What is a species? (2 Viewers)

[Jon.Bryant]

PSC, by definition, does not recognize subspecies.

I think this is the extreme end of PSC. As I stated earlier in the thread Birdlife suggested that a PSC list could have circa 16k species - there are circa 11k species and 19k subspecies, so 16k is well off considering all subspecies as species. The Dutch authority only looks at the Western Palearctic species (including vagrants such as Thayer’s Gull), which makes things a bit simpler. Last time I looked at their WP checklist they separated all obviously differentiated subspecies - so all the Yellow and White Wagtails, but they didn’t split subtle subspecies (say British Blue Tit) or subspecies mainly distinguishable on biometrics (say British Lesser Spotted Woodpecker).

 

[Mysticete]

I am skeptical that the Thayer's Gull situation can be resolved without detailed study of nesting birds in the Arctic. For example, maybe they start forming pair bonds in the south, but do these persist? or do they break up during migration or on nesting grounds. To verify that, you would probably need to radiotag the birds and see what those pairs are up to in the north. That again is going to be beyond the ability of the average birder. This doesn't even get at the issue that hybrids are probably diagnosed, since you could have a hybrid that mostly almost entirely looks like only one species, even if it has genes of both. So you really need to measure the gene flow going on.

Those are just two issues. Citizen science I think can help with some species. For instance, birders could provide valuable information on what is going on in hybrids zones of Bell's Vireo, which occurs in the lower 48 and has two vocally and morphological distinct populations. Defining the extent of mixed pairs in the hybrid zone or response to playback would be quite useful. Their is straight forward data here to collect. The Thayer's Gull situation is IMHO completely different.

 

[Mysticete]

I think this is the extreme end of PSC. As I stated earlier in the thread Birdlife suggested that a PSC list could have circa 16k species - there are circa 11k species and 19k subspecies, so 16k is well off considering all subspecies as species. The Dutch authority only looks at the Western Palearctic species (including vagrants such as Thayer’s Gull), which makes things a bit simpler. Last time I looked at their WP checklist they separated all obviously differentiated subspecies - so all the Yellow and White Wagtails, but they didn’t split subtle subspecies (say British Blue Tit) or subspecies mainly distinguishable on biometrics (say British Lesser Spotted Woodpecker).

This assumes that all subspecies are valid. A LOT of subspecies probably represent nondiagnostic populations or segments of a continuous cline. Subspecies just largely get ignored unless people think a certain subspecies should be treated as a species. I suspect that might factor into the Birdlife quote.

 

[Jon.Bryant]

I am skeptical that "citizen science" can help with the Thayer's Gull situation

The DB article stated that pair bonds are actually formed in late winter prior to spring migration, so mixed pairing can be reasonably determined in early spring, in the wintering grounds.

The article also stated that typical types have different vocals, which again could be established on the wintering grounds. It would be very interesting to determine the range of vocals of intermediates. Are vocals in gulls learnt or inherited? Assuming the former, vocals may be able to determine origin of intermediate bird - hybrids with west coast calls, or vagrants of the more variable Kumilein’s type.

 

[Jon.Bryant]

or segments of a continuous cline.

I was thinking about clines the other day. Does it really show that a subspecies is dubious, or might it just be a current snapshot of the slow advance of an advantageous form, with a wide band of hybridisation.

Some east-west clines do not seem to be able to be readily mapped against a similar variation in climate, or other environmental factors.

 

[Jon.Bryant]

since you could have a hybrid that mostly almost entirely looks like only one species, even if it has genes of both. So you really need to measure the gene flow going on.

This makes things even more complex and perhaps a little bias. As an example, we know that large Gull hybridise quite often. We have recorded that Marsh and Blyth’s Reed Warbler hybridise, but I think the warbler hybridisation has only been recorded through the encounter of mixed pairs. I think the field identification of hybrids post fledging would be extremely hard. So for one group we see the need to fully test that typical birds are not hybrids, and for the other we are satisfied that two species are involved without looking at gene flow and establishing that hybridisation may be much more widespread that we think - particularly as Blyth’s Reed has been expanding west.

It was interesting to read in DB that there is marked gene flow between Pallid and Common Swift, although prior to this revelation I had never heard of, or even thought about the possibility of hybridisation. I recall (I might be wrong) that typical birds in one species were pure, where as typical birds in the other were often not, suggesting that hybrids tended to look similar to one parent rather than intermediate. Field detection of a fast moving aerial hybrid, that tends to look like one parent, sounds highly unlikely, and we therefore assume everything we see is pure.

I suppose going back to the gulls, what do our studies of pair bonding in other species show? Birds of the Western Palearctic states for Herring Gull ‘Monogamy the rule, though polygamy recorded in 0.07% of 1690 territories in two colonies… Monogamous pair-bond considered life long, though divorce frequent [no data on frequency or timing provided]…. Established pairs often consort during winter, particularly if little post-breeding dispersal, and associate increasingly as the breeding season approaches…’.

I do find it odd that the majority of birds on one coast or the other are one type. I spent a week in Vancouver a long time ago and spent quite a bit of time studying and photographing the Thayer’s Gulls, but as far as I am aware I have yet to encounter a Kumilien’s Gull. It seems a tad odd, if from winter distribution we are suggesting that parent A migrates back and forth to the west coast and partner B to the east coast, but in the arctic they come together and hybridise, perhaps first having to shake off a same type winter pairing… and that obvious hybrid offspring seem to rarely follow dad or mum all the way back to the west coast. Feels like we are putting poor Thayer’s Gull through a higher lever of scrutiny and statistical rigour. If there was ever a need for a mathematical model, it is for this species.

 

[Kirk Roth]

Really? I am only thinking of supporting information, but if we look at Thayer’s Gull as an example

• Do we witness pair bonds being formed on the wintering grounds, as has been suggested?
• If so, what percentage of pair bonds observed during late winter/spring involve birds of mixed characters?
• What percentage of typical Thayer’s type birds do we encounter on the east coast, or of Kumlien’s types on the west coast during winter months?
• Are vocalisations of Thayer’s and Kumlien’s distinct as suggested?
• If long calls are distinct do playback experiments to the other form, illicit any form of response?
• Are there any behavioural differences, or when seen together, does one form dominate the other - say during aggressive displays over food resource?
• Does the age mix of migratory flocks, indicate migration routes are learnt or inherited? In particular do ‘fresh’ mixed character first-winter birds typically associate on migration with mixed character adults, or also with ‘pure’ adult types?
• What is the true variation in plumage pattern where one type or the other is the norm?
• What is the winter distribution and population estimates for mixed character birds - I think in a previous thread someone said mixed character birds are the norm in the Great Lakes area, whereas it seems from literature that mixed character birds are rare at least on the west coast.

More straightforward I have seen ‘decisions’ for other ‘failed’ species that say, lack of complete study of vocalisations, lack of knowledge of field characteristics in potential area of overlap, not demonstrated that morphological differences are consistent etc. Similarly, some reports on ‘passed’ species have stated that subtle morphological differences exist (which I then can’t find referenced to in any ID guides, presumably indicating these differences were previously undocumented/unknown). Surely these types of issues are all things we could help to answer, when we are in the right place and have been told what is wanted?

If Killian Millarny can solve the ID of Saunder’s Tern, starting from an online image search, then I am sure a good collation of relevant photos and sound recordings should be of value to researchers working on specific taxonomic issues.

I remember a long time ago corresponding with Xeno Canto about ‘projects’ to target and collate bird vocalisation info - I wasn’t very articulate with my idea, and they got a bit excited about something rather different instead. I wanted a system to mobilise amateur recordists to undertake specific recordings - my interest was particularly in the ‘rumour’ that Pale-legged and Sakhalin Leaf Warbler contact calls were at different frequencies. I really think that if they had set up a system where researcher could ask birdwatchers to target certain species in certain areas, issues like those two leaf warbler vocals would have been proven much earlier and much more easily.

Surely a system that could mobilise birdwatchers to try to seek answers to specific questions could only aid both taxonomy and identification research.

To suggest that taxonomy is so elevated, that ordinary birdwatchers cannot even assist seems perhaps a little pompous and narrow minded - I am only suggesting we could act as extra eyes are ears to undertake field observations after all.

I suspect that it would only work for papers that had failed due to lack of sufficient evidence, as I doubt researchers would initially want to divulge their research direction to others, and before they had submitted a paper.

I believe I misunderstood your initial aim and perspective here, so I can apologize for that. Of course citizen science can be used for supporting information - my point might more clearly be said that for the most part in ornithological taxonomy, citizen science is limited to that supporting role.

I'm not saying that it cannot be used to assist, and not saying that citizen science doesn't uncover useful, interesting, or inquiry-sparking information. But any amount of photography or observation-based information needs to reckon with genetic data, and while this wasn't always the case in this discipline, genetics are currently our bread and butter. Behavioral and distributional data (no matter whether from "mere citizens" or from an academic institution) are supplemental - as I think you are also saying.

 

[Kirk Roth]

I do find it odd that the majority of birds on one coast or the other are one type. I spent a week in Vancouver a long time ago and spent quite a bit of time studying and photographing the Thayer’s Gulls, but as far as I am aware I have yet to encounter a Kumilien’s Gull. It seems a tad odd, if from winter distribution we are suggesting that parent A migrates back and forth to the west coast and partner B to the east coast, but in the arctic they come together and hybridise, perhaps first having to shake off a same type winter pairing… and that obvious hybrid offspring seem to rarely follow dad or mum all the way back to the west coast. Feels like we are putting poor Thayer’s Gull through a higher lever of scrutiny and statistical rigour. If there was ever a need for a mathematical model, it is for this species.

I don't quite follow this part - I've not heard it suggested that Iceland Gulls (sensu lato) disperse widely east and west from their breeding grounds to their wintering grounds. My impression (which is somewhat anecdotal and certainly based heavily on "citizen science!") is that there is a swath of these gulls which progress from paler to darker as one goes from Western Europe to Western America, and that the pattern bears itself out in both the Arctic summers and temperate winters. These observations and your own would support the suggestion that north-south movement is much more prominent than the east-west during migration.

 

[D Halas]

I do find it odd that the majority of birds on one coast or the other are one type. I spent a week in Vancouver a long time ago and spent quite a bit of time studying and photographing the Thayer’s Gulls, but as far as I am aware I have yet to encounter a Kumilien’s Gull. It seems a tad odd, if from winter distribution we are suggesting that parent A migrates back and forth to the west coast and partner B to the east coast, but in the arctic they come together and hybridise, perhaps first having to shake off a same type winter pairing… and that obvious hybrid offspring seem to rarely follow dad or mum all the way back to the west coast. Feels like we are putting poor Thayer’s Gull through a higher lever of scrutiny and statistical rigour. If there was ever a need for a mathematical model, it is for this species.

It's not really odd at all, since it's entirely consistent with an east-to-west cline existing in the breeding grounds. Anyone who's spent time on the Great Lakes in winter trying to determine if a gull is a very dark Kumlien's or a lightly-marked Thayer's would, I guess, be much more likely to support the lumping of the species than anyone who's only birded on one of the coasts.

 

[Andrew Spencer]

The article also stated that typical types have different vocals, which again could be established on the wintering grounds. It would be very interesting to determine the range of vocals of intermediates. Are vocals in gulls learnt or inherited? Assuming the former, vocals may be able to determine origin of intermediate bird - hybrids with west coast calls, or vagrants of the more variable Kumilein’s type.

I personally am very skeptical about the supposed vocal differences. A number of the supposed "Thayer's Gull" recordings available online are, IMO, mis-identified. Of the ones I am certain about the ID of, the long calls are identical to those from Kumlien's Gull from the eastern US. Obviously a much larger, and more rigorously sourced pool of recordings will be necessary to actually get to the bottom of any supposed vocal differences between the forms. But until then I think caution should be used.

 

[Mysticete]

Yeah I am not a gull expert AT ALL, but I have never really heard voice being used to ID similar gull species, nor that voice was all that important within gulls anyway. So I was surprised to see this being used as evidence.

 

[Kirk Roth]

I personally am very skeptical about the supposed vocal differences. A number of the supposed "Thayer's Gull" recordings available online are, IMO, mis-identified. Of the ones I am certain about the ID of, the long calls are identical to those from Kumlien's Gull from the eastern US. Obviously a much larger, and more rigorously sourced pool of recordings will be necessary to actually get to the bottom of any supposed vocal differences between the forms. But until then I think caution should be used.

I recognize that I'm beating this same drum again, but its worth noting that all of Western, Glaucous-winged, Glaucous and American Herring Gulls have diagnosable differences in long call sound and posture, yet these are components of well-established hybrid swarms in various areas of the western Nearctic.

To top it off, the long calls seem to be given in more situations than just territorial situation (e.g. Bald Eagle disturbance). Perhaps these are agonistic calls, not used the same way as a songbird for example? Gull vocalizations with regard to pairing is still something of a mystery if you ask me - its as if they may pair independently of the Long Calls but prior to using any Courtship Begging calls. All of this said from someone who is not a gull researcher so take what I say for whatever little it is worth to you.

Large white-headed gulls are among the most recently diverged of bird species - they may be a better example of the messiness of the "What is a species" question rather than methods to resolve it. At the very least I think there are major questions to be answered regarding what these vocalizations do for gull species before they can inform our diagnosis of them.

 

[Jon.Bryant]

It's not really odd at all, since it's entirely consistent with an east-to-west cline existing in the breeding grounds. Anyone who's spent time on the Great Lakes in winter trying to determine if a gull is a very dark Kumlien's or a lightly-marked Thayer's would, I guess, be much more likely to support the lumping

I have not visited the Great Lakes in winter, so am not aware of the situation there. However, I don't really think that the situation in winter is really as straightforward as attributing the spread of birds to an east west cline. To quote Birds of the World.

• glaucoides - some Greenland birds (nominate) disperse to NE North America S to New England, but most move to Iceland, where very common in winter, and the Faeroes, and small numbers continue to NW Europe, especially the British Isles and Netherlands.
• kumlieni - migration of eastern Canadian populations unclear, presumed to be primarily at sea or coastal, often southward to northeastern North America. Unknown proportion presumed to overwinter in Arctic polynya. A probable kumlieni banded at Digges Sound recovered on Placentia Bay, Newfoundland, 3 January 1983. Occasionally to northern Atlantic, including Iceland, Faroe Islands, or perhaps northern Europe.
• thayeri - Migration patterns of western Nunavut and Northwest Territories populations of thayeri unclear. Very few banding records (see Nature of migration in the species, above). A very small portion of western populations may migrate inland.... Some thayeri reach the North Slope and Pacific coast in Alaska, others (mainly juveniles) moving into Chukchi Sea, then moving south.... Common along coastal British Columbia and Oregon; first arrivals September and October, peak January and February.... November–April south to coastal California.... Rare inland transient Ontario. Rare to uncommon on Great Lakes mid October to mid April west to Thunder Bay area, rare migratory transient on Ontario coast of Hudson Bay.

So the majority of Iceland Gulls winter in coastal maritime habitats, but they seem to have very divergent migration routes dependent on wing tip colour - the vast majority of dark wing tipped birds move south and west, whereas birds with white primaries move south and east, and more intermediate birds probably follow the latter course, but staying much further north. It is as if there is a dividing line somewhere in the cline - a switch is flicked and migration proceeds the other way.

The situation in the Great Lakes may well be a melting pot of mixed charcater birds, but this wintering site is rather atypical for a bird which generally winters in coastal maritime locations, and presumably is only a small proportion of the total population. The DB article stated that S.Howell et al. assessed the west coast population of 'thayeri' types, as including very low number of atypical birds (from recollection less than 1%). Definitely all the birds I saw many years ago in Vancouver all looked like typical 'thayeri' types to me.

Sorry got the following wrong, so I have edited out the stuff on Iceland Gull being treated as monotypic!

If we assume the cline is through hybridisation, If birds inherit an innate sense of migratory direction from there parents - an urge to move in a certain direction - then perhaps the south east urge predominates in hybrids, explaining why the paler birds are rare in the west and the greatest variation (kumilien types) occurs in the east. If migration routes are learnt by migrating with their parents, then the winter distribution would seem to me to point more towards a smaller hybridization zone, with hybrid birds migrating to the Great lakes and perhaps north east Canada, but also large 'pure populations' from the apparent extremes of the 'cline', consistently following the same route back and forth to preferring wintering areas in Iceland and the Pacific.

The situation would seem to be very similar to Lesser Black-backed Gull and Vega Gull (IOC classification, treated as Herring by Clements and American Herring by BirdLife). Lesser Black-backed Gull subspecies 'taimyrensis' is no longer recognised, and I understand is considered a hybrid form between Lesser Black-backed Gull and Vega Gull, being somewhat intermediate in mantle and upperwing colour. Vega Gull winters on the pacific coast, as do the 'taimyrensis' (they actually seem to winter a little further south - in Hong Kong 'taimyrensis' is common and Vega much less so). Contrary to this, Lesser Black-backed Gulls (subspecies heuglini in the contact zone) winters in the Arabian Sea. But in this case the the very different wintering grounds are accounted for by the fact that we think we have two species.

On the later point, it dawned on me that the paper presented early - in which a system was described to compare known sympatric species against similar allopatric species in order to determine 'assessment' criteria for the latter - could also be used when we have sympatric species, but the hybridisation zone is unstudied (aka Iceland Gull). In this case we could assess Thayer's and Iceland types as if they were potentially allopatric species (no hybridization), by assessing them against other criteria, we have already established to be of use in separation of other sympatric Larus species. My money is that Thayer's and Iceland types, stand out like a saw thumb and are being assessed very differently from all other sympatric gull species. My understanding is that the treatment of the species is because the conservative option was seen to fall back on historic precedence (which was presumable established on even less data than even the relatively limited amount we have now!).

 

[jurek]

I am skeptical that the Thayer's Gull situation can be resolved without detailed study of nesting birds in the Arctic.

Actually, level of hybridization would be easy to answer using DNA studies. It is possible that moulted gull primaries collected around Arctic settlements would be sufficient, but I am not an expert in DNA extraction methods. It would not answer why Thayer's Gulls hybridize or not, whether from pairing or different wintering grounds or whatever, but would answer if and how much.

 

[Jon.Bryant]

have never really heard voice being used to ID

Caspian Gull was only recently detected as visiting the U.K, with the first records in the 1990s. I am not suggesting that it was discovered due to voice, but it does have a distinctive voice and ‘albatross’ display. The Latin name (including the original subspecies name) is cachinans, meaning laughing, so although it is quite similar to Herring Gull, we have known it is vocally distinct for some time.

When I put my recordings of Yellow-legged Gull from Greece into Merlin (without any location data), it correctly identifies the species. Actually Yellow-legged Gull is quite complex - a paper in DB explained that different populations around the Mediterranean have slight plumage, size and vocal distinctions, with some nominate birds approaching the subspecies from the Canaries and Azores. Regardless I don’t think that any of these call variations approach Herring Gull vocals. I think the vocals of gulls is probably understudied as to most of us they sound pretty similar - I think the differences are most apparent in sonograms, when you can look at the spacing of the long call notes, the number of harmonics and the pitch. It would be interesting to know if pitch differences are to do with structure or acquired. For instance Yellow-Legged Gull has a slightly different head structure to Herring and Caspian is different again - I am not suggesting this will effect the spacing of notes, but perhaps have an impact on pitch/nasal qualities.

 

[Jon.Bryant]

moulted gull primaries collected around Arctic settlements would be sufficient

My understanding is that moult cycles in gulls are rather protracted and some feathers are shed almost year round - so feathers could also be collected at the wintering grounds. Also I think DNA can be obtained from poop, so another easy thing to collect in winter. I think a putative American Herring Gull in the U.K. was proven to be just a Herring from DNA collected in one or other way. A putative Slender-billed Curlew was definitely proven to be an Eurasian Curlew from a poo sample and several Stejneger’s Stonechats have been proven or disproven in the U.K. by the same method.

I have joked in Birdwatch magazine that perhaps we all need to carry a test tube and have guidance on how to collect, store and send off a sample for processing. But perhaps it is not a joke - what is the process for collecting a sample, how does it need to be stored, where do you send it for analysis and how much does it cost?

Also satellite tracking technologies are becoming much cheaper - and devices can be easily carried by a large bird such as Larus gull. So tagging a few wintering birds, would help to confirm the breeding grounds of the different forms and the probable areas of overlap. Typical wintering grounds of kumilien’s could also be explored by satellite tagging winter birds during ‘irruption’ years, when birds are more widespread south and east.

I suppose thought that study on the breeding grounds would ultimately provide much more data - but as stated before, I think our downgrading of the bird, makes interest and funding of such a trip unlikely.

 

[jurek]

we all need to carry a test tube and have guidance on how to collect, store and send off a sample for processing. But perhaps it is not a joke - what is the process for collecting a sample, how does it need to be stored, where do you send it for analysis and how much does it cost?

I would have a clean test tube, fill it with alcohol and store in a fridge freezer. But I am sure there are people on this forum who know better. I have no idea how expensive it is - somehow, for a major rarity, birdwatchers afiliated with museums or research institutes shoved the samples into larger science projects.

Interesting, and worrying, is how often a DNA test revealed that the bird was of different species or subspecies than observers believed. Atlas Flycatcher in Britain turned to be European Pied and so on. For me it is that over-splitting and over-keen ornithologists claim fake small identification characters, which are within the range of individual variation.

 

[Jon.Bryant]

For me it is that over-splitting and over-keen ornithologists claim fake small identification characters, which are within the range of individual variation.

I suppose this goes back to my original comment as to whether a species is a human construct, necessary for classification and cataloging.

If so, and we are not able to consistently place taxa in the right box, then it could be argued that the system has gone too far and is too granular.

But if species can be properly defined and may or may not be cryptic, then correct allocation is purely a human problem. The library system is right, we just struggle to use it!

I suppose that a big issue here is that birds obviously think at a different pace and in different ways to humans - what looks or sounds the same to us doesn’t to birds. If not, how does a Sandwich Tern chick detect and respond to the call of a returning parent. Cryptic to us may be very distinct to birds! An example would be another example from DB, where a ringer used a caged Meally Redpoll to catch more Meally Redpolls, but no Lesser Redpolls, although both forms were present. He didn’t screen the caged bird from view, so it was not determined if call or appearance was the a powerful lure to one form only.

 

[Jon.Bryant]

Atlas Flycatcher in Britain

I recall this was a ‘putative’ Atlas Fly, reported at Flamborough Head, not a definite claim, although it did attract some interest from twitchers for ‘insurance purposes’. A possible Med Fly likewise also attracted some interest, but was a much shorter stayer, so not so many twitchers finally lost their bottle, and dashed to see it just in case.

However, it could be argued that we have quite a lot of Pied and Spotted Flycatchers annually in Britain that don’t cause headaches. If say 1 in 1000 or 1 in 10,000 birds are atypical and approach the appearance of a sister species, does this mean the classification is flawed? This would not seem right, particularly if the appearance of Atlas of Med Fly (or any other species) is consistently different (or virtually so) in their own ranges. We have lost a few species recently as aberrant individuals, so we know aberrant individuals occur - does a Pied Fly with an aberrant amount of white on the forehead, consign an apparent similar looking species to the bin.

I think as humans (and particularly as men) we have rather an obsession to categorise and count things. If taxonomy and nature is messy, it may spoil our fun, but I am not sure our obsession should be the main focus for resolving the messiness. This is a bit like the earlier Thayer’s Gull comment, where it was stated that anyone who had struggled to tell a dark Kumilien’s from a Thayer’s would argue against the two being considered species.

I also think that for some birders, trying to untangle the messiness, and solve the ultimate ID challenge is a thrill. So by simplifying taxonomy, we are appeasing those who want to keep things simple, and disappointing those who enjoy the challenge.

Another article in DB looked at a new approach to ID, where characters were weighted, scored and combined. This type of system has already been used for birds such a Zino’s Petrel, ‘Atlantic’ and ‘American’ Herring Gull, ‘Russian’ Common Gull etc. where only birds showing all classic features can be assigned and many birds may well be unidentifiable - not so great for many of us who obsess over the need to know!

 

[MJB]

My money is that Thayer's and Iceland types, stand out like a saw thumb and are being assessed very differently from all other sympatric gull species.

Your autocorrect's made a tiny oops... but does suggest that something is finger-lickin' good...!
MJB