How should I go about IDing Unknown Observations?
I primarily ID salmonids and thus am not very good with other taxa, but I want to help ID some of the Unknown observations on the website. Is it better for me to ID to a specific species (likely using CV to do so), or to a broder taxa like class or even kingdom? On the one hand I don’t want to inadvertently create falsely RG observations as a result of the observer blindly agreeing with me, but on the other hand I imagine that most specialists (especially of larger taxon groups like plants/birds) will ID by more specific taxa than kingdom or even class and I want to make sure that they will see the observation.
2 Likes
It’s pretty standard to only ID as far as you can independently confirm. So probably don’t use the CV.
4 Likes
It might depend on where in the world you are doing the IDs. In general a lot of unknown observations are plants. In Africa there’s a strong belief that plants should be identified to family at minimum, or else left as unknown. In other parts of the world “Angiospermae” or “Plantae” are more acceptable… but whether or not another identifier will take things from there just depends on who is available in that area and what their search practices are. In my experience, observations more than three days old are usually lost in the pile unless you were able to ID them as a particularly popular taxon (for example orchids and cactus receive good attention in most areas.) The flip side of this is that some of the most recent unknowns are caused by observers batch uploading first and adding their IDs later; sometimes they don’t take kindly to getting notifications of your ID before they have had their chance to add their ID.
Many of these plants will also be cultivated, so up to you if you want to try to learn the distinction between wild and not wild.
I would not use the CV to put a species level ID if you don’t know it yourself. As you said the chance of blind agreements is too high. I would be okay with using the CV to guess at a family or maybe genus; you could apply some judgement to the CV results, checking that the suggestions are geographically reasonable and determining that yes, a lot of the top suggestions are in the same family, so that family seems like a good guess… Alternatively there’s a series of “yellow label” projects in which @jeanphilippeb has already grouped some unknowns by possible taxon using the CV suggestions. (Observations go into those projects as unknowns but are not removed when they are identified, so be aware many of of them have IDs already. The projects also include casual observations. )
2 Likes
My biggest piece of advice is that you want the process to be enjoyable for you, so don’t feel pressured to ID every single observation on a page if it is causing you frustration. For example, if you find bad photography or inscrutable bits of sea life frustrating/tedious/too difficult, you can always mark them reviewed and move on. Also be prepared for the occasional sassy comment from the observer–responses along the lines of “duh it’s a plant” definitely happen. If you want to try to prevent such remarks, you can use a copy-paste message such as:
" I’m not quite sure which species this is, but this general identification will help other people who might know the species find your observation. Many people helping identify observations on iNaturalist filter the observations by the group of species they know how to identify, like “plants” or “insects”, and this general ID will help them find it more quickly. If you’re interested in learning more about how identifications progress on iNaturalist, you can read more here: https://www.inaturalist.org/pages/getting+started#identify"
4 Likes
If you are generally knowledgable about fish orders/families you can try identifying these observations https://www.inaturalist.org/observations/identify?reviewed=any&iconic_taxa=unknown&order_by=random&order=asc&verifiable=any&project_id=unknown-actinopterygii
do be warned that there many that are not identifiable
this makes use of jean philippe’s unknown projects (https://www.inaturalist.org/journal/jeanphilippeb)more info here: https://www.inaturalist.org/journal/jeanphilippeb/73398-phylogenetic-projects-for-unknown-observations
There are a few other journal and form posts that give tricks on quickly identifying arthropods and mollusks, and other groups, you could probably find it on your own, or maybe someone else will link it
For plants you could work on unknowns in your local area as you might be familiar with common species and those that might be cultivated and can be removed from the unknown pool
2 Likes
It’s fine just IDing something to Arthropoda, Mammalia", or Insecta.
There are people (like me) who refine stuff stuck at taxonomic levels like that. For example here’s my Identify bookmark for refining observations stuck at Arthropoda.
5 Likes
There are a lot of existing threads on the forum about IDing Unknowns that would be good to check out, including:
https://forum.inaturalist.org/t/good-global-criteria-when-identifying-unknowns/47167
https://forum.inaturalist.org/t/whats-the-fastest-way-to-identify-unknown-obs/52513
https://forum.inaturalist.org/t/identifying-taxa-just-to-remove-them-from-unknown/33749
https://forum.inaturalist.org/t/useful-inaturalist-tasks-for-non-experts-wiki/35034 (more broad, but includes IDing Unknowns)
and many more too
3 Likes
When I realised that the Needs ID heap is about as big as the Unknown heap, I tackled my Cape Peninsula ones. I have used it as a long, slow and painful learning curve.
First peel off those stuck at Life - because Kingdom disagreement (20K sort by location?) Easy clicks of obvious vs. No Idea!
Then down the taxon levels, one by one.
I have finally dug down to Cape Peninsula Family with over 8K obs … the impossible will take a little longer, but the good ones are rewarding!
Find your slice of iNat, where you can enjoy IDing
PS for ones above Family that have since snuck in, the oldest is a fishy one ;~)
2 Likes
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.