18S gene trees are positively misleading for monocot/dicot phylogenetics

Monocots are consistently paraphyletic in 18S gene trees in all studies to date. This anomaly is generally expressed in the phylogenetic associations of two lineages, that of Acoraceae, which is excluded from the monocots, and Ceratophyllaceae, which sometimes clusters within the monocots. Six explanations for these unexpected results are proposed: (1) erroneous published sequences, (2) actual paraphyly of monocots with dicots, (3) insufficient taxon density among relevant taxa; (4) long-branch attraction between selected taxa, (5) an anomalous 18S paralogue in the basal species of monocots, and (6) differential lineage sorting or other molecular evolutionary events. Tests of each of these hypotheses were conducted. For the first five hypotheses, test results refuted the hypothesis. Support for the sixth hypothesis, differential lineage sorting, is the strongest. Since lineage sorting seems to be restricted to a few species, identification and removal of these species prior to performing separate or combined phylogenetic analysis of DNA data incorporating sequences from 18S may be advisable.