Criteria for default geo-privacy

I’ve read several threads on the topic of geo-privacy, (the most pertinent one closed, I think, last September) but, so far as I can tell, this issue has not been resolved. I offer a perspective here in reference to rare plants in New Jersey. Much of this may have broader implications. I think my perspective is close to that of cmcheatle.

All taxa that are listed as “Endangered Plant Species” and “Plant Species of Concern” by the NJ Natural Heritage Program are on that list because a professional judgment has been made about a level of imperilment for the respective species within the state of New Jersey. (Taxa ranked SX, presumed to be extirpated, may be considered in a separate category.)

The “S” (state) rank, for taxa on this list, is a very pertinent level of imperilment. That is to say, although the “N” (national) rank and the “G” (global) ranks are certainly important considerations, from the standpoint of New Jersey protecting its biodiversity, we need to pay attention to and make decisions taking the “S” rank into consideration. Many taxa that are rare in NJ are rare only because they are at the edge of their range, but this category of rarity is by no means one to be dismissed or taken lightly. (Ecological arguments are available.)

An “S” rank is an indication of a degree of imperilment, not merely an indication of rarity. In other words, if a taxon has a rank that qualifies it to be on the list, it’s because a level of threat has been determined. It’s possible that some of those ranks need to be revised, but one would need to provide evidence for such an argument, not merely an opinion.

Publishing location data for rare plants has several potential threats. For example: (1) Over-collection of entire plants or parts, including propagules, by unscrupulous, ignorant, or careless collectors. (2) Trampling by careless visitors. (3) Deposition of disease vectors by careless visitors. (4) Deliberate eradication by developers who don’t want the hassle of T&E species issues. (5) Creation of trails that may encourage increased deer browsing. Probably other direct threats could be added.

Additionally, our imperiled species are typically found in rather well-preserved ecological communities (comparatively speaking), many of which are imperiled in their own right, and many of which harbor other imperiled taxa. People visiting a site to see, collect data on, etc. one rare species may be totally unaware of the impacts they are having on the ecological community itself and the other rare species present.

This is not to say that these threats apply to every rare plant. It’s entirely possible that some, perhaps many of the taxa on the Heritage list, would NOT face increased threat if their exact locations are published.

However, we don’t have a ready means of evaluating, for each individual taxon, its degree of susceptibility to these threats. That being the case, it would be hard to argue that it is generally safe to publish exact locations, or that we should only obscure when we have demonstrable proof, on a case by case basis, that obscuring is good.

All of this leads me to recommend that we adopt a mechanism, as a default, to obscure, rather than publish the exact locations of imperiled species, as determined by the “S” rank.

3 Likes

It is my understanding that this already happens. Have you read this section of the curator guide?
https://www.inaturalist.org/pages/curator+guide#geoprivacy

You can flag any organism you think should have a different geoprivacy default.

6 Likes

Thanks for the reply!

Yes, I have read that carefully. The criteria listed there are substantially different from what I’m recommending.

“Out of an abundance of caution, iNaturalist initially obscured the locations of all species with an IUCN equivalent status of “Near Threatened” or “worse”. Over time, many of these taxon geo-privacy statuses have been updated from “obscured” to “open” when advisable. In situations where a species is thought to be in little danger from the public’s knowledge of its location, curators are encouraged to open a flag for discussion regarding the taxon geo-privacy, and setting (or changing) the geo-privacy status to “open” may be recommended.”

IUCN status, to the best of my understanding, does NOT take into consideration the degree of imperilment within any given state.

What I’m recommending is that the geo-privacy be tailored state by state, and that the default, in any given state, be based the “S” rank in that state. Wherever it can be persuasively argued that geo-privacy is irrelevant to conservation of the taxon, then, allow taxa to be unobscured.

2 Likes

While we use official IUCN designations, many of the statuses come from NatureServe and are listed as the IUCN equivalent. Edit: Nevermind not true. NatureServe does include state designations.

2 Likes

Sorry, but I’m not sure I understand. As I understand, IUCN designations may very well be equivalent to (on some level) or based on NatureServe, but NatureServe provides Global, National, and sub-National ranks. To the extent that IUCN designations reflect NatureServe ranks, if I’m not mistaken, they do not necessarily reflect sub-National (for us, state) ranks. Thus IUCN designations cannot be relied upon to reflect the level of imperilment in any particular state.

2 Likes

In north America, most, but not all statuses entered and obscuring is based on NatureServe provincial/state rankings. It is not 100 percent complete as it is taxing to extract and update, but a very good chunk are entered.

Despite what is written in the curator guide, there is no formal practice or procedure about which taxa are obscured. Any curator can change (in any direction) any taxa they wish.

4 Likes

I meant that NatureServe uses numbers (1–5) that are correlated with statuses on IUCN (Critically Endangered, Endangered, etc.). Edit: Nevermind not true. NatureServe does use state designations, so the IUCN statuses can be used even if they do not match the same exact political geography. Edit: Not necessary when NatureServe has its own labels.

1 Like

Thanks! I was hoping to hear from you. I guess it will take me some time to get acquainted with the INat jargon. Is there a description available to us “plain folk” as to how this process plays out?

1 Like

I agree that

And continue to disagree that

Conservation statuses have everything to do with the geography they encompass, whether or not we happen to assign the same labels (like “critically imperiled”) to them for different geographies. I continue to argue that S-ranks and N-ranks have no appropriate IUCN equivalent other than “Not Evaluated (NE).” In any case, existence of an IUCN equivalent status need bear no relevance to the decision to obscure or not obscure a taxon in any particular jurisdiction. That decision can stand on its own merits.

As an(other?) employee of a NatureServe Network member, I agree with this in principle. In practical terms, though, many of the existing ranks are still primarily rarity-based, because there is very little documentation of threats for the vast majority of species tracked by the Network – and/or, very limited capacity for plugging known threats data into revised rank calculations.

Even when we do have good threats data, the specific threats may have nothing to do with public knowledge of locations (climate change, for example), or may actually be increased by obscuration (habitat conversion/development, for example). S2 may mean very different things for different taxa, regarding the need to obscure locations. So, in the end, I don’t think we’ll be able to avoid taking obscuration decisions case by case in each jurisdiction.

That said, I do also agree that any default taxon geoprivacy should

Except for “special-interest” groups like cacti, orchids, carnivorous plants, herps, some butterflies, etc., I am generally much less concerned for S3 taxa in large, low-population-density states like Nevada, than in small high-density states like New Jersey. Ideally in my opinion, we would avoid “default” obscuration as much as possible, and make decisions taxon-by-taxon as much as possible. But either way, making this a reality will require much more coordination with and involvement on iNaturalist among NatureServe Network members (setting aside Canada, which already has their own evolving process).

2 Likes

Thank you! That sheds a lot of light. But, though I understand the need to obscure on a case-by-case basis, I’m still inclined to think that we ought to err on the side of conservation. So, since a data-based decision to obscure or not requires time and analysis, why not use the “S” rank as a default to obscure, until evidence is amassed to un-obscure, rather than unnecessarily put things at possible risk?

1 Like

I am also

but deciding the appropriate S-rank defaults for each state will still require the involvement of those best positioned to make that call – the NatureServe biologists in each state. Unless you are suggesting that the current taxon geoprivacy situation for NatureServe subnational jurisdictions is so egregiously messed up that iNat should impose a blanket default across all jurisdictions. I think that approach would get a lot of pushback from many quarters on iNat.

1 Like

(Backstory: I was involved in some of the discussion over email and social media that preceded this–the original poster noted that a showy and probably collecting-vulnerable NJ S3 taxon had no geoprivacy protection, which I fixed, but this stimulated more general discussion of iNat geoprivacy.)

Having lived through the great Canada Geoprivacy Debacle, and reviewed the thread again, which was painful, I don’t think the community will accept precautionary blanket obscuration of NJ SX-S3 taxa, followed by removal of some, maybe. (“Evidence is amassed to un-obscure” sounds to me like it reduces in practice to “prove a negative”.) Maybe it ought to, but I’m pretty sure it won’t, and I don’t have either the formal or the informal standing to cram it down on the community.

That said, some of the points made above about threats beyond collecting, which were elaborated in other fora…uh, strike home. Leaving No Trace is harder than it looks, even for iNat “power users” like me who might consider themselves ecologically responsible and discreet. (I think we’ve had a case like this involving a salamander site in Ontario, where iNat seemed to be driving increased visitation?) On the other hand, that also feels like the sort of situation where I might obscure the observation on iNat but be tempted to show the individual to other “trustworthy” people. In other words, I suspect that kind of “sub-collecting” damage may spread more through informal social networks than through public locations on iNat.

There are a total of 818 plant taxa listed by NJ DEP. That’s sizable, but it may be a chunk small enough to swallow. If there’s appetite for a community-driven attempt to assess risk of collection/overvisiting, how frequently these taxa are observed (in NJ), and which ones are or are not currently obscured, I’m happy to set up a shared spreadsheet tracking this and try to help identify species we should be obscuring.

I’ll close by noting that I know the OP, who taught me a lot of botany and has really dedicated his life to protecting species and habitat in New Jersey. I hope the community will give some attention to his concerns.

2 Likes

Sorry I suppose I was mixing up IUCN labels with NatureServe labels. Critically endangered (IUCN) is not the same as critically imperiled (NatureServe). I agree with that.

But I’m still confused as to where this is a problem. Are they actually ever inappropriately combined anywhere on the site? Where are state-based labels not possible (in response to the OP conclusions)?

1 Like

Fairly frequently I have seen G4-G5 but S1 species assigned an IUCN equivalent of Critically Endangered in its iNat conservation status entry, apparently so as to justify giving it taxon geoprivacy within a subnational jurisdiction. While geoprivacy may well be justified for such a species, depending on the particular threats behind its S1 status, it is not in fact an IUCN Critically Endangered species (nor would it ever qualify as such), and no such equivalency should be playing a role in deciding or justifying local taxon geoprivacy for such a species, in my opinion. Hope that clarifies where I’m coming from there…

Agreed, and I meant to concur with those points in my earlier response, before noting that such threats are more likely to be a problem for a given S-rank level where human population densities are higher. And also where locations are generally more accessible and less remote than in many of my western haunts.

In the end, I think those most familiar with the taxa and threats in New Jersey should make the calls on taxon geoprivacy (default or otherwise) within New Jersey. I would just hope that such decisions are always balanced against the potential downsides of obscuration for conservation of some species, and also against the effectiveness of obscuration when locations are already well-known or easily obtained outside of iNaturalist. (The examples of Sequoia trees, Buffalo herds, etc., come to mind.)

1 Like

Thanks for the thoughtful comments and kind words, Chris. I’ve learned a lot from you, as well, my head stuck, as it were, in the sand of the Pine Barrens.

I got to a point where I thought I understood this situation well enough to move the discussion here. I wonder if it might be useful to take, to begin with, maybe 100 NJ S3 taxa at random, and solicit opinions about their respective vulnerabilities to having their locations published. I think that may be a helpful way to evaluate which way the geo-privacy default should go.

I read through most of the Canada debate, and I DO see how much pushback there was to the idea of basing default geo-privacy on subnational ranks. On the other hand, I’m sure you would agree that pushback, itself, ought not to be the deciding factor, but rather, the logic and reasoning prompting the pushback.

From what I saw, the pushback was based primarily on two issues. The first was the opinion that edge-of-range occurrences are somehow less in need of geo-privacy, merely due to the fact they’re at edge of range. This is subject to various logical objections. The second was that obscurations introduce impediments to current projects. I’m less able to address that point, until I can learn some particulars.

1 Like

I should add, it would be extremely helpful to me and others, I’m sure, if we could know which taxa currently do have geo-privacy in NJ, and what are the reasons these were selected.

https://www.inaturalist.org/observations?place_id=51&view=species&iconic_taxa=Plantae&taxon_geoprivacy=obscured

2 notes

  • any species listed as obscured in the state but not yet observed in the state won’t be listed.
  • one minor quirk, not a bug before people start yelling is any observation that had any id of an obscured species even if the ID is wrong permanently retains the obscuring even if the record is corrected to a non-obscured species. That explains some of the ‘wtf is that there for’. It is an intentional security measure to stop people getting access to obscured data by deliberately entering an incorrect identification of a non obscured species.
2 Likes

I’m starting to spin up a spreadsheet, but this will take a little while, and I want to look at some of the Canada material to figure out how to best absorb comment about individual taxa. Of the first 52, only 50% have actually been observed in New Jersey. This is probably a combination of ~10% of listed taxa being SH/SX, taxa with a small number of sites in New Jersey which are not generally known or visited, and graminoids and other unspectacular taxa which are both overlooked and not generally photographed in sufficient detail for an accurate iNat identification.

1 Like

As one of the people most involved in the first attempt, and now the second attempt to centralize this in Canada, and who does actually live in Canada, my perception of the pushback was that it was not based on the objective, but rather the process mechanics.

I didn’t see much pushback on the notion that local conditions and states of the taxa should drive local obscuring decisions, people want for the most part to see that taxa that are locally at risk are locally obscured. If they are stable and safe in other places so be it. There are some taxa (herps, orchids etc) that face universal threats from collecting so maybe global obscuring is fine just to cut down on the work.

The issue most people in Canada had was the complete removal of the Canadian iNat community from the decision making process, to the point of certain provincial bodies even flat out saying they would not even read requests for review of species, their word was final.

The process has shown that it is utterly impossible to effectively crowd source this decision making. The Canadian attempt has had multiple attempts at spreadsheets, data collection, feedback mechanisms etc all of which have failed. I’m not even convinced crowd sourcing is the best approach, the voices to open stuff up will always be screaming loudest.

The question raised about the salamander site is an interesting one. With the caveat that I personally know where the location is, and did not need iNat to get that information, although I’ve never been to see them.

I’m not sure how much of the increased traffic is from people directly getting the information for the first time by figuring it out based on iNat vs. return visits from people who already knew the location who want to get it on their iNat list, year list, project list whatever vs. people who got the location independently through being given the info. Not sure how you can determine the split. I can only anecdotally say that as a passionate promoter of the site in the province, I constantly get pushback about how the site exposes too many taxa locations to risk.

The site badly needs a default framework so at least a consistent starting point and approach can be implemented, and ways to manage that. The default framework can be either

  • by default everything is open until a compelling case is made it needs to be obscured
  • by default everything that is conservation status listed (ideally with the caveat it applies to CR, EN listed taxa) is default obscured until a compelling case is made to open it.

I know which of the 2 is safer for wildlife, which if any gets chosen is a separate matter.

The thing is that the process for obscuring should be similar to what was done with taxonomy. Get a standard reference and at least use it as a common starting point. If people want to debate and agree to a deviation, fine, but start every discussion from the same place, and same criteria.

What drives people batty is seeing things like https://www.inaturalist.org/taxa/115031-Williamsonia-fletcheri where it is Critically Endangered in 1 province and open, and ‘merely’ Endangered in another province where it is obscured. How do you defend or even have a discussion about what to do with these kinds of inconsistencies.

2 Likes

This makes a lot of sense to me. But what is a situation or example of downsides of obscuration for conservation?

I also want to point out, as you no doubt know, that it has been a standard practice for professionals to guard specific site locations for imperiled taxa. Some users of INat may not be aware of this. I’m involved in a project right now where I would very much like to know at least what county of south Jersey the specimen was collected from, but I can’t get that information without making a specific request to the institution where the specimen is housed (or visit the herbarium). It’s an inconvenience, but it’s a standard that has been in place for a long time. INat practices are, in a sense, a rebellion against this protocol.