present (Muhammad Iqbal, personal communications).

It is still difficult for us to establish whether the bird came from the north or from the Indonesian population. Several waterbirds like the Pied Stilt and Javan Plover have made their way from the Indonesian Archipelago. At the same time, there was, interestingly, yet another sighting of a Black-headed Ibis near Malacca just 3.5 weeks after the Sungei Buloh bird. This sighting bolsters our decision that the Sungei Buloh bird was likely from a wild origin, and might provide hints to where it could have arrived from.

Wherever the bird flew over from, we hope that it’ll successfully find some friends and return to a colony to breed safely. And who knows, we might get more of them visiting us in the future!

Acknowledgements

We would like to express our gratitude to Art for sharing the sighting very promptly, and also thank Adrian, Bao Shen and Cheng Teng for photo contributions.

References

Lowe, K. W., & Richards, G. C. (1991). Morphological Variation in the Sacred Ibis Threskironis aethiopicus Superspecies Complex. Emu, 91(1), 41-45.

Puan, C. L., Davison, G., & Lim, K. C. (2020). Birds of Malaysia: Covering Peninsular Malaysia, Malaysian Borneo and Singapore. Lynx Edicions, Barcelona.

Treesucon, U., & Limparungpatthanakij, W. (2018). Birds of Thailand. Lynx Edicions, Barcelona.

Wang, L. K., & Hails, C. J. (2007). An annotated checklist of the birds of Singapore. Raffles Bulletin of Zoology Supplement, 15, 1–179. Link

By Laura Berman

A recent publication in the Journal of Ornithology has some interesting new data on the breeding seasons of Singapore’s birds, and if you are an avid contributor to eBird, your observations might be a part of it.

Singapore has an incredibly dedicated birding community, and many birders share their observations of nests, baby birds, and fledglings. All those observations were taken together with half a decade of mist-netting data collected by the Avian Evolution Lab at NUS and the National Parks Board to outline the breeding season in Singapore.

If you look hard enough you can find birds building nests in Singapore just about any month of the year, but you’re much more likely to succeed if you go looking in March – that’s when most insectivores like Malaysian Pied Fantails and frugivores like Coppersmith Barbets and Black-naped Orioles are constructing nests. Granivores like Baya Weavers nest a bit later in the year, the best chance to catch them constructing their charismatic hanging nests is in June. You’re most likely to come across active nests with unhatched eggs if you go looking in April or May. June to July is the best time to see parents feeding their fledglings, and very young birds still chasing their parents begging for food. By the time August comes around the breeding season is over, but if your field ID is particularly good you might be able to recognise juvenile birds still in their first set of flight feathers.

It’s always exciting to see nests and baby birds but remember not to get too close. Its stressful enough being a parent.

Another example of a very widespread frugivorous species, the Black-naped Oriole. They are very common in Singapore, but finding their nest takes a bit of effort! Photo credit: Goh Cheng Teng

Granivorous species like the Baya Weaver , known for their intricate nests, tend to breed later in the year. Photo credit: Keita Sin, taken in Malaysia

Insectivorous birds like the Malaysian Pied Fantail that likes to nest near waterbodies often nest around March. Photo credit: Keita Sin

Frugivorous birds like the Coppersmith Barbet tend to nest around March too. Photo credit: Goh Cheng Teng

References

Berman, L., Li, D., Yang, S., Kennewell, M., Rheindt, F. (2022) Breeding season linked to sunshine hours in a marginally seasonal equatorial climate. Journal of Ornithology. https://doi.org/10.1007/s10336-022-02009-9. Link

And so it’s been a year since we launched the Singapore Bird Database. Time really flies. This time a year ago, the Fairy Pitta at Hindhede Nature Park that we queued for was already gone, and many of us were busy chasing Singapore’s first proper documentation of the Grey-streaked Flycatcher.

Nostalgia feels good, but for now, let me share some updates on our database.

Our database was launched to provide a platform where anybody can easily find relevant information regarding rare bird species in Singapore. Since then, we’ve received over 25,000 views on our database website, and we’re glad that many of you have found our work useful. For those who are keen to find out how our platform came to life, you can check out our blog post or detailed write-up.

Information available in our database has almost doubled over the past year, and it now contains over 2000 records from more than 180 species. While many were compiled from social media archives, publications, and past eBird records, nearly 300 were submitted by none other than all of you – birders in the field – to which we would like to express our huge thanks to. 300 submissions, only for rare birds! All of these submitted records were evaluated by our Singapore Bird Records Committee team, whose latest works are constantly updated in our Recent RC Decisions page. We’ve been very busy – a happy problem because it means that our community has been finding a lot of good birds! 

Moving on to how the Singapore Bird Records Committee has evolved since then, first, we have developed an operational guideline that can be viewed in a Q&A format here. Second, we are renaming the term “Rarity” to “Review Species” in our checklist. It might sound like a matter of semantics, so allow us to explain. Our Records Committee makes a concerted effort to properly evaluate and document records of birds that one might not necessarily think is that rare. Say, Pied Harrier. This is an annually recorded raptor, but in numbers small enough that warrants celebration. This season, we seem to be witnessing a strange spike in their numbers for reasons unknown, and patterns like this will be missed if not for proper efforts. Another good example is the form of Barn Swallows with rufous underparts. we occasionally observe At present, we don’t even know which subspecies actually arrive here, but what we do know is that they are rare in Singapore, and that records keeping would be essential for future generations to have a point of reference when the ornithological community one day have a better understanding of these birds. In principle our workflow stays the same, and so we’d greatly appreciate it if you could submit your encounters. This terminology change will be reflected in our Species Pages as well in due time.

Our database has changed in more ways than one, and here is a comparison of how it looked 1 year ago versus today. Besides new features and useful tools, the layout now emphasises images and is more colourful and easy to use. An archived version of our website from Nov 2021 is available here.

Last, our database works have extended beyond collating records, and we’ve designed useful tools for everybody’s birding activity. Our Monthly Roundups provide accurate information on interesting species documented each month, and our bar charts can help you plan your birding itinerary to best target the species that you really want to see. If you’re hungry to find the next Mega in Singapore, you can also refer to our On This Day page to find gems from the past.

Moving forward, we aim to revisit all of our Species Pages for further improvements, and also plan to expand our database and website to include more species. If you have ideas for collaboration or are keen to help out with our work, don’t feel shy to get in touch with us!

I’d wrap up this update with the most important bit, the acknowledgements.

See all those Monthly Roundups and Bar Charts? They didn’t just fall out of the sky (though the birds did) – they were possible thanks to the tech-wizardry of Raghav, to whom we extend our utmost gratitude.

See those crazy numbers of historical records newly compiled in our database? They were collated and scrutinised by our amazing Singapore Bird Project teammates Clarice, Geraldine, Jenwei, Jing Ying, Hong Yao, Hui Zhen, Raghav, and Zachary. Massive, massive thank you for the hard work – checking through all those records is really not easy.

And lastly, to all the birders for your all-rounded support and for sharing and submitting your sightings – the Singapore Bird Records Committee thanks all of you.

See you all in the field and happy birding!

PS, IN CASE YOU MISSED IT:

We were invited by the Lee Kong Chian Natural History Museum to give a talk about our database in March 2022. Check it out too if you couldn’t join us!

In a recent post, we covered the why of our migrant bar charts – what motivated us to design a new way to present migration data? In this post, we will try to explain how our methodology works, and look at some examples of the algorithm in action.

Recently, a large volume of work in the field of ecology has been concentrated on generating robust avian population estimates, with a variety of different methods. One feature common to a number of the techniques has been a focus on applying citizen science data, especially from the eBird platform, as in Hansen et al. (2022). 

Over the last few months we have designed a straightforward technique to model migration patterns and trends in Singapore. Rather than making population-wide range and abundance estimates which require more sophisticated methods, like in eBird Status and Trends (Fink et al., 2021), we instead focus only on analyzing occurrences within Singapore, which are fine-grained and relatively well-represented in citizen science datasets such as eBird. 

The effect of confounding variables such as uneven effort distribution is relatively low due to Singapore’s small size and high observation density, so our approach only incorporates recorded observations and makes no attempts to estimate unknown quantities using extrapolation. 

We consolidate observation data using spatial and time-based constraints. Important terminology that we use to describe the process are defined:

• Sighting: one eBird observation of a species, extracted from the eBird dataset; the important information which is used is: species name, number of individuals, date, and location.
• Record: one or more eBird observations of a species where the observations in question all relate to the same individual(s) – one record may comprise sightings of different numbers of individuals
• Individual: a single bird of one species, may be seen multiple times or just once, and alone or together with other individuals

What this all means is that we need a method to map sightings to individuals. For instance, consider the following sightings of Yellow-rumped Flycatchers in late Sep 2021.

Simply choosing the high count of 4 is not a good idea, since there are clearly more than 4 unique birds in the table here. At the same time, adding them up won’t work because the 26 Sep sighting and the 30 Sep sighting at Tuas could possibly involve the same birds. 

So our approach is summarized as follows:

1. For each species, we deliberate and determine a reasonable “date threshold” which determines the maximum date difference between two sightings for them to be grouped together as one record. For raptors, the threshold is normally a shorter period because most sightings are of birds just passing through. Winter visitors will have longer thresholds.
2. For any species, our algorithm will “lump” sightings across multiple dates into a single record if the gap between them is less than or equal to the decided threshold and the sightings are less than 2km apart.
3. We repeat this process for every sighting until they are all categorized. Each record will have one or more sightings. 
4. To estimate the number of individuals observed during the timeframe, we take the highest count of individuals sighted in each record and add them all together. 

The below example is a color-coded example showing how each sighting is grouped according to location and date. The total number of individuals outputted by the algorithm, for the period 25 Sep to 30 Sep, would be 12, which is the sum of 1, 2, 3, 2, and 4 (highest count for each respective record, listed in bold in the table below).

After dividing the year into week-long chunks and taking the average over ten years, the result is quite smooth and matches our prior understanding quite well. In the case of Yellow-rumped Flycatcher, we can see the two clear peaks in fall and spring.

How our approach compares with other publicly-available statistics

The other main source of bar charts is on eBird. Because eBird is built for users around the world, it cannot take advantage of the unique features of data in Singapore: namely the high density of observations here compared to other places. We wrote about how our method differs from statistics available on eBird in our earlier post, which has a full explanation.

Limitations

This method of course only yields estimates, as some – not many – sightings which are from the same location are listed at different eBird hotspots which are more than 2km apart. Also, sometimes birds may hang around longer or shorter than the threshold decided for the species. The other problems that come from working with citizen science data are also reflected, like in the case of misidentification, incorrect location data, incorrect date information, and so on. Generally, the impact of these issues is not large and tends to cancel out, but are still problems to be aware of. 

The data is also not intended to be a population estimate but rather an estimate of the observed number of individuals. The difference is that we do not infer records where data is not available, which would allow for a complete population estimate. 

The algorithm is also optimized for Singapore’s uniquely high number of active observers, coupled with the tightly-packed hotspots here. In larger countries, the distance between observations would make it difficult to apply our approach. 

Further possible improvements and applications of the model

In future, we hope to further improve the model by considering low counts which may be the result of undersampling rather than actual low abundance. Particularly for pelagic species, higher perceived abundance may be due to a lower number of pelagic trips during certain times of the year.

We may also attempt to extend the model to display spatial abundance in addition to temporal abundance, to highlight the best areas for different species.

Conclusion

We wrote this post to highlight the key features of the algorithm which serves as the foundation for our newly-released migrant bar charts. Besides these charts, it also forms a basis for our internal analyses and discussions, which eventually lead to work such as our Records Committee’s rarities list.

We’ll continue to document any further updates to our methodology in a timely manner, and as always, constructive criticism is very valuable and will help us improve going forward. Please feel free to contact us!

References

Hansen, B. D., Rogers, D. I., Watkins, D., Weller, D. R., Clemens, R. S., Newman, M., Woehler, E. J., Mundkur, T., & Fuller, R. A. (2022). Generating population estimates for migratory shorebird species in the world’s largest flyway. Ibis, 164(3), 735–749. https://doi.org/10.1111/ibi.13042 

​​Fink, D., T. Auer, A. Johnston, M. Strimas-Mackey, O. Robinson, S. Ligocki, W. Hochachka, L. Jaromczyk, C. Wood, I. Davies, M. Iliff, L. Seitz. 2021. eBird Status and Trends, Data Version: 2020; Released: 2021. Cornell Lab of Ornithology, Ithaca, New York. https://doi.org/10.2173/ebirdst.2020

I’m sure many of us have used eBird bar charts in the past, and they’ve certainly helped me in deciding where to go birding or what to look for. So why did we make our own migrant bar charts here at SBP?

In this article, we’ll explore some of the drawbacks we noticed with eBird bar charts, and how our migrant bar charts stack up against eBird’s. We’ve also got a future article lined up in which we go more in-depth into the inner workings of the algorithm, so do stay tuned if you’re into that sort of thing! (Update: this post is now posted on our blog)

Because eBird is built for users around the world, it cannot take advantage of the unique features of data in Singapore: namely the high density of observations here compared to other places. 

eBird mainly uses two alternative methods to try to quantify abundance/frequency for its bar charts.

Method 1

This method uses the percent of checklists in which a species appears. This does not really tell us anything, because birders go to different places during different times of the year. Let’s say a certain wader is more common in November than in September. But because many people are at Henderson Waves instead of Sungei Buloh in November, the percent of checklists which contain that wader would obviously be much lower than in September. That would not reflect the species’s true status. Another problem with this method is that a checklist with 100 individuals of a certain species counts the same as a checklist with just 1 individual. And to make matters worse, an “incomplete” checklist – one where the observer says that not all birds that were seen are listed on the checklist – is not counted at all.

The difference between September and November is stark, despite just the two-month difference. It highlights the shift in observer coverage from shorebird areas in Sep (mainly Sungei Buloh) towards other habitats in Nov.

Method 2

The second indicator used by eBird is the highest count of a bird for a week in history. This also poses major problems. Consider the example of the Japanese Sparrowhawk Accipiter gularis. The highest count was on 13 Oct 2020, with 104 birds, and from the eBird chart, you would imagine Japanese Sparrowhawk is most common in the first half of October. But the highest count for a certain week is not the best way to gauge abundance. To take an extreme example, if we simply take the highest count, 1000 birds split into groups of 100 each would appear like a lower level of an abundance than 104 birds seen once. And indeed, this seems to be the problem with Japanese Sparrowhawks, which are far more common in end-Oct despite the high count telling us otherwise.

Our approach

From the above chart taken from our bar charts, you can see the peak for Japanese Sparrowhawk lies closer to the end of Oct, with rising numbers through October and falling numbers in November.

We’ve used eBird bar charts in the past too. We noticed these issues, and set out to design a more robust algorithm which uses all available data rather than only considering high counts or frequencies. The algorithm uses eBird’s dataset of observations too, but processes the sightings differently.

These new bar charts are the product of years of hard work from the Singapore birding community. Your observations have formed the basis for the raw data that we now have access to in eBird. Now, with your help, we have found a way of presenting information that is easily understood and has proper scientific grounding.

In the coming weeks, I’ll be writing about how our algorithm consolidates many thousands of individual eBird data points to estimate the abundance of each species over time – we’re always making an effort to ensure our approaches are firmly rooted in science, and striving to be transparent about it too. (Update: this second post is now posted on our blog as well)

Regardless of where you come from, you’re missing out on the Singaporean experience if have you never queued long long for food before. And it seems that our locally found parrots know that too! Here are some intriguing observations on the feeding behaviour of three parrot species – Tanimbar Corella, Coconut Lorikeet and Blue-crowned Hanging Parrot.

By Muhammad Nasry

Earlier this year, an alert that a stretch of mango trees around the Alexandra Village area was fruiting and attracting Coconut Lorikeet Trichoglossus haematodus began circulating online, featuring eye-level shots of this colourful parrot happily munching away at the fruit. Naturally, this attracted a whole horde of birders, myself included. Having lived near the area since birth but blithely unaware of the existence of these mango trees, I was a little embarrassed for not realising sooner, but excited by the discovery of a new birding spot just minutes from my doorstep.

And so I visited. Again. And again. And again.

Over the course of several visits I began to notice an odd pattern in terms of the foraging behaviour of some of the birds there. Consistently, this feeding pattern repeated itself over and over.

The first to arrive was usually the Tanimbar Corellas Cacatua goffiniana. These parrots seemed to prefer feeding on the outermost bits of the mango fruit, only occasionally reaching deeper into the mango for more of the pulp.

Second on the scene was typically the star attraction: the Coconut Lorikeets. They would pick up where the corellas had left off, peeling off what remained of the fruit skin. They would then proceed to rip apart the rest of the fruit (as parrots do) while eating, with the bits of fruit either being scattered on the surrounding foliage or dropping straight to the ground.

The lorikeets were often seen foraging in pairs, with the occasional dispute breaking out between pairs over mangoes. There were up to three pairs seen at the same area at any one time. In one instance, one of the pair was feeding while the other stood guard, screaming its head off to ward off any would-be usurpers of the coveted mangoes. 

Only after the lorikeets left would the Blue-crowned Hanging Parrots Loriculus galgulus move in on the fruits. A male and female were both seen feeding here, but never at the same time.

This feeding order was generally consistent – each parrot species usually came to feed only after the previous one had left. Competition due to size differences may explain this pecking order (pardon the pun), given that the corellas are the largest, and hanging parrots the smallest. The issue with this theory is that there were plenty of mangoes to go around, many of them a similar shade of green as the ones being eaten (which presumably indicates they were equally edible) and therefore competition was not necessary. However, the hanging parrots seemed to pass up these seemingly fine mangoes in favour of those that had already been opened by the two introduced parrot species. Perhaps letting the bigger parrots do the hard work of opening up the unripe mangoes may be more convenient for them – they may have to use a lot more effort with their comparatively smaller beaks to open up the mangoes on their own. 

The messy eating habits of the parrots did not go unnoticed by other animals; there were intermittent visits to the mangoes (or to the bits of mango strewn around on the branches) by other common urban species, which included:

• Olive-backed Sunbird Cinnyris jugularis
• Brown-throated Sunbird Anthreptes malacensis
• Yellow-vented Bulbul Pycnonotus goiavier
• Black-naped Oriole Oriolus chinensis
• Plantain Squirrel Callosciurus notatus
• Chocolate Pansy Junonia iphita

The following species were also seen in the vicinity, but had not been observed feeding on the mango fruits, be it directly or indirectly.

• Scarlet-backed Flowerpecker Dicaeum cruentatum
• Javan Myna Acridotheres javanicus
• Common Myna Acridotheres tristis
• Common Hill Myna Gracula religiosa
• Painted Jezebel Delias hyperete
• Grass Yellow Eurema sp.

It is interesting that the feeding behaviour of the introduced parrot species benefited other wildlife species (birds or otherwise). It may be possible that many more species may benefit from such interactions in other habitat types as well. Ground-dwelling fruit-eaters like the Asian Emerald Dove or Lesser Mousedeer may benefit from the bits of fruit scattered around the forest floor thanks to the parrots’ poor table manners. Of course, these interactions took place in a relatively urban habitat next to a busy road – generalisations to less developed habitats would not be wise without further study.

References

Nasry, M. (2022) Biodiversity Record: Observations of some urban birds feeding on mango fruit. Nature in Singapore, 15, e2022095. URL