Cyperales or Juncales?
skala at INCOMA.CZ
Thu Jul 11 09:50:24 CDT 2002
the same argumentation can be done in the case of other characters. Also flower colour can change in "any direction", can be subjected to reversals etc. Perhaps the problems with distributions-as-character are:
(1) intuitively strange type of "heritability"; should not make a problem if we simply consider character state as a possible trace of history within our model. Then you can see that distributions are/can be subjected to a kind of historical inertia in a similar way the other characters are, and, in this sense, distributions are lineage-dependent. The *mechanism* of this inertia is of secondary importance for the historical reconstruction.
(2) a fear that distributions can be greatly historically instable characters (e.g. when analyzing highly dispersable organisms). Similar to some another characters (e.g. hypervariable DNA segments or characters under high selection pressure) and will appear as such in the process of cladogram construction.
(3) problems with character-state definition (how much must two distributions overlap to be considered the same?). Similar to many other quantitative characters, IMO.
So, for a cladist, there should be no serious problem to take the distributions in a similar way as other characters.
From: Tom DiBenedetto [mailto:tdib at OCEANCONSERVANCY.ORG]
The organisms can move in any number of directions, and can move back - and the ground (or water) itself can move under them and alter the relationships of the areas. As a result, "place" is simply not a
manifestation of a lineage-dependant process, and thus does not qualify as evidence for discerning the contours of the lineage.
More information about the Taxacom