You are right that MIAPPE was conceived primarily with experimental plant phenotyping in mind, but:
- I don’t think there is a better alternative available. MIAPPE is the only metadata standard I know for plant phenotyping, and also likely the richest metadata standard overall (it’s minimum in name only).
- I don’t see any issue with ignoring the few MIAPPE fields that don’t apply to observations of natural populations and using the rest of it (most fields should still apply, and I for one would actually be interested in your feedback on this usage; we can propose a revision or derivation of MIAPPE for such use cases).
As for using the Plant Trait Ontology, I think it is a good starting point, but you should consider developing a controlled vocabulary for your observations that follow the Crop Ontology template (i.e., cover not only observed traits, but also the observation method and scale of the values), as this is the template with adopted in MIAPPE.
Let me know if you need any further input on using MIAPPE.