Unstable and Stable Classifications of Scombroid Fishes

Many cladists believe that a classification should strictly reflect a cladistic hypothesis. Consequently, they propose classifications that often differ markedly from existing ones and are potentially unstable due to phylogenetic uncertainty. This is problematic for economically or ecologically important organisms since changing classifications can cause confusion in their management as resources. The classification of the 44 genera of scombroid fishes (the mackerels, tunas, billfishes, and their relatives) illustrates this problem of instability. Previous cladistic analyses and analyses presented in this paper, using different data sets, result in many different cladistic hypotheses. In addition, the inferred cladograms are unstable because of different plausible interpretations of character coding. A slight change in coding of a single character, the presence of splint-like gill rakers, changes cladistic relationships substantially. These many alternative cladistic hypotheses for scombroids can be converted into various cladistic classifications, all of which are substantially different from the classification currently in use. In contrast, a quantitative evolutionary systematic method produces a classification that is unchanged despite variations in the cladistic hypothesis. The evolutionary classification has the advantage of being consistent with the classification currently in use, it summarizes anagenetic information, and it can be considered a new form of cladistic classification since a cladistic hypothesis can be unequivocally retrieved from an annotated form of the classification.