Updating to include terms relevant to restoration ecology and conservation

[timalamenciak]

Per meeting with @pbuttigieg @matentzn here is a prototype sentence template for restoration interventions and ecosystem degradation.

ENVO should be able to model this example:

A conservation organization is donated a former agricultural field. Analysis suggests that the area used to be a wetland before tile drainage was used to convert it to agriculture. A conservation practitioner wants to use a specific technique - crushing tile drains - to allow the wetland to rewet itself naturally.

Conservation Practitioner applies "crushing tile drains" which is an instance of ecosystem intervention class->abiotic intervention to address degradation type in area to transform it from an ecoregion->agricultural into ecoregion->wetland

It may be worth splitting these off into separate issues:

• ecosystem intervention - ENVO already has active ecosystem management process but its subclasses could use more development.

• degradation type - The SER Standards & Practices defines degradation as follows: "a level of deleterious human impact to ecosystems that results in the loss of biodiversity and simplification or disruption in their composition, structure, and functioning, and generally leads to a reduction in the flow of ecosystem services." I'm not fully sure how to model this in an ontology.

• area - In terrestrial ecology, the two major units of experimental division are plot (quadrat) and transect. None appear in ENVO. There is a well-developed definition of agricultural experimental plot in AGRO that could serve as a model.

• ecosystem types are being hotly discussed on the issues tracker right now: Rework ecoregions hierarchy  #1571 Clarify ecozone vs biogeographic realm #1574 Added spatio-temporal scale #1576 etc.

[pbuttigieg]

... here is a prototype sentence template for restoration interventions and ecosystem degradation.

ENVO should be able to model this example:

A conservation organization is donated a former agricultural field. Analysis suggests that the area used to be a wetland before tile drainage was used to convert it to agriculture. A conservation practitioner wants to use a specific technique - crushing tile drains - to allow the wetland to rewet itself naturally.

Conservation Practitioner applies "crushing tile drains" which is an instance of ecosystem intervention class->abiotic intervention to address degradation type in area to transform it from an ecoregion->agricultural into ecoregion->wetland

It may be worth splitting these off into separate issues:

I think they go together, let's leave it bundled for now.

Those examples aren't ecoregions, just ecosystems.

The term "abiotic intervention" isn't very useful - it's impossible to intervene abiotically in an ecosystem. I suppose they mean a process that doesn't use a lifeform to disrupt the ecosystem (like introducing a predator). If that's the case we can set up two classes that can be populated by inference:

"abiotic active ecosystem management process" =def. "An active ecosystem management process during which an agent uses non-living entities to alter the compositon, structure, or functioning of an ecosystem."

"biotic active ecosystem management process" =def. "An active ecosystem management process during which an agent uses living entities to alter the compositon, structure, or functioning of an ecosystem."

• ecosystem intervention - ENVO already has active ecosystem management process but its subclasses could use more development.

They sound synonymous. I'd add an exact synonym property to the existing term.

Tile drain crushing etc can go under that class.

• degradation type - The SER Standards & Practices defines degradation as follows: "a level of deleterious human impact to ecosystems that results in the loss of biodiversity and simplification or disruption in their composition, structure, and functioning, and generally leads to a reduction in the flow of ecosystem services." I'm not fully sure how to model this in an ontology.

There's a bunch of anthropocentric ideology in that definition, but we can unpack it.

'ecosystem degradation' =def. "An ecosystem process during which the composition, structure, and function of an ecosystem - existing in a desired state - are disrupted or destroyed."

The "desired state" is important - ecosystem restoration also messes up an existing ecosystem. The restoration or degradation aspects are subjective human projections.

'ecosystem restoration' =def. "An ecosystem process during which the composition, structure, and function of an ecosystem - existing in a undesirable state - are disrupted or destroyed to convert it into a desired state."

'anthropic ecosystem degradation' =def. "Ecosystem degradation which is caused by human activities."

We'll need some ES classes, but these should be populated by inference - we should not, for example, move photosynthesis under supporting ES.

'service' =def. "A process during which one entity provides benefit to another."

'ecosystem service' =def. "A service during which an ecosystem provides benefits to humans."

'provisioning ecosystem service' =def. "An ecosystem service during which an ecosystem produces materials that can be used as goods by humans."

'regulating ecosystem service' =def. "An ecosystem service during which the state of an environment is rendered more supportive of human life, health, and socio-economic activity."

'supporting ecosystem service' =def. "An ecosystem service which promotes the stability, continuity, and output of other ecosystem services."

'cultural ecosystem service' =def. "An ecosystem service which facilitates the exercise and development of the transgenerational behaviours, norms, and institutions of human societies."

We can use these as ingredients for more complex process classes as they come along, while users can compose statements with whichever combination works for them.

• area - In terrestrial ecology, the two major units of experimental division are plot (quadrat) and transect. None appear in ENVO. There is a well-developed definition of agricultural experimental plot in AGRO that could serve as a model.

I think I wrote that AGRO def. We can add these, please suggest some definitions. One has to decide whether you're talking about the areas (spatial entities ) or parts of an ecosystem (material entities).

[dr-shorthair]

(Ecosystem) intervention could be modelled as a subclass of Actuation from the SSN Ontology

[pbuttigieg]

(Ecosystem) intervention could be modelled as a subclass of Actuation from the SSN Ontology

These clases are outside the scope of SSN. Further, OBO semantics exist for all these cases.

You could create a SSSOM file to map SSN/SOSA to OBO classes.

[pbuttigieg]

Ecosystem services are relevant to #1515

Note I crested classes for these in SDGIO, but its state under UNEP management is unclear. It's wiser to maintain these in ENVO.

[dr-shorthair]

(Ecosystem) intervention could be modelled as a subclass of Actuation from the SSN Ontology

These clases are outside the scope of SSN. Further, OBO semantics exist for all these cases.

I was not intending to suggest that you adopt SSN.
Intervention or Treatment as a kind of Actuation seemed like one of the interesting areas to explore, as an extension or application of SSN.
Definitely not core SSN.

You could create a SSSOM file to map SSN/SOSA to OBO classes.

Yes. I did already create some SSSOM mappings: https://github.com/w3c/sdw-sosa-ssn/tree/187-sssom-versions-of-mappings (though I haven't touched it for a while). It is currently limited to the alignments mentioned in the spec document: https://w3c.github.io/sdw-sosa-ssn/ssn/#Alignments which does not include OBO. It does include PROV so I think OBO would be relatively straightforward.

[pbuttigieg]

@timalamenciak just like in #1579 it would be good to do this via a ROBOT template, so we can review in a more human friendly way prior to OWL generation