Phylogeny of Aulacoseira (Bacillariophyta)

Access full-text files

Date

2003

Authors

Edgar, Stacy McBride

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

The phylogeny of 67 populations representing 45 species of Aulacoseira is estimated by maximum parsimony methods using a combination of nucleotide sequence data and qualitative and quantitative morphological characteristics of the silica cell wall gathered primarily from original observation by LM and SEM. A new type of character employing continuous quantitative variables that describe the ontogenetic-allometric trajectories of cell wall characteristics over the life cycle (size range) of diatoms is introduced. In addition to the 45 Aulacoseira species, the phylogeny also incorporates one Alveolophora species, and two outgroup species (Melosira varians and Stephanopyxis cf. broschii). Fifteen species, represented by 24 populations, also contain molecular data from the chloroplast genome (rbcL) as well as the nuclear genome (18S), which were sequenced or downloaded from GenBank. The phylogeny of Aulacoseira is composed of five major clades: 1) an A. crenulata and A. italica clade, which is the most basal, 2) an A. subarctica and A. distans clade, 3) an A. granulata complex clade, 4) an A. ambigua clade, and 5) an A. islandica, A. skvorzowii, A. baicalensis, clade that also contains Alveolophora and many extinct Aulacoseira taxa. Monophyly of Aulacoseira is only achieved if Alveolophora, originally identified as Aulacoseira, is no longer given separate generic status. The choice of morphological characters, recognition of character states and explicit consideration of the states in coding are of great import in any phylogenetic study utilizing morphological data. Sensitivity of the Aulacoseira phylogeny to different coding methods was explored. Results indicate that use of step-matrix gap weighting utilizing the maximal number of character states allowable by phylogenetic software has two major advantages over other coding methods. Data are not manipulated in an effort to recognize gaps, i.e., the data remain in as raw a form as possible within the constraints of the requirements of phylogenetic software that they be in integer form. The maximal amount of potential phylogenetic signal contained within the data can contribute to the phylogenetic estimation. An additional methodological step is proposed in this study to allow large data sets to utilize the step-matrix gap weighting method.

Description

text

Keywords

LCSH Subject Headings

Citation