Automatically add `leaf_cycle` from `species:wikidata`

[afanasovich]

Description

Wikidata approved a new property for plant taxa - foliage type (P10906)

Any taxon can have one of the following meanings:

• evergreen plant
• semi-evergreen plant
• semi-deciduous plant
• deciduous plant

I propose that for natural=tree we do an auto-substitution from Wikidata of this parameter.

Screenshots

No response

[1ec5]

Is it common for mappers to be able to identify the species of a tree before they identify the foliage type? leaf_cycle is already present on 77% of the features tagged species:wikidata, so I’m not sure that there’s a strong need for the additional data, if any, that would come from automatically tagging leaf_cycle in response to species:wikidata.

[bgo-eiu]

I would use this as I often add tree species without foliage type - my city maintains a public domain database of nearly every street tree and its species, but not foliage type. I have been meaning to add these details but this would make it easier, especially for some less common species.

For example, it is not that hard to figure out that this one is broadleaved https://www.openstreetmap.org/node/9686799342#map=19/39.31142/-76.64069 but for the hundreds of trees I have added like this it would be quite time consuming. (I do line these up with aerial imagery and filter the data based on the notes about their condition so that I'm not adding trees that are likely to have been cut down or fallen over recently so it's not quite blind importing. Verifying that the city surveyors got the species right is something I'll worry about once these are Wikidata'd and I'm more competent at working out that sort of thing. Tentatively, one of the main advantages of adding foliage tags on OSM is I could then easily filter out the handful of evergreen street trees and check to see if those specifically were surveyed correctly.)

Also, living somewhere where nearly all the trees are broadleaved I do worry about absent-mindedly marking a tree that isn't as broadleaved (and wouldn't be surprised if I accidentally did this before). They definitely aren't non-existent, but you can miss them if you're not paying attention. I realize I should clarify I'm usually doing this on the train or bus so it's easier to make these mistakes when they disappear so quickly.

[bgo-eiu]

This is maybe a more obscure use case, but trees are one of the only objects I've found reason to use planned tags on, and often the species is decided beforehand if the municipality has a specific plan for where to put it. There's a note in this one explaining that the city replaced a dead silver linden with a black oak for example, where the fact that a black oak was going to go there likely would have been mappable had I been reading about this sort of thing in 2018: https://www.openstreetmap.org/node/9704388552

Here, they are going to cut this one down this year and not replace it with anything (they want there to be a clearer field of vision in the park). It seemed useful to indicate that in OSM otherwise I would never remember that this one is going to get cut down. https://www.openstreetmap.org/node/9662102748 (This is also one of those evergreen trees that does exist surrounded by broadleaved ones - I would not have noticed it even walking by though had I not read about the removal plan for it.) If they were to replace it with a particular species though, that would another situation besides surveyed databases where you might add the species without knowing the leaf cycle.

[1ec5]

To be clear, I’m not arguing against tagging leaf_cycle or for removing the existing Leaf Cycle field. This issue specifically requests automatically copying that information from the Wikidata item about the species tagged in species:wikidata. The operative word is automatically; you can already click the ↗️ button in the Wikidata field to jump to the item where this information would be present and decide for yourself whether OSM should state the same fact. That said, if you’re tagging trees en masse, a MapRoulette challenge or find-and-replace in JOSM would be a lot more efficient, except maybe if you’re micromapping an arboretum where no two trees are alike.

There have been previous suggestions to automatically copy details from a linked Wikidata item to the selected feature: #6975 #8028. Frankly, any such feature would face overall skepticism about this approach to tagging, justified or not.

[afanasovich]

It seems to me that the statement "pine is an evergreen" or "linden is a deciduous plant" is information more suitable for a wikidata than for a cartographic database. And now it turns out that with each description of a pine specimen, it is necessary to re-do the great biological discovery that this pine specimen does not shed its needles)))

I have read the skepticism about import from Wikidata. But now auto-substitution of an article from Wikipedia works when specifying the wikidata key (Wikidata key → Wikipedia link). Information about the deciduousness of taxa, as it seems to me, is the same trivial knowledge, and the correspondence is the wikidata key - a link to the Wikipedia article.

At least I took the first 30 most popular species:wikidata key values (https://taginfo.openstreetmap.org/keys/species:wikidata#values ) and listed the deciduousness values in Wikidata without using commercial sources.