The main photo for a taxon should “capture” the species as best as possible, as far as I see it.
I suspect that means a presentation of the species beneficial to average users who will use that first-facing photo as a guide, since experts will instead be guided by reference books.
So take a Common Daisy. It needs to show the leaves as well as flower.
The taxon photo for Bellis perennis had been just a full face flat on photo of the flower, which is useless for helping people identify, as there are many daisy-like flowers. Showing the flower face would be like showing the picture of a tiger’s eye to represent a tiger. It needs to show the whole beast.
I duly changed the picture to something that captured the leaves, conveying the ‘jizz’.
Now I see it’s changed back to the full-face flower again!
I wonder what can be done about this problem…?
Can I recommend that there be guidelines showing in highlight (e.g. red) when editing a taxon’s photos to explain best practice? this would be different for each class of taxon, so perhaps the mammal guideline will be different from the plant guideline.
I think with guidelines there it might prevent a lot of this sort of regression, where a photo chosen for identification is replaced with one chosen for photogenicity.
David
If you find that your photo choices are being reverted, it might be a good idea to flag the taxon in question and start a discussion there. A lot of this ends up being taxon-specific. Sometimes photos that show fine diagnostic details are better choices for second or third photos. The first photo serves as an icon and therefore needs to be clear, even when shrunken.
This actually is something that needs to be visible on the same page as the photo changes page. Not everyone is reading the flag or is being notified to it. So I agree there needs to be specific to the taxon guidelines on the same page.
In this article i propose to use a “taxonomic relevance index” to rate photos intended for scientific use in the context of floristic studies.
QT0: Photograph awaiting determination
QT1: Photograph illustrating the type of Raunkier
QT2: Photograph allowing the determination of the family
QT3: Photograph allowing gender determination
QT4: Photograph allowing the determination of the species or subspecies
QT5: Photograph illustrating a species at its different vegetative stages, usable as a model in a determination key
QTx: Photograph allowing to compare two close species, which can meet in the same environment and be confused
QT+: Assembly of photographs illustrating a species at its different vegetative stages, usable as a model in a determination key
When you choose the default taxon photo (and I often edit taxon photos) that first one needs to be meaningful when shrunk down to a thumbnail. At that size some pictures are an utter mystery! Also fuelling the homonym confusion.
Lovely picture of a flower with the focus on the pollinator? Not for a plant taxon picture.
Obviously planted in a manicured setting. Not for a ‘wild’ plant, thanks.
Dog or hiking group. Also, not.
If we click we can see the first few photos.
So for your daisy.
The flower face, yes.
The bracts.
Leaf detail
A wide view for scale and habitat.
But if This species has distinctive field marks, that takes top place.
This is all recorded, in painful detail on the taxon history page.
'Diana moved photo 7 to photo 3 …
Taking the Daisy as a working example, I’d expect the flower face (either one or several) with at least a glimpse of the underlying leaves, ideally at a slight angle down, giving more impression of height than a vertically-down look. It all gives a sense of the whole whilst working as a thumbnail. Probably there are many contenders for which is the very best one to choose, but the one thing I’d say is to be avoided is just a flat view of the flower face that fills the entire picture, as for an average person that could seem to be hundreds of taxa.
So I was envisaging a guidance note prominent for changers. I think that’s not replaced by a talk page, but additional, since you want to prevent, voluntarily, as many regressive changes as possible before they happen rather than catching issues after the event…
Something like:
“When choosing the first photo for the plant that will appear as its thumbnail, where possible try to catch several important characteristics that will be of most benefit to a non-expert eye.”
I agree tangible visible guidelines would be really helpful ( if they don’t already exist )
I´ve noticed several times on taxa where I have altered the photos somebody else has reverted or swiftly changed them one day later. At present I don’t want to flag taxa for discussion around this tbh, it feels like a minefield / time-sink given how subjective it is.
And I’m not sure taxon history is that helpful either, as it just makes it more frustrating in some ways.
In the end I decided changing taxon photos in the current system feels like a waste of time, so I’ve largely given up…at least at higher levels or with more common taxa.
Longer term perhaps there could be some sort of voting system to make good photos harder to flip so that one’s input feels like it has more longevity / meaning.
Sometimes it’s going to be impossible to “show the whole beast” in the first photo. I edit taxon photos for Lepidoptera. Practically every iNat photo shows either an adult or a caterpillar, except for a handful of lovely annotated photocollages showing the whole life cycle in one image that are submitted by a few people rearing caterpillars. (But those are way too detailed to make easily recognizable small icons.) What I generally shoot for is:
Adult photo with a good view of the wings in the most common natural position. If there is sexual dimorphism, I try to put a photo showing both sexes at the same time.
Ideally, a photo of the caterpillar at the most informative angle. But if there’s a lot of adult variation (there are only separate male and female photos, adults commonly display both dorsal and ventral wing surfaces in the field, etc.), another adult photo, because most people are submitting and trying to ID adults.
Either the first caterpillar photo or photos that start to show additional characters/variation in phenotypes. These might include alternate photo angles (dorsal vs. lateral vs. ventral), photos of pinned adults, caterpillar shelters, cocoons and/or pupae, eggs, etc.
If you feel it’s really important to have the flowers and the leaves, but photos of the whole plant don’t shrink well, you could always submit an observation that includes a close-up of a flower and a detached leaf and add that as a taxon photo
Two other features worth aiming for on the first photo would be a sense of scale and similarity to other photos used for the larger taxon (like the genus when we’re talking species photos). I’ve seen many cases where one distinctive species in a group has very small flowers, but since the photo chosen is a hugely blown-up macro, it looks like the plant has large flowers, making ID much harder. I’ve also seen lots of cases where five of six species in a group have flower photos as thumbnails, but the last has a fruit photo, which is then not helpul for initial comparison.
We have been asked not to use collage photos on iNat. I wouldn’t use a collage in the first few visible photos, but would put it within the 12 we are allowed.
If I want to compare leaf detail, I open the relevant photos in separate tabs.
I tweaked the photos for our most observed plant in CPT. Sturdy shrub, almost a small tree. Highly visible (and then observed!) when covered in More Yellow Daisies. Species name means wearing a necklace for the ring of berries - so typical of daisies, not. The history of the changed photos has been meticulously recorded by iNat.
Taxon photo changes have been recorded in the taxon history page for a few months now, so you can see who’s been changing them. You can reach out to them via message or flag the taxon and @ mention them.
There’s an FAQ about taxon photo guidelines. It’s not taxon specific, however. The issue for a default photo is, as others have said, it also gets shrunken down to a tiny icon that should be at least somewhat readable, if not completely diagnostic.
Actually on that front, on a PC you get the auto-suggest list.
It might be useful to -
Display a larger version of the thumb to the right (floating) for the picture a person hovers their mouse over, as there’s all that unused space there. (That should be relatively simple programmatically.)
Maybe retain a large version showing for an ID picked along with observation for a person when they create a new observation.
I notice a lot of people choose some strange IDs even though the auto-suggest is the correct one and with a well-matched picture, I think people would spot this if their choice taxon had the pictures showing bigger.
I think also the more people engage with and interactively think about the ID, the more they learn - if it’s just a small thumb it can promote less critical thought for many people.
Of course many people use a phone too, I don’t know how the possibilities stand in this area.