EDIT

Even in theory, not all specimenOrObservation types can be derived from all specimenOrObservation types.

We may want to define those combination which are allowed / forbidden in theory and maybe we want also define a default for a given context (e.g. in theory it may be allowed to derive OtherSpecimen from any of the given types but in practice we want to avoid the usage of OtherSpecimen and therefore it should not be allowed by default).

Some rules for derivations not allowed in theory are:

• "real" material units can not be derived from media based units and from pure observations
• field units can not be derived from preserved specimen and most other types
• ...

In a next step we should make the possible combinations configurable in TaxEditor which is another ticket.

The base default should be defined by the requirements for classical higher vascular plant treatments as these are the most common usecases for the CDM.