Species limits and taxonomy in birds (1 Viewer)

[jts1882]

The DOI is working now for me. The pdf is behind a paywall.

 

[Björn Bergenholtz]

You can try to "Request full-text" from Ignacio De la Riva (on of the Authors) himself, on ResearchGate, here.

 

[jurek]

Joseph A Tobias, Paul F Donald, Rob W Martin, Stuart H M Butchart, Nigel J Collar, Performance of a points-based scoring system for assessing species limits in birds, Ornithology, 2021;, ukab016, https://doi.org/10.1093/ornithology/ukab016

Abstract:

Species are fundamental to biology, conservation, and environmental legislation; yet, there is often disagreement on how and where species limits should be drawn. Even sophisticated molecular methods have limitations, particularly in the context of geographically isolated lineages or inadequate sampling of loci. With extinction rates rising, methods are needed to assess species limits rapidly but robustly. Tobias et al. devised a points-based system to compare phenotypic divergence between taxa against the level of divergence in sympatric species, establishing a threshold to guide taxonomic assessments at a global scale. The method has received a mixed reception. To evaluate its performance, we identified 397 novel taxonomic splits from 328 parent taxa made by application of the criteria (in 2014‒2016) and searched for subsequent publications investigating the same taxa with molecular and/or phenotypic data. Only 71 (18%) novel splits from 60 parent taxa have since been investigated by independent studies, suggesting that publication of splits underpinned by the criteria in 2014–2016 accelerated taxonomic decisions by at least 33 years. In the evaluated cases, independent analyses explicitly or implicitly supported species status in 62 (87.3%) of 71 splits, with the level of support increasing to 97.2% when excluding subsequent studies limited only to molecular data, and reaching 100% when the points-based criteria were applied using recommended sample sizes. Despite the fact that the training set used to calibrate the criteria was heavily weighted toward passerines, splits of passerines and non-passerines received equally strong support from independent research. We conclude that the method provides a useful tool for quantifying phenotypic divergence and fast-tracking robust taxonomic decisions at a global scale.

How this reasoning treats the fact that not splitting, as a negative result, is likely to be underpublished, and that there is a general trend to oversplit birds as it costs nothing (it is an opinion) but makes the article more read?

 

[Mysticete]

How this reasoning treats the fact that not splitting, as a negative result, is likely to be underpublished, and that there is a general trend to oversplit birds as it costs nothing (it is an opinion) but makes the article more read?

is it underpublished? At least with genetics, you often still can say something interesting about phylogeography even if a study doesn't support splits

 

[THE_FERN]

That is why genetic distance is measured by percentage of difference, not single genetic characters (which can be as misleading as phenological characters).

Most people now view this as wrong. It's essentially similar to the genome-wide DNA: DNA hybridisation approach. An obvious problem is where you get saturation or strong convergence (long branches attract problem where distantly related things appear closer). Strong stabilising selection can make for little divergence, even among ancient splits. Even if this were not true, not all genes are equal and most DNA is probably junk making simple distance misleading.

To paraphrase the saying about democracy: genetic difference is not perfect, but still more fair than other methods.

Essentially the point is there's no way that's perfectly fair or "objective". If the aim is to recover phylogenies there are better methods than distance. If the aim is to delineate taxa you have to apply arbitrary (=subjective) thresholds. What is a species? In some sense whatever you like

While culture is interesting, it is not usually relevant to birds.

Simply not true (sorry). Obvious example is song dialects. Famous example = tits pecking milk bottle caps, or herons using lures like bread to catch prey.

 

[THE_FERN]

is it underpublished? At least with genetics, you often still can say something interesting about phylogeography even if a study doesn't support splits

I think one problem is it's often unclear what they're saying (don't make specific taxonomic proposals). But yes there's still a strong bias towards suggesting splits. Another issue is that different papers apply different speciation models (PSD Vs biological etc). The better ones say "if it's PSD it means this, if..." Dunno how Tobias et al deal with this (but if course they've no point to prove!). As for advancing studies by 33 years or whatever, suggest various single publications by Ridgeley, Collar etc have stronger claims!

 

[Peter Kovalik]

Carla Cicero, Nicholas A Mason, Rosa Alicia Jiménez, Daniel R Wait, Cynthia Y Wang-Claypool, Rauri C K Bowie, Integrative taxonomy and geographic sampling underlie successful species delimitation, Ornithology, 2021;, ukab009, https://doi.org/10.1093/ornithology/ukab009

Abstract:

Species delimitation requires a broad assessment of population-level variation using multiple lines of evidence, a process known as integrative taxonomy. More specifically, studies of species limits must address underlying questions of what limits the distribution of populations, how traits vary in association with different environments, and whether the observed trait differences may lead to speciation through reproductive isolation. While genomic data have revolutionized the process of delimiting species, such data should be analyzed along with phenotypic, behavioral, and ecological traits that shape individuals across geographic and environmental space. The integration of multiple traits promotes taxonomic stability and should be a major guiding principle for species delimitation. Equally important, however, is thorough geographic sampling to adequately represent population-level variation—both in allopatry and across putative contact zones. We discuss the importance of both of these factors in the context of species concepts and traits and present different examples from birds that illustrate criteria for species delimitation. In addition, we review a decade of proposals for species-level taxonomic revisions considered by the American Ornithological Society’s North American Classification Committee, and summarize the basis for decisions on whether to split or lump species. Finally, we present recommendations and discuss challenges (specifically permits, time, and funding) for species delimitation studies. This is an exciting time to be studying species delimitation in birds: many species-level questions remain, and methodological advances along with increased access to data enable new approaches to studying age-old problems in avian taxonomy.

 

[Peter Kovalik]

Frank E Rheindt, Elize Y X Ng, Avian taxonomy in turmoil: The 7-point rule is poorly reproducible and may overlook substantial cryptic diversity, Ornithology, 2021;, ukab010, https://doi.org/10.1093/ornithology/ukab010

Abstract:

The ornithological world has 4 global checklists (as of early 2020). While 3 follow the results of peer-reviewed research at varying pace and conservatism, the HBW/BirdLife checklist, which is adopted by the global Red List authority, has implemented Tobias et al.’s (2010) 7-point scoring system to overhaul global ornithological treatment. Critically received in some academic quarters, this scoring system is lauded by other ornithologists for its simplicity and reproducibility, a claim that remains to be tested. We subjected 26 ornithologists to a set of 48 bird skins belonging to 20 controversial taxonomic complexes and observed a wide variance in scoring results, in most cases straddling anywhere from far below to above the species threshold of the 7-point rule and casting doubt on claims of high reproducibility. For a detailed assessment of genuine taxonomic discord, we compared the taxonomic coverage of the avifauna of the Indonesian Archipelago (comprising ~1,400 species) between the HBW/BirdLife checklist, other major authorities, and the peer-reviewed literature. We detected that controversial treatments supported by the 7-point rule but at odds with the peer-reviewed literature predominantly refer to lumps, not splits, which are the usual subject of modern taxonomic quarrels. Notably, the method tends to unite morphologically (and sometimes vocally) cryptic forms into single larger species because of its inability to accommodate molecular and massive bioacoustic datasets that would indicate otherwise. On the other hand, the 7-point rule has produced numerous novel proposals for splits that may or may not be corroborated by future peer-reviewed inquiry. We recommend the 7-point rule as one of the multiple unofficial exploratory tools to flag cases of potentially cryptic species requiring further inquiry, but we advise against its adoption by other taxonomic authorities and the ornithological community.

 

[jurek]

Most people now view this as wrong. It's essentially similar to the genome-wide DNA: DNA hybridisation approach. An obvious problem is where you get saturation or strong convergence (long branches attract problem where distantly related things appear closer).Strong stabilising selection can make for little divergence, even among ancient splits. Even if this were not true, not all genes are equal and most DNA is probably junk making simple distance misleading.

Traditional character-based taxonomy suffers from the same problem but to much larger extent. Poor sampling or scoring of morphological characters will give errors. Strong stabilizing selection will produce visually similar but long split species. This particular case is not an error at all, in my opinion - if two things do not differ, there is no reason to protect them separately. Not all morphological characters are equal, either. Striking visual differences in coloration can be irrelevant to taxonomy - see mantle color of gulls or any of color morphs of other birds.

By the way, non-coding or so-called 'junk DNA' is currently believed to be important, as there hide regulatory sequences regulating switching or and off nearby genes. Much of variation of living systems is generated not by new protein-coding genes, but by new switching on/off of them.

Simply not true (sorry). Obvious example is song dialects. Famous example = tits pecking milk bottle caps, or herons using lures like bread to catch prey.

Because culture is not genetic, there are two reasons not to protect it. First, it is ephemeral. Bird song dialects are studied since several decades at most, there is no idea whether they persist longer. Culture of tits pecking milk bottle caps is already extinct, together with the extinction of British human culture of leaving bottles at doorsteps. Was it an important extinction by any objective merit?

Second, because culture is not genetic, it is easily reconstituted. Perhaps a bird 'culture' with most relevance to bird conservation is fidelity of seabirds to specific breeding sites, which can persist for millenia, despite the seabirds being perfectly physically able to visit alternative breeding sites. However, it is proven that human can reconstitute an extinct 'culture' by transporting chicks. A first successful example I know of was returning atlantic puffins to Maine, close to a century ago.

Essentially the point is there's no way that's perfectly fair or "objective".

This is the most important piece. If you claim that there is no objective taxonomy, it is not hard science but opinion or social science, and conservation funds should not be wasted on it.

I am interested in data and knowledge management, so I see bird taxonomy in a bit different light.

While I consider bird taxonomy to be hard science, it is a very difficult one. Objective facts in bird taxonomy are not easily refuted practically, therefore errors can persist for decades unchecked. If you doubt it, consider bird species which turned to be invalid (e.g. Vaurie's Nightjar). Bird taxonomy is such a weak discipline, that nothing in it prevented such a major error from being discovered and refuted for decades. Therefore bird taxonomy needs especially strong methods to separate facts from fiction, and is constantly in danger of being flooded by false claims and pure unscientific opinion.

The article is a perfect example of confusing facts, claims and opinions. On top, it is based on a circular reasoning: we changed the definition [of species], so our results are better by out own definition. There are further sub-errors it it.

Most important practical purpose of bird taxonomy, and perhaps the only objective way how possible errors could be practically verified, is biodiversity conservation. Humanity wants to preserve biodiversity, and needs a list of taxa to design conservation actions. And here weakness of Tobias criteria, and much of bird taxonomy, is shown: they offer no way to check whether their method results in better protection of biodiversity. Few papers which deals with bird taxonomy vs bird conservation generally show that modern bird taxonomy creates problems for conservation by changing what should be conserved. Other papers generally offer a circular reasoning [our definition of delimiting species is better, because we better protect species by our definition].

 

[THE_FERN]

By the way, non-coding or so-called 'junk DNA' is currently believed to be important,

Some "junk" sites perform a role. But by no means all. There's no clear way in which the Protopterus aethiopicus (lungfish) genome "needs" 44x as much DNA as ours. Much of the difference is going to be due to repetitive "junk" DNA which plays no coding, regulatory or structural function.

Because culture is not genetic, there are two reasons not to protect it.

This is certainly not true for human cultures... ...Arguably not true for any others.

Second, because culture is not genetic, it is easily reconstituted.

Certainly not true and quite offensive when applied to human culture (I'm unclear to what extent human cultures are more distinct or valuable than non-human). Our knowledge of non-human cultures is such that it's quite difficult to make any overarching statement as to their distinctiveness or prevalence.

The statement is clearly not necessarily true for non-humans. We can see this because, for example, it can be extraordinarily difficult to get captive-reared individuals to adapt to life in the wild.

This is the most important piece. If you claim that there is no objective taxonomy, it is not hard science but opinion or social science, and conservation funds should not be wasted on it.

That is a very strange view which doesn't apply to other areas of human activity. For example, most would suggest it's valuable to spend money promoting and conserving "culture" or "art" even though this is impossible to define objectively. We operate a legal system even though its basis is subjective and subject to change (is slavery right or wrong? Both have been true under the law at different times). Those unfamiliar with physics and the philosophy of science expect science to be different and "objective". This is a category error (as is any distinction between hard and soft science [Heisenberg anyone, or Gödel? What about the critical p value in significance testing?])

I think an important point that genetic sequence analysis makes clear is that even if you're able to reconstruct relatedness correctly, it's essential to name the frame of reference. The tree and hence taxonomy may well be different for different genes. Whether it makes sense to take some kind of whole-organismal "average" depends on the circumstances. [So a mitochondrial tree which is at odds with a nuclear tree is not wrong, just different: it tells you different things]

Taxonomies are categorical impositions on continuously varying entities. Because of this they will always be subjective.

 

[jurek]

That is a very strange view which doesn't apply to other areas of human activity. For example, most would suggest it's valuable to spend money promoting and conserving "culture" or "art" even though this is impossible to define objectively.

This is human cultural artifacts which has little to do with discussion about bird conservation. Nobody seems to be alarmed that Bluetit habit of pecking milk bottles appeared in the 1921 and died out later because milk bottles with tinfoil tops are no longer left in front of British houses. Or that Bewick Swans, Red-breasted Geese and multiple other birds changed their wintering quarters and migration routes within the last 100 years.

(By the way, British culture from the 1920s - clothing styles, habits etc, does not exist outside museums either).

By the way, an interesting study could be made by comparing old available bird recordings and today calls. I guess, many song variants would no longer be recognizable. Like human language becomes unrecognizable after low tens of generations.

Taxonomies are categorical impositions on continuously varying entities. Because of this they will always be subjective.

This only applies if one views taxonomy as categories. Which is not necessary. I joke that obsession with splitting species accelerated when ornithologists learned Microsoft Excel, which lists things only in discrete rows. So two birds must be listed either in the same row or two different rows. Modern software no longer forces people to use discrete tables.

Bird conservation can stop using categories and stop being subjective, and actually not care about taxonomic changes. By applying genetic uniqueness one can set priorities numerically and be more objective, and independent on opinionated and agenda-promoting individual scientists. One could perfectly say (numbers are examples): spending money on average unique species saves 2% of genetic diversity, spending money on conserving a localpopulation of Chiffchaffs saves 0.001% genetic diversity, and spending money on a truly unique species like Kakapo saves 10%. In this scenario, it does not matter whether a population with 1.7% genetic difference is a distinct subspecies or weak species - its importance stays the same.

 

[THE_FERN]

This is human cultural artifacts which has little to do with discussion about bird conservation. Nobody seems to be alarmed...

Not just cultural artefacts (things) but also languages, philosophies and religions. The fact that nobody is alarmed doesn't mean that these things are definitely not important. To a first approximation, nobody cares about biodiversity preservation (I do!); that does not mean it's not important... (Actually people are concerned about changing patterns in overwintering waterfowl, but mostly because it's an indicator of climate change)

This only applies if one views taxonomy as categories. Which is not necessary...

By definition it is... Perhaps we mean quantifying difference which varies continually (like % distance under some metric). That isn't the same as imposing a taxonomy. There are practical difficulties in such approaches, and various different ways of doing it (e.g. ultrametric trees, how you calculate the distance metric, whether you use some sensible model to compensate for multiple hits and what that is etc)

Bird conservation can stop using categories and stop being subjective, and actually not care about taxonomic changes. By applying genetic uniqueness one can set priorities numerically...

Are we advocating cash in proportion to genetic distance? This completely ignores ecosystem role (e.g. predator-prey dynamics and keystone species), for example.

At the end of the day, careful consideration demonstrates there's no single simple, objective way of assigning funding. This is true even if calculating the metric didn't suffer from subjectivity (it does) and you were only deciding how to assign funds in line with it. You'll note that good phylogenetic analyses employ multiple methods and report variant result in each for these reasons.

 

[jurek]

Are we advocating cash in proportion to genetic distance? This completely ignores ecosystem role (e.g. predator-prey dynamics and keystone species), for example.

I am not saying the system is perfect, only that it is incomparably better than the current one. Counting of species using Tobias or whatever other criteria also ignores their ecosystem role.

 

[THE_FERN]

I am not saying the system is perfect, only that it is incomparably better than the current one. Counting of species using Tobias or whatever other criteria also ignores their ecosystem role.

Well that's certainly one view. Doubtless he would disagree. Suggest the important one is mine: that there's no perfect approach. I think it's valuable to understand the subjectivity involved in taxonomy. Gives the lie to listing.

 

[njlarsen]

Well that's certainly one view. Doubtless he would disagree. Suggest the important one is mine: that there's no perfect approach. I think it's valuable to understand the subjectivity involved in taxonomy. Gives the lie to listing.

Listing is a subjective item in its own right -- only one person decides what that person has seen/heard/identified well enough to put on that list. That does not mean that it cannot be a fun exercise to do
Niels

 

[THE_FERN]

Listing is a subjective item in its own right -- only one person decides what that person has seen/heard/identified well enough to put on that list. That does not mean that it cannot be a fun exercise to do
Niels

Yes well I do it... But I object a bit to those who ignore the elephant standing next to the cisticola entirely...

...Especially since many of them don't understand the subjectivity involved (at least in my experience)

 

[jurek]

I think it's valuable to understand the subjectivity involved in taxonomy.

The only conclusion is that taxonomy which cannot eliminate its subjective part is not a hard science. Therefore this kind of 'taxonomy' should not be admitted in scientific journals, and should be used as little as possible in conservation decisions.

It is, however, perfect for birdwatching magazines, which contain the hard science part (like identification criteria and records or rare birds) and the soft part (like personal stories of birdwatching experiences and lists producing constantly new armchair ticks for the readers).

 

[Mysticete]

The Fern is absolutely correct. I don't know why people are arguing about taxonomy not having a degree of subjectivity. Scientists have discussed subjectivity in what is or isn't a species for hundreds of years. Charles Darwin even wrote about it in the Origin of Species.

As speciation is a constant and continuous process, there are going to be species that phenotypically (or genetically, or ecologically) reside in a gray zone between "clearly not distinguishable" and "absolutely a distinct and obvious separate species". Name an absolute system of species ranking and I can almost certainly find areas where delimitation is difficult. BSC? How much hybridization is too much not to mention what level of variation would be consistent with different species status in allopatric populations. PSC? How different do two clades need to be and what measure do we use to define those differences?

 

[MJB]

The Fern is absolutely correct. I don't know why people are arguing about taxonomy not having a degree of subjectivity. Scientists have discussed subjectivity in what is or isn't a species for hundreds of years. Charles Darwin even wrote about it in the Origin of Species.

As speciation is a constant and continuous process, there are going to be species that phenotypically (or genetically, or ecologically) reside in a gray zone between "clearly not distinguishable" and "absolutely a distinct and obvious separate species". Name an absolute system of species ranking and I can almost certainly find areas where delimitation is difficult. BSC? How much hybridization is too much not to mention what level of variation would be consistent with different species status in allopatric populations. PSC? How different do two clades need to be and what measure do we use to define those differences?

Hence the quotations on each of my posts...
MJB
PS In quantum physics, few of the conceptual criteria relate to classical physics, and so the boundary between objectivity and subjectivity is blurred because definitions in classical physics don't apply. The observer affects the results... Quantum physics arguably is the hardest of hard sciences...

 

[THE_FERN]

Hence the quotations on each of my posts...
MJB
PS In quantum physics, few of the conceptual criteria relate to classical physics, and so the boundary between objectivity and subjectivity is blurred because definitions in classical physics don't apply. The observer affects the results... Quantum physics arguably is the hardest of hard sciences...

Yes that was one of my points (Heisenberg). Also Gödel who demonstrates that you can't prove some fundamental tenets in science/maths: you have make assumptions.

(My other example was the probability threshold in frequentist stats which is entirely arbitrary, subjective and varies somewhat by discipline. Bayesians embrace this, of course)

"Hard" vs "Soft" science is sooo early/mid- 20th century...