Analysis of Mitochondrial nad1b-c Intron Sequences in Orchidaceae: Utility and Coding of Length-change Characters

Abstract

Abstract DNA sequences for 100 species representing all subfamilies of Orchidaceae and Hypoxis as outgroup were produced for the intron between the b and c exons of subunit 1 of mitochondrial NADH dehydrogenase (nad1). Although the proportion of variable sites is relatively low (slightly greater than rbcL in the same taxa), length variation is common, such that the intron ranges from 905 bp to approximately 2100 bp among the species. The low variability means that the sequences are fairly easily aligned, and provides an opportunity to consider the use of indels where they are a significant proportion of the data set. Critical examination of indel-coding schemes indicates that unordered multistate coding is most useful when nested indels are present. Cladistic analysis of base changes and indels alone, as well as a combined data set, reveals the importance of the indels as characters, since they provide significant additional resolution and branch support beyond that for base changes alone and have higher consistency and retention indices than base changes. The phylogenetic pattern obtained from this locus is consistent with recognition of either four or five subfamilies, and places the vanilloid orchids away from the epidendroids, in agreement with other molecular data. Communicating Editor: Matt Lavin