Premium
MORPHOLOGICAL AND ECOLOGICAL VARIATION WITHIN THE ACHNANTHIDIUM MINUTISSIMUM (BACILLARIOPHYCEAE) SPECIES COMPLEX 1
Author(s) -
Potapova Marina,
Hamilton Paul B.
Publication year - 2007
Publication title -
journal of phycology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.85
H-Index - 127
eISSN - 1529-8817
pISSN - 0022-3646
DOI - 10.1111/j.1529-8817.2007.00332.x
Subject(s) - biology , taxon , discriminant function analysis , morphological analysis , ordination , multivariate statistics , morphology (biology) , species complex , diatom , variation (astronomy) , zoology , evolutionary biology , ecology , statistics , artificial intelligence , mathematics , biochemistry , computer science , gene , phylogenetic tree , physics , astrophysics
Variation of frustular morphology within the Achnanthidium minutissimum (Kütz.) Czarn. species complex was studied in type populations of 12 described taxa and in 30 recent North American river samples. The SEM observations in this study and other publications showed that ultrastructural characters on their own do not discriminate among taxa within the A. minutissimum complex. Therefore, an attempt was made to use other characters, such as valve shape and striation pattern, to delineate morphological groups. The sliding‐landmarks method was used to obtain valve‐shape descriptors. These shape variables were combined with conventional morphological characters in multivariate analyses. It was shown that some historically recognized taxa are morphologically distinct, while others are difficult to differentiate. Morphological grouping of “old” taxa most similar to A. minutissimum did not correspond to their taxonomic hierarchy in contemporary diatom floras. Morphometric analysis of a data set of 728 specimens from North American rivers revealed six morphological groups, although it was impossible to draw clear boundaries among them. These morphological groups differed significantly in their ecological characteristics and could be recommended as indicators of water quality. Application of the discriminant function analysis based on shape variables and striation pattern showed that North American specimens could be more consistently classified into the six groups identified in our analysis than into historically recognized taxa.