How Taxon Changes work

The Tree

Identifications on iNaturalist reference particular taxa on the iNaturalist taxonomy. The iNaturalist taxonomy is a single global taxonomic tree. Each taxon, or node, on the tree has four essential components: the scientific name, taxonomic rank, the parent node to which it is grafted and the state of being active or inactive. Ranks are constrained to an ordered set that includes the standard ranks (e.g. Kingdom, Phylum, Class, Order, Family, Genus, and Species) as well as several non-standard ranks (Subfamily, etc.). For all active nodes, parent nodes must be of coarser rank than child nodes (e.g. a node of rank genus cannot be grafted to a node of rank species). Active nodes must have an active parents. Taxa that lack parents are referred to as ‘ungrafted’.

Curators can make new taxa unconditionally if they are inactive or ungrafted (ignoring a few constraints on the structure of the scientific name, e.g. no special characters). Active, grafted new taxa must follow the above constraints regarding rank order and active parents. Also, active, grafted taxa can't have the same name as any active siblings. Additionally, curators cannot activate or graft taxa into positions covered by curated taxon frameworks unless they have appropriate taxon curator permissions (more on taxon frameworks later on).

Curators can edit taxa by changing the parent, rank, or state of an existing taxon provided that the above constraints. iNaturalist does not allow the name to be edited in order to minimize the extent to which taxonomy curation changes what was intended by identifiers and loose the record of their original identifications.

Taxon Swaps

To change the name of a taxon, curators must draft a Taxon Swap. A Taxon Swap is a type of Taxon Change that takes a single input taxon and a single output taxon. When a draft Taxon Swap is committed, the input taxon is inactivated and all of the identifications referencing the input taxon are made not-current. The output taxon is activated (if it isn’t already) and new analogous current identifications referencing the output taxon are created to accompany each identification that was made not-current. Furthermore, all content associated with the input taxon (e.g. taxon names, taxon photos, taxon ranges, etc.) are copied over Estrilda the output taxon.

For example, to change the name of the species Estrilda caerulescens to Estrilda coerulescens, a curator would first create a new inactive taxon with the name Estrilda coerulescens and rank species grafted to genus Estrilda. For the purposes of this example, let’s assume that the taxon Estrilda caerulescens has a Taxon Photo, a Taxon Name, and an Identification on an observation associated with it.

Next, the curator would create a draft Taxon Swap with Estrilda caerulescens as the input taxon and Estrilda coerulescens as the output taxon.

Committing the swap inactivates Estrilda caerulescens and activates Estrilda coerulescens. The Taxon Photo and Taxon Name are duplicated and associated with the output taxon. The identification of Estrilda caerulescens is made non-current. A new current identification associated with Estrilda coerulescens is added to the observation.

To streamline the figures, let’s switch to a format with the pre-change tree on the left and the post-change tree on the right. The green Taxon Swap line now connects the input and output taxa bridging both trees. We omit inactive taxa, but we know that Estrilda coerulescens was inactivated since its the input of a Taxon Swap (Taxon Swap inputs are always inactivated). Since the states of Taxon Swap outputs don’t necessarily change from inactive to active (more on this later) the square indicates that Estrilda coerulescens was indeed activated from an inactive state by the Taxon Swap.

Our convention is also to show a root (in this case Estrilda) above the taxon involved in the swap to help orient us. But note that the root is altered by the swap. Note that even though there are many other species in the genus Estrilda (e.g. Estrilda astrild), we omit them because they are not involved in the swap (ie their names, ranks, and parent names aren’t altered pre and post change).

Here's a key that covers all the colors and shapes in these figures that we'll cover later in this document.

Taxon Swaps cannot be committed if the input taxon has active children unless the ‘move children’ box is checked. The move children box will automatically move the children before the taxon change is committed.

Note that in our figures we don’t explicitly point out moves with lines like we do swaps, but they can be implied by comparing the structures of the pre-change and post-change trees. Also note that we omit descendants such as the children of Casuarius since they aren’t involved in the swap (ie their names, ranks, and parent names don’t change)

If the children are species or subspecies then their names will change (e.g. Cholornis paradoxa taipaiensis becomes Cholornis paradoxus taipaiensis) to which means instead of moving them, iNaturalist will create associated swaps that will try to deal with these changes in genus and species names.

However, the ‘move children’ system is naive to many of rules such as gender agreement that govern taxonomic names. For example, note in the above figure that ‘move children’ would have naively swapped Cholornis paradoxa paradoxa with Cholornis paradoxus paradoxa. But according to the rules, the output taxon should be Cholornis paradoxus paradoxus. In general ‘move children’ should only be used if you’re absolutely sure it won’t create a mess and a better option is to manually handle the children first by moving them or swapping them.

If you are manually handling these changes, note that the order of operations is important. For example in the Casuariidae example, first create all new inactive taxa (e.g. Dromaiidae). Then move the appropriate children of Casuariidae (e.g. Casuarius and Dromaius) to this new inactive taxon Dromaiidae. Note that iNaturalist will allow you to move an active child to an inactive parent if its the output taxon of a draft taxon change which will subsequently activate the parent. But please only do this if it's to manually handle children immediately before committing a taxon change since grafting active children to inactive parents is otherwise problematic.

In the Cholornis proxima example, first make the new inactive taxa (e.g. Cholornis proximus, Cholornis proximus taipaiensis, and Cholornis proximus proximus) in their proper positions. Then create the 3 draft swaps. When committing the swaps, start with the swaps with the finest ranked inputs (e.g. Cholornis proxima taipaiensis to Cholornis proximus taipaiensis or Cholornis proxima proxima to Cholornis proximus proximus) and work your way up (e.g. Cholornis proxima to Cholornis proximus) to ensure that none of the swap inputs have active children when you commit them.

Lumps

Taxon Changes can also be used to synonymize one taxon with another taxon. For example, imagine that Anas crecca nimia and Anas crecca crecca were both active taxa and it was subsequently determined that Anas crecca nimia was a synonym of Anas crecca crecca. In this case, a taxon swap can be used as above to move or copy all content from Anas crecca nimia to Anas crecca crecca. The mechanics are identical to the name change example above, the only difference being that when two taxa are lumped the output taxon (Anas crecca crecca) is already active before the swap rather than inactive.This is why in the figure below there is a circle, not a square associated with the output taxon to indicate that the taxon wasn’t activated. The blue color means Anas crecca crecca was ‘broadened’ by absorbing Anas crecca nimia.

What happens when a taxon is broadened? Because all identifications of Anas crecca nimia are updated with identifications of Anas crecca crecca, there are no issues with identifications becoming misinterpreted. Likewise, the updated identifications will appropriately broaden the set of observations associated with Anas crecca crecca. However, there may be associated content that is broadened properly following a lump. For example, imagine Anas crecca crecca has an Endangered conservation status and a very localized taxon range imported from IUCN while Anas crecca nimia Least Concern status and a very broad range. After absorbing Anas crecca nimia, both the conservation status and taxon range of Anas crecca crecca would become misspecified and would need to be manually updated. We consider broadened taxa like Anas crecca crecca involved in the swap (and thus appears in the figure) even though its name, rank, or parent didn’t change because what we mean by the taxon changes and it’s a reminder to revisit its associated content.

As a rule of thumb, taxon ranges, conservation statuses, atlases, and listed taxa (and associated establishment means) should always be reviewed after a taxon is broadened. Taxon names should also be checked (for example, ‘eastern newt’ might not be appropriate after ‘western newt’ is lumped into it). These types of content are more commonly associated with taxa of rank species, but this is not always the case.

We should note that there is a different type of Taxon Change of type ‘Merge’ that is identical to changes of type ‘Swap’ except that it has multiple input taxa. It doesn’t do anything that can’t be achieved by more than one swap but sometimes it can reduce the number of steps needed to curate a taxon.

Splits

So far, we’ve discussed creating new taxa, editing existing taxa, swapping taxa to deal with name changes and swapping taxa to deal with lumps. With lumps, we’ve begun to wade into this issue of changing (ie broadening) what we mean by a taxon and some of the associated misspecification it can cause with taxon properties like taxon ranges.

Often, a new taxon appears because it is carved off from some other taxon. This is the opposite of a lump. For example, Rote Leaf Warbler (Phylloscopus rotiensis) was added to the 2019 Clements checklist. The map below shows the existing taxon range for Timor Leaf Warbler (Phylloscopus presbytes). Note that it includes Rote island. But according to the 2019 Clements checklist Timor Leaf Warbler only occurs on Timor and Flores. This is an indication that Rote Leaf Warbler was carved off from and subsequently narrowed what we mean by Timor Leaf Warbler.

Because there are no identifications or observations of Timor Leaf Warbler from Rote (at the time of this writing there is just a single observation from Flores), we do not need to split Timor Leaf Warbler (Phylloscopus presbytes) and can simply create a new taxon for Rote Leaf Warbler without splitting Timor Leaf Warbler. But just as with broadening taxa via lumps, we need to be aware that adding Timor Leaf Warbler narrowed what we mean by Timor Leaf Warbler and we may need to update associated content - specifically we would need to update the range map to remove Rote Island and check that any atlases are still correct. The orange color in the figure below indicates that Timor Leaf Warbler (Phylloscopus presbytes)

Now let’s consider a similar situation where there are identifications and observations of the narrowed taxon. The 2019 version of Clements splits the Iberian portion of European Green Woodpecker (Picus viridis) off as Iberian Green Woodpecker (Picus sharpei)

If we had followed the above procedure here of adding an active new taxon for Iberian Green Woodpecker (Picus sharpei) and narrowing the range and atlas of European Green Woodpecker (Picus viridis), we would have a problem because we’ve altered what was intended by all the identifiers of existing observations of the woodpecker from the Iberian peninsula. When they added identifications of European Green Woodpecker (Picus viridis), they meant the broader concept that includes Iberia. If we narrow the concept of European Green Woodpecker (Picus viridis) to exclude Iberia, iNaturalist will interpret these identifications as disagreements with Iberian Green Woodpecker (Picus sharpei) which is not what the identifiers intended. In these cases, where we wish to narrow the concept of a taxon and there are existing observations or identifications, we must make a Taxon Split.

A taxon split is a type of Taxon Change that takes a single input and multiple outputs. In this example, first create new inactive taxa for both Picus sharpei and this narrower concept of Picus viridis. Then create a Taxon Split with the active Picus viridis as the input and these two newly created taxa as outputs. In the figure below note that the output Picus viridis has a square associated with it since it is a new taxon that it will be activated by the change. The orange color is meant to remind you that even though it has the same name as the input Picus viridis, its concept is narrower.

Committing a split will inactivate the input taxon and activate the output taxa but unlike other taxon change types we’ve seen, it will not copy associated content like Taxon Photos and Taxon Names since its not clear which output taxon to copy them two. For similar reasons, there is also no option to automatically ‘move children’ (ie these must be manually handled before committing the split). As with other taxon changes, identifications of the input taxon are made non-current, but the new identifications that are added are of the common ancestor (e.g. Picus) of the input and output taxa (again since its not clear to which output taxon they are assigned).

This is why taxon splits can be destructive, as they can coarsen many identification and thus their associated observations. One way to minimize this is to use atlases to assign identifications to the proper output based on non-overlapping spatial distributions.

Note that there’s nothing wrong with using a Taxon Split when you want to narrow a taxon that doesn’t have observations or identifications from the geographic region in question (as in the Timor Leaf Warbler example). It’s perfectly fine to narrow a taxon this way, you just don’t have to since there are no identifications to worry about.

As noted above, when making a split, you manually have to handle the children since it’s unclear to which output taxon they should be moved. Remember that in the figures we omit drawing children who’s names, ranks, parent names and parent names don’t change, but remember that you have to move these from the split input to the narrowed split output with the same name. For example, in the following figure Anthreptes malacensis anambae isn’t shown but its implied that it is moved from the split input Anthreptes malacensis to the split output Anthreptes malacensis. Children who’s names, ranks, or parent names change are shown such as Anthreptes malacensis birgitae which is swapped into Anthreptes griseigularis birgitae.

Multi-change taxonomy alterations

We’ve already discussed how handling descendants (by moving them or swapping them) can mean that getting from ‘here’ (the current tree) to ‘there’ (the desired tree) requires more work beyond a single taxon change. There are many other situations where such ‘multi-change taxonomy alterations’ are required. In this section, we’ll describe a few such structures that you’ll frequently encounter using examples from the Clements 2019 update. Remember, when a taxon is narrowed you only have to split if the input taxon has identifications in the region that is split off - otherwise its possible to just manually update the narrowed taxon’s associated content. But in these examples we assume splits are necessary.

Swing-split

The most common alteration is something we can refer to as a ‘swing-split’ this is when a subspecies is elevated to species status (or more generally any child elevated to sibling status). For example, to elevate Amazilia saucerottei hoffmanni to species status:

  1. Create new inactive taxa for the narrowed Amazilia saucerottei and Amazilia hoffmanni.
  2. Create a draft taxon split from the active Amazilia saucerottei to these new taxa and create a draft taxon swap from Amazilia saucerottei hoffmanni to Amazilia hoffmanni.
  3. Move all other children of the active Amazilia saucerottei to the inactive Amazilia saucerottei.
  4. Create atlases for the split outputs
  5. Commit the swap and then the split

The figures are useful because they summarize all of this information into a single figure. They become less confusing to interpret with practice.

Swing-lump

The opposite of a swing-split is the less common ‘swing-lump’. Whereas swing-splits narrow the parent of the elevated taxon, swing-lumps broaden the parent of the sunk taxon. To sink Anthus latistriatus below Anthus cinnamomeus:

  1. Create new inactive taxa for Anthus cinnamomeus latistriatus.
  2. Create a draft taxon swap from Anthus latistriatus to Anthus cinnamomeus latistriatus
  3. Commit the swap
  4. Review content associated with Anthus cinnamomeus since it was broadened

Reshuffle

A ‘reshuffle’ is when one or more taxa are moved from one taxon to another. Note that in this scenario the taxon that loses the child is narrowed (and thus split) while the other is broadened. To reshuffle Anthus vaalensis saphiroi and Anthus vaalensis goodsoni from Anthus vaalensis over to Anthus leucophrys:

  1. Create new inactive taxa for the narrowed Anthus vaalensis, Anthus leucophrys saphiroi and Anthus leucophrys goodsoni
  2. Create a draft taxon split from Anthus vaalensis to Anthus leucophrys (which is already active) and the narrowed Anthus vaalensis. Also create draft taxon swaps from Anthus leucophrys saphiroi and Anthus leucophrys goodsoni to Anthus leucophrys saphiroi and Anthus leucophrys goodsoni respectively.
  3. Move all other children of the active Anthus vaalensis to the inactive Anthus vaalensis.
  4. Create atlases for the split outputs if they don’t already exist.
  5. Commit the swaps and then the split
  6. Review content associated with Anthus leucophrys since it was broadened

Dealing with nominates

Remember that if a species has subspecies, it must have what we call the ‘nominate’ subspecies (e.g. Serinus alario alario) which has the same specific and subspecific epithets. Likewise, species shouldn’t have a single subspecies. They should either have more than one (including the nominate) or none. To conform with these rules, sometimes we have to add additional steps to ‘bud’ or ‘envelope’ nominate subspecies onto the structures we’ve discussed above. For example, the swing-lump that places Serinus leucolaemus below Serinus alario as Serinus alario leucolaemus requires creating Serinus alario alario to comply with these rules. In the figures, we indicate budded species with as yellow squares. Note that these ‘budded’ taxa aren’t the outputs of taxon changes but have to be activated manually.

Whenever a taxon is broadened from a subspecies being moved beneath it via a swing-lump or a reshuffle always check to make sure you don’t need to bud off a nominate. Likewise, sometimes swing-splits or reshuffled can leave orphaned nominate subspecies that no longer have siblings. For example, splitting Batrachostomus mixtus off from Batrachostomus poliolophus requires also enveloping Batrachostomus poliolophus poliolophus into its parent with an additional swap.

You’ll also see more general instances of subspecies budding off

Or enveloping back into their species parents

Which can be handled by manually activating taxa and creating swaps accordingly.

Cross branch changes

In the examples we’ve looked at so far, the output taxa of splits are siblings of the input taxa. But sometimes, splits and lumps can act farther afield across branches of the tree. For example, Alcippe ludlowi was reshuffled into the genus Fulvetta. However, the genus Fulvetta is in a different family (Paradoxornithidae) than the genus Alcippe (Leiothrichidae). Just as Fulvetta is broadened by the reshuffle and Alcippe is narrowed, all ancestors of Fulvetta back to the common ancestor Passerformes (e.g. Paradoxornithidae) are broadened. Likewise, all ancestors of Alcippe back to the common ancestor Passerformes (e.g. Leiothrichidae) are narrowed. In this example, this means we need to make an extra split (splitting Leiothrichidae into Leiothrichidae and Paradoxornithidae).

Multi-step

So far we’ve discussed ‘name changes’, lumps, splits, swing-slits, swing-lumps, reshuffles, and associated budding and enveloping of taxa. Most changes you’ll encounter fall into one of these categories.

If you encounter more complex changes, sometimes it’s useful to think of them as two or more of these changes stitched together. For example, this change looks complex but really it’s just a split (Calandrella eremica split off from Calandrella blanfordi) followed by a swing-lump (Calandrella erlangeri sunk beneath from Calandrella blanfordi as a subspecies).

For even more complicated multi-step changes, we’ve found that the figures themselves are the best way to figure out and visualize what needs to be done. Start by drawing the ‘before’ tree with all the input taxa that are involved in the change - remember these are any taxa who’s names, ranks, or parent names are altered by the change or taxa who are broadened or narrowed. In this example, it’s the species Pitchui kirhocephalus and 15 of its child subspecies. Then, for the ‘after’ tree, think about where each of those input taxa went either via being moved or via taxon changes and draw those output nodes.

The completed figure will tell you exactly the taxon changes you need to do. In this example, you need to:

  1. Create inactive taxa for the narrowed Pitohui kirhocephalus, Pitohui cerviniventris, and Pitohui uropygialis as well as the subspecies whose names were changed: Pitohui cerviniventris pallidus, Pitohui cerviniventris cerviniventris, Pitohui uropygialis aruensis, Pitohui uropygialis brunneiceps, Pitohui uropygialis meridionalis, Pitohui uropygialis nigripectus, and Pitohui uropygialis uropygialis. Move all the remaining children of Pitohui kirhocephalus to the inactive output Pitohui kirhocephalus
  2. Create a draft split from the input Pitohui kirhocephalus to the narrowed Pitohui kirhocephalus, Pitohui cerviniventris, and Pitohui uropygialis and a draft split from Pitohui kirhocephalus salvadorii to Pitohui kirhocephalus kirhocephalus and Pitohui kirhocephalus dohertyi.
  3. Create a draft merge from Pitohui kirhocephalus carolinae, Pitohui kirhocephalus adiensis, and Pitohui kirhocephalus stramineipectus to Pitohui kirhocephalus rubiensis, and a draft merge from Pitohui kirhocephalus uropygialis and Pitohui kirhocephalus tibialis to Pitohui uropygialis uropygialis
  4. Create the 6 draft swaps from Pitohui kirhocephalus pallidus to Pitohui cerviniventris pallidus, Pitohui kirhocephalus cerviniventris to Pitohui cerviniventris cerviniventris, Pitohui kirhocephalus aruensis to Pitohui uropygialis aruensis, Pitohui kirhocephalus brunneiceps to Pitohui uropygialis brunneiceps, Pitohui kirhocephalus meridionalis to Pitohui uropygialis meridionalis, and Pitohui kirhocephalus nigripectus to Pitohui uropygialis nigripectus
  5. Make atlases for the split outputs if feasible (the species are probably higher priority then the subspecies)
  6. Commit the swaps, merges, and splits
  7. Check the associated content for broadened taxa (in this case they are subspecies so unlikely to have ranges etc.)

While these multi-step changes can get incredibly gnarly, reducing them to these basic steps outlined above will work in all but one case we’ve encountered where an additional step is needed. This case is when a taxon change will activate a taxon with the same name as a sibling which isn’t allowed (siblings in iNaturalist must have unique taxon names). For example, imagine the dual-reshuffle illustrated below where Oenanthe monticola is moved to Myrmecocichla and Myrmecocichla albifrons is moved to Oenanthe. The associated splits require new inactive taxa for both Myrmecocichla and Oenanthe.

Unfortunately, the split with Oenanthe as the input will attempt to activate a new output Myrmecocichla sibling to the input Myrmecocichla for the other split (which is still active). Because iNaturalist won’t let two active siblings to have the same name, this split won’t commit. The work-around is to manually inactivate the input Myrmecocichla before committing the first split.

For an example of how gnarly these multi-level changes can get, here’s what happened to several white-eyes (Zosterops) in the Clements 2019 revision.

Taxon Framework Relationships

Taxon Frameworks, help make it explicit what taxonomy we’re using and where we are deviating from that taxonomy. At the time of this writing, the bird taxonomy on iNaturalist is still using Clements 2018, but the taxon framework has been updated map to Clements 2019. This means that wherever iNaturalist/Clements 2019 matches Clements 2019 the relationship will be of type “Match”. For example, Passer domesticus is the same (e.g. it has the same name, rank, and parent name and its neither broadened or narrowed) in iNaturalist and Clements 2019.

But every place where taxa are not the same is currently mapped as a ‘deviation’. For example, we have a deviation with Alethe diademata, Alethe diademata castanea, Alethe diademata woosnami, and Alethe diademata diademata in iNaturalist (on the right). All of these are not represented exactly (e.g. by name, rank, and parent or are broadened or narrowed) in Clements 2019 which appears on the left.

You may notice that these are exactly the taxa we’ve been using in our figures (apologies that in these iNaturalist appears on the left). To update the iNaturalist taxonomy to match Clements 2019, we’ll need to design taxon changes to get us from the ‘before’ trees to the ‘after’ trees in each of these cases.

We should note that sometimes deviations are used to illustrate places where we are intentionally deviating from the reference (in this case Clements 2019). But in this case, all of these deviations just reflect that we haven’t yet gotten around to updating to Clements 2019. .

You may ask how we’ve decided where each pair of before/after trees start and stop for each deviation. For example, in theory we could just have one giant deviation with the entire before and after trees for all birds, but this would be intractable and hard to understand. Our preference is to break them down into the smallest possible pairs of branches that tell a complete story - for example, iNaturalist and Clements 2019 differ in this part of the tree in that Alethe castanea is split off in the latter from Alethe diademata involving several subspecies.

Clements 2019 update

The Taxon Framework for birds is currently showing 432 deviations from what’s currently in iNaturalist and its reference (Clements 2019). Of these 256 are ‘not external’ deviations to capture taxa like hybrids that are in iNat but are not included in Clements.

The remaining 176 deviations describe the difference between what’s currently on iNaturalist and what we’d need to update the tree to Clements 2019. For each of these 176 deviations, we’ve made figures like the ones described above which we’ll link to below. We’ve also made 162 groups of draft taxon changes required to ‘resolve’ each of these deviations. The remaining 14 are deviations which can be ‘resolved’ simply by adding taxa without any swaps or splits. They include subspecies budding off of a species that previously had no subspecies or a taxon that appears to be ‘wholly new’ - that is it can be added to the tree without affecting any other taxa.

Below we’ve listed links to figures of each of these 176 deviations and when applicable the taxon change group.

Alethe diademata swing-split (tree figure) (taxon change group)

Amazilia saucerottei swing-split (tree figure) (taxon change group)

Anas crecca crecca lump (tree figure) (taxon change group)

Anthreptes malacensis split (tree figure) (taxon change group)

Anthus cinnamomeus swing-lump (tree figure) (taxon change group)

Anthus vaalensis reshuffle (tree figure) (taxon change group)

Apalis thoracica swing-split (tree figure) (taxon change group)

Apalis porphyrolaema swing-split (tree figure) (taxon change group)

Arizelocichla nigriceps swing-split (tree figure) (taxon change group)

Artamidae lump (tree figure) (taxon change group)

Atlapetes lump (tree figure) (taxon change group)

Petrochelidon split (tree figure) (taxon change group)

Batis minor swing-split-bud (tree figure) (taxon change group)

Batrachostomus poliolophus swing-split (tree figure) (taxon change group)

Brachypteryx montana swing-split (tree figure) (taxon change group)

Buphagus erythrorynchus name change (tree figure) (taxon change group)

Calamanthus campestris swing-split (tree figure) (taxon change group)

Calamonastes undosus split (tree figure) (taxon change group)

Calidris pygmaea name change (tree figure) (taxon change group)

Camaroptera brachyura swing-split (tree figure) (taxon change group)

Cercotrichas galactotes lump (tree figure) (taxon change group)

Ceyx flumenicola name change (tree figure) (taxon change group)

Cholornis paradoxus name change (tree figure) (taxon change group)

Chroicocephalus novaehollandiae swing-lump (tree figure) (taxon change group)

Cinnyris manoensis swing-split (tree figure) (taxon change group)

Cinnyris mediocris split (tree figure) (taxon change group)

Circus maillardi swing-split (tree figure) (taxon change group)

Circus spilonotus swing-split (tree figure) (taxon change group)

Clytorhynchus nigrogularis swing-split (tree figure) (taxon change group)

Coccopygia melanotis swing-split (tree figure) (taxon change group)

Coracias benghalensis swing-split (tree figure) (taxon change group)

Corvus torquatus name change (tree figure) (taxon change group)

Corvus enca swing-split (tree figure) (taxon change group)

Crithagra gularis split (tree figure) (taxon change group)

Crossoptilon crossoptilon swing-split (tree figure) (taxon change group)

Cyornis tickelliae swing-split (tree figure) (taxon change group)

Daphoenositta chrysoptera split (tree figure) (taxon change group)

Delichon dasypus cashmeriense name change (tree figure) (taxon change group)

Dicaeum melanozanthum name change (tree figure) (taxon change group)

Dicrurus modestus swing-split (tree figure) (taxon change group)

Dinopium benghalense puncticolle split (tree figure) (taxon change group)

Dryocopus schulzi name change (tree figure) (taxon change group)

Ducula aenea consobrina lump (tree figure) (taxon change group)

Ducula spilorrhoa swing-split (tree figure) (taxon change group)

Edolisoma melan name change (tree figure) (taxon change group)

Edolisoma tenuirostre ultima name change (tree figure) (taxon change group)

Elaenia obscura swing-split (tree figure) (taxon change group)

Elaenia pallatangae swing-split (tree figure) (taxon change group)

Estrilda coerulescens name change (tree figure) (taxon change group)

Ficedula erithacus name change (tree figure) (taxon change group)

Foudia eminentissima swing-split (tree figure) (taxon change group)

Fringilla teydea swing-split (tree figure) (taxon change group)

Geokichla piaggiae swing-lump (tree figure) (taxon change group)

Gymnoris name change (tree figure) (taxon change group)

Hemixos flavala swing-split (tree figure) (taxon change group)

Ianthocincla lanceolata swing-split (tree figure) (taxon change group)

Arachnothera split (tree figure) (taxon change group)

Oreothlypis split (tree figure) (taxon change group)

Leptocoma aspasia name change (tree figure) (taxon change group)

Cranioleuca reshuffle (tree figure) (taxon change group)

Lonchura malacca reshuffle (tree figure) (taxon change group)

Macropygia mackinlayi arossi lump (tree figure) (taxon change group)

Megascops cooperi cooperi lump (tree figure) (taxon change group)

Melanitta deglandi swing-split (tree figure) (taxon change group)

Tiaris split (tree figure) (taxon change group)

Microcarbo pygmaeus name change (tree figure) (taxon change group)

Napothera danjoui swing-split (tree figure) (taxon change group)

Ninox novaeseelandiae split (tree figure) (taxon change group)

Nisaetus cirrhatus lump (tree figure) (taxon change group)

Nothura maculosa swing-lump (tree figure) (taxon change group)

Notopholia corusca name change (tree figure) (taxon change group)

Oenanthe xanthoprymna swing-split (tree figure) (taxon change group)

Oreolais pulcher name change (tree figure) (taxon change group)

Oreotrochilus estella swing-split (tree figure) (taxon change group)

Ortygospiza atricollis lump (tree figure) (taxon change group)

Otus manadensis swing-split (tree figure) (taxon change group)

Oxypogon stuebelii name change (tree figure) (taxon change group)

Claravis split (tree figure) (taxon change group)

Gallinula swing-split (tree figure) (taxon change group)

Passer simplex swing-split (tree figure) (taxon change group)

Passer simplex swing-split (tree figure) (taxon change group)

Petroica australis swing-split (tree figure) (taxon change group)

Philemon moluccensis swing-split (tree figure) (taxon change group)

Calliphlox split (tree figure) (taxon change group)

Phrygilus split (tree figure) (taxon change group)

Phylloscopus borealis env (tree figure) (taxon change group)

Phylloscopus montis barisanus name change (tree figure) (taxon change group)

Picus viridis swing-split (tree figure) (taxon change group)

Poicephalus robustus swing-split (tree figure) (taxon change group)

Polioptila albiloris swing-split (tree figure) (taxon change group)

Polioptila guianensis swing-split (tree figure) (taxon change group)

Pomatorhinus ruficollis hunanensis name change (tree figure) (taxon change group)

Poospizopsis split (tree figure) (taxon change group)

Psilopogon franklinii swing-split (tree figure) (taxon change group)

Psilopogon chrysopogon swing-lump (tree figure) (taxon change group)

Psittacara holochlorus swing-split (tree figure) (taxon change group)

Pternistis castaneicollis swing-split (tree figure) (taxon change group)

Ptilorrhoa caerulescens swing-split (tree figure) (taxon change group)

Puffinus bailloni swing-split (tree figure) (taxon change group)

Puffinus assimilis swing-split (tree figure) (taxon change group)

Pycnonotus simplex perplexus lump (tree figure) (taxon change group)

Pycnonotus bimaculatus swing-split (tree figure) (taxon change group)

Ramphocaenus melanurus swing-split (tree figure) (taxon change group)

Regulus regulus ellenthalerae name change (tree figure) (taxon change group)

Rhizothera longirostris swing-split (tree figure) (taxon change group)

Rhynchostruthus socotranus swing-split (tree figure) (taxon change group)

Saxicola split (tree figure) (taxon change group)

Sericulus aureus swing-split (tree figure) (taxon change group)

Serinus alario swing-lump (tree figure) (taxon change group)

Sylvia abyssinica swing-split (tree figure) (taxon change group)

Sylvia nigricapillus name change (tree figure) (taxon change group)

Sylvia subcoerulea name change (tree figure) (taxon change group)

Leptasthenura reshuffle (tree figure) (taxon change group)

Tangara split (tree figure) (taxon change group)

Tanysiptera sylvia swing-split (tree figure) (taxon change group)

Tyrannidae reshuffle (tree figure) (taxon change group)

Theristicus melanopis swing-split (tree figure) (taxon change group)

Threskiornis aethiopicus swing-split (tree figure) (taxon change group)

Topaza pyra reshuffle (tree figure) (taxon change group)

Treron australis swing-split (tree figure) (taxon change group)

Trogon collaris castaneus lump (tree figure) (taxon change group)

Turdus ignobilis swing-split (tree figure) (taxon change group)

Turnix hottentottus swing-split (tree figure) (taxon change group)

Vireo crassirostris reshuffle (tree figure) (taxon change group)

Xiphorhynchus fuscus swing-split (tree figure) (taxon change group)

Yuhina flavicollis reshuffle (tree figure) (taxon change group)

Zoothera dauma swing-split (tree figure) (taxon change group)

Zosterops chloris maxi lump (tree figure) (taxon change group)

Zosterops abyssinicus split (tree figure) (taxon change group)

Zosterops lateralis swing-lump (tree figure) (taxon change group)

Zosterops pallidus env (tree figure) (taxon change group)

Zosterops poliogastrus swing-split (tree figure) (taxon change group)

Zosterops virens capensis lump (tree figure) (taxon change group)

Zosterops virens virens lump (tree figure) (taxon change group)

Coracina cinerea swing-split-bud (tree figure) (taxon change group)

Dicrurus divaricatus swing-split-bud (tree figure) (taxon change group)

Pernis celebensis swing-split-bud (tree figure) (taxon change group)

Horornis diphone swing-split (tree figure) (taxon change group)

Calandrella blanfordi swing-split (tree figure) (taxon change group)

Cisticola galactotes swing-split (tree figure) (taxon change group)

Pharomachrus pavoninus reshuffle (tree figure) (taxon change group)

Trogon collaris lump (tree figure) (taxon change group)

Epinecrophylla haematonota swing-lump (tree figure) (taxon change group)

Sylvia curruca swing-lump (tree figure) (taxon change group)

Trichoglossus haematodus swing-split (tree figure) (taxon change group)

Pitohui kirhocephalus swing-split (tree figure) (taxon change group)

Colluricincla reshuffle (tree figure) (taxon change group)

Sarothruridae reshuffle (tree figure) (taxon change group)

Megalurus split (tree figure) (taxon change group)

Petroica multicolor split (tree figure) (taxon change group)

Alcippe split (tree figure) (taxon change group)

Dicrurus ludwigii swing-split (tree figure) (taxon change group)

Myrmecocichla reshuffle (tree figure) (taxon change group)

Zosterops palpebrosus swing-split (tree figure) (taxon change group)

Pycnonotus simplex split (tree figure) (taxon change group)

Phylloscopus presbytes split (tree figure) (taxon change group)

Bradypterus bangwaensis split (tree figure) (taxon change group)

Caprimulgus ruwenzorii ruwenzorii split (tree figure) (taxon change group)

Ducula aenea sylvatica split (tree figure) (taxon change group)

Ducula aenea paulina split (tree figure) (taxon change group)

Ducula aenea aenea split (tree figure) (taxon change group)

Heliangelus amethysticollis decolor split (tree figure) (taxon change group)

Oreotrochilus cyanolaemus wholly new

Dicrurus adsimilis jubaensis wholly new

Eudynamys orientalis hybrida wholly new

Grallaria guatimalensis binfordi wholly new

Zoothera dauma iriomotensis wholly new

Mecocerculus leucophrys montensis wholly new

Charadrius alexandrinus nihonensis wholly new

Hemiphaga novaeseelandiae novaeseelandiae wholly new

Zentrygon linearis bud

Pternistis squamatus bud

Phyllomyias griseiceps bud

Phalacrocorax chalconotus bud

Pelecanoides georgicus bud

Newtonia amphichroa bud

When to split

Unfortunately, we often aren’t told ‘this taxon was lumped into that one’ or ‘this one was split from that one’. Normally all we are told is that ‘this taxon is no longer valid’ and ‘this is a new taxon’. When a taxon is no longer valid, it’s usually straightforward to determine what taxon it has been lumped into and is now a synonym of. But when a new taxon appears it’s unfortunately extremely difficult to determine which, if any, other taxa it was split from.

We call taxa that arrive without narrowing any other taxa “wholly new”. For example, in 2018, Sornoza-Molina et al. described Blue-throated Hillstar (Oreotrochilus cyanolaemus) from southern Ecuador which was added to 2019 version of the Clements checklist Since this taxon occurs .

Blue-throated Hilllstar is restricted to a single mountain range outside the known ranges of its siblings Ecuadorian Hillstar and Green-headed Hillstar. And there are no iNaturalist observations or identifications of Hillstars from this mountain range. Thus adding this species is unlikely to alter the existing ranges of the sibling taxa or change the intention of any existing identifications. We call these taxa that don’t narrow the concept of any siblings ‘‘wholly new”. It is safe to add “wholly new” taxa without making adjustments to sibling taxa. The yellow color in the figure indicates that these taxa have no-analogue in the pre-change tree (where green indicates the same concept and blue indicates a broadened concept). You could make an argument that “wholly new” taxa broaden the concepts of their ancestors (hence the blue color) so it’s worth a check.

Other taxa are clearly split from other taxa. We’ve discussed above how to handle these. But usually there are a pile of names that we can’t determine if they are ‘wholly new’ or not. We generally do as much due-diligence as we can to rule out a split, but often a species is added to iNaturalist as ‘wholly new’ and only later someone from the community shines light on the split that should have happened. In these cases, it’s fine to split the taxon after the split-output was accidentally added as wholly new.

Here's an example of a taxon (Sporophila morelleti) that was added as 'wholly new' when it should have been split from Sporophila torqueola. As you can see, there are now many observations Sporophila torqueola that should have been carved off as Sporophila morelleti but weren't. And associated content such as the taxon range of Sporophila torqueola are now too broad. The solution is to create a new narrowed version of Sporophila torqueola and to retroactively split Sporophila torqueola into the existing Sporophila morelleti and this new narrowed Sporophila torqueola

Taxa can be thought of along a gradient running from ‘poorly studied to ‘well studied taxa. At the poorly studied end of the spectrum, all we have are published names that can be traced back to individual museum specimens. No one has done the work to determine the boundaries of the species attached to each specimen. At this end of the spectrum splits probably aren’t worth making as its just too unclear where one species stops and the other begins. At the other end of the spectrum are extremely well studied taxa such as birds and other vertebrates. Here wholly new taxa are rarer and splits are much more likely to be necessary.

In the middle of the spectrum you’ll often see a paper that describes a new species and says it was ‘formerly confused with’ some other sibling. While this is often a hint that we’re dealing with a split, it may be too unclear what is meant by the sibling, or which sibling to warrant a split. As a general rule of thumb, the new taxon should be split off rather than raised as ‘wholly new’ if you can determine which sibling may have been impacted and if the sibling in question has associated distribution data such as a taxon range or atlas, conservation statuses, more than 10 observations especially if it’s in a well studied clade (e.g. vertebrates) or a well studied location (e.g. the United States). But if we really don’t have much information (identifications, observations, ranges, etc.) to define where siblings start and stop, a split might be pointless.

Revised on December 30, 2019 19:45 by loarie loarie