hlapp
commented
3 years ago I am now arguing that this is pointless and should not be included in the semantic model. (This doesn't mean it can't or shouldn't be included in the Phyx format.)
Here's the current draft text from the nascent manuscript:
Crown clade. Phylogenetic nomenclature provides for a clade definition to designate a crown clade. In general, a crown clade includes the most recent common ancestor of all living (extant) species (or organisms) of a group, and all of their descendants. (Note that the descendants can include extinct species.) For a definition to designate a crown clade, the internal specifiers need to be extant.
In principle, it would be possible in the Phyloreference OWL model to include a condition for internal specifiers to be extant. However, there are numerous ways this could be represented (subclassing, property restriction, value restriction, to name a few), and most phylogenetic trees, to which a phyloreference would be applied for retrieving a clade and its content, do not include specific annotation as to which OTUs are extant and which are not. If an internal specifier were extant at the time of erecting a definition but subsequently went extinct, the result (if any) for its semantics would become problematic, at the very least for automatic reasoning. Therefore, we do not include a crown clade designation in the logical phyloreference model, although an associated verbatim clade definition in natural language may mention it.
Total clade. Phylogenetic nomenclature provides for a clade definition to designate a total (or pan-) clade. A total clade definition is a maximum clade definition in which at least one of the internal and all of the external specifiers are extant. Often, a total clade is defined in reference to a crown clade: it includes the earliest common ancestor of the crown clade from which no other extant organism is descended, and all of its (extinct or extant) descendants. For example, the closest extant relative to the crown group of birds (Neornithes) are crocodilians. Thus, the total clade of birds originates from their earliest ancestor from which crocodilians are not also descended, and therefore includes dinosaurs etc.
We do not include total clade semantics in the Phyloreference OWL model for the same reasons as crown clade semantics are not included. This effectively means that a maximum clade-based phyloreference will have implied total clade semantics if all of its external specifiers and an internal specifier are extant.
Clades, whether using a maximum or minimum clade definition, may be defined as crown-clades or total clades. These semantics need to be captured, even if it is practically impossible to convert these into logical constraints that can be applied when resolving the definition against a tree. (Doing so would require knowing which tips are extinct and which ones aren't.)