Updating IUCN Red List conservation statuses

As part of our ongoing work improving how iNaturalist restricts sensitive geographic information including processes for deciding when species are threatened by coordinate disclosure, we are revisiting the role of IUCN Red List statuses.

Almost 10 years ago, iNaturalist automatically imported global conservation statuses from the IUCN Red List. Complexity surrounding taxonomy and status vs. geoprivacy concerns have led us to not automatically update these statuses since then. This document describes a new plan for how we will automatically update/maintain these conservation statuses in light of these issues. We’re piloting this process with reptiles and amphibians (IUCN Red List statuses for amphibians and reptiles have now been updated via the process described below) and plan to apply this process to update the entire IUCN Red List next month if the pilot goes well. The IUCN Red List generally releases updates biannually and we plan to use this process to update these statuses on that schedule.

Taxonomic concerns

In order to fetch conservation statuses from external authorities, the taxon’s valid scientific name on iNaturalist (internal) must match the name on the external authority. If there is an external name that doesn’t match any internal name but uniquely matches an invalid scientific name, a one-to-one relationship can be used to fetch conservation statuses from external authorities.

Note that this process is naive to complex taxonomic mappings such as one-to-many (taxon broadened internally) and many-to-one (taxon narrowed internally) relationships. iNaturalist treats these as matches based on the name. This means that in the one-to-many case the applied conservation status may be too narrow. In the many-to-one case the applied conservation status may be too broad and there will be internal taxa that don’t have statuses (e.g. Agalychnis taylori).

In the many-to-one scenario, if curators want to duplicate a status (e.g. for Agalychnis callidryas) for a taxon that doesn’t have one (e.g. Agalychnis taylori) they are welcome to do so. But do not list “IUCN Red List” as the authority. Refer to some other authority if it exists (e.g. NatureServe G-ranks) or leave the authority field blank and describe in the conservation status description how it was manually derived. For example: “Agalychnis taylori was carved-off from Agalychnis callidryas sensu IUCN which has a Least Concern status so I manually applied that status here. See https://www.inaturalist.org/flags/XXXX” where https://www.inaturalist.org/flags/XXXX is a resolved flag you created on the Agalychnis taylori taxon page to host any discussion about this status.

When automatically updating statuses, iNaturalist will use matching and one-to-one names to update the status for any existing global statuses with the IUCN Red List Authority and create any new ones. For newly created statuses, iNaturalist will set geoprivacy to open if the status is LC (Least Concern), NE (Not Evaluated), or DD (Data Deficient) and to obscured if the status is NT (Near Threatened), VU (Vulnerable), EN (Endangered), CR (Critically Endangered), EW (Extinct in the Wild), or EX (Extinct). See sections below for how iNaturalist treats taxon geoprivacy of existing statuses and other global statuses not in the IUCN RedList.

Status vs geoprivacy concerns

The status associated with conservation status is not always a good proxy for whether the species is threatened by coordinate disclosure. Species may be threatened by other factors such as habitat disturbance. Our default position on iNaturalist is that if the status is secure (LC, NE, or DD) then geoprivacy should be open and if the status is vulnerable (NT, VU, EN, CR, EW, or EX) then geoprivacy should be obscured.

However, there are exceptions where the community may choose to set taxon geoprivacy to open for a vulnerable species or vice versa. When this occurs, a resolved flag should be created and linked to from the conservation status description. The community should work to establish a record that this deviation was intentional any why it is warranted in the flag.

For examples, see Sclerophrys pantherina

and Melanophryniscus stelzneri

If an existing conservation status has a link to a flag in the description, iNaturalist will not automatically update the status to avoid undoing this work done by the community.

As part of this reptiles and amphibians pilot, we’ve temporarily avoided updating geoprivacy on 933 global IUCN Red List conservation statuses where geoprivacy is not in the default position relative to status (e.g. open and threatened or obscured and secure) and there is no flag in the conservation decision explaining the deviation. We suspect that most of these are due to drift from updates on the IUCN Red List side and not community members intentionally deviating from the default position. But, if the latter is the case, please create a flag and include the flag link in the conservation status description by September 20th. After that date, we will update the geoprivacy for any of these that lack a flag. You can find the list of these defacto-deviations here.

Other global statuses

The automatic update process will delete any global statuses with the authority set to IUCN Red List on taxa no longer map to names in the IUCN Red List (e.g. in cases where a species was removed from the Red List in an update).

To avoid confusion, we’d like to have only one global conservation status on iNaturalist per taxon and for the IUCN Red List to be the primary source for these global statuses. Please limit adding other global authorities (e.g. Natureserve G-rankings) only to taxa that are not in IUCN. If a species is in IUCN, adding S or N rank Natureserve conservation statuses for each state or country is preferable. If you think another global authority suggests the IUCN default position geoprivacy is incorrect, rather than add a 2nd conflicting global status linked to this authority, please change the geoprivacy on the IUCN status and link to a flag in the description where you use this other authority to explain your reasoning (e.g. ‘while the IUCN Red List treats this species as secure, this other global authority thinks its threatened by location disclosure so I’ve set geoprivacy to obscured and included a link to this flag in the description so we can discuss this decision…’).

If a taxon is in the IUCN Red List and has another global conservation status on iNaturalist, the automatic process will merge those conservation statuses giving the IUCN Red List status and geoprivacy precedence. If the non-IUCN global conservation status has a flag in the description, iNaturalist will treat this as an intentional deviation and preserve this description and give the geoprivacy of the non-IUCN global conservation status precedence in the merge.

Conclusions
Thanks for helping us with this pilot. We’ve had many requests to update the IUCN conservation statuses over the years. What we’ve outlined here is a bit clumsy, but we think it will allow us to maintain these statuses automatically while properly handling taxonomic concerns and status vs. geoprivacy concerns and without squashing intentional ‘deviations’ made by the iNat community.

12 Likes

How do you post the link in the c.s. description?

1 Like

when creating or editing the conservation status (which only curators can do) paste the flag URL in the description field along with some context (e.g. ‘Not threatened by location disclosure despite threatened status, See https://www.inaturalist.org/flags/542784’) if you are not a curator, you can still make the flag and then mention a curator to get help creating/editing the conservation status itself

1 Like

Is it possible to please have a report on species that change their status with the updates. (as has been done for the 933 amphibian taxa in this test run - although it would really help if a list of countries was included for each species)
Keeping tabs on the Red List status of 20,000 plant species in South Africa (with >60% endemic) is quite onerous - esp. with our national plant red list having over quarter of species of Conservation Concern .
And we prefer to keep Sensitive Species obscured, and other species open, irrespective of the National or IUCN Red List status… That is a lot of flags to post.

Is it possible to merely have a flag at generic level? Or do all taxa need individual flags for the automatic updates to work?

2 Likes

This is excellent!

There is a massive update due in a few months for amphibians, that will be a good test of the automated updates.

As it regards amphibians, there will be a missmatch with the Hyla/Dryophytes species. In this case, and for other species in this case, is there a way to manually assign them?

Thank you!

1 Like

To check I understand correctly – deviations from the name / IUCN taxonomy should NOT be cited as based on IUCN, but deviations from the geoprivacy settings ARE to be cited as IUCN?

and another daft question, why obscure extinct species? If I understand correctly, a species has to be far and long gone before it gets that status, and it is rare indeed for one to be rediscovered.

3 Likes

Having worked with the Red List assessment teams in trying to get species information updated to reflect current and accurate information I have to warn that Red List information for species is all over the place in terms of accuracy.

In some cases it’s fully up to date, in other cases it’s decades out of date and relying on information that was incorrect even at the time.

While the Red List forms a useful foundation to work from, it’s best to not rely on the Red List as the sole source of information concerning any given species.

5 Likes

tonyrebelo, we can make a report that describes what the automatic update does after it runs. Thats a good idea. It would be possible to include IUCN’s list of countries for each species in that report. We’d need to have flags a the conservation_status level to know which ones should have non-default position geoprivacy (e.g. open-and-threatened or obscured-and-secure)

amarzee, the Hyla /Dryophytes species (e.g. Hyla gratiosa<=>Dryophytes gratiosus) are good examples of one-to-one mappings (because Dryophytes gratiosus as a taxon name exists uniquely on the Hyla gratiosa taxon on iNat, we were able to match the Dryophytes gratiosus IUCN assessment with the Hyla gratiosa taxon )

trh_blue, I think I’m understanding you and that thats correct. Since we’re referencing the IUCN redlist as an external provider of assessments and want to be able to automatically maintain/update that external data, its important that we match assessments to iNat species either through exact matches or through the one-to-one system I described. There shouldn’t be ‘IUCN Red List’ authority assessments on other taxa not mapped to the IUCN Redlist accordingly (e.g. Agalychnis taylori), but someone can still create a conservation status for it with a non-IUCN Redlist authority. And while we need the status part (LC, NT, etc.) of the conservation status to match whats on IUCN, we’re saying its ok to have the geoprivacy be different from the expectation derived from that status (e.g. open-an-secure or obscured-and-threatened) as long as the reasoning for that ‘deviation’ is laid out in a flag linked to in the conservation assessment. Extinct species are an edge case as there likely aren’t obs of extinct species so its kind of moot, I did it simply because the logic is easier since iNat treats IUCN ranks as increasing integers where 0 => “NE”, 5 => “DD”, 10 => “LC”, 20 => “NT”, 30 => “VU”, 40 => “EN”, 50 => “CR”, 60 => “EW”, 70 => “EX” so it was easiest just to say >10 assumed obscured.

earthknight, I agree. I think the Red List is unique because it provides a global foundation as you describe. Its also rather loose in its philosophy (e.g. many species that are Least Concern in IUCN are treated as Vulnerable in other global assessments from authorities like the NatureServe G-ranks) so I think it’s most useful for parts of the world where we don’t have good regional assessments. But when there are good regional assessments (e.g. Natureserve S and N ranks in the US and Canada, and various national systems that get on-boarded as part of the iNaturalist Network, such as Nom-059 in Mexico) those often add a more refined/recent level of assessment for at least the portion of the species range falling within those nations (e.g. if a species has a global IUCN LC ‘open’ status but is endemic to Mexico and has a Mexico Nom-059 VU 'obscured status, all the observations will be obscured as a result of the national status). But if a species has no regional/national assessment but is in IUCN, we still need the IUCN assessment to be included as a conservation_status and have the conservation_status.status match the published IUCN status. However, as described in the post above the geoprivacy doesn’t have to match the expected geoprivacy from the IUCN status as long as there’s a flag linked to in the description that describes the reasoning which is a great place to include other sources of information without having multiple conflicting global conservation statuses on the same species.

6 Likes

Again this should depend. If an extinct species would be rare and collectable if discovered, then it should best be obscured.
If the species is obscure and unexciting, then opening it would be best.

Many species have long-lived seed banks, and specific germination cues. Mimetes stokoei was declared extinct after it failed to emerge in several successive fires and over 40 years. And then after a hot fire, it reappeared :see https://www.inaturalist.org/observations/10782952. This should probably be obscured to protect if from gawpers.
But others are post-fire species and cryptic, and their rediscovery is probably just because someone looked: https://www.inaturalist.org/observations/11071901 - only the most ardent botanist would bother looking for this. And without knowing where to go (and knowing that you only had 1 year after a fire to do so) it could well take you another 75 years to find it again.

7 Likes

Will the IUCN ranges (the light pink blobs on the maps) be updated as part of this process?

We’re treating them as a separate issue to the statuses. We could update them, but we need to come up with a different process, so lets discuss on a separate forum thread. I can make that now if you’d like

Thanks for the reply. Haven’t noticed masses of ranges needing updates. It was a concern about ranges modified after splits being overwritten, if there is no plan to update then it is not an issue.

OK I took a shot at hashing out a somewhat analogous procedure for IUCN red list derived taxon ranges here https://forum.inaturalist.org/t/updating-iucn-red-list-rangemaps/25817

4 Likes

I am guessing that we will end up with some orphaned IUCN statuses in iNaturalist when they can’t be found by matching or one-to-one names. Are these going to be looked for and dealt with in some way?

I’m interested and excited to see how this goes. Lots of caveats but I do envision a future where iNaturalist observations are also feeding information into Red List assessments more seamlessly, since I do think that in many cases there are ranges verified (and correct) on here that are more accurate than Red List ranges for many reasons. As a conservationist I echo the points regarding data issues with the list however it is a global basis for decision-making and it ties into other important resources like the STAR metric. A good relationship with the Red List could be crucial for maximizing iNaturalist’s impact in the future.

2 Likes

Interesting to see how it will improve the current tags on reptiles and amphibians.
I am not sure, as the question is about threats, how "national " IUCN Redlist could be taken into account. As an example the IUCN French committee has been doing many evaluations for 10-15 years and they are recognized and supported by the government and the French taxonomic authority (MNHN). The categories can be different from global IUCN Redlist. Not to obscure the location of some of highly threatened taxa (classified LC on global Redlist) could be a threat. This may happen in French oversea’s territories for overcollected orchids or other plants or reptiles, in mainland for species of very large distribution but very rare here. These national redlist are published and regurlarly updated.

1 Like

Ah: this has now been implemented.

Please. When there are these dates (deadlines), can one please get a pre-implementation warning of at least one week. Things do slip through the cracks and deadlines do sneak up on one. A reminder or warning is essential please!

1 Like

Do the deviations have to be entered one-by-one, or can we provide a list of species and the reasons for deviation in a spreadsheet, that can be bulk uploaded, or alternatively handled via an api?

please create the flags one by one

That wont be feasible when we move to plants. We are talking over a thousand deviation flags for plants in South Africa!

1 Like