We assess the effects of gap characters on phylogenetic analyses and the relative levels of homoplasy of gap and base characters from a selection of published sequence-based matrices. Our results, based on structural rDNA, internal transcribed spacer (ITS) of rDNA, intron, and exon sequence-based matrices, demonstrate that: 1) gap characters can represent a considerable portion of the potential phylogenetic information in sequence-based matrices; 2) gap and base characters have similar levels of homoplasy; 3) including gap characters in sequence-based matrices often changes the tree topology and/or resolution of the most-parsimonious trees; and 4) including gap characters in sequence-based matrices often increases branch-support values. These empirical results, in combination with the theoretical bases given for using gap characters and rigorous methodologies with which to code gap characters, strongly support the use of gap characters in phylogenetic analyses that include sequence data from structural rDNA, ITS of rDNA, intron and/or exon regions. Our results suggest that longer gaps are not necessarily better phylogenetic characters than shorter gaps.

Key words: cladistics, deletions, gap characters, indels, insertions, molecular systematics