Wikidata:Property proposal/distribution map of taxon

distribution map of taxon edit

Originally proposed at Wikidata:Property proposal/Natural science

Done: distribution map of taxon (P8485) (Talk and documentation)

Description	(Dynamic) distribution map of taxon
Represents	should be similar to taxon range map image (P181), but not static. Perhaps an instance of distribution map (P1846).
Data type	Geographic shape
Domain	taxon (Q16521)
Allowed values	geoshapes
Example 1	~~Southwestern water vole (Q301535) → c:Data:Atlas de Mamíferos de Portugal/Arvicola sapidus.map~~
Example 2	Phyllostegia mollis (Q5488952) → c:Data:Ecos.fws.gov/Endangered habitat 58938/Phyllostegia mollis.map
Example 3	~~Spiny-footed Lizard (Q205053) → c:Data:Portugal/ICNF/Atlas de Anfíbios e Répteis de Portugal/Acanthodactylus erythrurus.map~~
Example 4	Alabama beach mouse (Q2636280) → c:Data:GAP/Peromyscus polionotus ammobates.map
Example 5	Alabama map turtle (Q40678) → c:Data:GAP/Graptemys pulchra.map
Planned use	I intend to add distribution maps for all species of mammals present in portugal, see c:User:GoEThe/Atlas de Mamíferos de Portugal
See also	(note the different datatype) taxon range map image (P181): range map of a taxon distribution map (P1846): distribution of item on a mapped area (for range map of taxa, use (P181).)

Motivation edit

I have recently started importing geojson data related to the distribution of mammals in Portugal to Commons. I was following mw:Help:Map Data and Wikidata:Map data to help me use this new format, when I realized I couldn't link this to the species item because a taxon is not a geoshape (P3896). taxon range map image (P181) is for static images of distribution maps, and distribution map (P1846) is not specific for biological taxa. So, I think this should be a new property. GoEThe (talk) 17:56, 10 January 2020 (UTC)[reply]

WikiProject Taxonomy has more than 50 participants and couldn't be pinged. Please post on the WikiProject's talk page instead.,

Notified participants of WikiProject Biology GoEThe (talk) 18:00, 10 January 2020 (UTC)[reply]

Discussion edit

We have taxon range map image (P181). --Succu (talk) 17:59, 10 January 2020 (UTC)[reply]

taxon range map image (P181) is not for geo-shape data. GoEThe (talk) 18:01, 10 January 2020 (UTC)[reply]

Could you please provide some examples other like the one from c:User:GoEThe/Atlas de Mamíferos de Portugal. --Succu (talk) 21:30, 10 January 2020 (UTC)[reply]
Use P181 (Taxon range map image). It has been used for many plants: see Q15287793, Q14861081, Q15366854 and many, many other plant species. As it is an image shape maps can be used. However, they are static. Linking to live maps would mean that such live maps MUST be CC-BY-0. Your example fits that requirement, but how many others do? MargaretRDonald (talk) 02:01, 11 January 2020 (UTC)[reply]

Provided the property were simply a link there is no issue with copyright. MargaretRDonald (talk) 15:30, 12 January 2020 (UTC)[reply]
And other examples would be: Cyperus dives occurrence data from GBIF, and Isolepis aucklandica from Australasian Virtual Herbarium MargaretRDonald (talk) 15:30, 12 January 2020 (UTC)[reply]

Strong support I think we should migrate all the image maps to geoshapes in the future. --Tinker Bell ★ ♥ 18:34, 12 January 2020 (UTC)[reply]
Reject (in its current form.) Geoshapes are NOT appropriate for occurrence data. They are not the form of the occurrence data given in the example, and they imply an (unjustified) inference by their maker of continuities which may well fail to exist. The proposal above only makes sense if the datatype proposed is a URL. MargaretRDonald (talk) 19:53, 12 January 2020 (UTC)[reply]

The Atlas I am using uses presence/absence in an area (in a 10x10km UTM grid) instead of point localities, but provide the coordinates in the centroid of each square. Still, I am trying to figure out how to transform the data to polygon data, in order to display it correctly, as it is done in the original publication. This is a common form of publishing Atlas data. Any distribution range map infers presence in a continuum, so I don't understand that objection. I also don't understand why such maps must be CC-BY-0 if they are hosted in Wikimedia Commons. The example I gave is CC-BY-SA-4.0. Here are other examples: [1] GoEThe (talk) 12:25, 13 January 2020 (UTC)[reply]

I don't think Wikimedia should include occurence data, such as every location where a given species has been observed. I think this property should be for polygon data such as IUCN's maps (unfortunately not CC). The example I gave is similar in that the presence is given per unit in a square grid. GoEThe (talk) 12:31, 13 January 2020 (UTC)[reply]

I see some more examples here now. GoEThe (talk) 16:03, 27 January 2020 (UTC)[reply]

Strong support This seems to be a really useful move forward. Images are nice for people, but terrible for machines. If we had GeoJSON files to represent distributions those files would be much more computer friendly, and could also lead to better visualisations by client apps using Wikidata (in other words, those apps could read the distribution data and display it in different ways, rather than rely on a fixed image). GeoJSON supports polygons, points, etc., so distributions could be represented in whatever way is appropriate. Sources might be problematic initially, but people could generate them from, say GBIF, or other sources and add them to WikiCommons. --Rdmpage (talk) 10:37, 9 May 2020 (UTC)[reply]

Just found another source of cc-by geoshapes https://www.nature.com/articles/s41559-020-1198-2. GoEThe (talk) 22:48, 28 May 2020 (UTC)[reply]

Seems useful but I wish there would be a way to smooth GeoJSON and use fewer points - https://phabricator.wikimedia.org/T194455 Shyamal (talk) 10:54, 29 June 2020 (UTC)[reply]
Comment @GoEThe, Tinker Bell, Rdmpage: I fixed the sample/domain above, please add two more samples. --- Jura 16:35, 15 July 2020 (UTC)[reply]

Thank you! I added two more samples. GoEThe (talk) 08:13, 16 July 2020 (UTC)[reply]

@GoEThe, Tinker Bell, Rdmpage: now that there are three, I'm less sure this can work out as expected. Only Phyllostegia mollis has a distribution limited to the area on the map (at least according to Wikipedia). For the other areas, countless additional statements would be needed. For these, I think a way needs to be found to associate structured data directly with the files, rather than through statements on a taxon.--- Jura 11:05, 16 July 2020 (UTC)[reply]

Hi @Jura1:. Yes, I guess that makes sense. It should only take one value, that of the most complete map data encompassing the entire range. There is also c:Data:Ecos.fws.gov/Endangered habitat 58938/Plethodon neomexicanus.map if it is a better example. The data that I uploaded to Commons is only for species that occur also elsewhere. But I can add some more maps that I have for complete distribution data, if it helps. GoEThe (talk) 12:48, 16 July 2020 (UTC)[reply]

I crossed-out these samples. --- Jura 13:40, 18 July 2020 (UTC)[reply]

Just to add another source of compatible map data: [2]. GoEThe (talk) 12:54, 16 July 2020 (UTC)[reply]

Support per Phyllostegia mollis sample. Maybe "dynamic" shouldn't be in the description either. I see it mainly as taxon range map image (P181) with a different datatype.--- Jura 13:39, 18 July 2020 (UTC)[reply]

I've added two more examples. GoEThe (talk) 13:42, 22 July 2020 (UTC)[reply]

@GoEThe: Excuse me, I do not understand what do you mean with "I couldn't link this to the species item because a taxon is not a geoshape (P3896)". P3896 is a generic property to link a JSON map stored in commons instead a flat image to an item. It has no constraint related to P31 or P279 of item that avoid use it on taxons. So, you can use a P3896 as a main property with a qualifier applies to part (P518) to indicate the specific meaning or content shown on the map. See exemple: Olympic Park Line (Q7427195). Thanks, Amadalvarez (talk) 04:38, 23 July 2020 (UTC)[reply]

@Amadalvarez:, interesting. I was able to add it here. Perhaps the constraint was removed between January and now. GoEThe (talk) 09:42, 23 July 2020 (UTC)[reply]

@GoEThe: Well, 3 minutes after your answer, the property has been created. Consider to keep or ask for deletion and use my solution before start using it. It's up tou you. Salut !Amadalvarez (talk) 09:49, 23 July 2020 (UTC)[reply]

@GoEThe, Succu, MargaretRDonald, Tinker Bell, Rdmpage: @Shyamal, Jura1, Amadalvarez:

Done distribution map of taxon (P8485) Pamputt (talk) 09:45, 23 July 2020 (UTC)[reply]