| Full text | |
| Author(s): |
Marques‚ A.C.
;
Gnaspini‚ P.
Total Authors: 2
|
| Document type: | Journal article |
| Source: | CLADISTICS; v. 17, n. 4, p. 371-381, 2001. |
| Abstract | |
We here propose a procedure to treat characters which are susceptible to parallel evolution (in this case, troglomorphisms) as a replacement for the two procedures used so far: to either completely consider or completely disregard these characters. These procedures may lead to one of two opposite errors, respectively, (1) to consider them as true synapomorphies when they are not or (2) to disregard them as true synapornorphies when they are. We suggest herein to recode the characters by splitting each troglomorphic character into as many as the number of taxa which show the given troglomorphic state. For each split character each taxon will have the derived state, while the others will be coded as missing data. We provide three real examples to test our procedure and conclude that it may give results different from those of the other two procedures. This means that our procedure does not repeat the two above "errors." In addition, we believe that the procedure filters the possible biases, resulting in trees in which the troglomorphisms do have phylogenetic signals. (C) 2001 The Willi Hennig Society. (AU) | |
| FAPESP's process: | 00/04686-4 - Comparative biology of insects and arachnids, with emphasis on groups including cave specimens |
| Grantee: | Pedro Gnaspini Netto |
| Support Opportunities: | Research Projects - Thematic Grants |