Dear Dr. Andrew,
thank you very much for your comments, they helped me to solve the problem. I used all the trees generated by PAUP, each one containing branch lengths, and ran the SplitsTree program to obtain the consensus tree with the median branch lengths.
Best wishes,
Miguel.
--
Miguel Vizoso. Universidad de La Coruña.
Grupo de Investigación en Biología Evolutiva
www.udc.es/dep/bm/genetica/gibe