This is the one part I have some concerns with. At some point this backlog of missing data needs to be dealt with. My reading of the curator guide (if someone from the site wants to give an official rebuttal etc great) is this:
the default state of a threatened species is obscured.
if it can be demonstrated and documented that doing so provides benefits and/or no risks then this may be reset to open for a specific species
the onus of proof is to demonstrate why a threatened species should be open, not the reverse.
What message does it send if we know a species is threatened and and not documented in our data and choose not to acknowledge it and at least review that ?
yes, but does âthreatenedâ apply to edge of range species that happen to get state/province status? Why obscure things on the edge of somethingâs range, when thereâs no collection risk, but do so only when the range happens to intersect with only a small part of an arbitrary political boundary? That doesnât make sense. If youâre just talking about globally threatened species, thatâs different.
I donât think it is an issue if 90 percent of the listed species end up as open for these or any other reasons. I just think that the process to review and document that should be a requirement.
to be clear we are talking about things like a super common oak species that clips a state line in its range here, not a rare orchid only found in one bog.
Please donât go on an obscuring binge. I understand your frustration with this system and have had frustrations too. But the lack of documentation means you may be undoing a lot of other peopleâs work anyway, and obscuring edge of range common species is really harmful for conservation planning and climate change tracking, for instance.
In the very least please leave the Vermont auto obscuring alone. Our NHI has already gone through and recommended which species should be obscured and if you change it you will undo all of that work. I understand you donât like the current system, but this is something everyone on iNat relies on to work, so I donât think unilateral action is appropriate here, at least not without staff go-ahead.
The concern I have is that there is no evidence any of these have been previously reviewed and determined as appropriate for being open. The site has noted above they have never made an effort to integrate these data for NatureServe and this is certainly true for ex North America national Red Lists etc.
Yes the âhotspotâ areas of most heavy use for the site like parts of the US, Canada, NZ, ZA have probably done decent reviews, but large parts of the world have not.
When or how would you propose that this backlog be dealt with and the species lists get documented and reviews done as to which are appropriate for obscured and which for open?
Were i still living in Europe and knew this situation I dont think I would be a site user because Iâd be concerned the site is not taking the protection of sensitive species seriously. For example I recently finished updating part of a national red list for a European list where they maintain a list of specific species that by law it is illegal to disclose the locations of. That should not have been pending still in 2020 on the site.
No one wants to see species that dont need it get obscured. But when or how do we tackle the ones that need it ? Waiting for a flag to get raised is not a very proactive solution and anyone who would abuse the current situation is certainly not going to put one in.
@carrieseltzer
Do you have any suggestions for how users might validate that the list of names they provide matches exactly to an entry in the iNat taxonomy list ?
For example, it is an easy 2 minutes of excel work to do the matching, but if you try and download a large checklist (to do this I tried to download the Europe checklist), it just times out.
Iâm happy to provide a test case for you, I have the Danish national red list ready to go. It has 1,798 species listed as NT or worse. Iâm not really interested in investing the time to manually check all those one at a time. Please note that by no means do all 1,798 need to be obscured, but having lived there, I can tell you there certainly are ones that should be.
If the import encounters a name not in the database will it cause an error or be ignored.
Iâd certainly be interested in broader community feedback about how to clear this backlog up. For this test case, we could :
load the statuses for all 1,798 as open. I donât really see the point to this, it may be nice to have documented, but if someone wants to see what is on the Danish Red List, they can go to the Danish Red List
load them all as initially obscured, knowing full well there are things here that do not need to be obscured
dont do it at all, just revert back to the wait for someone to flag it approach, in which case, realistically we are saying it wont get done in most places around the world.
Iâm not a curator, so maybe this is a question with an obvious answer: is it impossible to update the conservation statuses in bulk without changing the current geoprivacy settings? It looks as though all the taxa flags you referenced above were requests to update the conservation status, without an associated request to change the taxon geoprivacy, and importing e.g. the Danish Red List in bulk should not be done if it erases all the current geoprivacy settings. So are the two things inextricably linked? (Logically, they shouldnât be.)
Right now only site staff can do bulk updates. All others including curators are limited to one at a time manual updates.
Currently it seems (at least using the recent case of the updates done by the site to the IUCN Birds) it is an overwrite.
But in this specific case, there is no data to overwrite, the entries would be specific to Denmark (or whatever other nation/state etc was done) are not in the database at all.
I cant speak for any other individual adding flags about missing conservation statuses, and their motivation, but they do seem to be selective in what they are flagging, otherwise why not just request that all statues for Arizona etc get added ?
this is true, but going through and undoing this work isnât a good idea just because you donât know it has been done.
I propose if itâs globally threatened and youâre not sure go ahead and obscure it but do not obscure state threatened anything just based on its status (if there are other collection risk factor yes you should do it).
Please at least wait to hear back from staff before obscuring any large group of observations. @tiwane@carrieseltzer . You could do a lot of damage and i donât see any reason why it needs to be done without making sure first. Most likely the iNat admins have more info on what places might need changes. If you want to import the status without obscuring that seems fine as long as it is from a reputable source.
I might refine this to say, without overwriting existing geoprivacy status. Assume that current geoprivacy (or lack thereof) is intentional, until someone flags it requesting a change. (And such flag can come from the implementing curator if/when there is obvious urgent need for a change.)
I have done this by just creating a temporary personal list using the batch upload option. The description field is populated with an arbitrary unique value prior to upload, for later comparison with the original batch. Non-matching (or ambiguously matching) names are immediately flagged on-screen in the resulting upload, with options for finding the equivalent iNat name. Then I can re-download the resulting âcross-walkedâ list and compare to the original in excel or equivalent data tool. The temporary personal list can be deleted if no longer needed.
If one needs to check more than 1000 names, they would just have to be divided into separate batches to test.
Trying to catch up on all the good comments and questions from the last day, roughly in order.
To the best of my knowledge, the primary purpose of the âIUCN equivalentâ field in the template is to determine the color of the flag and use a controlled vocabulary that can be translated like âvulnerableâ, âcritically imperiledâ or âpossibly extinctâ. For example, the purple pitcher plant has several different examples of associated âlevelsâ with different colors of flags.
I agree that edge-of-range species might not be labeled clearly under this system and that local rarity (e.g. S1 status) should not strictly lead to obscured taxon geoprivacy.
On the topic of keeping track of changes to conservation statuses:
I also very much want a better system for tracking these changes and Iâve done some thinking on it, but I donât expect itâs a feature weâll be able to tackle soon. There are a couple of related feature requests (here and here).
As @jdmore says, flags are the best tool we currently have for recording some context for these decisions, though definitely imperfect since flags are not a complete history.
A couple of times a year we do export all of the conservation statuses in iNaturalist. The one Iâm looking at has 157,118. It is not very easy to work with on its own because it contains taxon ids rather than taxon names, but I am happy to share it by request if it would be useful because it doesnât contain any sensitive information.
Revisions, clarifications, and additional guidance in the curator guide around conservation statuses and taxon geoprivacy are definitely needed.
Based especially on what we went through last year with Nature Serve Canada (for the few of you reading this who werenât in the thick of that, itâs in this thread), I think weâve evolved to a place that is more nuanced than currently articulated in the curatorâs guide. I would paraphrase it as âas open as possible but as closed as necessaryâ.
In the best case scenario, there will be not only a list of threatened/endangered/protected species, but also a list of âspecies subject to persecution or harmâ or sometimes âsensitive speciesâ that reflects local/national consensus of the species most at risk from sharing locations of populations or individuals, not just their threat due to all causes. If such a list exists, that list should be used to guide decisions about taxon geoprivacy. The remaining threatened/endangered/protected species that are not on the âpersecution or harmâ list can still be included for a bulk upload of conservation statuses with the taxon geoprivacy set to open.
For places that do not have a list specifically for âspecies subject to persecution or harmâ, then maybe the best approach is to obscure everything ânear threatenedâ or worse (or perhaps a higher threshold?), expecting that over time the community will then help curate and open up the taxon geoprivacy for species where the cost of hiding observations will likely outweigh the benefit. Iâm interested to hear what others think about how to best approach places where we have little if any conservation statuses.
It is possible to update the conservation status without changing taxon geoprivacy. However, it is more complicated to do it that way and I donât think it is a great idea in our current system of bulk updates done by staff.
You could use the API to look for exact matches in your list. OrâŚ
That is a very clever approach, @jdmore. I almost think you should make a new thread specifically about your method.
Thanks for the wikipedia reference. Iâll just note that it only compares IUCN and NatureServe G ranks, not S ranks, and this is correct, since G ranks are also global and range-wide in scope. NatureServe S ranks are a totally different beast, referring only to local political jurisdictions, not to overall species ranges. These are the ones that I would assert do not have any IUCN equivalents, and should not be the sole basis for automatic obscuration.
Thinking about this there is actually a quick and easy way - use the batch load tool already there to try and load the list to the relevant place checklist.
This will result in:
species already on list in which case it is obviously ok
species not on list but gets added which is also obviously good, plus the side benefit of improving the checklist data
species can not be found which are the ones you need to look at
Fyi I have had some offline discussions with Carrie, we have a proposal for how to deal with the backlog to share for comment. I will write it up later today when not on my mobile
As mentioned above, I have been engaged in some direct communication with @carrieseltzer about how we can move forward on trying to deal with the backlog of national and regional conservation statuses that exist. We would like to propose the following guidelines be added to the curator guide, specifically for adding net new statuses (ie only for cases where the data has apparently not been entered before). These guidelines would not apply to existing entries or call for any changes to data, conservation statuses or obscuring already done.
We would propose, for new entries:
any species regionally or nationally classified as Critically Endangered (CR) or Endangered (EN) be initially loaded as obscured
any species regionally or nationally classified as Vulnerable (VU), Near Threatened (NT), Least Concern (LC) or Data Deficient (DD) be initially loaded as open
that in all cases, the existing process by which a species may be reviewed to change its obscuring still applies, if there is a justifiable argument that there are benefits and/or no risk to unobscuring a species, that remains acceptable, likewise there may be arguments to obscure a species with a less serious status
if a national or regional list happens to use a different terminology than the IUCN codes, the curator should use their best judgement to map the ones used to an IUCN equivalent and then apply the above
the curator doing the load should not unilaterally decide there is an acceptable argument to change the obscuring and simply change it. Rather, in keeping with the current process for existing data, they should initiate a community review/discussion via a flag on the relevant species page.
I would add that if something is obscured without the status loaded yet please donât uh obscure it without community discussion. I just donât want any past work to get undone.
Can you clarify what you mean by âin keeping with the current process for existing dataâ? There is no policy to âinitiate a community reviewâ. There is a process to initiate a non-community review for Canadian provinces other than Quebec, and the only somewhat related thing for places other than Canada in the Curator Guide is âbe prepared to support why you are doing thisâ, as in, if someone asks you, itâs helpful if you have a justification, but thereâs no requirement to have a discussion first. I know youâve proposed a more formal decision process in the past, but itâs not a current policy.
We only need discussions for the non-obvious situations. There are still thousands of obscure plants, fungi, arthropods, and other taxa that donât need to be obscured, and in most of these cases the decision to unobscure is pretty easy to make.
Isnât that what a flag is? An opportunity for the community to review and discuss the question.
If we say that there are many obscure species that dont need obscuring, then we may in my mind as well just eliminate any attempt to have a standard process and just say obscuring will be left up to the unilateral decision making of any curator who chooses to do so, using whatever individual decision making process they want or feel is justifiable. And any other curator can undo that for the exact same reason.