Sensitive species data

[Bill Parker]

Good day all,

First off, this is a great initiative and I hope to contribute as much as possible in the future and use some of this information to provide direction for field research. With my field research team we will be researching and confirming rare and sensitive species observations here in the Eastern Aegean. 

I was hoping that for sensitive species for example Chaemeleo chaemeleon, targeted by smugglers, we may be able to hide specific details on their locations so we do not aid in the targeting and collection of these. 

Perhaps the locations can be put in the sea, as someone has already done, and the true location available only to those who we can be sure are not smugglers and are using the knowledge for good.

Any direction in this matter would be much appreciated.

Happy herping!

Yours in conservation,

Bill

William Parker

Terrestrial Team Supervisor

B.Sc. Natural Resources Management

Wildlife and Fisheries

Archipelagos Institute of Marine Conservation

www.archipelago.gr

parkerwk...@gmail.com

Marine Research Base:

P.O. Box 42, Pythagorio, Samos 83103, Greece

Telephone: +30 22730 61191 Fax: +30 22730 37533

                                                                                

           

     

[Charlie Hohn]

There is already a mechanism to do that, if you come across species that should be obscured but are not, it would be good to contact admin about it. 

You should never intentionally create wrong data such as mapping something where you know it is not. Instead, use the built in obscuring mechanism to make the location obscured or private. (This may cause some things to display in the ocean, as the exact location is obscured, but would have a round rather than pointy icon indicating this). Rare species at risk of collection or poaching will be obscured automatically once they are in the system as such. 

Go to the species page https://www.inaturalist.org/taxa/32864-Chamaeleo-chamaeleon and click on status to see which conservation statuses have been entered. If you see that one is missing it should be added. Right now the species shows up as least concern or no status in any of the places I see, but it may have a conservation status elsewhere. If there is a population with no status but still has a credible threat it might still be possible to obscure in that area. For instance I see you are in Greece and the info on the species page says " In Greece the species is now only found on the island of Samos, having disappeared from the other islands in recent years. Conservation efforts on Samos are being led by the institute of Marine Biology Archipelagos to preserve the remaining populations, but the species is not currently evaluated under the IUCN redlist. [1]". So perhaps this species should be obscured in greece even though there isn't yet a status defined. I will defer to other curators/admins on this since i don't really know much about chameleons.

[Chris Cheatle]

Just a note if you use it since it catches users off-guard until they understand it. If you mark a sighting as Obscured or Private, you will still see the exact location and details if you open it back up. But it is hidden to everyone else. For instance if you try to open this observation of mine, you wont see the actual information : http://inaturalist.ca/observations/6437780 which is hidden as it is critically endangered in my home province, or even this one which I have chosen to obscure even though it has no conservation status (the location formerly had open access when I observed it, but the owners have since closed that access, so I have hidden it) : http://inaturalist.ca/observations/5352232

[tony rebelo]

The problem is that it is not "hidden" - it is incorrectly placed in  a different locality.  This catches up lots of users who are unaware of this fact and start using the wrong data for analyses and summaries.  The number of people who are fooled (and who should no better than to use data without checking the terms and conditions)  is rather embarrassing. 

I do not approve of iNat giving Wrong localities just to make it maps look prettier.  It is no better than users doing the same!

[Tony Iwane]

Tony, can you explain what you mean by "just to make it maps look prettier"? You're correct that the location of an obscured observation is displayed on a random point within a 0.2 x 0.2 degree area, but its status as obscured is clearly marked and is defined in the Map Legend. It's on the map not to make it look prettier but to show the observation's existence. Should the observation not appear at all? I'm not asking to start an argument, I'm just curious as to how you feel this should be handled. If there's a better method, we're always interested. 

Tony

[John B]

I am not sure that the obscure feature is really good enough for species that do not move much, plants for example.  As an exercise, I looked for an observation of a locally rare orchid that a user uploaded to iNat.  The location was obscured, but looking at the un-obscured observations immediately before and after the rare plant, I was able to see the exact location of the observer a couple minutes before and after the rare plant observation.  It would not take much effort to go in the field and locate the orchid.  Ouch.

[Charlie Hohn]

I thought that the date of observation and upload were obscured as well, there were plans to do so. if it isn't implemented yet hopefully it is soon.

[Upupa epops]

Even so, can't you still go and look at the list of observations by that user, find the obscured observation, and look at the observations before and after that observation on the list?

[Charlie Hohn]

The obscuring is never going to be perfect. There are some things people should just not post. 

[Reuven Martin]

I'd like time to be automatically obscured, but I think automatically obscuring date would do more harm than good. The option to manually obscure by date (i.e. to only show month and year) would be nice though.

90% of the stuff that is obscured isn't a huge deal. But I've uploaded a few truly sensitive sightings, and in those cases:

1. Modified the date (to the 15th of the same month)
   
2. Uploaded not in sequence with the associated observations
3. Given a location that is well removed from the actual sighting, but with a very large accuracy circle that encompasses it (I don't feel comfortable putting the real coordinates online and leaving iNat's obscuring as the single line of defense)

As Charlie say, you can never stop someone from inadvertently giving away too much information. iNat does better than Flickr or Facebook!

[Chris Cheatle]

Maybe one tool (and has been pointed out there is no magic solution) is for private observations to mean just that - private. Visible only to the submitter, not findable via searches, extracts etc. If there is concern about sharing or research, create an option on the user account side, with the default set to enabling the privitization, for all profiles to allow it to be shared in the way private or obscured ones are now. Similar to tools like Flickr, where I can hide completely (and do so), or elect only to make visible to people I allow.

[Charlie Hohn]

I actually really like the idea of 'secret' observations because it would expand the sue of iNat for conservation organizations and such who are not able to share things for whatever reason. Last time it came up the devs were not into the idea at all though, not sure if that has changed.

[tony rebelo]

"Given a location that is well removed from the actual sighting, but with a very large accuracy circle "

That is unethical and you can expect litigation from a properly owner on whose land you inadvertently placed the circle should it result in any restrictions on development.

Similarly, you could result in a useless area being designated for conservation at the expense of the area in which the orchid occurs, or alternatively as the site containing the orchid being considered as not so important as there are "other" sites available nearby.   

Not the thing to do!

[tony rebelo]

I would oppose that.  The whole idea of a Citizen Science site is that the data are used for science, for conservation planning - including zonation and priority conservation sites - and vetting Environmental Impact Assessments and informing development proposals.

If the data are not available for this, then what is its value.

As for "Privating" a site: for heavens sake - dont post it anywhere: in fact, dont even take a picture.  Why do half a job if you dont want people to know about it?

[Charlie Hohn]

no, you can't get sued for having a point on a map that clearly states it's something in the general area. People sometime bring that up but i'e never heard of any cases where it happened. If you trespass on someone's land and document it on inaturalist, you are being dumb and may end up getting in trouble for it, but that isn't what happened here.
I agree that the 'within circle' thing isn't a great proxy for obscuring but to be honest I've done it too (not sticking the point on private land).

[Chris Cheatle]

Respectfully, this feels like overkill. No conservation, or valuation decision is going to be made based on a single unverified (or even a few) observation on some, with respect to the great community here, random website. The burden of evidence and regulatory framework to do either is far above this as a decision-making standard. And any case that was brought would be tossed in 30 seconds once it was evident the "evidence" was an artificial buffer.

In response to : "

As for "Privating" a site: for heavens sake - dont post it anywhere: in fact, dont even take a picture.  Why do half a job if you dont want people to know about it?"

Why do it, because this thread and others are centred around how data that can be used for conservation etc can be collected, but minimizing the risk of enabling poachers, collectors and unethical behaviour. It's a circular argument to say data needs to be collected to prioritize conservation and then say dont collect the data if you are concerned about issues it raises. What the discussion should be is how those issues can be addressed.

The discussion is that right now the "obscured" and "private" settings leave loopholes, which I wont repeat here (including at least 1 I can think of that even the solutions discussed would leave open) enable folks looking to exploit the data for nefarious purposes, and as such are there ways to resolve those issues.

If the only solution is to not collect it so be it, but I'd rather have a discussion if there are ways the loopholes can be minimized, yet still allow the valuable information to be collected.

[Charlie Hohn]

I agree. I know a couple of loopholes too that i won't post here for obvious reasons. The system will never be perfect but I agree it's good to work to make it better! It's already gotten better and as a consequence i post some obscured things I wouldn't have posted in the past.

[Tony Iwane]

Reuven, I think if you're worried enough to do all those things, it's probably best to not post those observations on iNaturalist at all.  At its core iNat is about sharing things, and if you're falsifying data (the date, in this case) to do so, this probably isn't the platform for the observation.

The team has had some lengthy discussions about how to improve obscuration (eg time/date), but doing that would open up a big can of worms as far as database/programming/etc so it's not a trivial thing. 

Tony

[Reuven Martin]

Tony - I see your point although I don't really feel like I'm putting in an unreasonable effort - I see it as just doing the bare minimum to prevent someone from figuring out the location with minimal effort.
I guess the solution given iNat's current architecture would be to leave the date off altogether, but that relegates the observation to "unverifiable" where nobody will ever see it. Which doesn't seem necessary when the only piece of information I'm not able to share, beyond the usual obscuring, is the exact date. But I guess I can live with that.

[tony rebelo]

"Respectfully, this feels like overkill. No conservation, or valuation decision is going to be made based on a single unverified (or even a few) observation on some, "

Maybe not in the USA, but spare a thought for us countries with serious biodiversity challenges, where hundreds of species are known from less than a handful of records, where localities may be remote or simply inaccessible, and species occur in very small ranges.  Getting to verify these is often not a trivial exercise (even when the data are not "false").  Consequently, conservation decisions ARE made using exactly these data: there is no choice. You either 'pretend' that the species does not exist by discarding unverified data, or presume that people collecting the locality data have some competence and integrity and accept it.

In a decade or two when resources allow verification it may be too late (one way or the other) ...

[Colin Meurk]

further to 'how to obscure sensitive record details without resorting to misleading deception' - as Tony R has alluded to, or leaving date off as someone suggested; another more ethical solution perhaps is to counter 'deception' with transparency.  in other words 'fess up to your deceptions.  So, as well as using the built in 'obscuring' function, if one tries to hide location details by not sandwiching the record between other unobscured ones, placing the record away from the actual site with a large ring of inaccuracy, and changing the date to say beginning, middle or end of month - just say that you have done this in the 'comments/details' - not specifically of course, but that you have changed the time and place to protect the innocent.  this way people are not misled into believing it is exactly where and when stated, and of course if they have a legitimate reason for having the information they can always message the observer for the hidden information.

[bouteloua]

It's especially important to be transparent about your obscuration when the location is both obscured with a large ring of inaccuracy *and* via the geoprivacy settings.

For example, I encountered observations that appeared to occur south of Chicago, where no specimens are known, and would represent significant findings. But it turns out that it was just "further obscured" by the user (illustrated in the attached).

cassi

[Upupa epops]

I've seen quite a few cases where an observation of a sensitive species is deliberately misplaced (usually at the location where Google places the label for the nearest city, so downtown where the organism wouldn't actually be), and a large ring of accuracy is drawn, but the location is not obscured. As a result there is a point on the species map that seems precise until you click on it.

[Charlie Hohn]

Well, the problem of things mapped with huge uncertainty circles isn't limited to sensitive species. Any species that is mapped so broadly causes the same problems, and it's pretty common. This is why I wish there were a better way on the species maps to display that something had a very large uncertainty buffer. I think if it's super huge they don't show up on the map at all. Ideal would be that after you zoom in a bit anything over a few hundred meters uncertainty disappears from the map entirely, obscured or not.

[tony rebelo]

"they can always message the observer for the hidden information"

What?: the average user lasts a few years on this site!.  How on earth is someone  going to ask them in 300 years time?  

Clearly many users how no conception of what accurate data means, or how it is used.  Suffice it to say that for southern Africa there are lists of collectors goling back to the 1800s whose localities are not to be trusted.  If it is eternal notoriety that you are striving for, this is one way of achieving it.

[Colin Meurk]

Well tony you raise an important issue which I’ve tried to get discussed before – the simple fact that even the youngest of us are not immortal.  There needs to be some policy generated around what to do with records from those that have passed, or are indisposed, or heaven forbid end up suffering irrational dementure.  We are all prone to this at some time.  there should be a requirement when signing up, that data is not corrupted or taken down inadvertently and that a representative governance group can deliberate on and reveal to legitimate requests the true geolocational data.  It’s not a pleasant topic, but hey I know I’m approaching the zone myself.

Tony, I don’t think you read or quoted my whole statement – which was really about transparency.  If there are valid reasons for obscuring a record then it is important that people at least know that the geolocation etc is innacurate.  But this is why obscuring through the existing function is preferable to making ones’ own decision about deliberately moving the flag from the real location with a large encompassing inaccuracy.  Because then at least the accurate geolocation is archived in the system and it only requires some suitably authorised persons to release that information to appropriate people.  But going back to what I said – there needs to be developed a policy around this, and an acceptable mechanism for administering it.

c

                                                                                

           

     

 

--
You received this message because you are subscribed to a topic in the Google Groups "iNaturalist" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/inaturalist/azZvFgv7GQM/unsubscribe.
To unsubscribe from this group and all its topics, send an email to inaturalist...@googlegroups.com.
To post to this group, send email to inatu...@googlegroups.com.
Visit this group at https://groups.google.com/group/inaturalist.
For more options, visit https://groups.google.com/d/optout.

---

Please consider the environment before printing this email
Warning: This electronic message together with any attachments is confidential. If you receive it in error: (i) you must not read, use, disclose, copy or retain it; (ii) please contact the sender immediately by reply email and then delete the emails.
The views expressed in this email may not be those of Landcare Research New Zealand Limited. http://www.landcareresearch.co.nz

[Jane Widness]

Can we do anything about the issue Cassi brought up?  That is, an obscuring rectangle on an observation with a wide uncertainty in location can be very misleading.  Like maybe if the uncertainty circle is larger than 400 square km (so larger than the obscuring rectangle would be), then don't use the obscuring rectangle as the user has effectively obscured it with the large uncertainty anyway.  Or maybe something else to indicate a large potential inaccuracy in location?

[Charlie Hohn]

i'd argue that something with an uncertainty that large shouldn't show up on range maps at all, in fact i'd probably place the cutoff at a few hundred meters and otherwise exclude from the range maps (would still be searchable and such). It's really annoying to have stuff mapped in the wrong place even if the issue is precision not accuracy. I think there may already be some sort of filter to do this though? I also think the obscured observations should look different at a further zoomed out scale, and perhaps should display differently on range maps. Ideally, they would only exist as the square itself, darker with more observations, kinda like eBird, and that would overlay along with the exact points of observations that were mapped with greater than 1 km accuracy or whatever. But maybe that's hard to do. And there are lots of obs with no accuracy recorded at all, what do you do with those? Usually they aren't too far off but sometimes...

On Saturday, February 24, 2018 at 3:44:56 PM UTC-5, bouteloua wrote:

[cassi saari]

The screenshots from my message on February 24th were from the individual observation's page, not the taxon page or in a filtered observation search.

cassi

--

You received this message because you are subscribed to a topic in the Google Groups "iNaturalist" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/inaturalist/azZvFgv7GQM/unsubscribe.

To unsubscribe from this group and all its topics, send an email to inaturalist+unsubscribe@googlegroups.com.

[Jane Widness]

Yes, sorry I was unclear.  I meant I would like a better indication of uncertainty *on the individual observation page*.

On Wednesday, March 28, 2018 at 3:07:57 PM UTC-4, bouteloua wrote:

The screenshots from my message on February 24th were from the individual observation's page, not the taxon page or in a filtered observation search.

cassi

On Wed, Mar 28, 2018 at 2:05 PM, Charlie Hohn <naturalis...@gmail.com> wrote:

i'd argue that something with an uncertainty that large shouldn't show up on range maps at all, in fact i'd probably place the cutoff at a few hundred meters and otherwise exclude from the range maps (would still be searchable and such). It's really annoying to have stuff mapped in the wrong place even if the issue is precision not accuracy. I think there may already be some sort of filter to do this though? I also think the obscured observations should look different at a further zoomed out scale, and perhaps should display differently on range maps. Ideally, they would only exist as the square itself, darker with more observations, kinda like eBird, and that would overlay along with the exact points of observations that were mapped with greater than 1 km accuracy or whatever. But maybe that's hard to do. And there are lots of obs with no accuracy recorded at all, what do you do with those? Usually they aren't too far off but sometimes...

On Saturday, February 24, 2018 at 3:44:56 PM UTC-5, bouteloua wrote:

It's especially important to be transparent about your obscuration when the location is both obscured with a large ring of inaccuracy *and* via the geoprivacy settings.

For example, I encountered observations that appeared to occur south of Chicago, where no specimens are known, and would represent significant findings. But it turns out that it was just "further obscured" by the user (illustrated in the attached).

cassi

--
You received this message because you are subscribed to a topic in the Google Groups "iNaturalist" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/inaturalist/azZvFgv7GQM/unsubscribe.

To unsubscribe from this group and all its topics, send an email to inaturalist...@googlegroups.com.