• Welcome to BirdForum, the internet's largest birding community with thousands of members from all over the world. The forums are dedicated to wild birds, birding, binoculars and equipment and all that goes with it.

    Please register for an account to take part in the discussions in the forum, post your pictures in the gallery and more.
ZEISS DTI thermal imaging cameras. For more discoveries at night, and during the day.

Pan Species Listers anywhere? (3 Viewers)

Do pan-listers need to get down to species level for all taxa? I would think given arthropod diversity and difficulty with idea, getting down to even just family would be specific enough.
For most arthropods, I generally consider getting down to genus a meaningful identification – at least when there is not a comprehensive guide to a group (describing all extant species), as there always is with birds. (That is only the case for groups like butterflies and odonates, and only on a regional basis, e.g. for the U.S. or Europe). Family level or above usually doesn't provide useful information about distribution, etc. to anyone.

As to the general question of the thread, I am sure I am over 7500 taxa, and have entered a lot into iNaturalist. But I can't use it to generate a complete list. No way am I going to add all my eBird sightings into iNaturalist. I'm over 4000 worldwide in bird species and I know of no way to add them that would not take an enormous amount of time – it looks like you need to add each observation without media individually noting location and date.
 
For most arthropods, I generally consider getting down to genus a meaningful identification – at least when there is not a comprehensive guide to a group (describing all extant species), as there always is with birds. (That is only the case for groups like butterflies and odonates, and only on a regional basis, e.g. for the U.S. or Europe). Family level or above usually doesn't provide useful information about distribution, etc. to anyone.

As to the general question of the thread, I am sure I am over 7500 taxa, and have entered a lot into iNaturalist. But I can't use it to generate a complete list. No way am I going to add all my eBird sightings into iNaturalist. I'm over 4000 worldwide in bird species and I know of no way to add them that would not take an enormous amount of time – it looks like you need to add each observation without media individually noting location and date.
I suspect it can be done in bulk, from ebird to scythebill, then scythebill to inaturalist.

I suspect this, as I'm sure I bulk loaded mammals, flowers etc from scythebill to inaturalist, and scythebill works well with ebird.

However, what state your records would be in at the end of that journey is anyone's guess!

I think ebird is more useful for birding data. I'm split between ebird for birds, and inat for everything else. Both platforms are contributing to science/conservation projects, so it's all good.
 
I finally broke 6500 "species" (more exactly unique end taxa) of animals. At the same time I am missing only 6 invertebrates for them to overtake birds. I would not have thought that this is gonna happen so soon two years ago when we started with them, but there is simply a lot of diversity in insects. Also, I am lousier and lousier in birding with every coming year.
 
I finally broke 6500 "species" (more exactly unique end taxa) of animals. At the same time I am missing only 6 invertebrates for them to overtake birds. I would not have thought that this is gonna happen so soon two years ago when we started with them, but there is simply a lot of diversity in insects. Also, I am lousier and lousier in birding with every coming year.
Well done. These milestones are good for the soul?! Sanity?! Insanity?!!

I find some fields can compliment each other. My evenings with a bat detector helped a little to describe and differentiate bird calls.
Anything that has us looking close at something sitting on a leaf, or a flower, can't be good for the birding though!
 
Yep, walking around with extension tubes on my camera, I'd need the birds to come closer than about a meter and then I would only get the head at best. I do sometimes walk wound with two cameras nowadays, but it's not deeply comfortable.

Also since I started mammalwatching I am seeing far less sunrises than before.
 
Well done. These milestones are good for the soul?! Sanity?! Insanity?!!

I find some fields can compliment each other. My evenings with a bat detector helped a little to describe and differentiate bird calls.
Anything that has us looking close at something sitting on a leaf, or a flower, can't be good for the birding though!

No. My attention has largely been on insects and plants for a few months and I’m finding I hardly look to the sky or worry about sounds at all. I enjoy there being birds around but am not really focusing on them at all. Take my phone and Pentax papillo bins far more often than full size stuff at present.

Am going on a birding trip (although hopefully with plenty of insects) tomorrow and am wondering how long it’ll take to shift my attention from leaves and flower heads
 
It's the first year I've semi-seriously looked at plants and am throughly enjoying it. Pretty impressed with the iNaturalist recognition too.

The recognition is really good in specific places, but if you take it somewhere less-travelled it'll struggle. It's not good with more technical IDs either (like spiders, wasps, so on). Basically their system requires 200 "confirmed" sightings to include the species in the algorithm, and most insects are still falling short of that (or only go to genus).
 
I used iNaturalist quite heavily in planning for a trip to France recently.
The web version allows you to search records, species in geographical areas, and download, so I had a list of waypoints for each species I was keen to see.
The idea was to delete the whole group of waypoints once I'd found each one, or at least have a good idea of the regions where each species was most regular.
It worked to a degree, but I think I had more luck just stumbling across species in decent looking areas, plus knowledge of preferred habitat etc.
Would be useful for a very limited number of target species I suspect.
 
The recognition is really good in specific places, but if you take it somewhere less-travelled it'll struggle. It's not good with more technical IDs either (like spiders, wasps, so on). Basically their system requires 200 "confirmed" sightings to include the species in the algorithm, and most insects are still falling short of that (or only go to genus).
They update the model frequently and it now includes 77,276 taxa.
Computer Vision Update July 2023
Where I live (Wash. DC area), the model is very good with odonates, butterflies, and grasshoppers--even with difficult IDs. More than once it has alerted me to a rare sighting that I hadn't focused on yet. I think high qualilty recognition happens when you have very knowledgeable people repeatedly correcting/confirming IDs for specific groups. For other groups it can usually help narrow things down and save time.
 
Last edited:
The main problem with the computer vision on iNat is the missing species - when you have similar species of which some are common and some are so rare that they don't get yet included, you necessarily get the common one. Also, it's AI, so it doesn't "understand" the ID - I am afraid that it can train itself on externalities of observations. I am generally accepting its IDs to genus/family when it's "pretty sure" and it doesn't look outrageous, but I only put species-level IDs on things I personally understand.
 
It's helped me out in some specific cases where I might never have found a lead. Notably some new-for-country introduced plants, and a handful of insects that I had no clue on. It's just a starting point, but I feel all of us here can be trusted with doing a bit of due diligence rather than blind acceptance of the suggested ID.
 
James, a few of us are coming up with a devious plan to get you a girlfriend ;)

You're welcome here for a week, anytime btw.

Certainly hope so!! On both counts ;)

I am overdue for my great "UK redux" visit though...it's been a very long time, and I've learned much since then!
 
Do pan-listers need to get down to species level for all taxa? I would think given arthropod diversity and difficulty with idea, getting down to even just family would be specific enough.
As one of the "founder members" of the PSL concept in the UK I can say that as originally conceived it was important to only count at a species level, that is what I do except in cases where it just isn't currently feasible to do so because molecular sequencing is the only way to do it. I would suggest you treat all taxonomic groups exactly the same as you would birds, if you are happy with a genus level bird on your list then ok, do that for plants, otherwise stick to species level.

I would also suggest caution using iNat and the various AI apps, they are brilliant and a great place to start, but only as a place to start, always invest in the ID books to corroborate your IDs. The internet feeds off itself, consuming and churning out the mis-identifications as much as the correct identifications. It is obvious on iNat that many people just accept the AI ID uncritically, so you see numerous records of plants for the UK which do not occur here at all. Unfortunately the AI is not intelligent enough to take geographical range into account it seems!
 
Unfortunately the AI is not intelligent enough to take geographical range into account it seems!

The AI is, the users aren't :) It shows you with suggestions whether the species has been "seen nearby". Obviously this can be poisoned by enough misplaced observations, but it works pretty nicely.
 
The AI is, the users aren't :) It shows you with suggestions whether the species has been "seen nearby". Obviously this can be poisoned by enough misplaced observations, but it works pretty nicely.
But it only takes one person to accept an ID of a species not in range, and one more to agree making it "research grade", and then the next person using it "properly" will see that the species is found locally. Eg. do a search for Campanula - United Kingdom and you will see presented 3 species not known in GB even as garden escapes. One of these, Campanula giesekiana, has 15 separate records, from 15 different recorders, 12 of them "research grade". These must all be errors for Campanula rotundifolia. I am sure similar instances can be found across most taxonomic groups.
 
As one of the "founder members" of the PSL concept in the UK I can say that as originally conceived it was important to only count at a species level, that is what I do except in cases where it just isn't currently feasible to do so because molecular sequencing is the only way to do it. I would suggest you treat all taxonomic groups exactly the same as you would birds, if you are happy with a genus level bird on your list then ok, do that for plants, otherwise stick to species level.

I would also suggest caution using iNat and the various AI apps, they are brilliant and a great place to start, but only as a place to start, always invest in the ID books to corroborate your IDs. The internet feeds off itself, consuming and churning out the mis-identifications as much as the correct identifications. It is obvious on iNat that many people just accept the AI ID uncritically, so you see numerous records of plants for the UK which do not occur here at all. Unfortunately the AI is not intelligent enough to take geographical range into account it seems!
How anyone keeps a list is up to them–provided they honestly disclose what they do when comparing. I don't recognize anyone as being authoritative on the subject. But treating all taxonomic groups like birds seems arbitrary and ornithocentric (okay I'm making that word up)–why arbitrarily impose standards from one group on another? There are few if any bird species that can't be distinguished in the field – either by appearance, voice, or range. Many many insect species can only be reliably identified by dissection, often of the male genitalia. (But in the interest of full disclosure, I don't personally keep a list – other than what various websites tell me I have seen.)

But I completely agree AI needs to be employed cautiously. I have had photos of insects with busy backgrounds be identified as northern right whale! But iNaturalist is starting to help with the due diligence process as well – if you go to the page for a taxon, there is a "similar taxa" tab which shows common mis-identifications of the taxon. In any event, if I do follow an iNaturalist suggestion (which I only do in unusual circumstances), I always indicate "tentative ID" in the notes if I can't confirm it by other means.
 
But it only takes one person to accept an ID of a species not in range, and one more to agree making it "research grade", and then the next person using it "properly" will see that the species is found locally. Eg. do a search for Campanula - United Kingdom and you will see presented 3 species not known in GB even as garden escapes. One of these, Campanula giesekiana, has 15 separate records, from 15 different recorders, 12 of them "research grade". These must all be errors for Campanula rotundifolia. I am sure similar instances can be found across most taxonomic groups.
Good point – but it also usually only takes one person to disagree with an ID to set things right. (A disagreement often prevents research grade status – though I know this is not always the case.) So if you know of any such instances, you can start the corrective process yourself.
 
But it only takes one person to accept an ID of a species not in range, and one more to agree making it "research grade", and then the next person using it "properly" will see that the species is found locally. Eg. do a search for Campanula - United Kingdom and you will see presented 3 species not known in GB even as garden escapes. One of these, Campanula giesekiana, has 15 separate records, from 15 different recorders, 12 of them "research grade". These must all be errors for Campanula rotundifolia. I am sure similar instances can be found across most taxonomic groups.

A small enough amount of records that someone with the knowledge could go in right now and fix them all in about 5 minutes or less, dare I say...!
 

Users who are viewing this thread

Back
Top