it creates a relatively arbitrary way for people to decide your observation is invalid, which we already have enough of
it creates retroactive confusion and concern for past observations (which can include several thousands or tens of thousands per person)
it interferes with the relatively open freedom of iNaturalist in that not everyone wants to post every individual from 20 they saw that day separately
it solves arguably little with the annotation concerns, because you still have plenty of cases where one plant may have both fruit or flowers, or where a flock of birds (if that is still considered valid) has different ages present
In this case I would see the photo of the reared adult as âevidence supporting the identification of the larvaâ. Having said that, I have have done exactly what you suggest - hereâs a reared adult marked captive https://www.inaturalist.org/observations/102376334
Personally, I withdraw (not delete as @cthawley said) the disagreeing ID, and add another ID as appropriate (even if it is a non-disagreeing ID identical to the one I withdrew)
Great! I think we are in agreement that both approaches are valid. The approach illustrated in your example is more complex, and you need to have your iâs dotted and tâs crossed. Many iNat observers are going to find it difficult to pull that off. The other approach is less formal, and could be viewed as a âbaby stepâ stop along the way to the approach you have used. The best need not be the enemy of the good.
Adding a photo of the same individual in the same condition even if taken later
Adding a photo of genitalia dissected later
are proper and have no issue to me.
However, I disagree that they are analogous to a situation where multiple life stage photos have been uploaded to the same observation from different points in time. In the examples above, the condition/life stage of the organism has not changed. The photos will not be treated erroneously by iNat or via export - eg, an egg being shown as annotated for adult or a life stage photo being given a date that may be off by months from when it was actually observed.
I donât think that thereâs anything wrong with linking to the photos stored somewhere off iNaturalist, though it isnât ideal - the link might change or the site might go offline, etc⌠Linking to another iNat observation would be preferred imo. Linking to the photos, either via another iNat observation or off site storage does avoid the issue with incorrect data being entered. Using another iNat observation has the advantage of making the data for the other life stage available.
All of that said, based on the current guidance for this new DQA, itâs clear it should not be used to address observations with this issue, though it clarifies that observations with multiple life stages of the same individual are not encouraged on iNat:
"Please do not vote âNoâ to the âEvidence related to a single subjectâ DQA condition in these scenarios:âŚ
When there are different photos of the same individual at different points in their life history (e.g. photo 1 is an egg, photo 2 is a caterpillar, and photo 3 is an adult butterfly).
These last three situations are not encouraged on iNaturalist, but please do not use the âEvidence related to a single subjectâ DQA condition to flag them."
Given that the convention is to use the original place and time of collection for e.g. microscope pictures, pinned insects or herbarium specimens to count for RG, this should not cause a problem at all. If you have an in situ picture for the specimen before preparation to document location and time at collection from the wild, even better!
Hello,
I can see there are many issues about this. Nature never being clear cut. How will this affect the Hungry Parrots project where the intention was to show a bird eating various plants?
Thank you Julie
Or do you have - this parrot - eats all these plants - in one obs? That would need to be split. Also location and timestamp will vary, so need to be separate obs.
There shouldnât be any conflict as long as the parrots are pictured in each photo. I checked out the project. While I didnât check every observation, I did check many and none of them were in conflict with the new DQA.
Some ideas proposed in the comment you are replying to were nixed; see here for the current DQA description. An observation should have a single subject species that is ideally pictured in each photo. It is perfectly ok to have photos that contain multiple species as long as the subject species is in each one (thought please add clarifying comments if the intended subject is counterintuitive from what a random iNat user might assume).
Observations are only potentially problematic if an observation was IDed for the parrot but contained some photos of the plant without the parrot in it. (And as I said, this was not the case in any of the observations I checked). And even then, it would still not conflict with the DQA because it would fall under the following situation:
Please do not vote âNoâ to the âEvidence related to a single subjectâ DQA condition in these scenarios:
[âŚ]
When there are photos related to the subject that donât show the subject (e.g. a picture of a bat sonogram, or a picture of the habitat the subject is in, or a picture of a drawing of the subject).
[These situations] are not encouraged on iNaturalist, but please do not use the âEvidence related to a single subjectâ DQA condition to flag them.
For what it is worth, I believe that iNaturalist should stick with the requirement that all observation data include must contain the same individual organism in the same (relatively brief) period of observation. Or, as an alternative, provide a way to subdivide observations with ROIs (regions of interest) to account for the different individuals, being able to specify to which organism or object an annotation belongs (e.g. obs #200580384.1, #200580384.24, etcâŚ)
I asked about this about 3 years ago and the responses were generally inadequate:
The DQA or Annotation ought include the condition of the same individual organism, in the same period of observation.
I would also note that the draft guidelines here
Please do not vote âNoâ to the âEvidence related to a single subjectâ Data Quality Assessment condition in these scenarios: [âŚ]
When there are different photos of the same individual at different points in their life history (e.g. photo 1 is an egg, photo 2 is a caterpillar, and photo 3 is an adult butterfly).
also go against past advice of iNat staff (@tiwane in this response) in response to this question:
Otherwise tools like that leverage annotations like life stage phenology or sex etc will be muddled by these observations (eg will return caterpillars during times of year when only adults are present or photos of juveniles in response to queries for adults etc)
Wouldnât this only in the case where the photos are taking at (significantly) different times (e.g. months apart), and not in the second example (âdifferent organisms of the same speciesâ). If both larval and adult stage appear at the same time, then it wouldnât be incorrect if they were tagged either way. Of course this isnât possible to verify by the IDâer that they are taken at the same time, but in the end you have to trust that an observation is uploaded with the correct date/time.
The fact that you can only tag as either larva or adult, and not both, is a good reason to separate them so they can be better annotated, but that seems like a less severe issue than the one you mention?
If the observation is not obscured, the original metadata is displayed on the pages for the individual photos, so it is often possible to check the date/time associated with the files the user uploaded.
(I refer to the timestamp a lot â it is helpful when trying to confirm whether the observer accidentally conflated multiple bee visitors on a busy plant.)
I frequently check the metadata for suspicious observations, but more and more, Iâm finding that the EXIF is not visible, even though the observation itself is not obscured. I donât know if there is some setting that folks are using to make EXIF data invisible, but it seems to have become fashionable in my region.
Iâm starting to think that I need to create a project where folks have to follow strict rules if they want me to ID their observations. Violate the rules, and you get kicked out of the project. TANSTAAFL
Some tech-savvy, privacy-, or conservation-aware persons perform metadata sanitation (including but not limited to EXIF, btw). Various tools allow clearing, sanitizing, or altering all sorts of metadata.
Some image processing software does this automatically. E.g. âsave for webâ in Photoshop will generate a smaller file size optimized for email/uploading to social media but also strips all EXIF. Uploading in itself will strip some EXIF data originally associated with the image.
If the observation was made with the iOS app, the metadata arenât included in what iNat gets from the system. Itâs an Apple thing. Plus, yes, some people either (intentionally or not) strip metadata before uploading or are maybe getting screenshots from video files/using similar methods that wonât generate useful metadata in the first place.
Yes, I realize all this. Thanks. Iâm talking about folks that Iâm somewhat familiar with, who donât strike me as being tech savvy enough to know what metadata is, much less edit it (I am often surprised that they manage to tie their own shoes). I thought there might be a setting in iNat that abbreviates the metadata, but perhaps itâs iOS app mentioned by tiwane thatâs to blame. Or it may be that they have the kind of busy-body friends who would provide them with the tools, instruction, and encouragement required to do this kind of thing.
Youâll see there isnât really any EXIF data on the photo, aside from basic size stuff: https://www.inaturalist.org/photos/60072403 So you can usually check to see if that was the cause.
If someone posted their photo to a site like Facebook first and then downloaded it from there, the photo would also be stripped of metadata. Same thing with a screen capture of video.
