Repository navigation
Conversation
There was a problem hiding this comment.
Pull request overview
Adds the ENVO class “wetland biome” (ENVO:03605010) to the core ENVO edit ontology using the established biome design-pattern axiom (wetland ecosystem and participates in some climactic ecological succession), so it can classify appropriately under the broader biome hierarchy.
Changes:
- Declares new class
ENVO_03605010inenvo-edit.owl. - Adds label, textual definition, creator/date annotations for
wetland biome. - Adds the design-pattern axiom (
EquivalentClasses) and assertsSubClassOf wetland ecosystem.
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
|
@turbomam heads up: with #1676 merged, ENVO:03605010 is taken again. It's the restored "NMDC environmental context value sets" annotation property, the root of the NMDC value-set hierarchy, with 03605011/12/16/20 as its sub-properties. It was in the 2025-02-05 release. This PR declares the same IRI as the class "wetland biome", so after merging we'd have a punned IRI with two labels. Can you mint wetland biome from a free ID in your range instead? Everything from 03605010 to 03605026 is used by the value sets. |
f53fac0 to
c7fb716
Compare
turbomam
changed the title
2634939 to
d481185
Compare
|
Done, as you asked. Both branches are rebased onto master after the restore and the release, and the new class is now ENVO:03606010, the next free identifier in my range after the restored ENVO:03606009. The axioms are unchanged. I also set the base of this pull request to master so that it gets a CI run; until the re-parent pull request below merges, the diff shows its one line as well. |
d481185 to
68dab49
Compare
matentzn
left a comment
matentzn
left a comment
There was a problem hiding this comment.
The ID clash is fixed, the biome follows the pattern, and both versions reason cleanly. One modelling consequence of the stacked re-parent (#1667) I'd like @pbuttigieg to sign off on. (request changes here mostly means request review)
Thanks for moving to ENVO:03606010. I checked it: it's in your range (3605000–3619999) and appears nowhere on master or anywhere in the git history apart from this commit.
Since this branch contains #1667's commit, merge #1667 first as you said.
| Declaration(Class(<http://purl.obolibrary.org/obo/ENVO_03620006>)) | ||
| Declaration(Class(<http://purl.obolibrary.org/obo/ENVO_03620007>)) | ||
| Declaration(Class(<http://purl.obolibrary.org/obo/ENVO_03620008>)) | ||
| Declaration(Class(<http://purl.obolibrary.org/obo/ENVO_03606010>)) |
There was a problem hiding this comment.
this declaration (and the # Class: block at L46587, which also has no blank line before it) was put in by hand after ENVO_03620008, so it's out of ID order. The next Protégé/ROBOT save will move both and leave a noisy diff in someone else's PR. A robot convert -f ofn round-trip before merging avoids that.
| AnnotationAssertion(Annotation(<http://www.geneontology.org/formats/oboInOwl#hasDbXref> "USGS:SDTS") <http://www.geneontology.org/formats/oboInOwl#hasRelatedSynonym> <http://purl.obolibrary.org/obo/ENVO_01001209> "muskeg") | ||
| AnnotationAssertion(rdfs:label <http://purl.obolibrary.org/obo/ENVO_01001209> "wetland ecosystem"@en) | ||
| SubClassOf(<http://purl.obolibrary.org/obo/ENVO_01001209> <http://purl.obolibrary.org/obo/ENVO_01001110>) | ||
| SubClassOf(<http://purl.obolibrary.org/obo/ENVO_01001209> <http://purl.obolibrary.org/obo/ENVO_01001790>) |
There was a problem hiding this comment.
The class itself matches patterns/biome.yaml: the equivalence is wetland ecosystem and ('participates in' some 'climactic ecological succession'), and the definition is the pattern text (with "A" instead of the template's "An", which is right). CI is green.
I ran ELK on current master and on master + this PR. It's coherent, and no class loses an inferred parent. But putting wetland ecosystem under terrestrial ecosystem makes 37 classes terrestrial ecosystems. Most of them are fine (peatland, bogs, fens, mires, swamps, dambos). A few are less obviously "an ecosystem which primarily occurs on land":
- intertidal ecosystem (ENVO:01001468, via coastal wetland ecosystem)
- tidal mudflat, mudflat
- saline marsh, saline wetland, coastal wetland ecosystem
- mangrove swamp, and mangrove biome, which also becomes a terrestrial biome (and the only inferred child of wetland biome)
None of them currently sits under anything marine or aquatic, so there's no double classification and nothing breaks. The wetland ecosystem definition already says "A terrestrial ecosystem which…", so you're making the axioms match the text rather than inventing something new. WWF also counts mangroves as a terrestrial biome. I'm OK with it, but intertidal ecosystem becoming terrestrial is the kind of thing @pbuttigieg will have an opinion on.
There was a problem hiding this comment.
Isnt this the same as #1667
Yes. It shows here because this branch is stacked on the re-parent.
There was a problem hiding this comment.
They are land-based, just rather wet - wetLANDS after all. I'll take a closer look but this should be okay
|
I did some more analysis of which would, under reasoning, make nine classes liquid astronomical body parts, which is dubious for some of them. That re-parent was mainly to make more classes eligible for NMDC's expectation that When I come back I'll do the |
|
We can deal with the Liquid ABP issue later - it is also probably incorrect to claim they are primarily composed of water |
|
remove the primarily composed of axiom? |
Adds
wetland biome(ENVO:03605010), a DOSDP-compliant biome on the biome design pattern: equivalent towetland ecosystem and ('participates in' some 'climactic ecological succession'),SubClassOf wetland ecosystem. Resolves #1666. Stacked on the re-parent PR, merge that first.env_broad_scalevalue for peat-forming wetlands.marine salt marsh biome.travis_testpasses