In article ,
Stewart Robert Hinsley writes:
| |
| | Molecular systematics work doesn't support the split genera, but the
| | clade as a whole doesn't look well resolved. For the latest (April) word
| | on the topic see
| |
| | http://www.plantsystematics.com/qika...g/jse08050.pdf
|
| Thanks. But it seems to be yet another single gene/whatever analysis,
| and therefore as likely to be completely misleading as not!
|
| I did say that it "doesn't look well resolved". However, as a point of
| order, two genes (ndhF and ITS) were used. (From what I've seen
| elsewhere ndhF is too conservative to be ideal at this level.)

Oh, it wasn't a criticism of you! And my point is statistical, not
molecular - the chances of a single characteristic being misleading
about the ancestry of an organism are very high indeed. It's not
possible to give numbers without more data, but I could explain why
fairly easily.

Of course, uk.rec.gardening is not precisely the group for fairly
advanced statistical concepts, but what the hell? :-)


Regards,
Nick Maclaren.