fix and maybe deprecate phenology observing process

[ramonawalls]

phenology observing process (PPO:0007004) has an EQ axiom that doesn't work and specifies:

• ((OBI:0000293 some PO:0000003) and (OBI:0000299 some (PPO_0000007 some (IAO_0000136 some PPO:0002000))))
• has specified input some whole plant and has specified output some (is quality datum of some plant phenological trait)

I think the problem is saying that it has some output that is a process (is quality datum). I changed it to:

• has specified input some whole plant and has specified output some (data item that (is about some plant phenological trait))

which makes more sense to me. However, it still does not work. The syntax validator in ROBOT will not accept clauses like:

• (has specified output some (data item and (is about some plant phenological trait))
• (quality of some (whole plant and has visible part some plant structure))

even though they are in the current ontology release AND they work fine in DOSDP patterns.

[ramonawalls]

Do we really need this relationship? I don't think we need to do any reasoning over the axiom, so it may be better to just use BCO:observing process, or even OBI:planned process.

@jdeck88 are you aware of a step in the pipeline that uses this axiom?

[ramonawalls]

I have struggled with this long enough. I'll submit an issue in the ROBOT repo. For now, I will change this axiom to

• has specified input some whole plant and has specified output some data item

which should not mess up anything in our reasoning, since we aren't doing any reasoning over the observing processes now that we have removed GCIs from the trait present and absent classes.

[ramonawalls]

See also #89